ENA is cheaper. HYPE is the higher-quality asset today, but you’re paying a very large premium for that quality.

Using current DeFiLlama data:

Hyperliquid currently generates roughly $721M annualized revenue, and critically, DeFiLlama classifies essentially all of that as holder revenue because 99% of eligible trading fees flow to the Assistance Fund to buy HYPE.

Ethena’s accounting is messier. DeFiLlama shows ~$310M annualized fees and a much smaller formal “revenue” figure because of how Ethena distributes yield, but its income statement shows materially larger gross protocol revenue. Ethena generated $65.1M in Q1, $51.6M in Q2 and $29.5M so far in Q3 2026.

Subscribe now

Get Actionable Reports in Real-time by Subscribing

Here’s how I think about it

HYPE = better business, more expensive token.

Hyperliquid is extraordinary economically. It has nearly $13B of open interest, $219B of 30-day perp volume, ~$960M annualized fees and ~$721M annualized revenue. And the token-value-accrual mechanism is already operating today.

But HYPE now has an ~$19B circulating valuation and is essentially at its all-time high around $85–86. So investors are already paying approximately:

$18.94B ÷ $721M = 26.3× revenue

And there is an even bigger valuation issue lurking behind it:

HYPE FDV = ~$81B.

That’s enormous.

ENA = inferior value capture today, dramatically cheaper optionality.

At ~$1.48B market cap, Ethena is only about 1/13th the market capitalization of HYPE.

Yet its underlying economic engine is nowhere near 1/13th as large.

If we conservatively use ~$175M of Ethena economic revenue:

ENA = ~8.5× revenue

versus:

HYPE = ~26× revenue

So HYPE is approximately 3× more expensive on revenue.

And ENA’s FDV is only ~$2.26B versus HYPE’s ~$81B. That’s a dramatic difference.

But there’s One Huge Distinction

Read more

It could write an email, analyze a spreadsheet, generate software, search through a database, create an image, or hold a conversation. The intelligence was increasingly impressive, but it was still largely trapped inside the digital world.

That boundary is beginning to disappear.

The next major phase of artificial intelligence is Physical AI: machines that can perceive the physical world, reason about it, make decisions, and then act.

And in 2026, we are beginning to see the pieces come together.

Robotics companies are building increasingly capable bodies. AI companies are building increasingly capable brains. Semiconductor companies are building the computing platforms that allow those brains to operate locally and in real time. Simulation platforms are creating virtual environments where robots can practice millions of times before performing a task in the real world.

Perhaps most importantly, large companies are beginning to put these systems to work.

This is a significant development in what I call the Age of Autonomy.

AI is no longer simply learning how to think.

AI is getting a body.

From Language Models to World Models

Large language models taught machines something enormously valuable: how to understand context.

That may sound straightforward, but context is what separates a traditional software program from something approaching intelligence.

Traditional software follows instructions.

If this happens, do that.

AI models can instead evaluate an unfamiliar situation, interpret what is happening, and determine an appropriate response.

Robots need exactly that ability.

A conventional industrial robot might be extraordinarily precise, but it generally operates within a tightly controlled environment. A robotic arm on an automobile assembly line may perform the same weld thousands of times.

Move the component six inches, however, and the system may have a problem.

Physical AI changes the paradigm.

Instead of programming every movement, engineers are increasingly building models that allow robots to understand the world around them.

Google DeepMind calls one important class of these systems Vision-Language-Action models, or VLAs. A VLA connects what a machine sees with language-based reasoning and ultimately with physical actions.

Google’s newest Gemini Robotics 2 models take the idea further, combining visual understanding, physical reasoning, dexterity and whole-body control. DeepMind describes the goal as allowing robots of different shapes and sizes to intelligently interact with the physical world rather than merely executing predetermined sequences.

Consider how different that is from traditional automation.

You can tell a machine:

“Pick up the blue container from the table and place it on the shelf.”

• The robot has to identify the table.

• It has to determine which container is blue.

• It has to understand where the shelf is.

• It has to estimate the object’s dimensions and weight.

• It has to determine how to grasp it.

• It has to navigate around whatever happens to be between the table and the shelf.

• And if something changes while it is moving, it has to adjust.

These are trivial tasks for a human being.

For a robot, they require perception, spatial reasoning, planning, motor control, and continuous feedback. That is why the AI revolution and the robotics revolution are becoming the same revolution.

Robots Are Starting to Learn Instead of Being Programmed

There is another important change taking place.

Robots increasingly don’t have to learn everything in the real world.

They can learn in simulation.

NVIDIA has made this a central part of its Physical AI strategy through its Isaac robotics platform and Cosmos world models. Developers can construct virtual environments, generate synthetic data, train robot policies, test edge cases and then deploy those capabilities onto physical machines.

This matters because real-world training is expensive.

If you want a robot to learn how to pick up thousands of different objects, you could place those objects in front of a physical robot and have it practice repeatedly.

• Or you could create millions of variations inside a simulation.

• Change the lighting.

• Change the object’s position.

• Change its shape.

• Change the floor.

• Put another object in the way.

• Have the robot fail.

• Reset the simulation instantly.

• Try again.

The result is something analogous to the development of autonomous vehicles. Before a system encounters an unusual situation in the real world, developers can expose it to enormous numbers of simulated situations.

The combination of AI models + simulation + synthetic data + robotics hardware dramatically accelerates the learning curve.

And that creates a potentially important feedback loop.

More deployed robots generate more real-world data.

More data creates better models.

Better models make robots more useful.

More useful robots lead to more deployments.

Those deployments generate still more data.

This is exactly the kind of flywheel that has driven other major computing platforms.

The Humanoids Are Clocking In

The most visible manifestation of Physical AI is the humanoid robot.

There is a practical reason companies are pursuing the humanoid form.

• Our world has already been built for humans.

• Factories have stairs.

• Warehouses have shelves.

• Buildings have doors.

• Tools have handles.

• Vehicles have controls.

• Workstations are positioned for people with two arms, two hands and roughly human proportions.

• Rather than rebuilding every environment around machines, one strategy is to build machines that can operate inside environments we already have.

• And in 2026, humanoids are beginning to move beyond carefully staged demonstrations.

BMW provides one of the clearest examples.

At BMW’s Spartanburg factory, Figure’s Figure 02 humanoid participated in the production of more than 30,000 BMW X3 vehicles during a roughly ten-month deployment. According to BMW, the robot moved more than 90,000 components, accumulated approximately 1,250 operating hours and walked roughly 1.2 million steps.

That is considerably more interesting than a robot folding a towel on YouTube.

It means the machine operated inside an actual production environment.

Figure has now returned to BMW with its Figure 03 robot and a more sophisticated logistics application. Its newer Helix AI system coordinates hands, arms, torso and legs to perform tasks requiring whole-body movement rather than a simple fixed pick-and-place routine.

Toyota is moving in a similar direction.

In February, Toyota Motor Manufacturing Canada signed a commercial Robots-as-a-Service agreement with Agility Robotics after testing Agility’s humanoid robot, Digit. The machines are intended to support manufacturing, logistics and supply-chain operations.

That distinction matters. It isn’t another pilot announcement. It is a commercial agreement following a pilot.

Boston Dynamics Is Finally Turning Atlas Into a Product

Boston Dynamics has been producing some of the world’s most astonishing robotics demonstrations for years.

The old Atlas could run, jump, climb and perform movements that sometimes seemed more appropriate for an action movie than a factory.

But impressive demonstrations aren’t necessarily businesses.

The transition occurring now is much more significant.

In January, Boston Dynamics unveiled the production version of Atlas, its fully electric humanoid designed specifically for industrial work. The company says it began manufacturing the robot immediately, with its 2026 deployments already committed to Hyundai and Google DeepMind.

Hyundai provides Boston Dynamics with something robotics startups rarely possess: an enormous industrial testing ground.

Hyundai plans to introduce Atlas into manufacturing operations gradually, beginning with parts-sequencing applications at its Georgia Metaplant in 2028 and moving into more complicated assembly applications later. Hyundai is also building the Robot Metaplant Application Center, essentially an environment for training, testing and validating robots before putting them onto production lines.

There is an even more interesting relationship developing underneath this.

Boston Dynamics and Google DeepMind are collaborating.

One company has spent decades building some of the world’s most sophisticated robotic bodies.

The other is building increasingly sophisticated AI brains.

Hyundai describes the thesis nicely: combining a general-purpose brain with a highly capable generalist body.

That may ultimately be one of the defining technological combinations of the next decade.

Subscribe now

For Actionable Alerts of Marketing Timing, Asset Selection and Portfolio Construction, Subscribe!

NVIDIA Wants to Build the Operating System for Physical AI

There is another company appearing almost everywhere in this story. NVIDIA. That should not be surprising.

The first stage of the AI boom required enormous amounts of computing power for training foundation models.

Physical AI requires that compute in even more places.

Robots need to process cameras, lidar, microphones and other sensors. They need to reason about what they see. They need to generate actions. And many of those calculations need to happen in milliseconds rather than waiting for a distant cloud server.

NVIDIA increasingly describes its opportunity as extending AI from the data center into the physical world.

Its stack now stretches across much of the robotics development process:

• GPUs train the models.

• Omniverse and Cosmos simulate the world.

• Isaac trains and develops robotic behavior.

• Jetson brings inference onto the robot itself.

And NVIDIA is building safety systems around the entire stack.

In June, NVIDIA introduced Halos for Robotics, a full-stack safety architecture intended for machines that sense, decide and act in the physical world. Agility Robotics was announced as its first humanoid robotics partner.

Then this summer NVIDIA expanded its Jetson Thor family for advanced robotics and edge AI. And just this week it announced the Jetson Orin Nano 2, offering twice the inference performance of its predecessor while consuming substantially less power at equivalent performance. NVIDIA says more than three million developers have now built on its robotics stack.

This is an important point for investors. The robotics opportunity will not belong exclusively to companies selling robots.

Just as the smartphone boom created enormous businesses in chips, operating systems, sensors, networking, software and applications, the Physical AI economy will have an entire technology stack.

The robot may be the visible product. The larger investment opportunity could exist across the infrastructure underneath it.

Intelligence Is Moving to the Edge

There is another subtle shift worth watching closely.

• Robots cannot depend entirely on the cloud.

• A chatbot can sometimes wait a second before answering.

• A robot carrying a 40-pound component near a human worker cannot.

• It has to perceive and react immediately.

• That makes edge computing increasingly important.

• Google DeepMind’s Gemini Robotics On-Device 2 is designed specifically to run robotic intelligence locally across multiple kinds of robots.

• This moves us toward a computing architecture where intelligence exists everywhere.

• Some AI will live in giant data centers.

• Some will live in vehicles.

• Some will live inside factories.

• Some will live inside cameras and sensors.

• And some will literally walk around.

• The internet connected computers.

The smartphone connected people.

IoT began connecting physical objects.

Physical AI gives those objects intelligence and agency.

Why This Matters for the Age of Autonomy

I have argued for years that the Age of Autonomy will emerge from the convergence of several technologies rather than from any single breakthrough.

• AI provides intelligence.

• IoT provides perception.

• Robotics provides physical action.

• Blockchain and programmable money can provide economic coordination and settlement.

• Edge computing allows decisions to happen locally.

• We can now begin to see how those components fit together.

• Imagine a warehouse ten years from now.

• Sensors constantly observe inventory and equipment.

• AI systems forecast demand.

• Autonomous software agents order supplies.

• Robots unload trucks, move inventory and perform maintenance.

• Autonomous vehicles transport goods between facilities.

• Machines purchase electricity or computing resources dynamically.

• Payments settle automatically.

• Humans supervise the overall system, solve higher-level problems and determine objectives.

• This isn’t simply automation.

• It is an autonomous economic system.

And that is what makes the intersection of AI and robotics so important.

We are moving from machines that perform predefined actions toward machines that can observe their environments, reason about what is happening and determine what action to take next.

That is a fundamentally different capability.

Subscribe now

For actionable news, subscribe!

The Next Great Computing Platform May Walk Around

The early internet gave computers the ability to communicate.

Cloud computing gave companies effectively unlimited computing resources.

Smartphones placed a computer in billions of pockets.

Generative AI gave computers the ability to understand and create.

Physical AI gives computers something we have never deployed at scale before:

the ability to interact intelligently with the physical world.

We are still early.

Today’s humanoids remain expensive. Reliability has to improve. Dexterity remains difficult. Battery life matters. Safety matters enormously. Many impressive demonstrations will fail to become viable products.

There will almost certainly be a shakeout among today’s robotics companies.

But those caveats should not obscure what is changing.

A Figure robot has already worked on tens of thousands of BMWs.

Toyota has moved from piloting humanoid robots into a commercial deployment agreement.

Boston Dynamics is manufacturing a product version of Atlas.

Google DeepMind is building foundation models designed specifically for physical reasoning and action.

NVIDIA is building an increasingly complete computing stack around the emerging Physical AI industry.

These are no longer disconnected robotics experiments.

They are pieces of an emerging computing platform.

The Age of Autonomy has always been about what happens when intelligence escapes the computer screen and begins participating directly in the economy.

That transition is now becoming visible.

• AI learned to speak.

• Then it learned to reason.

• Now it is learning to act.

And once intelligence can reliably act in the physical world, the definition of what a machine can do changes permanently.

Happy Hunting,

Jake Ryan

Founder and CIO, Tradecraft Capital
Author of Crypto Investing in the Age of Autonomy and Crypto Decrypted

When we published the Wildcatters 5, the idea was not to chase whatever was hot for a week or two. It was to assemble a small basket of companies that fit a broader thesis: a world that needs more reliable energy, more domestic supply, and more strategic control over critical resources. So far, that thesis has been working.

Since our last published price update, the equal-weight Wildcatters 5 basket is up approximately +12.8%. Four of the five names are positive, and two of them — Perpetua Resources and Energy Fuels — have turned in particularly strong gains. Liberty Energy has been the one notable laggard, and that gives us something useful to think through as we evaluate how the basket is evolving.

Subscribe now

Subscribe and Get Actionable Intel!

Below is a simple scorecard showing where each stock stood when we last published it, where it trades now, and the percentage move since then.

Wildcatters 5 Performance Table

Read more

They have had something resembling a brain—but not a nervous system. That is beginning to change. A new generation of intelligent machines is being connected to cameras, microphones, radar, lidar, vibration monitors, temperature gauges, GPS devices, wearables, industrial-control systems, and billions of other sensors. These devices allow AI to observe physical conditions continuously rather than waiting for a person to describe what happened.

Once AI can perceive the physical world, it can begin to interpret what it sees. Once it can interpret those conditions, it can make decisions. And once it is connected to software, vehicles, industrial equipment, or robots, it can increasingly act on those decisions.

This creates a progression that sits at the heart of the Age of Autonomy:

Sense. Interpret. Decide. Act.

The Internet of Things (IoT) created the sensory layer. Artificial intelligence is giving that sensory layer understanding. Edge computing allows decisions to be made close to where events occur. AI agents can coordinate the response. Robotics and automated machinery can then carry out the action.

This is how AI begins to move beyond cyberspace and into the physical economy.

From Connected to Intelligent Machines

The original promise of the Internet of Things was relatively straightforward: connect physical devices to the internet and collect information from them.

A connected motor could report its temperature. A truck could transmit its location. A camera could record activity inside a warehouse. A thermostat could measure conditions in a building. An industrial machine could send vibration readings to a maintenance dashboard.

That was useful, but it did not necessarily make the device intelligent.

Much of the first generation of IoT consisted of sensors feeding data into dashboards, where humans still had to determine what the information meant and what should be done about it. In many cases, companies created more telemetry without creating better decisions.

AI changes that equation.

A traditional sensor might report that a motor is vibrating more than usual. An AI system can compare that vibration pattern against billions of historical measurements and determine that a bearing is deteriorating.

A connected camera can record traffic. Computer vision can recognize that a driver is distracted, that a forklift is approaching a worker, or that a package has been placed on the wrong conveyor.

A temperature sensor can record that a machine is running hot. AI can combine that information with maintenance records, production schedules, operating conditions, and energy usage to determine whether the machine should be serviced, slowed down, or temporarily removed from production.

The progression looks something like this:

• Connected: “The motor is vibrating.”

• Intelligent: “The bearing is likely deteriorating.”

• Agentic: “I have created a maintenance order, identified the replacement part, adjusted the production schedule, and alerted the supervisor.”

The difference between these stages is the difference between observation and autonomy.

The Industrial World Is Becoming Software-Defined

One of the clearest examples of this transition emerged in January 2026, when Siemens and NVIDIA expanded their partnership to develop what they call an Industrial AI Operating System.

The objective is larger than adding an AI assistant to a factory. Siemens and NVIDIA want to integrate artificial intelligence throughout the industrial lifecycle—from product design and engineering to manufacturing, production, operations, and supply-chain management.

The partnership combines Siemens’ industrial software, automation systems, and digital twins with NVIDIA’s accelerated computing, simulation technology, and AI platforms. The two companies are also working toward AI-driven, adaptive factories that can continuously optimize their own operations.

The operating-system analogy is important.

Factories have traditionally consisted of hundreds or thousands of individual machines, control systems, databases, and specialized software applications. Much of this infrastructure was developed at different times and operates in separate silos.

An industrial AI operating system attempts to create a common intelligence layer across the physical plant.

Sensors observe machines, production lines, inventory, workers, energy consumption, and environmental conditions. Digital twins organize those inputs into a virtual representation of the facility. AI models identify patterns, predict failures, and simulate alternative decisions. Software agents coordinate workflows across systems. Automated equipment or robots can eventually execute the selected response.

The factory becomes less like a collection of isolated machines and more like a living system with perception, memory, analysis, and coordinated action.

The Factory That Can Rehearse the Future

PepsiCo provides one of the most interesting examples of this convergence.

Working with Siemens and NVIDIA, PepsiCo is creating high-fidelity digital replicas of manufacturing plants, warehouses, equipment, conveyor systems, pallet routes, and worker movements.

These are not merely static three-dimensional models. Siemens’ Digital Twin Composer connects engineering data, operational information, computer vision, simulation models, and real-time sensor streams inside a virtual environment built partly on NVIDIA Omniverse.

AI agents can then run thousands of simulated scenarios before PepsiCo changes the physical facility. The company can test new equipment layouts, identify bottlenecks, model material flows, evaluate automation decisions, and anticipate problems before spending money or interrupting production.

Siemens reports that the technology has allowed PepsiCo to identify as much as 90% of potential problems before physical modifications are made. Another Siemens account cites a 20% throughput improvement on repurposed production lines. Because these figures come from the companies involved, they should be understood as company-reported results, but they nonetheless demonstrate the commercial ambition behind the technology.

The deeper significance is not simply that PepsiCo has a better simulation tool.

A continuously updated digital twin creates a machine-readable representation of physical reality. AI can use that representation to rehearse decisions before executing them in the real world.

The digital twin becomes a kind of parallel environment where autonomous systems can ask:

• What happens if we move this conveyor?

• What happens if a machine fails during peak production?

• What happens if demand rises by 20%?

• What happens if a robot takes over this task?

• What happens if we reroute inventory through a different facility?

In the past, companies often discovered the answers after making the physical change. Increasingly, they can discover them first in simulation.

The factory can build twice: once in the virtual world and once in reality.

Samsara and the Digitization of Physical Operations

While the Siemens–NVIDIA partnership illustrates the architecture of industrial autonomy, Samsara demonstrates that sensor-driven intelligence is already becoming a substantial commercial business.

Samsara connects vehicles, cameras, trailers, equipment, workers, and physical assets across industries such as trucking, construction, utilities, logistics, field services, manufacturing, and government.

Its platform gathers information from GPS devices, vehicle telematics, AI-enabled dash cameras, equipment monitors, environmental sensors, and mobile applications. Artificial intelligence then interprets that data to improve driver safety, maintenance, routing, fuel consumption, regulatory compliance, and workforce productivity.

Samsara ended its 2026 fiscal year with approximately $1.9 billion in annual recurring revenue, representing 30% year-over-year growth. It reported $432 million in net new recurring revenue during the year and $1.2 billion of recurring revenue from customers spending more than $100,000 annually.

The company continued that momentum in the first quarter of fiscal 2027, reported in June 2026, with revenue of approximately $479 million, up 31% from the previous year, and annual recurring revenue approaching $2 billion.

These numbers matter because they show that the convergence of sensors and AI is not confined to laboratories or futuristic robot demonstrations. Businesses are already paying significant amounts to make physical operations more observable and more intelligent.

Samsara’s long-term ambition is to become an operating system for the physical economy.

Its newest AI capabilities are moving beyond historical reporting toward systems that can communicate with workers, recognize operational risks, generate reports, automate repetitive processes, and coordinate responses. At its 2026 Beyond conference, the company introduced additional AI tools for fleet safety, shipping, communications, and customized operational workflows.

Samsara also says its platform helped customers prevent approximately 380,000 accidents during fiscal 2026. That is a company-reported estimate, but it provides a useful illustration of the potential value created when sensors and AI are connected to real-time behavior.

A sensor network becomes more valuable when it does more than record an accident. It becomes valuable when it recognizes the conditions that precede an accident and intervenes before one occurs.

Subscribe now

For actionable reports focused on market timing, portfolio construction and asset selection in digital assets, join us!

The Industrial AI Workforce

Augury represents another important layer of this emerging system.

The company places sensors on industrial machinery and analyzes signals such as vibration, temperature, magnetic activity, and ultrasonic behavior. These measurements can reveal subtle changes that indicate misalignment, wear, lubrication problems, mechanical stress, or an approaching equipment failure.

Predictive maintenance is not a new concept, but the technology is becoming considerably more capable.

Early systems generated alerts. More sophisticated systems diagnosed likely mechanical problems. The next generation is beginning to combine sensor readings with production data, maintenance history, environmental conditions, and operational context.

In 2026, Augury began describing this architecture as an Industrial AI Workforce.

At the center of the system is what Augury calls an Industrial Context Graph—a continuously updated information layer that connects machine-health signals to process, operational, and environmental data. AI agents can use this context to reason about cause and effect across a production system rather than evaluating each machine in isolation.

This matters because a machine does not operate in a vacuum.

A pump may be vibrating because it is damaged. It may also be vibrating because another part of the production process has changed, because the material flowing through it is different, or because an upstream machine is operating incorrectly.

Industrial intelligence requires context.

Augury says its platform is built on more than 1.1 billion hours of machine readings. Its newer systems are intended to move from predicting problems to helping maintenance and operations teams prioritize work, initiate processes, and coordinate action across other systems.

This is a subtle but important transition. AI is moving from being a tool that provides an answer to becoming a participant in an operational workflow.

Intelligence Is Moving to the Edge

Another critical development is taking place inside the sensors themselves.

Historically, connected devices often sent raw information to a centralized data center or cloud platform for analysis. That architecture works for many applications, but it has limitations.

Physical systems frequently operate under strict requirements for latency, reliability, privacy, bandwidth, and energy consumption. A vehicle approaching an obstacle cannot always wait for a cloud server to interpret the camera feed. A factory safety system cannot become useless when internet connectivity is interrupted. A wearable device should not necessarily transmit every raw measurement about its user to a remote database.

For these reasons, more AI inference is moving to the edge—closer to the sensor and the physical event.

An intelligent camera may determine locally whether something unusual has occurred and transmit only the relevant event. A machine sensor may recognize an abnormal vibration signature before sending an alert. A vehicle may fuse information from radar, cameras, and other sensors without waiting for the cloud.

This makes companies such as Bosch and NXP important participants in the Age of Autonomy.

Bosch is integrating sensing, microcontrollers, software, and artificial intelligence into automotive systems, industrial equipment, robotics, and consumer devices. In January 2026, Bosch highlighted plans to invest more than €2.5 billion in AI through the end of 2027.

The investment includes AI applications in manufacturing, supply chains, driver assistance, robotics, and autonomous trucking. Bosch is also working with Microsoft on agentic AI for factories and with Kodiak AI on autonomous-trucking systems involving sensors, steering, and vehicle hardware.

NXP Semiconductors, meanwhile, is seeing what its CEO calls physical AI begin to affect actual semiconductor demand. The company is also developing edge-computing and sensor-fusion systems intended for robotics and other physical-AI applications.

In February 2026, NXP said it was experiencing stronger demand for chips used in logistics, workplace safety, robotics, drones, and factory automation. Technologies originally developed for automotive radar, sensing, and embedded control are increasingly being adapted to industrial applications. NXP projected approximately 20% year-over-year growth in its industrial-chip business for the first quarter.

By the second quarter, the company was continuing to cite physical-AI demand as a contributor to growth across automotive and industrial markets.

This is one of the most important signs of traction.

A theme becomes more credible when it begins to appear not only in conference presentations but also in semiconductor orders, recurring software revenue, industrial capital spending, and customer deployments.

Safety Becomes Part of the Technology Stack

As AI systems become capable of acting in the physical world, the consequences of failure become more serious.

A chatbot can produce an incorrect answer. A robot, autonomous vehicle, or industrial-control system can cause physical damage.

That means physical AI requires a safety architecture in addition to an intelligence architecture.

In June 2026, NVIDIA announced Halos for Robotics, a full-stack safety system for robots and physical-AI applications. The platform brings together sensor connectivity, AI computing, system software, safety applications, validation, and support for third-party certification.

Agility Robotics became the first announced company to incorporate elements of the system into its humanoid-robot safety architecture. Agility’s robots are being used or evaluated in manufacturing and logistics environments associated with companies including Amazon, GXO, Schaeffler, and Toyota Motor Manufacturing Canada.

The existence of this safety layer tells us something about the maturation of the industry. Early technology cycles ask whether something is possible. Commercial technology cycles must ask whether it is reliable, secure, repeatable, certifiable, and safe around people.

Physical AI is beginning to move from the first question toward the second.

The Missing Ingredient Is Context

Sensors generate enormous amounts of information, but raw information alone is not enough.

For AI to manage a physical asset intelligently, it must understand more than the asset’s current sensor readings. It may also need to know:

• How was the machine designed?

• Which components are installed?

• How old are they?

• What maintenance has previously been performed?

• What operating conditions are normal?

• What production process depends on the machine?

• Is a replacement part available?

• How much would downtime cost?

Much of this information currently lives in separate systems: engineering databases, maintenance platforms, enterprise-resource-planning software, customer-service systems, IoT platforms, and handwritten field notes.

The next phase of industrial autonomy therefore requires systems that can assemble a coherent history and identity for each physical asset.

This contextual layer may ultimately become as important as the AI model itself.

The model supplies intelligence, but the operating data, sensor history, asset records, and domain-specific knowledge tell the model what it is reasoning about.

Subscribe now

For actionable reports, join us!

Why This Matters for the Age of Autonomy

The Age of Autonomy is not simply an AI cycle.

It is the convergence of artificial intelligence, the Internet of Things, robotics, and blockchain technology. Each layer solves a different part of the autonomy problem.

• AI provides intelligence.

• Sensors provide perception.

• IoT networks provide communication.

• Digital twins provide a structured representation of physical reality.

• Agents provide coordination and execution.

• Robotics provides physical action.

• Programmable money (blockchain technology) allows machines and software agents to transact with one another.

The systems emerging in 2026 are not yet fully autonomous. Human operators remain essential. Most factories cannot redesign themselves, order every needed part, negotiate with suppliers, and deploy robots without supervision.

But the relevant component technologies are beginning to converge into commercially deployable operating systems. That is the more credible and meaningful milestone.

The sensors already existed. The industrial machines already existed. The connectivity largely existed. AI models existed. What is changing is that perception, simulation, reasoning, agents, and physical action are beginning to form a continuous feedback loop.

All together, autonomous operations generate economic activity without the need for human intervention (or minimized intervention).

The Investment Implication

The most obvious beneficiaries of physical AI may appear to be robotics companies, but the opportunity is considerably broader.

Every autonomous system requires a technology stack beneath it:

• Sensors, cameras, radar, lidar, and MEMS devices

• Microcontrollers and edge processors

• Industrial connectivity and networking

• AI accelerators and embedded computing

• Digital-twin and simulation software

• Machine-health and predictive-maintenance platforms

• Fleet and physical-operations software

• Robotics platforms and safety systems

• Operational-technology cybersecurity

• Data-center, electrical, and communications infrastructure

The central investment question may not be simply:

• Who has the most powerful AI model?

It may be:

• Who owns the physical-world data, the installed sensor network, the operating context, and the permission to act?

General-purpose AI models will continue to become more capable, but an industrial model cannot diagnose a machine it cannot observe. It cannot optimize a factory whose systems it cannot access. It cannot manage a fleet without real-time information from the vehicles. It cannot safely control a robot without trusted perception and low-latency computing.

The companies with entrenched access to physical assets may possess a form of competitive advantage that is difficult to replicate from the cloud alone.

Their moats may include installed hardware, proprietary sensor data, domain expertise, customer workflows, integration with legacy equipment, regulatory approvals, and years of operational history.

In the Age of Autonomy, the most valuable data may increasingly be generated not by what people say online, but by what machines experience in the physical world.

In Closing

The internet connected people.

The Internet of Things connected machines.

Artificial intelligence is now teaching those machines to understand what they sense.

AI agents will increasingly allow them to decide and coordinate. Robotics will allow them to manipulate the physical environment. Programmable money will eventually allow them to purchase services, pay other machines, and participate directly in economic activity.

This transformation will not happen all at once. Nor will every connected device suddenly become autonomous. But the direction is becoming clearer.

AI is growing a nervous system.

Its sensors are spreading across factories, vehicles, warehouses, cities, farms, homes, and infrastructure. Its intelligence is moving from centralized data centers toward the edge. Its models are being connected to digital twins, operational systems, agents, and machines.

For the first time, artificial intelligence is beginning to close the loop between understanding the world and acting within it. That is when AI stops being merely a tool we use. It begins to become an economic actor.

And that is when the Age of Autonomy truly enters the physical world.

Best,

Jake Ryan
Founder and CIO, Tradecraft Capital
Author of Crypto Investing in the Age of Autonomy and Crypto Decrypted

The War with Iran and the extreme demand being created by AI creates a unique opportunity in the energy space. Tech companies are buying up all the supply of nuclear facilities. Some are looking to build their own with new types of small capacity nuclear reactors being constructed. Natural gas is being pumped like never before. The demand for energy is increasing. Where are those companies looking for new energy sources?

Why Now?

The war with Iran and instability around the Strait of Hormuz are forcing Asian buyers to rethink where their energy comes from. The United States is emerging as a more strategically important source of oil and gas—not necessarily because every American barrel is cheapest at the dock, but because reliable supply outside the world’s most dangerous energy chokepoint can be more attractive on a risk-adjusted basis.

Energy Security Is Becoming As Important As Price

For decades, Asian economies optimized energy procurement around cost, proximity, and long-standing Middle Eastern supply relationships. The Iran conflict has exposed the vulnerability in that model. When tankers face attack, traffic slows, insurance premiums rise, delivery schedules become uncertain, and governments worry about strategic reserves, the lowest quoted barrel may no longer be the lowest-cost barrel in practice.

This is not simply a temporary oil-price trade. It is a reminder that supply reliability has economic value—and that buyers will pay for diversification when a critical route becomes politically or militarily fragile.

Asia Carries Most of the Hormuz Risk

The Strait of Hormuz is the world’s most important oil transit chokepoint. In the first half of 2025, the U.S. Energy Information Administration estimated that 89% of the crude oil and condensate passing through Hormuz was destined for Asian markets. China, India, Japan, and South Korea alone accounted for roughly three-quarters of those flows.

LNG exposure is also concentrated in Asia. Qatar’s exports and other Persian Gulf gas shipments depend on the same narrow waterway, making the conflict relevant not only to crude oil but also to electricity generation, industrial demand, and heating across the region.

America’s New Role

The United States is exporting far more energy. America has moved from energy scarcity to energy abundance. U.S. LNG exports rose from roughly 0.5 billion cubic feet per day in 2016 to 15.0 billion cubic feet per day in 2025. Then, as Hormuz disruptions tightened global supply, total U.S. petroleum exports reached a record 13.6 million barrels per day in April 2026—15% above the previous monthly record.

The Key Distinction

U.S. energy will not always win on headline spot price or shipping distance. But the true delivered cost of energy includes more than the commodity itself. It includes marine insurance, security premiums, rerouting, delays, inventory buffers, financing costs, and the possibility that a cargo does not arrive when needed. Once those risks are included, American supply can become more attractive even when its quoted price is higher.

COMMODITY + SHIPPING + INSURANCE + DELAY RISK + SUPPLY SECURITY

Age of Autonomy Intersecting w/ the Geopolitical World

We’ve researched the sector based on the Age of Autonomy thesis and the current geopolitical backdrop. We’ve found 5 energy companies poised for huge growth over the next 1-3 years.

Read more

For the last several years, artificial intelligence has been framed mostly as a story about intelligence. Better models. Better reasoning. Better coding. Better text, images, video, and voice.

That was the first phase.

The next phase is more important. It is not just about whether AI can answer questions. It is about whether AI can take action.

That is the real heart of the Age of Autonomy. In this next technological cycle, AI is not merely a tool that waits for a human prompt. It becomes an economic actor: something that can perceive, decide, transact, coordinate, and execute. That shift requires more than better models. It requires payments, identity, data access, APIs, wallets, compliance, auditability, robotics, and physical-world infrastructure.

Over the last month, we saw meaningful progress across several of those layers. Some of it came from AI companies. Some came from crypto infrastructure. Some came from stablecoin regulation. Some came from robotics. Taken together, the signal is becoming clearer: the Age of Autonomy is moving from narrative to infrastructure.

AI Agents Are Becoming a Real Workflow Layer

One of the most important recent data points came from OpenAI’s research on Codex, its agentic coding system. OpenAI reported that Codex usage grew rapidly in the first half of 2026, with active users increasing more than fivefold. The paper also noted that adoption has expanded beyond the original developer audience and that users are beginning to manage multiple concurrent agents and delegate more complex work. A particularly important finding was that the share of individual Codex users submitting at least one task estimated to require more than eight hours of experienced human work increased nearly tenfold since the beginning of the year.

That matters because coding agents are one of the clearest early examples of AI moving from “assistant” to “delegated worker.” A chatbot gives you an answer. An agent can take a task, break it into steps, interact with tools, generate output, revise its own work, and increasingly operate over longer time horizons.

This is an early but important transition. Software development is likely to be one of the first major labor markets reshaped by agentic AI because the work is already digital, tool-based, and measurable. The same pattern will eventually extend into legal research, financial analysis, customer service, operations, marketing, compliance, and investment research.

The investment implication is straightforward: the value will not accrue only to the model companies. It will also accrue to the infrastructure surrounding agents — workflow orchestration, permissions, identity, security, compliance, data access, observability, and payments.

Agents are becoming workers. Workers need tools. Economic actors need rails.

The Web Is Starting to Accept Software as a Customer

The most directly relevant development for the Age of Autonomy came from Coinbase and AWS. In June, Coinbase announced that it is working with AWS CloudFront and AWS WAF to let publishers and API providers accept AI agents as customers through x402. The idea is simple but powerful: an AI agent requests content or an API resource, receives a payment instruction through the HTTP 402 status code, pays programmatically, and then receives access.

This is exactly the kind of infrastructure the internet has been missing.

The original web was built around information exchange, not value exchange. We solved communication through protocols like HTTP, SMTP, and TCP/IP. But payments were bolted on later through cards, banks, merchant processors, app stores, and platform intermediaries. That worked reasonably well for human commerce, but it is poorly suited for autonomous software.

Agents will need to pay for data, APIs, compute, storage, identity checks, financial information, media, access rights, and eventually other agents. Many of those transactions will be small, frequent, automated, and global. Legacy payment systems were not designed for that kind of activity.

That is why x402 is important. Earlier this year, the Linux Foundation launched the x402 Foundation, with Coinbase contributing the protocol and with support from companies including Stripe, Cloudflare, and AWS. The foundation described x402 as a standard for embedding payments directly into web interactions, enabling AI agents, APIs, and applications to transact value as seamlessly as they exchange data.

This is one of the clearest examples of the Age of Autonomy thesis becoming concrete. AI agents do not just need intelligence. They need economic agency. Economic agency requires payments. And programmable money is the natural substrate for machine-native commerce.

Stablecoin Regulation Is Becoming the Settlement Layer Story

The stablecoin story also continued to mature over the last month. The OCC issued additional June guidance and proposed rulemaking materials related to the GENIUS Act, including reporting requirements and anti-money-laundering / countering-the-financing-of-terrorism rules for payment stablecoin issuers.

This may sound dry, but it is central to the Age of Autonomy.

If AI agents are going to transact, they need a settlement asset. If tokenized assets are going to move across programmable rails, they need a cash leg. If APIs are going to charge software directly, they need a low-friction payment mechanism. If robots eventually coordinate with digital services, inventory systems, supply chains, or financial accounts, they need a way to settle value.

Regulated stablecoins are the most likely candidate.

The political fight around stablecoins also tells us how important this has become. Community banks have pushed back against stablecoin legislation, warning that stablecoins could draw deposits away from traditional banks and reshape local credit creation.

That opposition is understandable. Stablecoins are not just another crypto product. They are a potential re-architecture of deposits, payments, settlement, and dollar distribution. If dollar-backed tokens become embedded into apps, wallets, exchanges, fintech platforms, AI agents, and tokenized asset markets, then the competitive landscape for money changes.

In the Age of Autonomy, stablecoins are not simply “crypto cash.” They are a programmable settlement layer for autonomous economic activity.

Market Structure Is Moving Toward Rules, but the Fight Is Not Over

Digital asset market structure also remains highly relevant. In May, the Senate Banking Committee advanced the Digital Asset Market Clarity Act, a broad market-structure bill that addresses asset classification, DeFi oversight, stablecoin yield, developer protections, tokenization standards, and customer-property protections.

That matters because the Age of Autonomy will require more than stablecoins. It will require programmable networks, decentralized applications, tokenized assets, and protocols that can coordinate value without relying entirely on closed platforms.

The key regulatory question is whether the United States creates a framework that allows those systems to be built domestically, or whether the next generation of financial and autonomous infrastructure migrates offshore.

For years, crypto builders operated in a fog of uncertainty. The core question was not just whether Bitcoin was legal. It was whether developers could launch networks, tokens, governance systems, applications, and financial infrastructure without constantly wondering whether the rules would change after the fact.

That uncertainty has mattered. It has shaped where founders build, where capital flows, and how aggressively institutions engage with the space.

Market structure legislation is not exciting in the same way as a new AI model or a humanoid robot demo. But it may be more important. The Age of Autonomy needs legal infrastructure as much as technical infrastructure.

Tokenized Assets Are Growing Up

Real-world asset tokenization also continued to mature. According to RWA.xyz, tokenized U.S. Treasuries now represent roughly $14.85 billion in distributed value. That’s big and growing.

That is still small relative to traditional fixed income markets, but it is large enough to matter. More importantly, the conversation around tokenized assets is changing. It is no longer just about putting assets “on chain.” It is increasingly about market quality, liquidity, legal claims, custody, transfer restrictions, redemption mechanics, reserve verification, and governance.

Recent academic work has made this point clearly. One June paper argued that tokenization and liquidity should be analyzed as distinct outcomes, because putting an asset on chain does not automatically create deep secondary-market liquidity. Another paper developed a risk framework for tokenized RWAs that looks beyond headline TVL and evaluates liquidity, concentration, and market-quality risk. A third paper proposed a taxonomy for RWA systems, emphasizing that most real-world asset tokens remain hybrid structures: blockchain tokens provide transferability and composability, while legal guarantees still depend on off-chain wrappers, custodians, compliance processes, and reserve verification.

This is exactly what institutionalization looks like. The early phase was about proof of concept. The next phase is about risk, liquidity, structure, and integration.

For the Age of Autonomy, tokenized assets matter because autonomous agents will eventually need access not just to payments, but to collateral, yield, credit, insurance, settlement instruments, and capital markets. A machine-native economy will not run only on stablecoins. It will also need programmable financial assets.

Robotics Is Becoming the Physical Layer of Autonomy

The physical side of the Age of Autonomy also had an important month. Crunchbase reported that robotics startups had already raised $18.8 billion globally in 2026, exceeding the $15 billion raised in all of 2025 and surpassing the prior 2021 peak.

That level of funding is important because robotics is where AI leaves the screen.

For most of the last two years, the market has focused heavily on software-based AI: chatbots, copilots, coding agents, research assistants, and enterprise automation. But the broader Age of Autonomy includes embodied intelligence — machines that can operate in factories, warehouses, homes, hospitals, restaurants, farms, and logistics networks.

Robotics is still early. Humanoid robots remain expensive. Reliability is hard. Safety is hard. Unit economics are not yet obvious in many consumer categories. But the direction of travel is clear. AI is becoming more capable, sensors are improving, batteries are improving, edge compute is improving, and the cost curve should eventually move in the right direction.

The near-term winners may not be the most futuristic humanoid robot companies. They may be the suppliers of the robotics stack: actuators, sensors, chips, batteries, simulation software, edge compute, safety systems, industrial automation, and integration services.

That is often how platform shifts begin. The first investable opportunities are not always in the most visible product. They are often in the enabling infrastructure.

The Caveat: Autonomy Is Harder Than the Demos Suggest

It is worth noting that not every recent signal was bullish in a straight line. Reuters reported that Meta CEO Mark Zuckerberg told employees that progress on AI agents had been slower than expected, even after major restructuring and significant AI investment.

That is an important reminder.

The Age of Autonomy is real, but it will not arrive all at once. Agents still make mistakes. Long-horizon planning is difficult. Tool use can be brittle. Permissions and security are unsolved. Payment systems need better standards. Stablecoin rules are still being implemented. Tokenized assets still face liquidity and legal-structure challenges. Robots remain capital-intensive and physically constrained.

But this is exactly why the infrastructure layer matters.

The market tends to get excited about demos. Investors should pay more attention to bottlenecks. Every bottleneck is a potential company, protocol, network, or investment category.

If agents cannot safely take action, we need identity, permissions, and audit trails. If agents cannot pay, we need programmable payment rails. If agents cannot access capital markets, we need tokenized assets. If autonomous systems cannot operate in the real world, we need robotics hardware, sensors, edge compute, and safety systems. If institutions cannot participate, we need regulation, custody, compliance, and reporting.

The Age of Autonomy is not one technology. It is a stack.

The Investment Implication

The last month reinforced a core point: the Age of Autonomy is moving from intelligence to action.

AI models are the intelligence layer. Agents are the action layer. Stablecoins are the settlement layer. x402 and similar protocols are the transaction layer. Tokenized assets are the capital-markets layer. Robotics is the physical-world layer. Regulation is the institutional adoption layer.

Each layer is developing at a different speed, but the direction is increasingly clear.

The biggest mistake investors can make is to think about AI, crypto, stablecoins, tokenization, and robotics as separate themes. They are not separate. They are converging.

AI creates autonomous decision-makers. Crypto and stablecoins give them a way to transact. Tokenization gives them assets and collateral. DePIN and decentralized infrastructure give them physical-world data and networks. Robotics gives them bodies. Regulation determines whether all of this can scale inside the formal financial system.

That is why I continue to believe the Age of Autonomy is not just an AI cycle. It is a broader technological restructuring of the economy.

• The internet connected information.

• Crypto connected value.

• AI is connecting intelligence.

• Robotics will connect intelligence to the physical world.

The next step is connecting all of these into autonomous economic systems.

We are not fully there yet. But over the last month, the pieces became a little more visible. Agents are starting to work. The web is starting to accept software as a customer. Stablecoins are moving toward regulated settlement infrastructure. Tokenized assets are becoming more institutional. Robotics funding is accelerating.

The Age of Autonomy will not arrive as a single product launch. It will arrive as infrastructure.

And that infrastructure is now being built in plain sight.

In Gratitude,

-Jake Ryan

CIO Tradecraft Capital

Author, Crypto Investing in the Age of Autonomy

Author, Crypto Decrypted

For most of Bitcoin’s history, investors were told they had to make a choice. They could own Bitcoin for upside, volatility, scarcity, and long-term monetary debasement protection. Or they could own income-producing assets for yield, cash flow, and portfolio stability. Bitcoin was the growth asset. Bonds, dividend stocks, REITs, preferreds, and covered-call funds were the income assets. The two categories lived in different mental buckets.

That separation is starting to break down.

A new category of Bitcoin-related income products is emerging. These products are not all the same, and investors should be careful not to lump them together simply because they all have a Bitcoin narrative attached. But they are aiming at a similar investor problem: how do you participate in the Bitcoin ecosystem while generating a meaningful cash yield?

Let’s review 3 Bitcoin income-related investments now on the scene!

Subscribe now

Get Actionable Reports by Becoming a Member!

Read more

For the last several years, I have argued that we are moving into a new technological revolution: the Age of Autonomy. This is not simply an “AI cycle,” just as the internet was not simply a “software cycle.” It is a broader restructuring of the economy around autonomous systems: artificial intelligence, robotics, IoT, decentralized infrastructure, and programmable money.

In the early stages of any long-wave technological shift, the narrative always moves faster than the infrastructure. The 1990s had websites before it had scalable payments, cloud computing, mobile broadband, or mature cybersecurity. Crypto had tokens before it had regulatory clarity, institutional custody, or real-world utility. AI had chatbots before it had agentic commerce, reliable identity, permissioning, payment rails, and real-world embodiment.

That is why Q2 of 2026 feels important. The Age of Autonomy is still early, and there is still plenty of hype. But the most meaningful progress is no longer happening only at the level of model demos, token speculation, or futuristic robotics videos. It is increasingly showing up in the plumbing.

We are beginning to see the pieces of an autonomous economy assemble themselves: agents that can take action, stablecoins that can settle value instantly, protocols that allow software to pay other software, robots that can operate in physical environments, and tokenized assets that can move through programmable financial rails.

This is the transition from “AI as a tool” to “AI as an economic actor.”

And that distinction matters.

A tool waits for a human to use it. An economic actor can perceive, decide, transact, coordinate, and execute. Once software agents can do those things reliably, the internet changes from a network of information into a network of autonomous economic activity.

That is the core of the Age of Autonomy.

1. Agentic Payments Are Becoming the Missing Commercial Layer

The most important progress in Q2 may be the least visually dramatic: the emergence of machine-native payment infrastructure.

For years, the internet has had an awkward relationship with payments. Content, APIs, data, compute, and digital services all have value, but payment systems were never built into the architecture of the web itself. The result was the rise of advertising, subscriptions, walled gardens, payment processors, app stores, and platform gatekeepers.

That worked well enough for human commerce. It works poorly for autonomous agents.

An AI agent cannot efficiently open a bank account, fill out a credit card form, wait for a human to approve a transaction, or negotiate a SaaS contract every time it needs access to a dataset, API, model, storage layer, verification service, or another agent. If agents are going to perform economically useful work, they need the ability to pay for resources programmatically.

That is why protocols such as x402 matter.

The idea is simple but powerful: revive the old HTTP 402 “Payment Required” status code and turn it into a machine-readable payment layer for the internet. Instead of a website or API merely denying access, it can quote a price. An agent can read the quote, authorize payment, and continue the transaction.

This is where stablecoins become more than speculative crypto infrastructure. They become the settlement layer for autonomous software.

In the human internet, payments were bolted on after the fact. In the agentic internet, payments need to be native. Agents will need to pay for inference, data, compute, bandwidth, identity, storage, digital labor, and eventually real-world services. A machine-native payment layer allows those interactions to happen in real time, at small denominations, across borders, and without the friction of legacy financial rails.

This is one of the clearest examples of crypto’s role in the Age of Autonomy. The winning use case is not “crypto replaces money” in a simplistic sense. It is that autonomous systems need programmable money because legacy payments were designed for humans, institutions, and batch settlement.

Agents need value transfer that looks more like an API call.

Of course, the infrastructure is not mature yet. Early security research around x402 has already identified issues around authorization, replay protection, synchronization between web requests and blockchain settlement, and the possibility of unpaid service usage or paid-but-denied outcomes. That should not be surprising. Every important financial protocol goes through a hardening phase. The fact that researchers are already stress-testing the design is a sign that the ecosystem is moving from concept to implementation.

The investment implication is that the first large value pools may not accrue only to the agents themselves. They may accrue to the payment, wallet, identity, custody, compliance, and developer infrastructure that allows agents to operate safely.

In other words, the agent economy needs financial plumbing.

And in Q2 2026, that plumbing is becoming visible.

Subscribe now

Get the Actionable Intelligence by Becoming a Paid Subscriber.

2. Stablecoins Are Becoming the Monetary Substrate

The second major area of progress is stablecoins.

Stablecoins have been around for years, but the market’s understanding of them is changing. They are no longer just a crypto trading tool. They are becoming the base money layer for digital commerce, cross-border settlement, tokenized markets, and autonomous agents.

That shift is important because the Age of Autonomy cannot run only on intelligence. It needs settlement.

AI can generate text, write code, analyze markets, negotiate workflows, and coordinate tasks. But once an agent needs to procure something in the real world — compute, data, electricity, storage, hardware access, logistics, or another service — it needs money. Not metaphorical money. Actual transferable value.

Stablecoins are emerging as the most obvious bridge between the existing dollar system and the programmable economy. They carry the unit of account the world already uses, but they move on networks that are more compatible with software.

The GENIUS Act and related implementation work are important in this context because they suggest stablecoins are moving from regulatory gray zone to regulated financial infrastructure. That does not remove every risk. It may even introduce new competitive dynamics, particularly as banks, fintech companies, payment processors, and large technology platforms enter the market.

But it does change the conversation.

For a long time, the question was whether stablecoins would be allowed to exist at scale. The question now is who will issue them, who will custody them, which chains will settle them, which wallets will distribute them, and which applications will create demand for them.

That is a much more investable question.

In the Age of Autonomy, stablecoins may function like the monetary layer of the machine economy. Agents will not want to hold volatile assets for ordinary commercial activity. They will need predictable units of account, low transaction costs, fast settlement, and programmable permissions.

That points toward stablecoins, not Bitcoin or high-beta tokens, as the everyday transactional substrate.

Bitcoin remains the premier monetary asset of the digital age. But stablecoins may become the working capital of autonomous software.

3. Robotics and Physical AI Are Moving Toward Commercial Deployment

The most visible part of the Age of Autonomy is robotics. The least appreciated part is how slowly physical autonomy develops compared with digital autonomy.

Software can scale globally almost instantly. A model update can be pushed to millions of users. A new agent can be deployed in the cloud. A payment protocol can be integrated into APIs.

Robots are different. They must operate in the physical world, where every environment is messy, every object has geometry, every task has safety implications, and every mistake can damage equipment or harm people.

That is why recent progress in humanoid robotics and industrial automation should be interpreted carefully. We are not yet at the “robot in every home” stage. But we are moving closer to the “robot as industrial labor platform” stage.

This is a major distinction.

The first economically significant use cases for humanoid and semi-humanoid robots are unlikely to be general household chores. They are more likely to be structured industrial environments: factories, warehouses, logistics facilities, retail operations, and repetitive physical workflows where the environment can be mapped and the task set can be constrained.

This is why announcements from companies like Boston Dynamics, Hyundai, Tesla, Xpeng, and others are worth watching. The progress is not just in robot movement. The progress is in manufacturability, deployment plans, task training, sensor integration, and the use of AI models to improve robot decision-making.

In Q2 2026, the theme is no longer simply “robots can walk.” It is that robots are being prepared for commercial environments.

That is a much bigger deal.

Physical AI represents the embodiment of autonomy. It is where artificial intelligence leaves the screen and begins interacting with the world. Once that happens, the economic stakes expand dramatically. Software eats information markets first. Robotics goes after labor, logistics, manufacturing, and physical services.

The timing will be uneven. There will be setbacks. Many humanoid robotics companies will overpromise. Unit economics will matter. Safety will matter. Maintenance will matter. Battery life, dexterity, perception, and task reliability will matter.

But the direction of travel is clear: AI is becoming physical.

That is one of the defining features of the Age of Autonomy.

4. Tokenized Real-World Assets Are Building the Financial Bridge

Another area of meaningful progress is tokenized real-world assets, or RWAs.

At first glance, RWAs can feel less exciting than AI agents or humanoid robots. Tokenized Treasuries do not make for viral videos. On-chain money market funds do not look like science fiction. But they may be one of the most important bridges between traditional finance and the autonomous economy.

Autonomous systems need more than payment rails. They need collateral, yield, liquidity, credit, settlement assets, and financial primitives that can be composed programmatically.

That is where RWAs become important.

If stablecoins are the checking account of the autonomous economy, tokenized Treasuries and money market funds may become the savings account, collateral layer, and low-risk yield substrate. Eventually, tokenized credit, commodities, equities, and other assets could expand the range of financial instruments available to autonomous software and on-chain applications.

But this is an area where nuance matters.

Tokenization does not automatically create liquidity. Putting an asset on-chain does not magically make it trade actively, reduce credit risk, or solve legal enforceability. Recent research is making this point more clearly: tokenization and liquidity are distinct outcomes. Legal structure, custody, redemption rights, transfer restrictions, market makers, disclosure, and investor access all matter.

That is healthy for the market.

The early RWA narrative sometimes sounded like every asset in the world would simply move on-chain and become liquid overnight. The more mature version is more specific: high-quality financial assets will gradually become programmable, composable, and interoperable with digital settlement networks.

That is still a massive opportunity.

The Age of Autonomy needs financial instruments that software can understand and use. RWAs are one way the traditional financial system becomes legible to autonomous agents.

Subscribe now

Get the Actionable Intelligence by Becoming a Paid Subscriber.

5. DePIN Is the Hardest but Most Strategically Important Layer

DePIN — decentralized physical infrastructure networks — remains one of the more complicated parts of the Age of Autonomy, but also one of the most strategically important.

The basic idea is that physical infrastructure can be built, operated, or coordinated through token incentives. Instead of relying only on centralized companies to deploy wireless networks, compute resources, sensors, mapping data, energy assets, or storage infrastructure, DePIN networks use economic incentives to coordinate distributed participants.

This matters because autonomy requires infrastructure everywhere.

Autonomous vehicles need maps, sensors, connectivity, and compute. Robots need data, energy, and maintenance networks. AI agents need compute and storage. IoT devices need connectivity. Smart cities need sensing and coordination. The more autonomy expands into the physical world, the more valuable real-world infrastructure becomes.

DePIN tries to answer a difficult question: can open networks coordinate physical infrastructure more efficiently than centralized firms in certain markets?

The answer will vary by category.

Some DePIN models are more promising than others. Digital resource networks such as compute, storage, bandwidth, or GPU access may scale differently than physical networks requiring hardware placement, maintenance, local density, and real-world demand. The economics of a decentralized wireless network are different from a decentralized mapping network, which are different from decentralized energy or compute.

But as a category, DePIN sits at the intersection of crypto, AI, IoT, and robotics. That makes it central to the Age of Autonomy.

If AI is the intelligence layer, stablecoins are the payment layer, and RWAs are the financial layer, DePIN may become part of the infrastructure layer.

The challenge is that DePIN networks must prove real demand. Token incentives can bootstrap supply, but they cannot substitute for sustainable usage. The best DePIN projects will be those where token incentives help solve a real coordination problem and where the network eventually generates demand from customers who need the service, not just investors who want the token to go up.

This is where the sector is moving from story to substance.

The question is no longer, “Can we incentivize people to deploy hardware?”

The question is, “Can the network produce a service that customers actually buy?”

That is the right question.

6. Regulation Is Becoming an Enabler, Not Just a Risk

The final area of progress is regulatory clarity.

For most of crypto’s history, regulation has been treated primarily as a threat. In many cases, that was justified. Regulation by enforcement created uncertainty, pushed projects offshore, distorted token design, and made it harder for serious institutions to build.

But the Age of Autonomy requires regulated interfaces with the real world.

Stablecoins need reserve rules. Tokenized assets need legal claims and custody structures. Agentic payments need compliance and identity frameworks. DePIN networks need rules around telecom, energy, data, and hardware. AI agents will need permissioning, auditability, and accountability. Autonomous financial systems cannot scale indefinitely inside regulatory fog.

That does not mean all regulation is good. Bad regulation can still crush innovation. But clear rules can unlock participation from banks, asset managers, payment companies, enterprises, and developers who have been waiting on the sidelines.

This is especially true for stablecoins and tokenized assets. Once the legal perimeter becomes clearer, the competition shifts from “can this exist?” to “who can build the best product?”

That is the transition crypto has needed for years.

Regulatory clarity also changes tokenomics. If builders can design networks with more confidence around what is a security, what is a commodity, what is a payment instrument, and what kind of value accrual is permissible, then token design can become more rational.

The next cycle of crypto will not be led by vague narratives alone. It will be led by networks with product-market fit, real revenue, strong unit economics, and tokens that have a defensible reason to exist.

That is a healthier market.

The Bigger Picture: Autonomy Requires a Stack

What we are seeing in Q2 2026 is not one breakthrough. It is the assembly of a stack.

At the intelligence layer, AI models are becoming more capable and agentic.

At the payment layer, stablecoins and protocols like x402 are giving software a way to transact.

At the financial layer, RWAs are beginning to bring traditional assets into programmable markets.

At the infrastructure layer, DePIN is experimenting with decentralized coordination of real-world resources.

At the physical layer, robotics is moving from demonstration to early commercial deployment.

At the regulatory layer, stablecoin and digital asset frameworks are beginning to give institutions a path to participate.

That is what makes this moment interesting.

Each individual piece is still imperfect. AI agents are unreliable. Agentic payments need better security. Stablecoin regulation will create winners and losers. RWAs still have legal and liquidity constraints. DePIN needs real demand. Robotics will take longer than the market wants. Regulation remains politically fragile.

But the pieces are no longer theoretical. They are beginning to connect.

And when technologies connect, new markets emerge.

The Investment Implication

The Age of Autonomy will not unfold evenly. Some sectors will move quickly; others will disappoint. Many tokens and companies using the autonomy narrative will fail. That is normal. Every major technology cycle produces both extraordinary winners and enormous speculative excess.

The key is to separate narrative exposure from infrastructure exposure.

Narrative exposure is buying whatever has “AI,” “agent,” “robot,” or “DePIN” attached to it.

Infrastructure exposure is identifying the systems that autonomy actually needs in order to function: compute, payments, wallets, identity, data, security, settlement, tokenized collateral, machine coordination, and physical deployment networks.

The strongest opportunities are likely to appear where three things overlap:

First, there must be real demand.

Second, the product must become more valuable as autonomous systems grow.

Third, the token or equity must have a credible mechanism for capturing that value.

That last point matters. Not every important technology produces a good investment. And not every useful protocol creates value for its token. In crypto especially, the distinction between network usage and token value accrual is critical.

The Age of Autonomy will create enormous economic value. But investors still need to ask where that value accrues.

In Closing

Q2 2026 may be remembered as one of the quarters when the Age of Autonomy began to feel less like a futuristic thesis and more like an emerging operating system for the economy.

• Agents are beginning to transact.

• Stablecoins are becoming regulated payment infrastructure.

• Robotics is moving toward commercial deployment.

• RWAs are building the bridge between traditional finance and programmable markets.

• DePIN is experimenting with decentralized infrastructure for the physical world.

• And regulation is slowly turning from a source of pure uncertainty into a framework that serious builders can work within.

This does not mean the future has arrived. It means the foundations are being poured.

The Age of Autonomy will not be built by AI models alone. Intelligence is necessary, but not sufficient. Autonomous systems also need money, identity, data, compute, infrastructure, settlement, and the ability to act in the physical world.

That is what is now coming into view.

The internet gave humans access to information.

Crypto gave software access to value.

AI gives software the ability to reason and act.

Robotics gives software a body.

The Age of Autonomy begins when all of those systems start working together.

And in Q2 2026, that is exactly where we are starting to see the most progress.

In Gratitude,

-Jake Ryan

CIO Tradecraft Capital

Author, Crypto Investing in the Age of Autonomy

Author, Crypto Decrypted

For most of the last cycle, our base case was simple:

Liquidity drives crypto markets.

That was not just a slogan. It was the core macro framework behind our crypto thesis. When global M2 expands, capital eventually finds its way into risk assets. Crypto, as one of the highest-beta expressions of global liquidity, tends to respond early and aggressively. When liquidity expands, Bitcoin usually benefits first. Ethereum and the rest of the market typically follow once confidence broadens.

That framework still matters. But it needs to be refined.

Our updated view is this: liquidity drives markets over the medium and long term, but news and events can dominate the near term.

That distinction matters because it helps explain why crypto has felt so frustrating lately. The macro backdrop has improved. Global M2 has turned into a tailwind. The economy is still expanding. Several of our cycle indicators remain green. Bitcoin has stabilized. Ethereum has stopped behaving like it is in freefall. And yet the market has not fully transitioned into a clean, broad-based, liquidity-driven expansion.

Why?

Because liquidity does not operate in a vacuum.

When uncertainty is low, liquidity can drive prices higher quickly. Capital gets more comfortable taking risk. Investors extend their time horizons. Bitcoin leads, Ethereum follows, and eventually capital rotates into higher-beta segments of the market.

But when uncertainty is high, liquidity becomes less effective. It does not disappear. It just stops being the lead actor. It sits in the background while the market reacts to headlines, policy uncertainty, geopolitical risk, regulatory developments, hacks, ETF flows, and macro surprises.

That is the environment we are in now.

The Old Base Case Was Not Wrong. It Was Incomplete.

In 2025, the liquidity framework worked extremely well. The expansion of global money supply, the weakening dollar, the turn toward easier monetary conditions, and the growth in institutional crypto infrastructure all supported the view that crypto was entering another liquidity-led advance.

That is why we argued that global M2 mattered more than tariffs.

Tariffs were important, but they were discrete shocks. They could hit sentiment for days or weeks. They could create volatility. They could disrupt supply chains and spook investors. But they were not the same as a sustained contraction in global liquidity.

Liquidity was the tide. And when the tide rises, risk assets tend to rise with it.

That is still true. But Q1 and April reminded us that even a rising tide can be interrupted by storms.

In Q1, crypto markets did not behave as though liquidity was fully in control. They behaved as though uncertainty was in control. Investors were dealing with too many unresolved questions at once: monetary policy, inflation persistence, economic growth, geopolitics, regulatory direction, cross-asset volatility, and broader policy risk.

In that environment, crypto was not being repriced based on improving background liquidity. It was being repriced based on hesitation.

That is the key shift in our thinking.

Liquidity matters over the cycle. But in the near term, markets trade on what investors are afraid of.

Subscribe now

We are a reader-supported substack. Upgrade for actionable thoughts on portfolio construction, market timing and asset selection.

April Was Better, But It Was Not an “All Clear”

April was a constructive month for crypto, but it was not a clean risk-on breakout.

Bitcoin led the recovery. ETF demand improved. Institutional demand returned. Ethereum participated. The broader market stopped acting like it was in freefall and started acting like it was trying to rebuild trend.

That is important.

Bitcoin finished April up roughly 12%, while Ethereum recovered meaningfully as well, rising roughly 13% from its prior-month level by the end of April. Bitcoin reasserted leadership, trading near the $76,000–$77,000 range at month-end after recovering from the roughly $60,000 area earlier in the year. This was not a retail-mania rally. It looked more institutional, more measured, and more technical.

That was encouraging.

But the recovery was still fragile.

Ethereum participated but did not decisively reclaim leadership. ETH ended April around $2,250–$2,265, but the ETH/BTC ratio remained weak compared with January levels. In plain English, capital was still favoring Bitcoin over higher-beta smart contract exposure.

That matters because true crypto bull markets usually broaden. Bitcoin leads first, but eventually Ethereum, Solana, DeFi, AI, DePIN, and other high-beta segments need to participate. April showed stabilization. It did not yet show full market rotation.

And beneath the surface, there were still real risks.

Macro uncertainty remained. Geopolitical risk remained. Inflation remained sticky. Yields remained elevated. And DeFi suffered another confidence shock after exploit activity spilled into lending markets and reminded investors that composability is both crypto’s superpower and its systemic vulnerability.

So April was not the beginning of a guaranteed straight line higher.

It was a transition month.

Crypto stopped acting like a market in collapse. It started acting like a market trying to rebuild trend.

That is progress. But it is not confirmation.

Subscribe now

Read more

Founders could launch protocols, but not always know whether their token would later be treated as a security. Investors could underwrite networks, but not always know whether value accrual at the token level would be punished by regulators. Builders could design useful economic systems, but the more directly a token captured value, the more legal risk it seemed to create.

That may be changing.

The CLARITY Act has suddenly become one of the most important catalysts in crypto. This week, Polymarket odds reportedly moved to roughly 65% that the Act becomes law in 2026, up sharply from earlier levels near the mid-40s. Prediction markets are not perfect, but they are useful because they force people to put money behind their expectations. In this case, the market is telling us something important: Washington may finally be moving from regulation by enforcement to regulation by architecture.

And that matters because the next phase of crypto is not just humans buying tokens.

It is software moving value.

It is AI agents paying for data, compute, APIs, identity, storage, execution, and eventually real-world goods and services.

It is the Age of Autonomy becoming economically real.

Subscribe now

The Missing Layer for AI Agents

AI agents are going to need money.

Not metaphorical money. Not “credits” trapped inside one company’s closed platform. Not a prepaid balance that only works in one app.

Actual programmable money.

An autonomous agent that can research a market, book a trip, pay for compute, buy data, execute a transaction, or hire another agent needs a payment rail that is global, instant, low-friction, machine-readable, and programmable.

That is not the credit card system.

Credit cards were built for humans, merchants, chargebacks, bank intermediaries, and 20th-century commerce. They were not built for autonomous software negotiating with other autonomous software in real time.

Stablecoins are.

This is why protocols like x402 matter. HTTP has always had a “402 Payment Required” status code sitting dormant inside the internet’s architecture. Now that idea is being revived for machine-native payments. With x402, an AI agent can request a resource, receive a payment-required response, pay in stablecoins, and continue the transaction without opening an account, entering a credit card, or waiting on a human approval loop.

That is a small technical change with enormous economic implications.

The internet was built to move information.

Blockchains were built to move value.

AI agents are the software layer that will increasingly decide when, where, and why value should move.

From Human Commerce to Agentic Commerce

Think about how online commerce works today.

You search. You compare. You read reviews. You check delivery dates. You decide whether the price is fair. You type in your payment details or approve a wallet transaction.

That entire flow is going to be compressed.

In the future, you may tell an agent: “Find me the best flight to New York next Thursday, under $600, aisle seat preferred, avoid red-eyes, and use my airline miles if it makes sense.”

The agent will search, compare, negotiate, verify, and pay.

The same logic applies to buying software, paying for data, renting compute, purchasing insurance, optimizing supply chains, ordering inventory, booking hotels, or paying other agents.

This is the machine-to-machine economy.

And the machine-to-machine economy needs three things: identity, trust, and money.

Crypto is the only technology stack that plausibly provides all three in an open, programmable environment.

Why CLARITY Matters

The CLARITY Act is important because it starts to define the rules of the road for digital assets in the United States.

The Senate Banking Committee’s section-by-section summary says the bill creates disclosure requirements for certain transactions involving “ancillary assets” and treats the tokens themselves as commodities. That is a meaningful shift because it separates the token from the fundraising transaction around the token. In plain English: a project may still need to comply with disclosure and fundraising rules, but the token itself may not automatically live forever under securities law simply because it began life inside a capital formation event.

Reuters reports that the bill also addresses SEC fundraising exemptions, DeFi definitions, AML requirements, stablecoin rewards, and the treatment of tokenized securities. Tokenized stocks and bonds remain securities, but many network tokens may be handled under a different framework.

That distinction is critical.

For years, crypto builders were trapped in a paradox.

If a token had no value accrual, investors questioned why it existed.

If a token had too much value accrual, lawyers worried it looked like a security.

So builders often designed tokens around governance, access, emissions, staking, or vague “utility,” while avoiding more direct mechanisms that might return value to token holders.

That made many tokens weaker as assets.

The CLARITY Act could begin to change that.

The Return of Tokenomics

The next generation of tokenomics, how tokens are structured within its ecosystem to create and accrue value, will not look like the last generation.

The last cycle was often about narrative, emissions, liquidity mining, staking and governance tokens that did not govern much.

The next cycle will have to be about real value capture.

If a network earns revenue from agent transactions, API payments, data markets, compute usage, settlement fees, or autonomous commerce, then the token should have a rational claim on that economic activity.

That does not necessarily mean every token pays a traditional dividend. Builders still need legal guidance, and the final regulatory framework will matter. But the direction of travel is clear: the industry is moving toward a world where tokens can be designed as economically meaningful assets rather than legal liabilities.

That is a major change.

If builders can design tokens where value is created at the protocol level and accrues transparently at the token level, then digital assets become easier to underwrite.

Investors can ask better questions:

• What does the network do?

• Who pays to use it?

• How much revenue does it generate?

• How does that value flow through the system?

• Does the token capture any of it?

• Is the token necessary to the network’s operation?

That is how real asset classes mature.

AI Agents Need Economic Substrate

This is where the Age of Autonomy comes in.

The Age of Autonomy is not just about smarter chatbots. It is about autonomous operations: systems that sense, decide, act, coordinate, and transact with less human involvement.

Robots will need to pay for charging, maintenance, mapping data, and logistics.

AI agents will need to pay for APIs, compute, identity verification, private data, and other agents.

Autonomous vehicles will need to pay tolls, insurance, energy, routing services, and fleet-management systems.

DePIN networks will need to compensate physical infrastructure providers.

Onchain finance will need settlement assets, collateral assets, routing layers, liquidity venues, and compliance-aware execution.

All of this requires value transfer.

And value transfer requires rules.

That is why CLARITY matters. It is not just a crypto bill. It is an infrastructure bill for the autonomous economy.

The Bigger Investment Implication

The market is still mostly focused on the obvious beneficiaries: exchanges, custodians, stablecoin issuers, and large public crypto companies.

That makes sense. Regulatory clarity helps the incumbents first.

But the larger opportunity may be further out on the curve.

If the United States creates a workable framework for digital commodities, stablecoin payments, and compliant token issuance, then builders can start designing protocols for autonomous commerce with more confidence.

That opens the door for several major categories:

Agent wallets, where autonomous software can hold assets, sign transactions, enforce budgets, and operate under human-defined permissions.

Payment protocols, where agents can pay for resources instantly without accounts, invoices, or credit cards.

Tokenized service networks, where data, compute, storage, bandwidth, and real-world infrastructure can be bought and sold programmatically.

Revenue-generating protocols, where tokenomics can be designed around actual economic throughput instead of speculative emissions.

Compliance-aware DeFi, where autonomous agents can interact with regulated financial infrastructure without every transaction becoming a legal gray zone.

This is how crypto moves from speculation to infrastructure.

Not all at once. Not without setbacks. Not without bad designs, bad actors, and regulatory fights.

But the direction is becoming clearer.

The Internet Got Payments Wrong the First Time

The original internet was brilliant at moving information but terrible at moving money.

That is why the business model of the internet became advertising.

If payments had been native to the web from the beginning, the internet might have evolved very differently. Content, data, APIs, software, and digital services could have been monetized through micropayments instead of surveillance-based advertising.

AI agents may give the internet a second chance.

Agents do not want to watch ads.

Agents do not click banner ads.

Agents do not need loyalty points, email marketing funnels, or retargeting campaigns.

Agents need clean economic instructions: price, permission, payment, delivery, verification.

That is a stablecoin-native world.

And if CLARITY helps make that world legally legible, the implications are much bigger than one bill, one market structure debate, or one prediction market moving from 45% to 65%.

It means the United States may finally be creating a legal foundation for autonomous value transfer.

In Closing

The Age of Autonomy needs an economic rail.

• AI gives software the ability to decide.

• Blockchains give software the ability to transact.

• Stablecoins give software a dollar-denominated medium of exchange.

And regulatory clarity gives builders the confidence to combine all three.

That is why the CLARITY Act matters. It is not simply about whether crypto gets friendlier treatment in Washington. It is about whether the next generation of autonomous software can be built on open, programmable financial infrastructure in the United States.

The first era of crypto was about humans buying tokens. The next era will be about machines using networks.

And when machines begin using networks, paying for services, coordinating resources, and settling value at internet speed, the tokens that power those networks become more than speculative assets.

They become economic infrastructure.

That is the real story.

The Age of Autonomy will not just run on intelligence.

It will run on programmable money.

In Gratitude,

-Jake Ryan

CIO Tradecraft Capital

Author, Crypto Investing in the Age of Autonomy

Author, Crypto Decrypted

The world became obsessed with prompts, copilots, chatbots, image generators, and productivity tools. That phase mattered. It proved that software could reason, summarize, generate, and assist. But it was still mostly a screen-based revolution.

Q2 2026 is showing us the next phase.

The Age of Autonomy is no longer just about AI answering questions. It is about AI taking action — inside companies, inside robots, inside payment systems, inside logistics networks, and eventually inside the physical economy itself.

This is the transition we have been watching for: from intelligence as a tool to intelligence as an operating layer.

The four pillars remain the same: IoT, AI, robotics, and blockchain. But what is changing is how tightly they are beginning to converge.

AI provides the reasoning layer. Sensors and IoT provide real-world awareness. Robotics provides physical execution. Blockchain and stablecoins provide programmable settlement. Together, they form the early architecture of autonomous operations.

That is the key theme of Q2 2026: autonomy is becoming infrastructure.

1. Enterprise AI Is Moving From “Copilot” to “Autonomous Operations”

One of the clearest signals came from SAP, which announced an “Autonomous Enterprise” suite designed to embed AI agents into core business functions like finance, HR, procurement, supply chains, and customer engagement. The important point is not simply that SAP added AI. Everyone is adding AI. The important point is that enterprise software is moving from dashboards and workflows toward systems that can execute business processes with less human intervention.

That matters because the modern corporation is essentially a network of decisions: buy this, route that, approve this invoice, forecast that demand, restock this warehouse, rebalance that team, respond to this customer.

Historically, software helped humans make those decisions. In the Age of Autonomy, software begins making more of those decisions itself — within rules, permissions, and oversight.

This is how autonomy will likely enter the economy first: not through humanoid robots walking into every home, but through invisible automation inside finance departments, supply chains, procurement systems, customer service organizations, and logistics networks.

The first autonomous enterprises will not look like science fiction. They will look like companies where the back office suddenly gets faster, cheaper, and more adaptive.

2. Robotics Is Becoming Physical AI

The second major Q2 signal is the acceleration of “physical AI” — AI systems that do not just generate text or code, but perceive, plan, and act in the physical world.

NVIDIA has been highlighting robot foundation models that unify perception, planning, and control, using simulation environments where robots can practice large numbers of task variations before being deployed in the real world. The implication is powerful: robotics development is beginning to look more like software development. Train in simulation, test edge cases, improve models, deploy into the field, collect data, and iterate.

That loop is the autonomy flywheel.

The better the simulation, the better the model. The better the model, the more useful the robot. The more useful the robot, the more real-world data it produces. The more data it produces, the better the next model becomes.

This is why robotics could move faster than many investors expect. The bottleneck has always been the physical world: messy environments, irregular objects, safety issues, edge cases, and limited training data. But if robotics can borrow the scaling dynamics of AI — foundation models, synthetic data, simulation, and continuous learning — then the industry starts to look less like traditional industrial automation and more like a platform shift.

3. Dexterous Robots Are Getting Closer to Useful Work

A particularly interesting Q2 development came from Genesis AI, a French robotics startup that unveiled GENE-26.5 and a dexterous robotic hand designed for complex manipulation tasks. Reuters reported that the system is aimed at sectors like automotive, electronics, pharmaceuticals, and logistics, with the company also building robotics datasets from industrial worker interactions using sensor-equipped gloves.

This is important because hands are one of the hardest problems in robotics.

Moving a box is useful. Moving through a warehouse is useful. But the human hand is the interface for much of the physical economy. Assembly, repair, sorting, packaging, inspection, cooking, healthcare support, lab work, and manufacturing all require dexterity.

If robots can become meaningfully better at manipulation, the addressable market expands dramatically.

This is where sensors and AI intersect. A robot does not just need a hand. It needs tactile feedback, visual understanding, spatial awareness, motion planning, and the ability to adapt when the real world does not match the training environment. In other words, dexterity is not just a robotics problem. It is an Age of Autonomy problem.

The companies that solve it will not merely build machines. They will build embodied labor platforms.

4. Blockchain Is Becoming the Settlement Layer for AI Agents

The most crypto-native Q2 development came from Circle, which launched Circle Agent Stack, a set of tools designed for the “agentic economy.” The stack includes agent wallets, an agent marketplace, a command-line interface, nanopayments, and other infrastructure designed to let autonomous AI agents hold funds, discover services, and transact programmatically using USDC within permissions and guardrails.

This is exactly where blockchain becomes interesting again.

Crypto has spent years searching for mainstream users. But humans are not always the ideal users of blockchains. Humans dislike private keys, gas fees, bridges, wallets, and transaction complexity.

AI agents are different.

Software does not mind using software. Agents can sign transactions, call APIs, manage wallets, enforce rules, transact globally, and settle value programmatically. If autonomous agents become economic actors, then they need financial rails built for machines, not just humans.

That is where stablecoins, wallets, smart contracts, and onchain settlement begin to make sense.

In this world, blockchain is not just a speculative asset class. It becomes a coordination layer for autonomous economic activity. Agents will need identity, permissions, payments, auditability, reputation, and settlement. Traditional financial rails were built for people and institutions. Blockchain rails may be better suited for software entities operating at internet speed.

This does not mean every AI agent needs a token. Most probably do not.

But it does mean that stablecoins and programmable payments could become foundational infrastructure for the machine economy.

5. The Investment Lens: Autonomy Compounds

The most important idea for investors is that autonomy compounds.

A company that uses AI to reduce back-office labor gets more efficient. A warehouse that uses robotics gets faster. A logistics network that uses sensors gets more adaptive. A software agent that can pay, contract, and coordinate with other agents becomes more economically useful.

But the real power comes when these systems connect.

An autonomous enterprise can forecast demand. An IoT network can sense inventory. A robotic system can move goods. An AI agent can order supplies. A stablecoin rail can settle payment. A blockchain can record the transaction. A human manager can supervise exceptions rather than manually coordinate every step.

That is the Age of Autonomy.

It is not one technology. It is a stack.

And once the stack starts working, the advantage can become self-reinforcing. More automation creates more data. More data improves models. Better models improve operations. Better operations lower costs and increase speed. Lower costs and higher speed create competitive pressure on everyone else.

This is how moats are formed in the next cycle.

Not merely through brand. Not merely through scale. Not merely through distribution.

But through operational intelligence that continuously improves.

What Q2 2026 Is Telling Us

Q2 is still underway, but the early signal is clear: autonomy is moving from narrative to deployment.

SAP is pushing agentic automation into enterprise workflows. NVIDIA is building the tools for physical AI. Robotics startups like Genesis AI are attacking dexterous manipulation. Circle is building financial infrastructure for autonomous agents.

These are not isolated stories. They are pieces of the same puzzle.

The economy is beginning to absorb AI not as a novelty, but as an operating system. The next phase will not be defined by who has the best chatbot. It will be defined by who can turn intelligence into action.

• That means action inside the enterprise.

• Action inside supply chains.

• Action inside warehouses.

• Action inside robots.

• Action inside financial networks.

• And eventually, action across autonomous systems coordinating with one another.

For investors, this is the bigger picture. The Age of Autonomy is not about one product launch or one hype cycle. It is about a long-wave technological shift where intelligence becomes embedded into the physical and financial infrastructure of the economy.

• The internet connected information.

• Mobile connected people.

• Crypto connected value.

• AI connects intelligence.

• Autonomy connects intelligence to action.

• Robotics in the physical realm and blockchain in the digital realm.

• And that is when the real transformation begins.

In Gratitude,

-Jake Ryan

CIO Tradecraft Capital

Author, Crypto Investing in the Age of Autonomy

Author, Crypto Decrypted

For most of this cycle, investors have been trained to distrust rallies.

Every bounce has been treated as temporary. Every breakout has been met with skepticism. Every move higher has been framed as a bear-market rally, a liquidity head fake, or a macro trap.

But markets do not wait for perfect clarity.

They move first. The narrative catches up later.

These 4 crypto assets have great technical charts and show good risk/reward potential.

And right now, several important crypto charts are beginning to move in the same direction: higher. Bitcoin is back above the psychologically important $80,000 level. Bitcoin recently traded above $81,000, its highest level since late January, while extending a six-day winning streak. Ethereum is stabilizing and beginning to reclaim key technical territory.

This does not mean risk has disappeared. It does not mean every token goes straight up. It does not mean macro no longer matters.

But it does mean the market is giving us a signal.

When multiple high-quality crypto assets begin reclaiming their 50-day moving averages, momentum is changing. The 50-day moving average is not magic, but it is one of the simplest and most widely watched measures of intermediate trend. When price is below the 50-day, rallies often fail. When price gets back above it, the market begins to shift from defensive to opportunistic.

Right now, crypto is beginning to look opportunistic again.

Subscribe now

Get Actionable Data

Bitcoin: The Market Leader Is Back Above $80,000

Bitcoin remains the primary signal for the entire crypto market.

When Bitcoin is weak, everything else struggles. When Bitcoin is range-bound, altcoins fight for oxygen. But when Bitcoin begins to reclaim key technical levels, the entire market structure changes.

That is what appears to be happening now.

Bitcoin has pushed back above $80,000 and recently traded around the $81,000–$82,000 area. MarketPulse noted that a close above roughly $82,133 would strengthen the case for a resumed bull move, while the $80,000 breakout zone now becomes an important support area.

This is important because Bitcoin is not merely bouncing from oversold conditions. It is beginning to rebuild trend.

Binance’s technical summary showed Bitcoin’s daily 50-day moving average below current price, meaning the 50-day now has the potential to act as support on pullbacks rather than resistance on rallies. That is a meaningful shift. It tells us buyers are no longer waiting for dramatically lower levels. They are beginning to step in higher.

Bitcoin also just closed its best month in a year, rising more than 12% in April, helped by improving ETF flows and renewed institutional participation. That matters because Bitcoin’s strongest moves are usually not driven by retail enthusiasm alone. They are driven by institutional allocation, liquidity conditions, and the repricing of risk.

The chart is telling us the same thing the structure is telling us: Bitcoin is no longer acting like a broken asset. It is acting like a market leader trying to turn higher.

Ethereum: Quietly Reclaiming Its Setup

Ethereum has not had the same clean narrative as Bitcoin.

Bitcoin has the ETF bid, the institutional treasury story, the digital gold narrative, and the clearest macro identity in crypto. Ethereum, by contrast, has spent much of the past year fighting questions about fee compression, L2 value capture, Solana competition, and whether ETH still accrues enough value from the broader Ethereum ecosystem.

But price often improves before sentiment does.

Ethereum is now trading around the mid-$2,300s, with CoinMarketCap showing ETH near $2,360. Binance’s technical view also showed Ethereum’s daily 50-day moving average below current price, meaning ETH has begun to reclaim an important intermediate trend level.

That is the key point.

Ethereum does not need everyone to love it again for the chart to improve. It needs sellers to lose control. It needs buyers to defend higher lows. It needs price to reclaim moving averages that previously acted as resistance.

That process appears to be underway.

ETH is still not as clean as Bitcoin. It has more narrative baggage. It has more competition. It has more investor fatigue. But that is also what creates opportunity. When an asset with deep liquidity, institutional relevance, ETF infrastructure, and massive developer mindshare begins to move above its 50-day moving average while sentiment remains cautious, the risk/reward starts to improve.

Ethereum may not be leading the market yet, but it is no longer being left for dead.

HYPE: Hyperliquid’s Chart Is Starting to Wake Up

Hyperliquid remains one of the most important crypto stories of this cycle.

It is not just another exchange token. It is a vertically integrated onchain financial venue: a high-performance Layer 1, a decentralized perpetuals exchange, an onchain order book, and an emerging liquidity ecosystem wrapped into one asset.

That combination matters.

In the last cycle, DeFi proved that financial activity could move onchain. In this cycle, Hyperliquid is showing that sophisticated trading infrastructure can move onchain without sacrificing performance. That is a major Age of Autonomy theme: financial venues becoming programmable, transparent, composable, and increasingly autonomous.

The chart is now reflecting renewed interest.

HYPE recently traded above $44 and extended an upward trend for six straight days, with FXStreet noting that rising open interest and TVL were helping support the move. CoinGecko showed HYPE up roughly 10% over the past seven days, with a market cap above $10 billion.

Technically, HYPE is above 50-day moving average zone.

That makes HYPE especially interesting.

Bitcoin is the market leader. Ethereum is the institutional smart-contract platform. But HYPE is one of the purest expressions of crypto-native market structure innovation. If the market continues to move higher, assets with real usage, real revenue, and strong community ownership could attract outsized attention.

HYPE is one of those assets.

NAORIS: A Smaller-Cap Momentum Setup With an Age of Autonomy Angle

NAORIS is the speculative name on this list — and that should be stated clearly.

Bitcoin and Ethereum are core network assets. HYPE is now a major liquid crypto asset with significant market structure relevance. NAORIS is much smaller, earlier, and more volatile.

But that is also why the chart matters.

Naoris Protocol describes itself as a “Sub-Zero Layer 1” focused on decentralized cybersecurity, using post-quantum cryptography, Swarm AI, and its dPoSec consensus model to secure devices, smart contracts, and infrastructure. That puts NAORIS directly inside the Age of Autonomy theme: AI, devices, cybersecurity, cryptographic trust, and decentralized infrastructure.

As more machines, agents, sensors, and autonomous systems come online, cybersecurity becomes less of an IT department function and more of a foundational economic layer. Autonomous systems need identity. They need verification. They need resilience. They need security that can operate at machine speed.

That is the big-picture reason NAORIS is worth watching.

The near-term reason is simpler: the token is moving.

CoinGecko recently showed NAORIS up nearly 39% over seven days, with a market cap around $92 million. CoinMarketCap similarly showed NAORIS up more than 15% over 24 hours, with circulating supply of roughly 599 million tokens.

Because NAORIS is a newer and smaller token, its 50-day moving average should be checked directly on TradingView or your preferred charting platform before publication. But the current structure is clearly improving: price is moving higher, momentum is accelerating, and the market is beginning to reward the cybersecurity-plus-autonomy narrative.

This is exactly the type of chart that can move quickly when liquidity returns to smaller-cap tokens.

Why the 50-Day Moving Average Matters Here

The 50-day moving average is not a prediction tool.

It does not tell us the future. It does not guarantee a breakout. It does not eliminate risk.

But it does help us measure market behavior.

When price is below the 50-day moving average, rallies often run into sellers. The asset is still fighting its recent trend. When price moves above the 50-day, the psychology begins to change. Traders who were waiting for confirmation start to pay attention. Momentum strategies begin to re-engage. Pullbacks are increasingly viewed as entries rather than exits.

That is why this moment matters.

Bitcoin is back above key trend levels. Ethereum is reclaiming its intermediate structure. HYPE is pushing into a fresh momentum window. NAORIS is showing high-beta strength in a theme that connects directly to the future of autonomous infrastructure.

This is not yet a euphoric market.

That is the opportunity.

The best setups often happen when charts improve before sentiment does. By the time everyone agrees the bull market is back, the easy money has usually already been made.

The Bigger Picture: From Defensive to Opportunistic

The market is still dealing with major risks.

Geopolitics remain unstable. Fed policy remains important. Liquidity transmission is uneven. Bitcoin still needs to hold above key levels and prove that this move is more than a short-covering rally. Altcoin breadth still needs to improve.

But market timing is not about waiting until every risk is gone.

It is about recognizing when the balance of probabilities begins to shift.

Right now, the balance is shifting.

Bitcoin is no longer breaking down. Ethereum is no longer being ignored. HYPE is showing leadership from the crypto-native infrastructure side of the market. NAORIS is showing speculative momentum in one of the more important long-term themes: decentralized cybersecurity for an autonomous world.

That is enough to pay attention.

Not blindly. Not recklessly. Not with leverage just because a chart looks good.

But with the awareness that crypto may be transitioning from a defensive tape to an offensive one.

And when that happens, the assets above their 50-day moving averages are often the first place capital starts to flow.

Closing: The Market Moves Before the Story Is Obvious

The next phase of crypto will not be driven by one narrative alone.

It will be driven by macro liquidity, institutional adoption, onchain financial infrastructure, AI, cybersecurity, tokenized assets, and the continued buildout of the Age of Autonomy.

Bitcoin gives us the macro signal.

Ethereum gives us the programmable settlement layer.

Hyperliquid gives us the onchain financial venue.

NAORIS gives us a speculative window into decentralized cybersecurity for a machine-driven world.

Four different assets. Four different narratives. One shared technical message:

The charts are improving.

Crypto is on the move again.

In Gratitude,

-Jake Ryan

CIO Tradecraft Capital

Author, Crypto Investing in the Age of Autonomy

Author, Crypto Decrypted

That future needs neutral digital collateral.

Bitcoin is not an AI token. It is not a DePIN network. It is not a robotics platform. But it may be one of the foundational monetary assets for an increasingly autonomous economy because it provides something this new era will require: a scarce, permissionless, globally liquid store of value that exists outside any single institution, company, or sovereign balance sheet.

As AI agents become more capable, as machines begin coordinating commerce, as autonomous systems manage capital and infrastructure, the demand for digitally native settlement assets should grow. Stablecoins may handle much of the transactional layer. Tokenized real-world assets may bring traditional collateral onchain. Smart-contract platforms may host the application layer. But Bitcoin remains the pristine reserve asset of the crypto economy.

That matters for investors.

The current Bitcoin setup is not only about whether price can move from the mid-$70,000s toward $85,000. That is the tactical trade. The larger question is whether Bitcoin is becoming more deeply embedded in the financial architecture of the next technological cycle.

Our answer is yes.

In the Age of Autonomy, the world does not become less digital. It becomes more digital. It does not become less automated. It becomes more automated. It does not become less dependent on open, programmable, global networks. It becomes more dependent on them.

Bitcoin sits at the base of that transition.

Subscribe now

That is why the current bearishness is so interesting. The market is focused on the next Fed meeting, the next inflation print, the next ETF flow report, and the next geopolitical headline. Those things matter. They can absolutely move price over the near term.

But they do not change the long-term direction of travel.

The economy is becoming more autonomous. Capital is becoming more digital. Settlement is becoming more global. And monetary credibility is becoming harder to manufacture.

Bitcoin is still the asset most directly levered to that shift.

Why We Are Bullish

We are bullish because the market is not.

We are bullish because Bitcoin has recovered from the February lows while the narrative remains cautious.

We are bullish because price is holding near the 50-day trend area and has a clear path toward the $80,000–$85,000 resistance zone if momentum continues.

We are bullish because ETF demand has turned more constructive, even if flows remain volatile.

We are bullish because the macro picture is not perfect — but perfection is not required. Stabilization is enough.

And we are bullish because Bitcoin is no longer merely a speculative asset trading on liquidity cycles. It is becoming part of the monetary foundation of the Age of Autonomy.

In a world increasingly shaped by AI, robotics, sensors, autonomous agents, and blockchain-based coordination, Bitcoin remains the hardest, most liquid, most globally recognized form of digital collateral. That does not eliminate volatility. It does not immunize Bitcoin from Fed policy, oil shocks, ETF outflows, or geopolitical stress. But it does create a larger structural bid beneath the tactical setup.

This is how Bitcoin often behaves before the crowd changes its mind:

• First, the market stops going down.

• Then, it starts making higher lows.

• Then, it reclaims key moving averages.

• Then, resistance becomes the next magnet.

• Then, the same investors who were waiting for “one more dip” are forced to buy higher.

That is the setup today.

Bitcoin is not risk-free. The world is not calm. The Fed is not yet aggressively dovish. Geopolitical risk is real. Oil remains a serious macro variable. A liquidity shock could still create another drawdown.

But markets do not reward comfort. They reward asymmetry.

And right now, the asymmetry is improving.

Everyone is bearish.

That is why we are bullish.

And in the Age of Autonomy, Bitcoin’s role as neutral digital collateral may be more important than the market currently understands.

In Gratitude,

-Jake Ryan

CIO Tradecraft Capital

Author, Crypto Investing in the Age of Autonomy

Author, Crypto Decrypted

The quarter was defined less by a collapse in liquidity and more by a rise in uncertainty. That distinction matters. For much of the last cycle, we emphasized that expanding liquidity was the dominant force behind crypto price appreciation. We still believe liquidity matters over the long run. But Q1 challenged that framework in the short run. Markets did not behave as though liquidity was in control. Instead, they behaved as though uncertainty was.

This quarter, investors faced an unusual concentration of macro, political, and geopolitical uncertainty. Questions around monetary policy, growth, inflation persistence, conflict risk, regulatory direction, and broader cross-asset volatility combined to create an environment where capital stayed cautious. In that kind of market, liquidity does not disappear, but it does become less effective. It sits in the background while uncertainty drives positioning.

Here, we’ll explain why we are shifting our near-term emphasis away from liquidity as the primary active driver of crypto markets. Our current view is that until uncertainty meaningfully declines, liquidity will not be sufficient on its own to restart the next major leg higher. Put simply: news and events are driving the cycle right now, not liquidity.

Global Macro

Global Macro Summary for Q1

Q1 2026 was a reminder that markets do not respond to liquidity in a vacuum. Liquidity can improve in the background, but if uncertainty rises faster than confidence, capital will not behave as expected.

That is what defined the quarter. Rather than focusing on liquidity as the lead variable, investors spent Q1 reacting to uncertainty across multiple fronts. The market faced uncertainty around monetary policy, uncertainty around the durability of economic growth, uncertainty around inflation, uncertainty around geopolitics, and uncertainty around the policy environment more broadly. When many forms of uncertainty hit at once, investors tend to shorten time horizons, reduce risk, and avoid high-beta exposure. Crypto sits directly in the line of fire when that happens.

This is the key shift in our framework: liquidity did not disappear in Q1, but it failed to dominate. It failed to assert itself as the main driver because uncertainty was stronger. In practical terms, that meant crypto did not trade on improving background conditions. It traded on fear, hesitation, and an unwillingness by investors to fully embrace risk.

Bottom line: Q1 macro was not mainly a story of tightening liquidity. It was a story of elevated uncertainty overwhelming liquidity’s bullish impulse. Until that changes, crypto is unlikely to behave as though liquidity is the dominant driver of the cycle.

Subscribe now

Bitcoin Summary - Cautiously Bullish

Bitcoin in Q1

Bitcoin’s Q1 decline was not just a simple rejection of liquidity-driven optimism. It was a repricing of uncertainty.

In prior periods, Bitcoin often responded strongly to improving liquidity conditions and acted as an early beneficiary of monetary expansion. In Q1, that pattern broke down. Even where liquidity conditions were not clearly deteriorating, Bitcoin still traded lower because the market was focused elsewhere. It was focused on uncertainty.

That matters because it changes how we interpret price weakness. The takeaway is not necessarily that the longer-term bull case has ended. The takeaway is that the market is not yet willing to reward that bull case while uncertainty remains elevated. In other words, Bitcoin is being discounted not because liquidity is irrelevant, but because uncertainty is currently more relevant.

This also explains why rallies struggled to sustain momentum. In a confidence-rich environment, liquidity can push Bitcoin higher quickly. In an uncertainty-rich environment, rallies are sold, exposure is trimmed, and investors wait for clearer signals. That was the dominant pattern of Q1.

Our view is that Bitcoin can still recover meaningfully later in the cycle, but it likely needs a shift in narrative before it can do so. The next durable advance probably requires some combination of lower event risk, better policy clarity, calmer macro conditions, and reduced cross-asset stress. Until then, Bitcoin may remain more reactive to headlines than to liquidity.

Q1 Market Commentary

Layer 1s (and the execution layer renaissance)

The biggest story in Layer 1s was not simply token price. It was execution quality. Ethereum’s post-Pectra direction continues to center on making wallets and transactions behave more like applications, with features such as transaction batching, gas sponsorship, and smarter account behavior pushing the chain closer to mainstream usability. Solana, meanwhile, kept pressing its edge in raw throughput and activity: its February ecosystem update highlighted all-time highs in SOL-denominated TVL, roughly $95 billion in DEX volume for the month, and 3.4 billion transactions excluding votes. That is why the L1 race increasingly looks less like a branding contest and more like a contest over who becomes the best execution layer for actual economic activity. Hyperliquid is one of the most important new Layer 1s because it is proving that a chain built specifically for financial execution can win real market share. It has a current run rate of ~$1B in revenue! This is the highest revenue blockchain out there. They’re aiming to be the “market for everything” blockchain and they are executing on that vision.

DeFi (from protocols to onchain financial venues)

The most important DeFi shift in Q1 was that the category kept maturing from “protocols” into full-blown onchain financial venues. Aave remained one of the clearest examples: Defi Llama’s Q1 2026 view shows Aave V3 generating roughly $197 million in gross protocol revenue for the quarter, while broader protocol rankings placed Aave among the largest DeFi platforms by TVL at the end of Q1. The takeaway is that DeFi is no longer just about novel primitives. The winners are becoming scaled marketplaces for borrowing, liquidity, collateral, and increasingly institution-adjacent financial flow.

AI (agents, coordination, and decentralized compute)

The most important AI development in Q1 was the move from narrative to production rails. Bittensor’s own documentation makes clear that its subnet architecture is designed to create incentive-based marketplaces for AI-related digital commodities, including inference, training, storage, and other machine services. Akash is showing the same directional shift from idea to workload, positioning itself as an open marketplace for compute and highlighting real AI deployments on its network. The key point is that crypto AI is becoming less about “AI tokens” and more about coordination layers for agents, inference, and decentralized compute supply.

DePIN (real networks, real usage, real revenue pressure)

DePIN’s Q1 message was simple: the sector is being forced to prove real-world demand. Helium’s official mobile site emphasizes that its network is being used as a carrier offload solution by major U.S. carriers, while also expanding through Helium Plus so existing Wi-Fi infrastructure can join the network. Hivemapper, for its part, continues to show measurable mapping coverage and region-level progress through its live network metrics. In other words, the DePIN conversation is becoming less theoretical. Investors increasingly want to know which networks are attracting real usage, producing real service value, and facing real revenue pressure instead of relying on token incentives alone.

RWAs (the onchain cash stack takes shape)

The defining RWA theme in Q1 was the continued buildout of the onchain cash stack. Tokenized Treasury and money-market products kept scaling, with Defi Llama’s tokenized funds category showing more than $11.2 billion in active market cap, while Talos highlighted tokenized equities and 24/7 access to traditional assets as one of the quarter’s central structural trends. The result is that RWAs are no longer just a future promise. They are becoming the low-volatility yield layer, settlement layer, and institutional bridge layer that can anchor a much larger portion of onchain finance over time.

Subscribe now

In Closing

Q1 was a reminder that markets do not move on liquidity alone.

For much of the last year, we have argued that expanding liquidity was the dominant force behind crypto’s advance. We still believe liquidity matters over the medium and long term. But this quarter made something clear: when uncertainty becomes too elevated, liquidity stops being the lead actor. It remains in the background, but it does not drive price the way it otherwise might.

That is where we are today.

Crypto is not lacking for innovation. Layer 1s are improving execution. DeFi is maturing into real onchain financial infrastructure. AI and decentralized compute are moving from narrative toward coordination and utility. DePIN networks are being tested by real-world demand. RWAs are laying the foundation for an onchain cash and collateral stack. The industry continues to build.

What has changed is not the underlying direction of the space. What has changed is the market’s willingness to reward that progress in the presence of persistent uncertainty.

That is why we are adjusting our framework. For now, the most important question is not simply whether liquidity is rising. It is whether uncertainty is falling enough for liquidity to matter again. Until that happens, we should expect more hesitation, more false starts, and more sensitivity to headlines than many investors would prefer.

Still, periods like this are often where the next leaders separate themselves from the rest. When the market is no longer willing to reward everything, fundamentals begin to matter more. Real usage matters more. Real revenue matters more. Real resilience matters more.

That is where we remain focused.

We continue to believe that crypto is building the financial and coordination rails for the Age of Autonomy. But in the near term, we also recognize that markets need clarity before they can fully price that future. Until uncertainty breaks, discipline matters. Selectivity matters. Patience matters.

The long-term opportunity remains intact. The short-term path has simply become less linear.

In Gratitude,

-Jake Ryan

CIO Tradecraft Capital

Author, Crypto Investing in the Age of Autonomy

Author, Crypto Decrypted

They don’t.

The ugliest market declines often begin with a change in liquidity.

That is why the MOVE Index matters so much. The MOVE is the market’s benchmark for U.S. Treasury volatility. On the surface, that sounds like a rates-market statistic. In practice, it is much more important than that. Treasuries sit at the heart of the global collateral system. When volatility in that market jumps, the plumbing of finance starts to tighten. Hedging gets more expensive. Margin calls rise. Dealer balance sheets get less flexible. Funding conditions worsen. And suddenly investors are no longer asking, “What do I want to own?” They are asking, “What can I sell right now to raise cash?”

That is the real message of the MOVE Index.

Market Timing Alert! This one’s free.

See today’s:

It is not just the “VIX of bonds.” It is one of the clearest warning lights that the market may be shifting from a thesis-driven regime to a liquidity-driven regime. In a thesis-driven market, investors debate valuation, growth, adoption, earnings, tokenomics, and macro. In a liquidity-driven market, all of that temporarily takes a back seat. Good assets go down with bad ones. Long-term winners get sold with speculative junk. Not because investors have lost conviction, but because they need cash.

That distinction matters right now.

As the Iran conflict has intensified, markets have started to behave like the problem is no longer isolated to geopolitics. It is becoming a cross-asset liquidity event. Reuters reported on March 27 that investors from Shanghai to New York were scrambling to cut exposure amid war-driven volatility, with oil surging above $100 per barrel and traders describing a market with “no place to hide.” Reuters also reported that the bond market has already begun to register the shock, with the MOVE rising sharply in March as Treasury liquidity deteriorated.

This is where the macro chain reaction becomes critical.

If conflict around Iran broadens into a serious battle over the Strait of Hormuz, the consequences would extend far beyond the Middle East. The U.S. Energy Information Administration estimates that in 2024 about 20 million barrels per day of oil moved through the strait, accounting for roughly 20% of global petroleum liquids consumption. It also estimates that 84% of crude and condensate flows and 83% of LNG flows through Hormuz went to Asian markets, making Asia especially vulnerable to a disruption. The International Energy Agency likewise describes Hormuz as one of the world’s most critical oil chokepoints, with limited alternatives if flows are materially interrupted.

That is why this is not merely an oil story.

It is an inflation shock, a growth shock, and a liquidity shock all at once.

If Hormuz is materially disrupted, oil rises. If oil rises, inflation expectations rise. If inflation expectations rise, central banks become less able to ease. If rate-cut expectations get pulled back while war uncertainty rises, Treasury yields and Treasury volatility can jump together. And when Treasury volatility rises sharply, the MOVE tends to reflect that stress before many investors fully appreciate what it means for broader markets.

That is when the forced selling begins.

This is one of the most underappreciated features of modern markets: investors often sell risk assets even when they still like them. They sell because they are managing VaR. They sell because their funding costs just changed. They sell because a rates hedge moved violently against them. They sell because a prime broker wants more collateral. They sell because another part of the book is under stress and they need to free up cash immediately. In those moments, the market is not a voting machine on long-term quality. It is a liquidation mechanism.

This is also why crypto investors, equity investors, and even gold bulls should care about the MOVE.

When the market is forced to raise cash, correlations tend to rise. Assets that normally trade on different narratives begin to move together. Gold can get sold. Equities can get sold. Credit can get sold. Crypto can get sold. The first move in a genuine liquidity event is often not elegant. It is not selective. It is blunt. Investors dump liquid exposures first and ask questions later. Reuters’ March 27 reporting captured exactly that kind of environment, with investors describing broad stress and difficulty finding traditional places to hide.

So the MOVE should be viewed as more than a bond-volatility gauge. It is a practical indicator of whether the system is approaching a cash-raising phase.

When it stays contained, risk markets can usually absorb geopolitical noise. But when it spikes alongside surging oil, higher yields, and stress around a chokepoint as important as Hormuz, the risk is that markets stop trading on preference and start trading on necessity. That is the moment investors need to understand. Because once the market enters necessity mode, asset prices can fall hard even when the underlying long-term story remains intact.

And that may be the most important lesson here.

In calm markets, investors buy stories.

In stressed markets, they sell assets to raise cash, so caution should be considered. We should expect central banks to come to the rescue by providing liquidity. Becoming more defensive until a settling of the MOVE Index can be prudent.

Investors do that by raising cash. This is key.

That is why the MOVE Index matters. It tells you when the market’s core collateral base is becoming unstable. And when that happens during an oil shock, a war scare, or a Strait of Hormuz disruption, you should not assume risk assets are falling because investors suddenly hate them. More often, they are falling because the system has moved into cash-conservation mode.

In other words: when the MOVE screams, the market is no longer debating upside.

It is searching for liquidity.

Happy Hunting,

— Jake Ryan

CIO Tradecraft Capital

Author, Crypto Investing in the Age of Autonomy

Author, Crypto Decrypted

But the deeper story is happening off-screen.

AI is moving into the physical world. It is being paired with cameras, radar, industrial telemetry, tactile sensing, wireless modules, and machine vision. In other words, it is being fused with the sensory layer of the real economy. And once that happens, AI is no longer just helping people work. It starts helping machines perceive, decide, and act. That is where autonomy begins.

This matters because autonomy is not just another productivity upgrade. It can become a competitive weapon.

When a business embeds AI directly into operations, it does more than automate a task. It creates a feedback loop. More deployed systems generate more real-world data. More real-world data improves the models. Better models improve uptime, safety, speed, and cost. Those gains fund wider deployment, which generates still more data. Over time, this can create an advantage that is increasingly hard for rivals to copy. In the Age of Autonomy, the moat is no longer just brand, scale, or distribution. It is the ability to build systems that learn from the physical world faster than competitors do.

Here are four recent examples that show what that looks like in practice.

1) Siemens and PepsiCo: when a factory becomes a learning system

One of the clearest examples came at CES 2026, when Siemens introduced its Digital Twin Composer, a system designed to combine industrial AI, simulation, and real-time operational data inside a virtual environment. Siemens is not pitching this as a flashy demo. It is pitching it as a way for manufacturers to model production environments, test changes virtually, and improve operations before making costly physical changes on the factory floor.

The PepsiCo use case is what makes this especially interesting. PepsiCo announced that it is working with Siemens and NVIDIA on an AI and digital twin collaboration for industrial operations. Siemens said the system can help create detailed digital replicas of factories and warehouses, enabling teams to simulate workflows and identify operational problems before they happen in the real world. Siemens specifically said the approach can identify up to 90% of potential issues before physical modifications occur.

That is not just “using AI.” That is learning faster than your competitors.

A company that can test layout changes, reduce downtime risk, improve throughput, and refine workflows in a virtual environment has an edge that compounds. Every production cycle creates more data. Every improvement makes the next improvement easier. And because that system is trained on a company’s own facilities, process flows, bottlenecks, and historical operating data, it becomes more proprietary over time. The result is a quiet but powerful moat: the factory itself becomes smarter, more adaptive, and harder to catch.

2) Kodiak and Bosch: autonomous trucking is becoming real infrastructure

Autonomous driving has spent years in the realm of promise. But a more practical version of the story is now emerging in freight.

In January, Reuters reported that Kodiak AI partnered with Bosch to scale production hardware and sensors for self-driving trucks. Under the agreement, Bosch will provide automotive-grade components including sensors and actuation systems, while the two companies develop a production-ready redundant platform for Kodiak’s autonomous trucking stack. Reuters also noted that freight is seen as one of the clearest paths to commercial viability because the routes are more predictable and the economics are easier to prove than in consumer robotaxis.

This is important because the moat in autonomous trucking is not just the driving model. It is the operational stack around it.

If a company can run freight routes more safely, more consistently, and with higher vehicle utilization, it gains an economic edge. But it also gains a data edge. Every mile driven creates more training data. Every real-world edge case improves the system. Every commercial deployment makes the platform more credible to fleets and shippers. Over time, that can create a gap that is difficult for late entrants to close, because the leader is not just improving its software in theory. It is improving through live operations in the field.

That is a hallmark of the Age of Autonomy: the best systems do not merely automate work. They learn from doing the work.

Actionable Crypto News, Save 50%

3) Hyundai and Boston Dynamics: robotics moves from spectacle to plant economics

Another strong signal came from Hyundai Motor Group, which said it plans to deploy Boston Dynamics’ humanoid robot Atlas at its Georgia factory beginning in 2028. Reuters reported that Atlas is intended to start with parts sequencing, then gradually expand into more complex manufacturing tasks. Hyundai also said the robot is designed for industrial environments, can lift up to 50 kilograms, and includes tactile sensing in its hands.

That announcement matters because it marks a transition in how robots are being framed. For years, humanoid robots lived mostly in videos and research labs. Now they are being discussed in terms of plant deployment, production targets, and operational roles. Reuters reported that Hyundai aims to build a factory capable of producing 30,000 robots annually by 2028 and sees humanoid robotics as part of a larger push into “physical AI.”

The defensive corporate moat here comes in two forms.

First, there is the direct operating benefit. If robots can reliably handle repetitive or physically demanding tasks, companies can improve consistency, reduce strain on workers, and keep production running more smoothly. Second, there is the learning loop. Hyundai is not simply buying robots from a third party. Through Boston Dynamics, it is helping shape the hardware, software, and deployment path. That means every factory rollout can feed back into the next generation of capability. And once a company has years of real-world robotics data tied to its own workflows, training regimes, and environments, it becomes much harder for a competitor to buy a similar advantage off the shelf.

This is how a technology trend turns into a moat: when deployment itself becomes a source of proprietary learning.

4) Qualcomm and Wayve: integrated autonomy stacks are starting to matter more than point solutions

The final example is more subtle, but just as important.

Last week, Reuters reported that Qualcomm and Wayve are partnering to combine Wayve’s AI Driver software with Qualcomm’s Snapdragon Ride chips and active safety software. The goal is to help automakers roll out advanced driver-assistance and automated-driving systems more quickly across a range of vehicles, from entry-level platforms to more advanced driving capabilities.

Why does this matter?

Because the race in autonomy is no longer only about who has the smartest model. It is increasingly about who can deliver a complete system that is easier to deploy, validate, and improve over time. Automakers do not just need intelligence. They need sensing, compute, safety, software integration, and update-ability across an entire vehicle fleet. Reuters said the combined Qualcomm-Wayve platform is meant to help carmakers standardize this more efficiently across models and geographies.

That kind of integration can become a moat because it reduces friction. The easier autonomy is to deploy, the faster it scales. The faster it scales, the more real-world data the system collects. And the more data it collects, the stronger the performance can become. This is one reason the winners in the Age of Autonomy may not be the companies with the most impressive standalone demos, but the ones that can build complete, real-world stacks that improve continuously after deployment.

Spring Sale, Save 50%

The Bigger Picture

Put these examples together and a broader pattern emerges.

The next wave of competitive advantage is not just about having AI. It is about embedding AI into systems that can sense the world and act within it. A factory that can simulate and optimize itself. A trucking network that learns from every mile. A robot fleet that improves with every deployment. A vehicle platform that can perceive, decide, and update over time.

That is why the intersection of IoT and AI matters so much. Sensors are what allow a machine to observe reality. AI is what allows it to interpret what it sees. Connectivity and software are what allow those lessons to spread across a fleet, a factory, or a network. When those pieces come together, operations themselves become intelligent. And once operations become intelligent, the companies that deploy autonomy well can begin to widen the gap between themselves and everyone else.

In the Age of Autonomy, the ultimate prize is not simply automation.

It is adaptive operations.

It is building businesses that do not just run, but learn.

And over the coming decade, that may prove to be one of the most powerful moats in the global economy.

Age of Autonomy
The future belongs to businesses that can sense, decide, and act faster than the world around them changes.

Happy Hunting,

— Jake Ryan

CIO Tradecraft Capital

Author, Crypto Investing in the Age of Autonomy

Author, Crypto Decrypted

That made sense for a while. Growth looked shaky. Liquidity was improving, but not forcefully enough to overwhelm market anxiety around rates, geopolitics, positioning, and policy uncertainty. In that environment, investors leaned defensive, and crypto often behaved as though it were waiting for clearer confirmation before repricing higher.

But the data has changed.

The most important shift is this: U.S. ISM Manufacturing has now been above 50 for two straight months. January printed 52.6 and February followed at 52.4, both in expansion territory. ISM also says the February reading implies the overall economy continued expanding, extending a long run of growth rather than signaling contraction.

At the same time, liquidity is still moving in the right direction. U.S. M2 rose to $22.442 trillion in January 2026 from $22.366 trillion in December 2025, while FRED’s year-over-year M2 growth reading was 4.3% in January. MacroMicro’s major-central-bank aggregate also showed January 2026 M2 at about $99.562 trillion, with 11.09% year-over-year growth.

That combination matters.

When growth is stabilizing or re-accelerating and liquidity is expanding, crypto usually has a stronger foundation. Not because every day goes up, and not because volatility disappears, but because speculative assets tend to perform best when money supply and economic momentum are no longer working against them.

The Market Is Still Pricing Too Much Caution

To me, the issue is not whether macro has improved. It has.

The issue is whether crypto has fully priced that improvement. I do not think it has.

Bitcoin is currently trading around $74,133, with an intraday range of roughly $71,459 to $74,420 on March 16, 2026. That is a respectable price, but it still looks conservative relative to a backdrop in which manufacturing is expanding again and liquidity is still climbing.

That is the disconnect.

The market is still behaving as though the economy is slipping deeper into a late-cycle slowdown. Yet two straight ISM expansion prints argue otherwise. Rising M2 argues otherwise. The macro data is no longer painting a clean “Fall” picture. Instead, it is beginning to look more like a reversion back toward late Summer.

And that distinction matters.

Late Summer is still a supportive regime for crypto. It is not the earliest phase of the cycle, when everything is washed out and violently rebounds. It is the stage where growth is still good enough, liquidity is supportive enough, and risk assets can continue climbing even if the move becomes more selective and narrative-driven.

That is why I think prices should be higher.

Upgrade for Actionable Intel, Save 50%!

Why ISM Above 50 Is So Important

ISM is one of the best real-economy signals because it tells us whether the industrial side of the economy is expanding or contracting.

For months, the market narrative has focused on rates, ETF flows, election-related policy risk, and central bank speculation. Those things do matter. But they can obscure the bigger truth: crypto tends to do best when growth and liquidity are aligned. Right now, that alignment is improving.

January’s 52.6 reading could have been dismissed as a one-off rebound. February’s 52.4 makes that much harder to do. Two straight months above 50 suggests this is not just a false start. It suggests expansion has returned to manufacturing.

That does not mean the economy is booming. It does mean the economy is not rolling over the way many investors feared.

And once that becomes clear, the bearish case gets harder to sustain.

Because if growth is expanding and money supply is also expanding, then why should crypto remain priced as though macro were still deteriorating?

Global M2 Is Still Driving the Train

I have argued repeatedly that global M2 drives the train.

Not perfectly. Not every week. Not every headline. But over intermediate timeframes, expanding liquidity remains the dominant macro force for crypto. If more money is entering the global system, scarce and high-beta assets tend to benefit. Crypto sits near the front of that line.

That is why the January increase in U.S. M2 matters. It is also why the major-central-bank M2 aggregate matters. Even if the path is uneven, the directional message remains constructive: liquidity is still expanding.

This is the kind of macro backdrop in which Bitcoin usually stops feeling “stuck” and starts repricing higher.

The market may not do it immediately. Positioning can delay it. Rates can delay it. A geopolitical shock can delay it. But delay is not the same thing as invalidation. Macro often turns before price fully responds.

That may be exactly where we are now.

From Fall Back to Late Summer?

This is where the Four Economic Seasons framework becomes especially useful.

A few months ago, it was reasonable to argue that the economy was moving from Summer toward Fall. Growth looked softer. Inflation was no longer falling cleanly enough to reassure markets. Risk assets were becoming more selective. That interpretation fit the available evidence.

But fresh evidence can change the season call.

If ISM is back above 50 for two consecutive months and liquidity continues to expand, then the cleaner interpretation may be that we are not moving deeper into Fall after all. We may be reverting back toward late Summer.

That does not mean every asset rallies equally. It does not mean there is no volatility left in the cycle. It means the macro regime may be more favorable for crypto than consensus currently believes.

And if that is true, then the market’s current pricing is too pessimistic.

Upgrade, 50% Off for Spring!

Bitcoin Support and Resistance Levels

From a market-structure perspective, Bitcoin is giving us a useful map.

The first key zone is support in the low-$70,000s. Based on today’s range, that support looks roughly centered around $70,000. As long as Bitcoin can defend that area, the broader constructive thesis remains intact.

The next key zone is resistance near $74,000. Bitcoin has pushed into that area, but it still needs to break and hold above it convincingly. Then, we need to look at $90k level. That will be a key area of resistance. A clean move through that band would suggest the market is finally beginning to price in the better macro setup.

So the current framework is fairly straightforward.

Hold the low-$70,000s, and Bitcoin has a base. Clear the mid-$74,000 area, and the market has room to reprice. Look to test in the low $90ks. If that happens, leadership should broaden and higher-beta sectors of crypto could start participating more aggressively.

That is why this moment matters.

We are not looking at a market in freefall. We are looking at a market that may still be underpriced relative to what the macro data is now saying. Key areas of risk are geopolitical risk, tariff news risk and any news from the Fed that’s hawkish. If this war with Iran lasts longer than a month more, that would raise concerns.

Why Crypto Prices Should Be Higher

Put it all together and the thesis becomes pretty clear. While we do need to manage short-term risk, the macro backdrop is supportive.

ISM has been above 50 for two straight months. U.S. M2 is rising. Major-central-bank M2 is rising. The economy looks less like it is sliding into Fall and more like it is drifting back toward late Summer. Bitcoin is holding in the mid-$70,000s instead of breaking apart.

That combination should be producing higher crypto prices.

Not in a straight line. Not without volatility. Not without pullbacks and narrative scares. But the macro foundation is stronger than current pricing suggests.

To me, that means the market is still catching up.

And when the market finally does catch up, the move may feel obvious in hindsight. Growth was improving. Liquidity was expanding. The season had shifted. The setup was there.

Price just had not fully recognized it yet.

Final Thoughts

For months, investors have been conditioned to interpret every rally with suspicion. That instinct was understandable while the macro picture was muddled.

It is less understandable now.

The data is improving. Manufacturing is expanding. Liquidity is still flowing. The economic season appears more supportive than it did just a few months ago. And in that kind of environment, Bitcoin and crypto broadly should not stay sleepy forever.

The macro backdrop is leaning bullish.

Now price needs to catch up.

The Age of Autonomy is not waiting. Capital will eventually flow toward the systems, networks, and digital assets best positioned to absorb expanding liquidity and accelerating economic complexity. The only real question is whether investors recognize the shift before the market does it for them.

Happy Hunting,

Jake Ryan

CIO Tradecraft Capital

Author, Crypto Investing in the Age of Autonomy

Author, Crypto Decrypted

The “Age of Autonomy” isn’t one breakthrough. It’s the latest technological revolution. Just like the Age of Information and the Internet was in the last long-wave economic cycle. It’s the stack coming online at once:

• IoT (sensing + telemetry: the world becomes machine-readable)

• AI (perception + decision-making: software learns to operate)

• Robotics (actuation: machines do the work in the physical realm)

• Blockchain (coordination + incentives: autonomous software can transact, hire, and generate economic activity in the digital realm)

What changed in 2025–2026 is that autonomy stopped being “cool demos” and started looking like repeatable operations: fleets, routes, deployments, marketplaces, and enterprise agent platforms.

Below are 10 businesses that (in different ways) demonstrate the shift — plus the “why it matters” for investors and builders.

1) Waabi — autonomy for trucking, expanding to robotaxis

Tech stack: IoT ✅ | AI ✅ | Robotics ✅ | Blockchain ◻️
Summary: Waabi builds a “virtual driver” that learns in simulation and then transfers to real-world driving. In early 2026, it highlighted major funding and partnerships focused on scaling deployments.
Why it matters:
Autonomous driving is a software-defined labor replacement in one of the largest cost centers on earth: logistics + mobility. The near-term unlock isn’t “all roads, all conditions.” It’s commercial scaling in constrained environments (known cities/routes, known vehicle types). If Waabi (and peers) succeed, the prize is massive: lower cost-per-mile, higher fleet utilization, and a new AI-robotics flywheel where every mile improves the model.

2) Gatik — fully driverless middle-mile logistics

Tech stack: IoT ✅ | AI ✅ | Robotics ✅ | Blockchain ◻️
Summary: Gatik focuses on the middle-mile: predictable, repeating routes between distribution hubs and retail locations — the “boring” part of autonomy that’s actually easiest to scale.
Why it matters:
Middle-mile is where autonomy can win first because it’s repeatable. That matters because repeatability is what converts AI from a research project into a cash-flowing business. Every time autonomy moves from “pilot” to “daily commercial operation,” it validates the operating model: safety framework, remote ops, insurance, maintenance, unit economics. This is how autonomy becomes infrastructure.

3) Serve Robotics — sidewalk delivery at scale

Tech stack: IoT ✅ | AI ✅ | Robotics ✅ | Blockchain ◻️
Summary: Serve deploys autonomous sidewalk delivery robots for local logistics (think food and small parcels). 2025–2026 was about visible deployment milestones and scaling fleets in real neighborhoods.
Why it matters:
Last-mile delivery is a margin war. If robots take even a slice of the short-distance, high-frequency runs, you get:

• Lower delivery cost,

• More deliveries per hour,

• Higher throughput at peak times.
  It’s also a key autonomy pattern: small, cheap, many. Those fleets create data loops and operational moats that resemble rideshare dynamics — but with robots.

4) Starship Technologies — autonomous delivery + partnerships

Tech stack: IoT ✅ | AI ✅ | Robotics ✅ | Blockchain ◻️
Summary: Starship has been one of the earliest at-scale sidewalk autonomy operators and expanded partnerships/markets during 2025–2026 (including app-based delivery integrations).
Why it matters:
Starship shows the difference between “a robot” and a robot network. The moat isn’t the chassis — it’s: mapping, operations, charging, maintenance, routing, customer behavior, city-level permissions. As these networks grow, they start to resemble a new logistics layer: hyperlocal, automated, always-on.

Act Now! Save 50%!

5) Greenfield Robotics — autonomous weeding + farm operations

Tech stack: IoT ✅ | AI ✅ | Robotics ✅ | Blockchain ◻️
Summary: Greenfield’s model is practical autonomy in agriculture: robots that reduce chemical usage and automate repetitive fieldwork, with humans supervising exceptions.
Why it matters:
Agriculture is one of the best early markets for autonomy because labor is scarce, fields are semi-structured, and ROI can be obvious. The bigger theme: autonomy enables decentralized production (smaller farms can compete) and drives a multi-decade shift toward precision farming. This also sets up a broader “autonomous ops” template: many low-cost agents in the field, supervised by a small human team.

6) Carbon Robotics — LaserWeeder + autonomous tractor kit

Tech stack: IoT ✅ | AI ✅ | Robotics ✅ | Blockchain ◻️
Summary: Carbon uses computer vision to identify weeds and eliminate them with lasers; it’s a flagship example of “AI + robotics” replacing chemicals with computation. It also pushes autonomy hardware kits for farm machinery.
Why it matters:
This is a clean illustration of the autonomy economy: turn inputs into intelligence. If lasers replace herbicide and AI replaces manual labor, the farm becomes a compute-driven system. That has two knock-on effects investors should care about:

1. autonomy creates recurring software-like economics on top of equipment, and

2. it expands TAM because “robotic capability” becomes upgradable, like iOS updates — not a one-time machine purchase.

7) Agile Robots — industrial humanoids moving toward real work

Tech stack: IoT ✅ | AI ✅ | Robotics ✅ | Blockchain ◻️
Summary: Agile Robots introduced a humanoid platform for industrial settings (logistics/manufacturing tasks like pick-and-place, machine tending, material movement).
Why it matters:
Humanoids are the controversial frontier: skeptics call them demos; believers call them the “general-purpose robot.” The investable takeaway isn’t whether humanoids replace all warehouse workers tomorrow — it’s that factories and warehouses are finally getting robots that can handle variability. If even limited humanoid deployments stick (one shift, one aisle, one set of tasks), we get a new labor curve: the marginal cost of work drops as robotics gets more general.

8) Addverb — Elixis-W wheeled humanoid for industry

Tech stack: IoT ✅ | AI ✅ | Robotics ✅ | Blockchain ◻️
Summary: Addverb’s wheeled humanoid announcement fits the broader trend: industrial humanoids designed for practical intralogistics rather than sci-fi spectacle.
Why it matters:
The “wheeled humanoid” concept is a pragmatic compromise: keep mobility stable, focus on manipulation + task execution. The bigger theme is global: humanoids aren’t a Silicon Valley-only bet. As more regions build them, we move toward a competitive manufacturing base for autonomy hardware — and that’s how adoption accelerates (cost curves drop, availability increases, iteration speeds up).

9) Olas / Autonolas — on-chain marketplace where AI agents hire AI agents

Tech stack: IoT ◻️ | AI ✅ | Robotics ◻️ | Blockchain ✅
Summary: Olas pushes agent-to-agent coordination: marketplaces where autonomous agents can offer services, hire other agents, and settle payments programmatically.
Why it matters:
This is the clearest example of “autonomous software businesses.” If agents can:

• discover services,

• negotiate/contract,

• pay, and

• verify outcomes,
  then you get machine-to-machine commerce. That’s not just “crypto + AI.” It’s the idea that future businesses may have no employees, just agent swarms coordinating via markets. The investable surfaces: token incentives, marketplace fees, infrastructure rails, and agent tooling.

Act Now! Save 50%!

10) OpenAI Frontier — enterprise agent operations layer

Tech stack: IoT ◻️ | AI ✅ | Robotics ◻️ | Blockchain ◻️
Summary: Frontier positions “agents” as deployable enterprise software: permissions, governance, onboarding, shared context, and management — the plumbing needed to run AI coworkers in real orgs.
Why it matters:
The next phase of AI isn’t better chat. It’s AI that does work: triage, reconcile, schedule, research, draft, follow-up, execute. That requires governance (what can it do?), tooling (what can it access?), and repeatability (can we trust it?). Platforms like Frontier matter because they turn agents into a controllable operating layer, which is what unlocks enterprise spend at scale.

The Pattern: Autonomy is graduating from “model” to “operating system”

Across all 10 examples, the winning formula looks like this:

1. Instrument the environment (IoT / telemetry)

2. Perceive + decide (AI)

3. Act in the real world (robotics) or act in the digital world (agents)

4. Coordinate and settle (blockchain in some domains; enterprise governance in others)

This is why “Age of Autonomy” is an investable mega-trend: it’s not one product category — it’s the automation of operations, everywhere.

Closing: What to watch next

If you want a simple scoreboard for 2026, watch three things:

• Fleet scale (number of robots/vehicles/agents deployed, not pilots)

• Unit economics (cost-per-task and utilization rates)

• Coordination layer (marketplaces, governance, permissions, settlement)

When those three line up, autonomy stops being a bet — and becomes a business. The Age of Autonomy is not a theme. It’s the next operating system of the global economy.

Happy Hunting,

— Jake Ryan

CIO Tradecraft Capital

Author, Crypto Investing in the Age of Autonomy

Author, Crypto Decrypted

In every cycle, the market peak doesn’t arrive roaring. It arrives whispering: subtle shifts in macro and on-chain behavior that most participants ignore until it’s too late.

For most of 2025, crypto observers have argued about narratives — halving, ETF flows, retail sentiment. But the real story is being written in macro liquidity, manufacturing momentum, valuation, and dominance rotation.

Below, we lay out the five key metrics we watch to understand when the market has moved from a constructive phase to a risk-management phase, and what the latest data is telling us right now.

Macro “Season” — Have We Entered Fall?

Traditional economic season models aren’t as widely reported as PMI figures, but the spirit of the metric is this:

• Spring / Summer: liquidity and risk appetite expand

• Fall: growth slows, risk appetite falters

• Winter: contraction sets in

Today’s data suggest we are not in a classic expansion phase — but neither are we at a full contraction. Manufacturing activity in the U.S. has slipped back into contraction territory, with the ISM PMI remaining below the neutral 50 level after briefly crossing above it earlier this year. This suggests weakening underlying economic momentum rather than acceleration.

So while the market still prices in growth, the macro backdrop on balance looks more like late-cycle drift than mid-cycle strength — a subtle nudge toward “Fall.”

Current signal: 🔴 warning - are we going back to summer or staying in cycle?

Act Now! Save 50%!

2. ISM Manufacturing Index — Not Overheating Yet

One of the few hard numbers we have on the broader U.S. economy is the ISM Manufacturing PMI — a monthly snapshot of industrial activity.

Bullish narratives often focus on readings above 50 as growth, and above 60 as “overheating.” But in 2025, the PMI briefly touched above 50 before reverting into contraction, reflecting ongoing structural weakness.

What this means: We are not seeing the kind of sustained, overheated expansion (ISM > 60) that typically precedes late-cycle liquidity tightening.

Current signal: 🟠 not in red (no ISM ≥ 60) however under 50 is warning.

3. Global M2 Liquidity — Liquidity Still Growing But With Mixed Signals

Liquidity — especially global M2 money supply — is the silent engine behind risk assets.

In early 2025, researchers pointed out that global M2 was surging past prior highs (~$112T), and that Bitcoin’s price historically moved in step with this liquidity expansion.

However, short-term liquidity growth and its impact can oscillate. Analysts have also flagged periods where global M2 has flattened or contracted, which tends to herald risk asset weakness after a delay.

The key takeaway: Liquidity is currently still above long-term trend, but recent trends show flattening or contraction in certain regions.

Current signal: 🟢 near warning, but still expanding.

4. Bitcoin MVRV — On-Chain Valuation Is Approaching Extreme

One of the most powerful on-chain valuation metrics is Bitcoin’s Market Value to Realized Value (MVRV).

At cycle peaks, MVRV tends to blow past 3.5 — a sign that price has stretched far beyond the average cost basis of investors, and distribution pressure rises sharply.

Right now, data suggests that long-term holder MVRV sits around ~3.11, which is below the highest extremes of past cycles but well into lofty territory, especially compared to historical norms.

In other words: valuation risk is elevated but not yet at classic cycle-top exhaustion.

Threshold: 🔴 MVRV > 3.5
Current: 🟢 Current Bitcoin MVRV ~1.9

5. Bitcoin Dominance — Capital Rotation Underway

Bitcoin dominance — the share of total crypto market cap that BTC represents — is a window into where capital is flowing.

In a classic late cycle, capital shifts out of Bitcoin and into altcoins as risk appetite becomes indiscriminate.

Recent data from mid-2025 showed that altcoins had been gaining share versus Bitcoin at certain points, although BTC dominance remained well above the critical danger zone (<40%) and closer to historical mid-cycle levels (e.g., ~55% reported earlier in 2025). Comma

Current signal: 🟢 no red (still above danger threshold)

Pulling It Together: Where We Are Now

Here’s the distilled view, metric by metric:

Metric

Overall picture: The market isn’t screaming “top” yet — but it’s quietly clustering toward caution. Multiple indicators are approaching risk thresholds without yet triggering them. We need to see expanding global money supply turn into economic activity.

Two Important Insights for Premium Readers

1. Macro and On-Chain Divergence Matters

The macro backdrop is weakening unevenly — manufacturing is soft, liquidity growth is mixed — while Bitcoin’s valuation metrics are creeping toward cycle-top levels. This divergence is often a hallmark of the late speculative phase, where narratives carry price beyond fundamentals, right before the reset.

2. Red is Not a Switch — It’s a Gradient

None of the metrics are in full red danger yet. But the trajectory matters. When multiple indicators shift toward red simultaneously, the probability of a sharp top increases.

In other words: the market feels more fragile now than at prior cycle peaks.

What Comes Next?

No one can declare a top weeks in advance with precision. But late-cycle risk is not some mystical endgame — it’s measurable.

Every top in crypto has come not during economic expansion, but during late-cycle conditions paired with stretched valuations and liquidity inflections.

Today’s framework suggests:

• Macro momentum is weakening

• Liquidity is plateauing

• On-chain valuations need to see more blockchain transactions

• Capital rotation is underway

That’s the quiet narrative before the crowd notices.

ARE WE ENTERING CRYPTO WINTER?

We are using 6 global metrics to help us gauge where the bull market “Top Finder” is. We are also watching other researcher and investors own top finders to see where we are in the cycle. Only 1 of our metrics is “flashing red.”

Here they are currently:

Key Metric:

Currently, the Fed cut rates in December and provided us with an updated “dot plot.” However, that’s almost irrelevant since Trump has promised to bring in a new Fed Chief in May who will cut rates and most likely announce the candidate by the end of January. Polymarket has a dovish Fed Chair candidate at 94%. That means the rate cutting cycle is going much longer than people are thinking.

Therefore, Global M2 Money Supply will be expanding for longer as well. They need this so they can roll the $8T of USTs from Covid that are coming due to maturity and need to roll them at the lowest possible rates. This should create growth which should push up the ISM metric and move us back to Summer/Fall in the Economic Seasons delaying the cycle.

This added liquidity will move into risk assets like it always does and push crypto prices higher. Remember, bitcoin is the asset most correlated to global M2 with an 83% correla@on. We should then also have an actual “alt season” which will push Bitcoin Dominance lower.

We’re currently in a trading range and need a catalyst to move higher. We could retest $85k in bitcoin and bounce. We could test the next lower major support level at $72k and that would be a retest of the current local all-@me low. But that needs to hold for the bull case to remain intact.

THIS IS KEY - In the past 10 years and 2 cycles, we’ve always had:

• contracting global M2, ISM > 60

• Late Fall for Economic Season

• BTC MVRV Ra@o > 3.5

• Bitcoin Dominnance < 40%

• BEFORE we saw the market top and began our descent into crypto winter!

We aren’t close to any of that and that’s our case for why we’re in a super cycle and this is going much longer than people think.

What would change my position?

• If ISM is not above 50 by Q1 2026

• If Global M2 Money Index started to contract

• If a “hawkish” Fed Chair was nominated

Otherwise, we’ve not seen the cycle highs just yet.

Happy Hunting,

Jake Ryan

CIO Tradecraft Capital

Author, Crypto Investing in the Age of Autonomy

Author, Crypto Decrypted

Every cycle, the crypto industry finds itself fighting the same battle: the gravitational pull of the old Bitcoin Halving narrative. It was a neat mental model when crypto was small, illiquid, and mostly retail-driven. But today—after spot ETFs, institutional flows, and macro integration—the halving has become a sideshow, not the main event.

The new story is liquidity.
The new clock is global M2.
The new cycle is not four years—it’s closer to five.

And now, with fresh ISM data showing a slowing economy and political shifts pointing toward a potentially more dovish Fed Chair in mid-2026, the liquidity-driven view of crypto is more important than ever.

Let’s break it down.

1. The ISM Just Confirmed We’re in a Slow-Growth, Late-Cycle Economy

The ISM numbers are in:

• Manufacturing PMI: 48.2, contracting (< 50)

• Services PMI: 52.4, still expanding (> 50)

This combination—manufacturing weakness with services resilience—is the classic signature of a late-cycle, slow-growth economy. It isn’t recession. It isn’t early expansion. It’s the plateau before the next major move.

What does this season typically deliver?

• Slower GDP growth

• Higher-for-longer rates

• Rising credit stress

• Consumer fatigue

• Moderating inflation

• Pressure on central banks to ease before the cycle fully rolls over

In other words:

The economy is weakening just enough to force a policy shift, but not enough to collapse outright.
And that is exactly the environment where liquidity starts to matter more than growth.

In crypto terms, this is the inflection point.

2. Why ISM Matters for Crypto

Slow growth puts the Fed in a bind:

• growth is too soft to keep tightening,

• but inflation is too sticky to cut aggressively.

Historically, when ISM sits in this “contracting manufacturing + positive services” zone, the Fed eventually pivots to support credit markets and prevent a deeper recession. That pivot comes through:

• rate cuts,

• balance sheet stabilization,

• or outright liquidity injections.

Crypto doesn’t care about the ISM directly.
But it cares deeply about the liquidity triggered by this macro regime.

Which brings us to the heart of your thesis.

3. The 4-Year Halving Cycle Is Dead. Long Live the 5-Year Liquidity Cycle.

Let’s be blunt:

The ISM is slowing.

The economy is late-cycle.

Liquidity pressures are rising.

Sovereign debt rollover is accelerating.

These forces don’t move on a 4-year halving schedule.
They move on liquidity waves—typically every 4.5 to 5 years.

Here’s how the cycles line up:

The halving didn’t drive these peaks. Global M2 did.

And the next global M2 expansion—already brewing beneath the surface—aligns not with the halving, but with the 2025–2026 sovereign debt refinancing window.

So the ISM slow-growth data is not a bearish signal.

It’s the early signal that liquidity expansion is coming.

Save 50% on the Real - Act Fast!

Get actionable articles that help you generate alpha through asset selection, portfolio construction and market timing.

4. Powell Leaving in May 2026 Could Change Everything

Now add a political wrinkle.

If Chair Powell steps down in May 2026, and President Trump appoints a more dovish, pro-growth, pro-liquidity Fed Chair, the timing of the liquidity wave changes dramatically.

A dovish Chair—especially one selected by a politically pro-growth administration—would likely do the following:

• Cut rates earlier than the data demands

• Support asset markets explicitly

• Ease financial conditions to avoid recession

• Re-expand the balance sheet if markets wobble

• Tolerate higher inflation if needed

This is essentially “Powell pivot 2.0,” but with a more aggressive easing bias.

And from a liquidity standpoint:

This could extend the liquidity wave deeper into 2027.

Crypto would front-run this by at least 6–9 months.

5. How a Dovish Fed Chair Might Reshape the Crypto Peak Timing

Let’s compare the cycle timelines.

Scenario 1: Powell stays → Normal liquidity cycle

• Liquidity bottom: now → Q1 2025

• Liquidity expansion: Q3 2025 – Q2 2026

• Crypto peak: March–June 2026

Scenario 2: Powell leaves → Dovish Chair arrives May 2026

• Liquidity expansion: restarts mid-2026

• Easier policy: persistent into 2027

• Global M2: rising for longer

• Rates: cut sooner and deeper

Crypto peak shifts to…

November 2026 and Beyond!

Best estimate: March 2027

This would produce:

• A longer bull market

• A higher terminal peak

• A broader institutional adoption phase

• A more pronounced blow-off top

It would also give you a perfect narrative for your Age of Autonomy thesis:

“Crypto is no longer a four-year asset.
Crypto is a liquidity asset.
And liquidity is about to be politically engineered.”

6. What the ISM + Fed Change Story Means for Investors

Here’s the takeaway your readers need:

1. ISM says “slow growth,” not recession.

This is where liquidity injections often begin.

2. Crypto no longer follows the halving.

It follows global M2 expansion.

3. The next Chair of the Fed could reshape the cycle.

A dovish appointment in mid-2026 would extend the bull run.

4. The ‘crypto peak’ may shift from early 2026 into 2027.

5. This could be the longest / strongest crypto bull market, ever.

This is the perfect moment to write about the transition from:

“Bitcoin halving logic” → “Liquidity cycle logic.”

Because the macro evidence is piling up—and the ISM data just added a new chapter.

Conclusion

So, to summarize the conclusion:

• The ISM is telling a late-cycle story.

• Sovereign debt is telling a liquidity story.

• Politics may soon tell a dovish Fed story.

• And crypto, as always, will move first.

….

If Powell steps down and a dovish Chair is installed in mid-2026, the liquidity wave that normally peaks in Q2 2026 could instead surge into 2027. That would shift the crypto-cycle top from a simple halving-based timeline into a politically driven liquidity superwave.

This is why the 4-year cycle is dead.
This is why the 5-year cycle is the real driver.
And this is why 2026–2027 could be the most important moment in crypto’s history.

Cheers,

Jake Ryan

CIO Tradecraft Capital

Author, Crypto Investing in the Age of Autonomy

Author, Crypto Decrypted

Buy My Book!