That is the whole argument for looking at the same numbers every week. One week of data makes odd things look like accidents. Several weeks make them look like habits.

In other news, ICP is minting less tokens than it was three months ago, and the fall is steady enough to call it a trend. But most of the missing issuance is just sitting in a pile, waiting to be claimed.

Executive Summary

• Cycle burn sits 19.5x below the Mission 70 target, up from 19.1x a week earlier;

• Node provider payouts have fallen every month since May, down about 24%;

• The network minted roughly 628,000 ICP and burned about 4,100;

• Measured inflation over the past year is 3.4%, while over the past quarter it runs at 2.2%;

• Unclaimed rewards sit at 95.7 million ICP, around 17% of all supply;

• Compute rose 22.8%, a second straight weekly jump, while cycle burn fell;

• Storage hit a 14-week low even as the network kept adding canisters;

• Trading volume rose 41% and cleared its four-week average for the first time since June;

• Six Saturdays in a row have carried mints above 25,000 ICP, against a normal Saturday near 9,700.

Read more

Somebody minted nearly 37,000 ICP on Saturday. That is more than half of everything minted all week. That figure landed nowhere near the monthly node provider payout, and I have no idea who did it or why.

That was not the only thing that moved. Compute execution jumped 14% after falling by half since May. Yesterday saw a 5% price move backed by a 62% jump in volume. And cycle burn crossed back above its 4-week average, which sounds like progress until you see the number it has to reach.

This is the first edition of a weekly report I will publish every Sunday. Everything in it is measured from official on-chain data. Nothing is estimated, and where I do not know something I will say so.

Executive Summary

• Cycle burn is 19x below the Mission 70 target. Up slightly on the week, down 13% over four weeks.

• The gap has not narrowed. It has moved between 17x and 28x over 12 weeks, and was closer to target in early July than it is now.

• Compute demand rose 14%, crossing above its 4-week average. It remains around half its May level.

• Roughly 12,000 ICP was burned against 71,000 minted, leaving supply up about 59,000 for the week. Burn covered 17% of minting, up from 9% last week.

• Measured inflation is 3.4% over the past year. This is not comparable to Mission 70’s headline figures, for reasons set out below.

• Nearly 37,000 ICP was minted on Saturday, the largest single day of the week and outside the monthly payout window. Unexplained.

• The XDR floor is binding, so node provider rewards are being minted at the floor rate rather than the market rate.

• Trading volume is down 36% over four weeks, but the final day saw volume jump 62% on a 5% price move.

• 157 proposals were adopted, almost all of them routine operations.

The Gap: 19x, No Trend

The Mission 70 whitepaper sets a demand-side target of 0.77 XDR per second of cycle burn. XDR is the IMF’s basket currency, and 1 trillion cycles (TC) is defined as 1 XDR. This week the network averaged 0.0399 XDR/s. That is a gap of 19.3x.

Burn rose 3% week on week and crossed back above its 4-week average, the first time it has done that in three weeks. Taken alone that reads as improvement.

It isn’t, and the 12-week series shows why. The gap has moved between 17x and 28x over that period. It sat at 17x in early July, meaningfully closer to target than today. Over the past four weeks burn is down 13%.

One correction worth making explicitly, since I published a different figure in this article. The gap was at 20.6x using a 90-day median. On that same basis today it reads 19.5x. That looks like the gap closed, but almost all of the move comes from the 90-day window rolling past late May, when weekly burn hit its 12-week low. As those readings drop out of the window, the median goes up on its own, without anything actually improving.

The honest summary is that cycle burn is bouncing around inside the same range it has been in for months, and going nowhere.

In money terms, the network consumed about 24,100 XDR of compute this week, which is roughly 15,000 ICP, or $33,000.

Note: Cycle burn is compute consumed by canisters. ICP burn is tokens destroyed when someone converts ICP into cycles. The two diverge day to day because users top up in batches, and converge over time.

Compute Reverses, But From a Long Way Down

The most interesting movement this week was in instruction execution, also known as compute demand. The rate rose 14% to 9.8 billion instructions per second and crossed above its 4-week average.

Context matters here. In late May this metric was running at 20 billion per second. It has halved over 12 weeks, bottoming around 7.6 billion in early July. This week’s jump is a bounce off that floor, not a return to prior levels.

What makes it worth watching is how far it has pulled apart from everything else. Canister count and identity users both rose steadily through the same period, while compute per canister fell hard. The network keeps adding canisters and accounts while running far fewer instructions inside them.

Supply: 71,000 Minted, 12,000 Burned

Minting ran about six times ahead of burning this week. Supply rose about 59,000 ICP, to 555.3 million.

The burn side improved sharply in relative terms. 12,025 ICP burned is up 63% on the week and has more than doubled over four weeks, lifting burn to 17% of minting from 9% the week before. Minting fell 10%.

The daily pattern is where this gets interesting, and it illustrates why the two burn metrics diverge. Thursday and Friday saw about 4,400 and 4,200 ICP burned. Saturday saw 206. Compute consumption barely moved across those three days, holding between roughly 3,200 and 3,800 XDR. Nobody converted much ICP on Saturday, but the network ran regardless.

The mint side has an outlier. 36,744 ICP was minted on Saturday, the largest single day of the week and more than half its total. Monthly node provider payouts land between the 13th and 16th, so this was not that. I do not know what it was. It is flagged and I will watch whether it recurs.

Inflation: 3.4% Measured, Why It Is Not the Whitepaper Number

ICP’s supply grew 3.4% over the past year. That is the real number, measured. It is not the number in the Mission 70 whitepaper, and it’s worth telling you why it differs.

The measurement first. Total supply went from 536.3 million ICP a year ago to 555.3 million now, a net increase of about 19 million ICP. That is measured straight from the ledger. Nothing is scaled up or estimated.

Shorter windows read lower: 2.4% over 13 weeks, 2.2% over four weeks. Both are shorter periods stretched out to a full-year figure, and both sit below the annual rate. Recent weeks have minted less than the past year averaged.

This figure cannot be compared to Mission 70’s targets, and I want to be direct about why. The whitepaper’s numbers are 9.72% for January 2026, 5.42% for January 2027 and 2.92% as the target. But the paper states plainly that these “refer to gross reward issuance under full maturity disbursement; realized inflation is lower in practice.” In plain terms: their figure assumes every reward ever earned gets converted into ICP. In reality most of it sits unconverted as maturity. Voting rewards are quoted “including undisbursed and disbursed maturity.”

That is DFINITY’s own forecast, calculated from how rewards are meant to work across every neuron. There is no way to rebuild that number from on-chain data. Setting 3.4% against 5.42% and concluding ICP is running ahead of schedule would be misleading, and it is a mistake I have seen made repeatedly.

The maturity pool is part of the explanation. Maturity is what voting rewards accumulate as before anyone converts them into actual ICP. The pool stands at 95.5 million ICP, up about 150,000 this week: rewards earned but not yet minted. That is roughly 17% of total supply sitting as a claim outside the ledger.

The valid Mission 70 scoreboard is the burn rate against 0.77 XDR/s, because that target is measured the same way in both places. It is 19x short.

Staking: Flat, Slow Drift Toward the Exit

Total staked rose slightly, to 242 million ICP. Nothing dramatic.

The mix shifted slightly too. Dissolving stake rose 0.5% to 26.9 million ICP, while non-dissolving was flat. The dissolving share now sits at 11.1%, up slightly over four weeks.

That is a small drift and one week does not make a trend, but it is the fourth consecutive week the dissolving share has risen. Worth tracking rather than concluding anything too soon.

Network Usage: Growing, Slowly

Canisters, identities and storage all moved by fractions of a percent.

Running canisters rose 500 to 1.16 million. Internet Identity users added 5,600 to 3.21 million. Storage fell slightly to 11.6 TB, and is essentially flat over four weeks.

The picture across 12 weeks is steady growth in canisters and identities, and storage sitting below where it was in May. Combined with the instruction rate collapse, the network is adding canisters faster than it is adding work for them to do.

Market: A Quiet Week That Ended Loudly

ICP closed the 7 day interval we measure at $2.20, up 9% week on week but still down 2% over four weeks.

Weekly trading volume was $186M, down 2% on the week and down 36% over four weeks. The 12-week trend is a steep decline from roughly $880M in early June.

The last day breaks the pattern though. Saturday saw price rise 5% on volume of $41M against $25M the day before, a 62% jump. Whether that carries into next week is next week’s question, but the yesterday ended on considerably more activity than the rest of the week ran on.

On-chain transfer volume tells a different story from exchange volume: 8.9M ICP moved, down 29% on the week, while the transfer count rose 14% to about 77,000. More transactions, smaller sizes. Main driver was exchange activity, perhaps derivatives.

The ICP/XDR floor set by proposal 142681 remains binding. The 30-day average sits at 1.58 against the 2.0 floor, meaning node provider rewards continue to be minted at the floor rate rather than the market rate.

Governance

157 proposals adopted, which sounds like a lot until you see the breakdown. Almost all were routine operational traffic: replica version deployments and subnet management. Fewer than ten covered anything else.

Internet Identity backend and frontend canisters were both upgraded on 3 August.

Currently open: an Internet Identity backend upgrade and a Subnet Rental Canister upgrade to version 0.7.0, both decided 11 August.

What I Am Watching Next Week

The Saturday mint. Nearly 37,000 ICP appeared in one day, outside the payout window, with no explanation I can find. If it recurs next week it is a pattern worth naming. If it does not, it was a one-off worth understanding anyway. If you know what it was, tell me in the comments or on Telegram. I would rather be corrected than guess.

Whether the volume spike holds. The week ended on 62% more volume than the day before.

The dissolving share. It has risen four weeks running. Still small, and well inside its normal range, but four weeks in one direction is the point at which I start paying attention.

This report goes out every Sunday. This first edition is free so you can decide whether it's worth paying for. Starting next week it goes to paid subscribers, at $5 a month. You can also find me on Telegram.

Quantum computing has become crypto’s loudest talking point, and almost nobody is explaining it in simple terms. Two things happened in the last few weeks, and DFINITY’s position has shifted throughout the years in a way worth understanding.

What quantum computers actually threaten

When you hold ICP, BTC, ETH, or any other token, what you really own is a private key. Your wallet proves you own your coins by signing with it, and the network checks that signature without ever seeing the private key itself. That works because deriving a private key from a public one requires solving a math problem called the “discrete logarithm”. On normal computers, it would take longer than the universe has existed.

A quantum computer solves it quickly, using Shor’s algorithm, published in 1994.

So the risk is not that someone “picks the lock”, but that they can work out your key from information that is already public, then sign as you. Your coins move and the network sees a perfectly valid transaction.

This applies to all of crypto. Bitcoin, Ethereum, and ICP, all rely on the same family of maths securing the private keys.

Where ICP stands, in DFINITY’s own words

ICP is NOT quantum-safe today, and DFINITY has never pretended it was.

In December 2021 the foundation put a research proposal to the NNS titled “Long Term R&D: PQ security”.

“Quantum algorithms can break some of the cryptographic assumptions the Internet Computer relies on, specifically the hardness of computing discrete logarithms. A quantum-capable adversary could thus forge signatures used in the IC.”

The proposal argued for moving early, noting that attackers can already harvest encrypted data today, and hold it for decryption later once quantum computers exist. It named lattice-based cryptography as the most likely replacement. Asked in the same thread when he would want ICP hardened, Jens Groth, who led the proposal, said his personal target would be around 2025. He added that post-quantum work was not where the foundation was investing much at that moment.

That was their 2021 position: start early, and aim for the middle of the decade.

The plan for replacing it just took two hits

In July, Anthropic’s Mythos model was pointed at HAWK, a post-quantum signature scheme that had passed two rounds of NIST review over two years of expert scrutiny. In roughly 60 hours, Mythos found a hidden mathematical structure nobody had noticed. A post-quantum specialist at Google summed it up as: “Basically with this paper, HAWK is dead”.

To be fair to everyone involved, Anthropic was careful to say the result is specific to HAWK and does not extend to other candidates or to lattice cryptography generally. But it does show that schemes surviving years of human review can fall in days once something new starts looking.

Then a bigger claim appeared. At the end of July, Daniel Simon, whose 1994 work directly inspired Shor’s algorithm, posted a draft paper describing a fast quantum method for a problem that, if it works, would undermine the maths beneath most lattice-based schemes. The family that NIST standardised and that DFINITY named as its likely path back in 2021.

Three mentions, all of them important. The paper says “Preliminary Draft” on the front. Several key proofs are marked as sketches rather than complete. And this has happened before: in April 2024 a similar preprint caused a nine-day panic before two researchers found a flaw the author could not fix.

Bjorn Tackmann from DFINITY, who flagged the paper publicly, put it as honestly as anyone can right now: “it could go either way”.

Why DFINITY is now saying slow down

Dominic Williams has been arguing that quantum has become crypto’s Y2K: a narrative that lets projects market themselves as advanced without much engineering behind it.

His technical point is the one to actually take away. The cryptography used today has been attacked by researchers for 40 years without breaking. Post-quantum schemes have not. Switching now means trading a risk we understand well for one we barely understand at all, in exchange for protection against a machine that does not exist yet. He argues the urgent threat is AI finding holes in ordinary code today, not quantum computers arriving in the 2030s.

The HAWK result supports him, which is I assume why he cited it.

It is worth noting the position has moved. In 2021 DFINITY told the NNS it made sense to start early. In 2026 the message is that rushing is the greater danger. Both can be reasonable given what has happened since, and the recent evidence genuinely argues for caution.

What I would actually take from all this

The threat is real, but it’s not close. Nobody’s ICP is at risk from a quantum computer for the time being.

Be sceptical of anyone selling “quantum safe” as a completed feature. It is a bet on schemes that have not been tested for long. It is still a bet, and a project marketing it as a solved problem is telling you something about its marketing, not its cryptography.

ICP is not currently quantum-safe. Neither is Bitcoin, nor Ethereum. Everyone will have to move eventually, and the honest disagreement is only about when and to what.

If you found this helpful, please consider becoming a subscriber!

Subscribe now

I made a post questioning whether CaffeineAI serves a real purpose, or whether it's just an app chasing a narrative. So instead of guessing, I decided to pull every Caffeine-related canister I could find on-chain and inspect what's actually inside them, to see whether people are building real things or just testing it once and shelving it.

Caffeine is DFINITY’s “build an app by describing it” tool. You type what you want, it deploys a working web app onto the Internet Computer. My assumption, which I shared about a week ago, was that this was a “toy”: a wave of launch-day demos that nobody actually kept. A neat trick that produced throwaways.

So I checked, because that’s the whole point of this Substack. Here is what I found, and where I had it wrong.

Every Caffeine app is traceable, and there are a lot of them

Caffeine apps share one on-chain fingerprint: every canister it deploys carries the same controller. That controller currently sits on 268,100 canisters, about 22% of everything on the Internet Computer. Strip out the empty and duplicate ones and you get roughly 86,000 deployed app front-ends.

That number alone doesn’t prove anything. Because that could be a big pile of dead pages. So I pulled a random sample of 1,500 of those apps and actually probed them.

They’re not dead

Of the 1,500 apps I checked, 99.7% were serving a live page. Only 1.1% were empty placeholders. I went a step further and pulled the JavaScript bundle for 300 of them to make sure they were real deployments and not blank shells. All 300 of them loaded. These are real, complete apps. Are they complex applications? No, they are simple websites for the most part.

That already killed my “abandoned demos” assumption. But the thing that actually changed my mind was reading the titles.

There are a lot of small businesses

Scattered through the sample were things nobody builds to test a toy:

• A diagnostic and orthopaedic clinic in Bhilai, India

• A pharmacy in Perundurai

• A pizzeria in Gorakhpur

• A textile export B2B tool

• A Polish FPV drone-event platform

• A Persian coffee shop

Nobody generates a website for a real orthopaedic clinic in Bhilai to “check the tyres” on an AI app builder. Somebody wanted a working site for an actual business, and Caffeine gave them one. Under a conservative count, about 42% of the live apps are recognizable business tools or local-business sites. The real figure might be higher, because a chunk of what my classifier couldn’t label turned out, on manual inspection, to be more of exactly this: real apps with local or non-English names.

I was judging Caffeine as a developer tool, against things like Claude Code, and by that standard it looks thin. But developers aren’t the users. The users are the pizzeria in Gorakhpur and the pharmacy in Perundurai. Caffeine isn’t competing with developer tools. It’s competing with WordPress, Wix, and Squarespace, the “I just need a website for my business” market. Judged against that, a globally distributed base of small-business sites deployed by non-technical people is exactly the point, not a disappointment.

Where I’m keeping my skepticism

Being wrong about Caffeine being a flop doesn’t mean flipping to the hype. Three things keep honest:

1. Serving is not the same as being used. A canister keeps answering as long as it holds cycles, and Caffeine appears to pay for that hosting. So 99.7% “live” measures Caffeine’s willingness to fund hosting, not whether anyone visits these sites. It’s the cheapest possible sign of life. I have no traffic data, visitor counts, or retention. A website can be live and have zero users. And unfortunately there is no way for me to check against that.

2. “86,000 apps” is inflated. Titles repeat heavily, and one app appears roughly 2,200 times across the network. About 19% of named apps share their name with another, which is the signature of redeploys or repeated tutorial prompts, not 86,000 distinct ideas. The honest distinct-app count is closer to 70,000, and that’s still an impressive number.

3. The junk is real too. Right next to the clinic and the pizzeria are Valentine’s-day novelty pages, a dead memory game, an “Amazon Clone,” and about 16% of apps that never even got a title.

So what’s the verdict

The honest read is roughly 70,000 distinct, near-universally hosted apps, with a real and geographically diverse small-business core mixed with a big layer of experiments and duplicates.

The open question I still can’t answer is whether that base is growing or stalling, because this is a single snapshot. That needs tracking over time, and it’s what I’m considering for a future article.

I’m on record saying “Caffeine has no reason to exist”, but the chain says otherwise, and I’m happy to take that back.

If you found value in this article, please consider subscribing!

Subscribe now

The core mechanics behind ICP’s cloud engines, the product DFINITY is betting its 2026 tokenomics on, were filed as patents in 2019 and 2020. The patents describe the technology as “routine load balancing”. They never once use the words enterprise, sovereign, or jurisdiction, which are canon terms for cloud engines today.

The filings are public, so I sat down and went over the 21 patents I could find. The full documents rather than just the summaries. I went in expecting one of two things: either a wall of protection around the network, or nothing much at all. What I found was more unexpected than either of those.

Here is what the patents actually show:

The cloud-engine machinery was filed years before the product

A cloud engine, in DFINITY’s 2026 pitch, is a private slice of the network that an enterprise rents and configures: pick your nodes, pick your countries, keep the same application running on top. The whole thing depends on one technical trick. The identity of a subnet has to be independent of the specific machines running it, so you can swap the hardware underneath without breaking what is deployed.

That trick is US patent 12,388,628, filed June 2020.

“each of the plurality of subnets has an individual, i.e. its own static verification key which remains constant over time... While the static verification key remains constant, the corresponding secret key shares may change over time. Furthermore, the set of nodes which form or represent a subnet may change over time.”

A subnet keeps one fixed public identity while the machines behind it come and go. That is the exact property a rent-your-own-infrastructure product needs, described in full, in mid-2020.

Its companion is a patent filed the same day, WO 2022/002427. It covers moving a running program, with its live state intact, onto an entirely different set of machines:

“These additional nodes are fresh nodes which may catch up with the states of the migrant computational unit... the migrant computational unit has been completely migrated from the nodes of the first subnet to another new set of nodes.”

The same patent even handles the security problem that arises when you move a workload between machines run by different operators. It calls this crossing "two different trust domains" and certifies the transfer with a joint signature so the destination can trust what it receives.

So by September 2020, roughly five years before cloud engines were announced, DFINITY had patented a subnet whose identity survives a total change of hardware, and a way to move a live workload onto fresh nodes across a trust boundary. If you were designing the plumbing for “choose your own node providers, keep your app running,” this is what you would file.

The economics were in there too, from 2019

The cloud-engine business model routes through associations of node providers who sell capacity and keep the revenue, minus a cut the network burns. The idea that you could rent out your infrastructure to a paying customer and share the proceeds, rather than run it yourself, is the commercial heart of the pitch.

That idea appears, as an aside, in US patent 12,063,308, filed March 2019. The patent is about node privacy. Buried in the description is this:

“such a separation of the permanent node identities from the stake identities allows stake holders to ‘rent out’ their stake e.g. to a data center... This allows the stake owner and the data center to share rewards for the stake, without having to hand over the stake.”

Rent out your position, share the rewards, keep control. That is the cloud-engine revenue concept in embryo, written down in 2019 as a throwaway line in a cryptography patent.

The patents do not say any of this was built for cloud engines. Read them cold and the stated purpose is always internal. The migration patent moves workloads "for load balancing reasons," and migration is triggered by the network itself, not by a customer. The vocabulary of the actual product is completely absent. Across all patents, the words enterprise, sovereign, jurisdiction, geography, tenant, and compliance appear zero times. The word cloud appears eight times, every one of them boilerplate from a generic computer-systems template.

DFINITY built and patented the core mechanics of cloud engines years before they productized them, and described those mechanics as ordinary network housekeeping. Whether the foresight was deliberate or the pieces simply turned out to fit, the documents do not say. But the pieces were there, and that’s very interesting, regardless.

The reverse-gas model, also 2019

One more piece of the current network shows up early. ICP’s headline economic feature is that developers can pay for computation so users do not have to, the so-called reverse-gas model. US patent 12,572,963, filed October 2019, describes a two-currency system where "gas" is:

“a complementary accounting unit that is used by the distributed network to charge for the use of its resources and is separate from the native currency of the network.”

Stable-priced, burned on use, with the conversion rate set by governance. That is the cycles model, roughly 19 months before mainnet.

The reverse-gas part is more nuanced than the marketing suggests. The patent’s default is actually that users hold gas accounts. But it then makes the payer a choice the developer declares in code:

“the respective payment method for the respective application may be specified in an interface description language (IDL)... and can hence be specified by the respective programmer/developer.”

So “the app pays, the user holds nothing” is one option inside the 2019 patent, not the headline design. If you have read that DFINITY patented reverse gas in 2019, that overstates it. What they patented was a metering system flexible enough to let the developer decide who pays, including the case that later became the selling point.

What is NOT patented is the bigger story

Here is the thing that changes how I read the whole patent portfolio.

Their patents cover the hard cryptographic and consensus core: distributed key generation, threshold signatures, the consensus protocol, the random beacon, secret resharing. Seven of the 21 patents are threshold cryptography alone. This is the genuinely difficult mathematics, and it is protected.

But almost nothing DFINITY has shipped since 2021 appears anywhere. There is no patent family directed at the NNS or on-chain governance. None for chain-key Bitcoin or Ethereum. None for HTTPS outcalls, the Wasm execution environment, Motoko, or the boundary nodes. None for the AI and self-writing-cloud direction the company now talks about constantly.

And the filing itself stopped. Measured by priority date, which is when an invention is actually claimed, the program runs from 2016 to September 2021 and then halts. One lone family has been filed since, in April 2024. Patents kept publishing and granting well into 2026, but that is the pipeline draining. The last time DFINITY filed a wave of patents was over four years ago.

The obvious explanation is that the R&D was simply done, so there was nothing left to patent. I can rule that out from the documents themselves.

They kept inventing. They just stopped filing.

DFINITY’s cryptographers published at least four substantial new results from 2023 onward. vetKeys, a way for a blockchain to hold secrets, in April 2023. A new threshold Schnorr toolkit. Lightweight asynchronous verifiable secret sharing. Fast batched asynchronous distributed key generation. vetKeys in particular rests on identity-based encryption, a primitive that does not appear anywhere in the entire patent portfolio.

I checked that directly. Across all patents, “identity-based encryption,” “IBE,” and “vetKD” appear zero times. The research was novel, but none of it was patented.

Before September 2021, DFINITY’s major cryptography papers each had a matching patent, usually filed within weeks. After 2023, the papers kept coming and the patents did not.

There is a second piece of evidence for the open posture, and it draws the same line the patents do. DFINITY’s code is public, but not under one license. Most of it is Apache 2.0, a permissive open-source license. A minority sits under a restrictive source-available license that the project’s own README says exists “to protect the Intellectual Property of the DFINITY Foundation”. The restrictive license clusters on the consensus protocol, the certification and state machinery, and the threshold-signature cryptography. The permissive license covers the product surface: the NNS, chain-key Bitcoin and Ethereum, HTTPS outcalls, Motoko. That is the same boundary the patents draw. Protect the hard cryptographic core, give away the product layer, in two separate instruments that happen to agree exactly.

vetKeys sits on the giveaway side, and that is the tell. Its implementation is public under the permissive license, and the source file cites the 2023 paper directly. It sits one directory away from threshold-signature code that is under the restrictive license. They had the option to protect it, and they chose not to.

That leaves a real question I cannot answer from patents alone. The most likely reading is a deliberate shift toward publishing over patenting, which fits a project that always published its research anyway. It could also be budget, or a change in legal strategy, or something else entirely, I can’t know for sure.

So what does the portfolio actually reveal?

DFINITY patented the part of the Internet Computer that is genuinely hard to reproduce, the threshold cryptography and consensus core, and left the rest to be protected by difficulty and by being given away in the open. The product surface, the governance, the cross-chain integration, the developer tools, they all sit behind no patents at all. And the mechanics of cloud engines, the thing the whole 2026 economic plan depends on, were quietly filed years early and labeled as routine.

If you are trying to judge whether ICP’s technology is defensible, the patents give a genuinely mixed answer. The products are not, and the company appears to have decided, sometime around 2021, that patenting them was not worth doing. Whether that is confidence, principle, or retreat is the one thing 21 patents cannot tell you.

What they can tell you is that the cloud-engine bet is not a pivot. The pieces have been sitting in DFINITY’s filing cabinet since 2019.

EDIT: Bjorn Tackmann from DFINITY replied to say they have not stopped filing, but the process simply takes a long time, and that there are "still some real gems waiting to be discovered." That is a fair correction. I measured priority dates, which is when an invention is actually claimed, and those only become visible when a filing publishes roughly 18 months later. So anything filed in the last year or so would not appear in my data no matter what it contained. What happens next is genuinely not visible yet, and by his account there is more coming.

Every quote above is word for word from the patent filings, which are public. I have the full text of all 21 families and I am glad to point anyone to the specific document. If this was useful, please consider subscribing.

I spent a day pulling ICP’s entire on-chain history from DFINITY’s public APIs. Burn rate back to Genesis, minting, staking behaviour, storage, canister counts, identity users. Every number in this article came from an endpoint you can query yourself. I’ve listed them at the bottom so you can check my work if you want. Data pulled 2 August 2026.

Here is what I found:

• Cycle burn is running at 0.037 XDR per second against the Mission 70 target of 0.77 XDR. That is a ~20x gap, and burn is lower today than it was a year ago.

• The target has been exceeded exactly twice, in January and February 2025, during a burn spike that then decayed to a fraction of its peak.

• Canister count and identity users grew straight through that bubble and kept climbing after it popped. Storage grew too, though it has slipped from an April peak.

• Mission 70 caused no staker exodus. Total staked ICP actually rose. The “8 Year Gang” was re-bucketed to the new cap rather than leaving.

• 95.4 million ICP of maturity is earned but unminted. That is 17.2% of total supply, and it is part of why measured inflation looks lower than it is.

• 99.85% of ICP burned is real cycle conversion, not transaction fees. The burn is small, but it is genuine demand.

Two of those findings are bearish, two are bullish, and two of them go both ways. That is roughly how the data came out, and I’m not going to pretend otherwise in either direction.

The gap is 20x, and it is DFINITY’s own number

Mission 70 is DFINITY’s plan to cut ICP inflation by 70% during 2026. The supply side does most of the work: sliced voting rewards, and reduced node provider rewards, taking annual inflation from 9.72% to 5.42% by January 2027. That part passed the NNS back in April and is already happening.

The demand side is the other half, and it needs the network’s cycle burn rate to reach 0.77 XDR per second. XDR is the IMF’s basket of currencies. 1 trillion cycles is defined as 1 XDR, worth about $1.36 today.

I measured the current rate at 0.0374 XDR per second, taking the median of the last 90 days. The most recent single sample was 0.0361 XDR per second.

That is 20.6x below the target.

In dollar terms, the whole network’s usage revenue is about $4,391 a day, or $1.60 million a year. Mission 70 requires roughly $90,511 a day, or $33.04 million a year.

*Worth noting: DFINITY’s own whitepaper stated the starting rate as 0.05 XDR per second, which produces the 15x figure you might see quoted. The measured rate is lower than that, so their own number understates the task.

The only time ICP hit the target, it was a bubble

Here is the part that might surprise some of you who are newer to ICP.

The target has been reached twice. Monthly average burn hit 1.1344 XDR per second in January 2025 and 0.8758 XDR in February 2025. Both well above the required 0.77 XDR.

But then it collapsed. Today’s rate is only about 3% of that January 2025 average, and under 1% of the spike’s single highest reading.

Pulling the full series back to Genesis makes the shape obvious. From 2022 through 2023 the network averaged 0.0035 XDR per second. Through the first 8 months of 2024 it averaged 0.0110. Then in September 2024 it jumped roughly 48x in a single month, stayed elevated for 11 months, and fell back.

The spike coincided with the BOB frenzy, a token where people burned cycles for a chance to mine it. I want to be precise here: I measured the spike and its decay, but I did not independently verify the cause. However, the BOB attribution is common ecosystem knowledge.

Either way, the mechanism matters for how you read it. That burn came from a token lottery. Mission 70’s burn is meant to come from enterprise cloud subscriptions. Those are different economic objects, and citing the first as evidence for the second doesn’t hold up.

What it does prove however, is that the network can physically process that throughput. The infrastructure handled 1.13 XDR per second without breaking. That is a real finding and it is an important one.

Adoption kept growing after the bubble popped

This is the section that changed my mind, and it’s the strongest bullish evidence in the whole data pull.

If the 2024-25 spike was the only thing holding ICP up, everything else should have fallen with it. But it didn’t.

Storage matters more than the others, and here’s why. Canisters pay a continuous rental fee for the memory they hold. Stored bytes are therefore contractual recurring burn. You cannot fake it with a token lottery, because the lottery doesn’t store anything.

Storage grew 49% during the spike and another 42% in the year after it. Though it peaked at 13.26 TB in April 2026 and has since slipped about 12% to 11.65 TB. Canister count did the same. Identity users are the sharpest case: they barely moved during the spike, up just 1.6%, then grew 21.5% after it, so that growth is not the bubble’s echo. Node operators added 45% more machines over three years and 46% more CPU cores. None of that follows a speculative burn that has since evaporated.

The counterweight, because there is one: instruction rate is at 4% of its peak, and 73% below where it sat before the spike even started. Actual compute executed has fallen hard while stored bytes rose. ICP is accumulating storage customers considerably better than compute customers, and storage is the cheaper of the two per unit of burn.

Also worth knowing: total network memory capacity is about 1,580.00 TB against 11.65 TB used. That’s 0.74% utilisation. Nothing here is capacity constrained.

Mission 70 didn’t break the stakers

The loudest objection to Mission 70 was that capping maximum dissolve delay from 8 years to 2 would break faith with long-term holders and trigger an exit. It’s worth checking whether that happened, because it’s the kind of prediction people make loudly and then never revisit.

It didn’t happen.

Total staked ICP rose 1.04%. Dissolving stake went up 19% in absolute terms, but the share of stake that is dissolving sits at 11.07%, which is still below its own January 2025 level of 13.22%.

The structural fingerprint is clearer than the headline numbers. Every dissolve-delay bucket above the new cap terminates on the same date, 19 April 2026. The bucket that held over 156 million ICP at 96+ months simply stops. The 24-month bucket, which is where a capped neuron now lives, reads over 183 million ICP today and is still growing.

The long-lock stake was re-bucketed to the new 2 year cap, not withdrawn.

One thing to not misread: total potential voting power jumped 67% in a single three-day step on 22 April, and deciding voting power jumped 65% on the same date. Every adjacent step is under 1%. That is the voting-power formula being re-parameterised, not stakers doing anything.

The catch: 95 million ICP is earned but unminted

Here’s where the picture gets less flattering, and it’s the finding I’d most want a perma-bull to sit with.

Measured minting over the last quarter annualises to about 2.5% of supply. Over the full trailing year it’s 3.59%. Either figure is below Mission 70’s stated supply-side target of 5.42%, which sounds like the plan is running ahead of schedule.

It isn’t, and part of the reason is maturity.

Voting rewards stack as maturity and only become minted ICP when a holder disburses them. The ledger records the disbursement, not the stacking itself. So the low measured minting rate is partly delayed issuance, not avoided issuance.

The maturity pool currently holds 95.4 million ICP-equivalent. That is 17.2% of total supply, sitting as a claim that hasn’t hit the ledger yet. It grew by 4.94 million ICP over the trailing 12 months, which is 0.89% of supply per year.

Does that explain the gap? Only about a third of it.

The gap between measured minting (2.5%) and the accrual target (5.42%) is 2.92 percentage points, or roughly 16.2 million ICP a year. Maturity growth covers 4.94 million of that, about 30%. On the longer 364-day basis the gap is narrower and maturity covers closer to half.

I can’t close the rest with this data, and I’m not going to invent an explanation for it. What I can say is that anyone quoting ICP’s low measured inflation as a bullish datapoint, is quoting a number with a 95 million ICP asterisk attached.

The burn that exists is real

One objection I expected to find support for, and didn’t.

A common criticism of low-burn networks is that most of what gets burned is transaction fee dust, and other “gotchas” rather than genuine demand for compute. On ICP that turns out to be false.

Splitting the burn by type across 364 days:

Essentially all of it is someone actually buying compute. The burn is small, but it is not noise.

One thing worth noting: mean daily burn is 2,055 ICP against a median of 741. That’s a 2.8x gap, which means burn is lumpy. It’s driven by periodic large conversions rather than a steady retail drip.

While we’re here: about “deflationary day” posts

A quick note, because you’ll see these. I’ve made this exact post myself in the past, which is part of why I wanted to actually check the data.

Over 364 days, 16 days were net deflationary. That’s 4.4%. So a post announcing that ICP burned more than it minted on some specific day is describing a real but roughly 1-in-23 event.

The reason is the payout calendar. Mean daily minting is 54,544 ICP against a median of 10,179. Large node provider payouts lands between the 13th and 16th of every single month. 78 days out of 363 account for 86% of all minting. Those are the days where minting ran more than 3x the median.

On the other 285 days, minting is low and a deflationary day is much easier to hit. It tells you about the payout schedule, not about burn strength.

The number that actually matters is simpler: burn offsets 3.8% of minting. Net annualised inflation is +3.45%. For the network to break even against current minting, minting would need to fall 96% or burn would need to rise 26.5x. That is a higher bar than the 20x gap to the Mission 70 target, because it must offset all issuance, not just hit the burn number.

What I’d watch instead of the price

Three things will tell you whether this is working, and all three are checkable.

The cycle burn rate on ICP Index. Is it genuinely climbing toward 0.77 XDR per second, or holding flat around 0.037? Over the last twelve months it has been flat to slightly down. If that changes, it changes everything else in this article.

Whether DFINITY ever publishes a cloud engine customer count. Cloud engines are the product Mission 70’s demand side is built on, and they are not generally available yet. A company winning enterprise deals eventually wants to talk about them. Too much silence past general availability would be its own answer.

Storage growth. It is the only demand metric here that is contractual, recurring, and immune to speculation. Storage compounded for two years before slipping from its April 2026 peak; whether it resumes climbing is the tell. If it does while burn stays flat, that says something specific: ICP is winning cheap storage customers faster than expensive compute ones.

My verdict

The technology claims survive head on contact with the data. The commercial ones don’t, yet.

Everything Mission 70 depends on is built and demonstrably working. The burn mechanism functions, supply-side cuts passed and are landing, stakers didn’t flee. Storage, canisters and users have compounded for three years straight, entirely independently of the speculative spike that made ICP’s burn chart look good for eleven months in 2024 and 2025.

What’s missing is customers. The gap isn’t 20% or 200%. It’s 2000%, against a burn rate that has fallen over the past year while the target stayed fixed. Five months left in the year the target was set for, on a product that hasn’t reached general availability.

My honest prediction is that Mission 70’s demand-side target will not be hit in 2026. I’d be glad to be wrong. Nothing in this article is an argument against ICP. It’s an argument against taking the target on pure faith, which is a different thing.

The adoption substrate is real though and compounding in a way I did not expect before I pulled this data. If cloud engines ship and sell into the European sovereignty demand that is currently being written into law, the burn side has a path. If they don’t, ICP is just a well-built network with 0.74% utilisation and a token that inflates at 3.45% a year.

If this was useful, subscribing is free and it helps more than you’d think.

Data Sources

1. metrics-api.internetcomputer.org,

2. ledger-api.internetcomputer.org,

3. ic-api.internetcomputer.org.

It all started back in 2022 when I got my hands on a book called The Fourth Industrial Revolution, written by Klaus Schwab, who back then was the Chairman of the World Economic Forum. You are probably familiar with this guy, and if not, he is the one who coined the phrase: “You will own nothing, and you will be happy”. I think most people misinterpreted his words and took them out of context, but nonetheless, I am not here to debate whether Klaus Schwab, his work, or if the World Economic Forum is either positive or negative. I am here to talk about his vision, and how the Internet Computer or a similar technology is poised to serve as the base layer for the future it promises. The connections I managed to find are staggering.

So without further ado, please leave your emotions aside for this one, and let’s get right into it.

Buy the book here: The Fourth Industrial Revolution

The Fourth Industrial Revolution

So back in 2016, Klaus Schwab published a book that made a lot of people feel good about themselves, like they understood the future.

The Fourth Industrial Revolution laid out a vision of the world to come: one where autonomous machines negotiate with infrastructure, where your genome powers personalized medicine without ever leaving your control, where votes are recorded on distributed ledgers that nobody can tamper with, where smart contracts execute automatically the moment specific conditions are met. A world where the fusion of physical, digital, and biological technologies transforms not just what we do, but who we are as humans.

The book sold millions of copies. It became required reading in boardrooms, government ministries, and business schools. Davos built entire annual themes around it. The phrase “Fourth Industrial Revolution” entered the standard vocabulary of anyone who wanted to sound like they were thinking seriously about the future.

And then the world handed the infrastructure to Amazon, Google, and Microsoft. Calling this a flop would be a massive understatement.

As of right now, those three companies control roughly 63% of all global cloud infrastructure spending, a market that crossed $400 billion in 2025 and is growing at 25% per year, driven almost entirely by the AI boom Schwab was already anticipating when he wrote the book, so probably long before 2016 when it was published. Every application, dataset, AI model, every connected device that needs to compute or store something almost certainly runs through one of these three corporate “landlords”.

Schwab described a revolution, but what the world got was a more efficient version of the existing power structure, running faster on better hardware, owned by even fewer people than before.

Klaus Schwab’s Vision

It is worth going back to the source, because the book is way more specific than you might think.

Schwab did not just describe self-driving cars and wearable technology in broad terms. He laid out specific transformations with great details across three clusters he called Physical, Digital, and Biological, and then spent his final chapters on Governance and the social questions that would determine whether those transformations benefited humanity broadly or concentrated power in fewer hands.

The Digital Cluster

He wrote directly about blockchain as a “distributed ledger” enabling people who do not know each other to collaborate without going through a neutral central authority. He listed the specific things he expected blockchain to eventually register:

1. Birth and Death Certificates;

2. Titles of Ownership;

3. Marriage Licenses;

4. Educational Degrees;

5. Insurance Claims;

6. Medical Procedures;

7. Voting Systems;

8. Etc.

He described smart contracts that self-execute, specifically mentioning a credit derivative that pays out automatically the moment a country or company defaults, without any trader required to process the transaction.

The Biological Cluster

He described a world where your genome can be sequenced in hours for under $1000, where AI matches your specific tumor’s genetic profile to a targeted therapy, where personalized medicine becomes possible because we can process vast amounts of individual health data. He raised these critical questions: who owns that data, who controls access to it, and what happens when insurance companies or employers get their hands on it.

The Physical Cluster

He spoke about machines that do not just operate autonomously but negotiate autonomously: vehicles routing themselves through infrastructure, drones delivering medical supplies in conflict zones, robots accessing shared knowledge through cloud networks and updating each other in real time.

Governance

And across all three clusters, he kept returning to a single underlying concern, which was the “platform effect”. He described how digitally driven organizations create networks that match buyers and sellers and enjoy increasing returns to scale, which produces a concentration of few but very powerful platforms dominating their markets. He argued that to prevent the concentration of value and power in just a few hands, society needed to find ways to ensure openness and opportunities for collaborative innovation.

Vision vs. Reality

Ten years after the book was published, let’s look at where Klaus Schwab’s specific predictions actually landed:

1. Blockchain as a registrar for birth certificates, land titles, and votes?

Largely still a pilot project or a proof of concept. The countries that tried it, Honduras for land and Estonia for identity, made meaningful progress but hit the same wall every time: the data still has to live somewhere, the computation still has to run somewhere, and that “somewhere” is almost always a centralized cloud server that a company controls, and is subject to United States laws. Blockchain became a ledger sitting on top of traditional infrastructure rather than replacing it.

2. Smart contracts that self-execute?

Real, but narrow. The financial DeFi ecosystem built genuinely autonomous contracts, but the applications, the front-ends, the oracles that feed real-world data into those contracts: almost all of it runs on AWS or similar services. The smart contract itself might be decentralized. Everything around it usually is not.

3. Personalized genomic medicine with patient-controlled data?

The sequencing got cheap, exactly as Schwab predicted. The data remained locked inside hospital systems, insurance databases, and research institutions with their own privacy concerns and institutional incentives. The patient still has no portable, sovereign record of their own biology.

4. Autonomous infrastructure that self-reports damage, manages maintenance contracts, negotiates with machines?

The sensors exist, and the deployments are real and growing. But the data those sensors generate flows into corporate platforms like AWS, Azure and Google Cloud, where it is processed, stored, and monetized by the platform operator.

So you can see that the technologies Schwab described actually arrived, the problem is that the ownership structure he was worried about arrived too.

Infrastructure Is Everything

This is the part that in my opinion, most coverage of The Fourth Industrial Revolution gets wrong, because it is easier to write about the applications than to write about the layer underneath them.

Think about what the internet actually is at the physical level: it’s a global network of interconnected computers following common protocols, with no single point of control. Anyone who follows the protocol can participate. The design is open, decentralized, and neutral.

Now think about what you actually use to access that internet. You connect through an internet service provider, you load applications hosted on AWS or Google Cloud, and you store your data on servers owned by Apple, Microsoft or Dropbox. The network itself is decentralized, but everything built on top of it is centralized.

Schwab identified this pattern and called it the “platform effect”. What he did not really name was the specific mechanism that makes it so hard to escape, which is that whoever runs the infrastructure makes the rules. The digital landlords decide what data gets stored, how it gets processed, who gets to access it, what applications are allowed to run, and what applications can be shut down if they become inconvenient.

This is why Dominic Williams, the founder of DFINITY, said that when most blockchains claim to be decentralized, what they actually mean is that the token and the core logic live on-chain, but the application itself runs on Amazon Web Services. The decentralization is real but only partial, as long as the infrastructure underneath still has a “landlord”.

The Internet Computer was built to solve exactly this problem. Not just to create another blockchain for financial transactions, but to build a computing platform that can host entire applications, front-end, back-end, data storage, AI inference, identity verification, all of this without any corporate server sitting underneath it. The Internet Computer is the attempt to build public infrastructure for the digital world: neutral, ownerless, governed by its community, and available to anyone who wants to build on it.

Three Clusters + Governance

The rest of this article series examines what the Fourth Industrial Revolution actually requires at the infrastructure level, to deliver on Schwab’s specific predictions, in a way that is both viable and fair to the whole world. Can this transition be done without a segment of the population influenced by different big industries opposing it? Of course not, because if you read about the Industrial Revolution of the 18-19th centuries, you will notice the same pattern that happens today. So nothing is new under the sun.

But nonetheless, I will take each of his three clusters he describes in the book, plus the governance discussion, and I will do an in depth analysis of why the Internet Computer is positioned to solve all the issues Klaus Schwab presents. One dedicated article for each.

The Physical Revolution requires machines that can transact with each other autonomously. Not just drive themselves, but pay for the road, negotiate the landing clearance, contract for maintenance, verify their own supply chains, and file their own compliance records, without any corporate intermediary taking a cut of every transaction or holding the power to shut them down. That requires a computing layer the machines can trust completely, because it is governed by math rather than by a company’s terms of service.

The Digital Revolution requires data infrastructure that treats humans as owners rather than products. Schwab described the “Internet of Things” as a world where everything generates data. He also described the risk that this data would flow entirely to platform operators who monetize it while the humans who created it receive nothing and control nothing. Building the Digital Revolution properly means rethinking who the data belongs to, who profits from processing it, and how AI gets trained without exposing the underlying private information of the people whose data makes it useful.

The Biological Revolution requires the most sensitive data humanity will ever produce to be handled with controls that no centralized institution can guarantee: your genome, neural activity, or personalized treatment history. The research data that could unlock cures for diseases that have resisted every previous approach. These applications require computation that happens inside verified secure environments, with results you can trust and access you can revoke. Trusted Execution Environments on the Internet Computer, what the ICP ecosystem calls TEE subnets, make this possible. Centralized hospital databases and corporate genomics platforms do not.

The Governance Challenge is about the question of who gets to decide how these technologies work, who benefits from them, and who gets excluded. The ICP answer to this is the Network Nervous System, a fully on-chain governance system where token holders vote on protocol upgrades, subnet configurations, and system-level parameters, with every decision recorded publicly and no board of directors overriding the community.

Trust me, if you read the book, the connection to the Internet Computer is uncanny. It almost feels like they are interconnected at a “molecular” level.

The Timing

We are still very early in the “infrastructure build” for The Fourth Industrial Revolution.

Companies and governments are making infrastructure decisions today that will shape who controls this data for decades. Every hospital that chooses an AWS-hosted electronic health record system, every city that deploys Azure for its traffic infrastructure, every company that stores its supply chain data on Google Cloud is making a long-term commitment to a particular ownership structure.

Once that infrastructure is in place and the switching costs are high, it becomes very difficult to change. The window to build it differently is now, while the deployments are still early and the standards are still being set.

The good news is that a very strong narrative against centralized cloud services is gaining momentum right before our eyes. Switzerland, the European Union, and many other countries already spoke against Big Tech, and it seems like they are looking for neutral alternatives like the Internet Computer.

Pakistan’s recent partnership with the DFINITY Foundation, which involves an MOU to deploy a sovereign subnet, is a sign that governments understand this. They are not outsourcing their digital sovereignty to American hyperscalers, but building infrastructure they can control themselves.

The broader build-out is still very much in progress, and the choices being made right now will determine whether The Fourth Industrial Revolution produces the neutral and autonomous world described in the book, or simply a more efficient version of the existing power structure running on better chips.

What To Expect In This Series

This was only the introduction. I’m sorry if I’m being repetitive at this point, but I want to make sure you get at least half of the excitement I feel when writing about this whole thing.

Over the next four articles, we will go deep into each of Klaus Schwab’s clusters and trace exactly how ICP’s technical capabilities map to the specific transformations he described.

The second article covers the “Physical Revolution”: autonomous vehicles, drones, advanced robotics, 3D printing, and smart infrastructure. We will look at what it actually takes for machines to transact with each other without a corporate intermediary, and why the Internet Computer’s architecture is the only blockchain that can realistically support those interactions at scale.

The third article covers the “Digital Revolution”: the new internet, AI, platform economy, and what data ownership actually means when it is enforced by protocol rather than by a company’s privacy policy. We will look at how canister architecture enables individuals to own their data and negotiate directly with the companies that want to use it.

The fourth article covers the “Biological Revolution”: genomics, personalized medicine, synthetic biology, neurotechnology, and the infrastructure requirements for the most sensitive data ever generated. We will look at vetKeys, TEE subnets, and federated learning as the technical foundation for a world where your biological data works for you rather than for the institution that holds it.

The fifth and final article covers the “Governance” layer that Schwab identified as the make or break question for the entire revolution: who decides, who benefits, and who gets excluded. We will look at the Network Nervous System as the first serious attempt to build governance infrastructure that matches the scale and openness of the technology it governs.

Klaus Schwab was right about what was coming. Whether you like him or not, I think it’s our duty as human beings to keep an open mind, and strive to be critical thinkers.

If this sparked your interest and want to go straight to the source by yourself, feel free to order the book here: The Fourth Industrial Revolution.

If you made it this far, thank you for reading. And if you want more ICP articles like this one delivered straight to your inbox, consider subscribing to my Substack.

Talk to you later. BasedGiant.

Subscribe now

Now let me ask you something: if that is the market’s reaction to an AI tool that patches holes in existing infrastructure, what would the reaction be to a technology that makes those holes structurally impossible in the first place?

The technology I’m talking about is of course the Internet Computer Protocol, and the fact that this question is not being asked across the crypto market right now tells you everything about where we are in the ICP adoption cycle.

Current Internet Issues

To understand why ICP’s security model is genuinely different, you need to understand what the traditional approach to cybersecurity actually is.

The entire cybersecurity industry, the one whose stocks just cratered, exists because of a fundamental design flaw baked into how the internet was built. Applications run on centralized servers, and those servers are owned by companies. Those companies have administrators, and those administrators have credentials which can be stolen, compromised, or abused.

Every layer of protection that CrowdStrike, Okta, Zscaler, and the rest of the industry sells is essentially a patch on top of this flawed foundation. Endpoint detection and response tools watch for attackers who have already gotten in. Identity management platforms try to make sure the person logging in is actually who they claim to be. Firewall and network security products attempt to keep the wrong people out of places they shouldn’t be.

These are all necessary products because the underlying architecture invites attack. Essentially, it’s like putting a better lock on a door that was designed to be opened.

The result is an arms race with no end. Forescout’s 2025 Threat Roundup, which analyzed over 900 million attacks recorded worldwide last year, found that cyberattacks became more globally distributed and cloud-enabled throughout 2025, with activity traced across 214 countries and territories. More than 30,000 vulnerabilities were disclosed in 2025 alone, a 17% increase from the year before. The cost of recovering from a ransomware attack now averages $2.73 million. The industry keeps growing because the problem keeps growing, not because it is being solved.

ICP is not trying to win this arms race, instead it’s trying to end it by completely changing the architecture.

What “Tamper-Proof” Actually Means

This word gets thrown around a lot in crypto, so let me be precise about what it means in ICP’s context, because it goes further than most people realize.

Applications on the Internet Computer run inside smart contracts called canisters. These canisters execute on a network of independent node machines distributed across the globe. No single company owns these nodes, and no administrator has privileged access to your canister. There is no central server to compromise, database to breach, or backdoor to exploit.

Every state change in the network goes through Byzantine Fault Tolerant consensus, which means every modification has to be cryptographically validated across thousands of independent machines simultaneously. To tamper with an application running on ICP, you would need to corrupt a supermajority of independent node providers across multiple jurisdictions at exactly the same moment. In practical terms, that is virtually impossible.

The DFINITY Foundation’s official positioning is direct: apps hosted on the Internet Computer are tamper-proof and immune to traditional cyber attacks. The United Nations is already using ICP to create tamper-proof, verifiable credentials. DFINITY has partnered a while ago with FEDERITALY to bring supply chain transparency to Italian goods on-chain. Pakistan’s Digital Authority signed an MOU with DFINITY earlier this month to build a sovereign AI cloud infrastructure using ICP.

These are governments and international organizations making infrastructure bets on the premise that ICP’s security model holds up under real-world conditions.

These Sectors Would Feel It

Here is where the thought experiment gets interesting, and where I think the ICP community has not been loud enough in articulating the case.

The cybersecurity industry is worth roughly $200 to $300 billion annually and growing rapidly. But cybersecurity is not the only industry whose business model rests on the insecurity of centralized infrastructure.

Identity and Access Management is a massive category. Okta alone is worth tens of billions. The entire premise is that you need a trusted intermediary to verify who someone is and what they are allowed to access. ICP has Internet Identity, a blockchain authentication system that creates cryptographic aliases for users across services. It doesn’t use passwords, or central databases of credentials, and there is no single company sitting in the middle of every login that becomes a target. If ICP becomes the authentication layer for internet applications, the identity and access management market gets structurally disrupted at the foundation.

Cloud Infrastructure is another one. Z-Scaler’s business is built on securing the connection between users and cloud applications. Cloudflare makes its money protecting web infrastructure from attack. Both of these markets exist because applications run on vulnerable centralized servers that need protection from the open internet. If your application runs natively on a decentralized, tamper-proof network, the threat model that justifies these businesses changes dramatically.

Enterprise Software more broadly faces a different kind of pressure. When ICP’s Caffeine and other tools on the network start generating fully functional on-chain applications from natural language descriptions, the purpose of every traditional Software as a Service vendor starts looking thinner. This is the same “vibe coding” fear that has been hitting software stocks across the board, but with an additional layer: the apps being generated live on infrastructure that doesn’t need a separate security stack layered on top.

The point is not that ICP will immediately destroy these industries. The point is that mainstream ICP adoption would force a fundamental risk across all of them simultaneously. And this is precisely something that this companies don’t want to happen. Their business is patching vulnerabilities, so they need those vulnerabilities to exist in order to remain relevant and rake in billions a year. They have no interest in fixing the underlying issues.

The Crypto Market Acts Like ICP Doesn’t Exist

The ICP community has been asking for years why the project gets ignored relative to its technical achievement, and I think the honest answer has a few parts.

Firstof all, the narrative is hard to sell quickly. “We are replacing the foundational architecture of the internet” is not a 30-second pitch. DeFi yields, meme coins, and NFT collections are easier to understand and faster to speculate on. ICP is building something that takes years to appreciate and even longer to reach the mainstream adoption that unlocks its full value proposition. That is genuinely difficult for a market that prices momentum.

Second, the token launched at the top of the 2021 bull market at an all-time high above $700 and has spent years recovering from that overhang. That history creates a psychological barrier that has nothing to do with the technology and everything to do with the chart. Whether that launch was engineered by competitors or not is not the purpose of this article. We’re here to explore the current reality as it stands.

Third, the security narrative specifically requires a level of technical depth that most crypto commentary does not go to. It is much easier to write about transaction speeds and yield rates than to explain chain key cryptography and Byzantine Fault Tolerance in a way that makes the investor case clear. So the case often doesn’t get made at all, or it gets made in technical documentation that reaches developers but not capital allocators.

But here is what is changing. The Anthropic announcement did something useful. It forced a very public conversation about AI automating away the value of traditional cybersecurity products. Analysts were out within days arguing the selloff was an overreaction. Others wrote that the announcement actually strengthens the case for established vendors.

Maybe, but none of those analysts are asking the next question, which is “what happens when you remove the vulnerability at the infrastructure level rather than the application level?”, and that is the question ICP answers.

ICP Keeps Building

ICP is not waiting for the mainstream adoption narrative to catch up. The network is building non-stop.

Canister smart contracts on ICP grew from roughly 373,000 in January 2024 to nearly 980,000 by late 2025. That is a real measure of developer activity, not price speculation. ICP has led all cryptocurrency projects in GitHub commits for an extended period, with over 3,100 commits and more than 100 active contributors as recently as early 2026.

The vetKeys feature, now live in production, enables canisters to manage encrypted keys for users without users needing to remember passwords or manage decryption keys themselves. That alone is a meaningful primitive for building privacy-preserving applications that have no equivalent in traditional web infrastructure.

The Mission 70 proposal to reduce ICP’s annual inflation from 9.72% to 2.92% by the end of 2026, is moving forward. If network usage continues growing while supply growth compresses, the tokenomics case starts to look significantly different from where they sit today.

Conclusion

The Anthropic story is interesting not because it tells us something new about artificial intelligence, but because it reveals how sensitive markets are to the idea that foundational infrastructure can shift.

One AI vulnerability scanner scared investors out of billions in cybersecurity stock value in a single afternoon. The fear was simple: if artificial intelligence can automate what these companies sell, what are they actually worth?

ICP is not proposing to automate what the cybersecurity industry sells, but is proposing something more disruptive and, for the patient investor, more interesting. It is proposing to build internet infrastructure where the attack surface that the entire cybersecurity industry was built to protect no longer exists in the same form.

The market does not have a framework for pricing that yet. The crypto market is distracted by charts and cycles, and the traditional financial market does not speak the blockchain language yet. And so ICP continues to build, largely ignored, while the problems it was designed to solve grow more expensive and more visible with every passing quarter.

Historically, this is exactly where the best asymmetric opportunities tend to live.

If you made it this far, thank you for reading. And if you want more ICP articles like this one delivered straight to your inbox, consider subscribing to my Substack.

Talk to you later. BasedGiant.

Subscribe now

If you’re new to Mission 70 or just need a refresher before you cast your vote, this is the breakdown you need. And if you’ve been following along since the whitepaper dropped, there are some details in here worth revisiting with fresh eyes, now that we’re actually on the doorstep.

Key Concepts

NNS: ICP’s on-chain governance system.

Dissolve Delay: How long your ICP is locked before you can unstake it.

Maturity Modulation: A mechanism that adjusts how much ICP you actually receive when you disburse staking rewards.

Cycles: The computational fuel of the Internet Computer.

Subscribe now

Why Mission 70 Exists in the First Place

The Internet Computer has had an inflation “problem” for a while, even though if you compare it to other blockchains, you can barely call it “inflation”. But nonetheless, ICP gets minted in two major ways:

1. Voting rewards paid to neuron holders who participate in governance;

2. Node provider rewards paid to the people running the physical infrastructure that makes the network function.

Those two streams have been pumping out new ICP at a rate that the current level of network usage simply hasn’t been able to offset through burning. The result is persistent sell pressure from both stakers disbursing maturity and node providers converting their monthly rewards.

The Mission 70 whitepaper outlines a plan to reduce ICP’s raw annual inflation rate from 9.72% to 2.92% by the end of 2026, combining supply-side reforms with demand-side acceleration through increased ICP burn from network usage.

The name itself is the goal: reduce net inflation by 70%. Here is how they plan to do it.

The Supply Side: Restructuring Staking From the Ground Up

This is the part that affects most ICP holders directly, so pay attention:

Maximum dissolve delay drops from 8 years to 2 years.

This is the headline change that everyone reacted to when the paper came out. All existing neurons would go through a one-time migration where their dissolve delays are capped at 2 years. If you have a 6-year delay today, you would wake up at 2 years. If you have 8, same result. Everyone gets capped.

Before you panic: if you’re a non-dissolving neuron, this migration does not touch your age bonus. Whatever you’ve accumulated over years of not dissolving, you keep it. That was a meaningful detail that got lost in some of the early community discourse.

On the other end, the minimum dissolve delay for voting drops dramatically, from 6 months down to just 2 weeks. This matters because the current 6-month barrier is genuinely high compared to other chains, and the last 6 months of dissolving don’t even earn rewards anyway. The new minimum is designed to lower the barrier to governance participation and attract more short-term holders into the system.

The reward curve goes quadratic.

Right now, the dissolve delay bonus increases in a straight line. Lock for twice as long, get proportionally more rewards. The new system uses a convex curve, specifically a quadratic function, which means short-term stakers get meaningful but modest rewards while long-term stakers get significantly better incentives. The curve gets steeper at the top. They are also raising the maximum dissolve delay bonus from 2x to 3x, so even though the maximum lockup is shorter, the bonus for hitting that max is actually higher than before.

Projected rewards under the new system puts a 2-year neuron at around 7%. That is down from the roughly 12.3% available to 8-year neurons today, but it also means you’re committing for six fewer years.

The 8-year gang gets recognized, not abandoned.

Neurons that have historically maintained an 8-year dissolve delay will receive a special flag and a dedicated 10% reward boost that stays active until the end of 2030. Nearly five years of elevated rewards as a thank-you for long-term commitment. The catch: if you start dissolving, you lose the flag and the boost. Same mechanic as the age bonus. You keep it by holding, you lose it by leaving.

The daily voting reward pool gets capped.

This one hasn’t gotten as much attention as the staking changes, but it might matter more for long-term inflation control. Right now, the reward rate decreases from 10% at Genesis down to 5% over 8 years post-Genesis. We are nearly at that mark. Once we hit it, the rate becomes constant at 5%, which means the absolute amount of ICP minted for rewards would keep growing indefinitely as total supply grows. The fix is a cap on the daily reward pool, frozen at the level it reaches at the 8-year mark. After that, it becomes a governance parameter that can be adjusted over time rather than something that grows on autopilot.

A dissolution queue to manage the transition.

When all those neurons suddenly migrate to a 2-year cap, there will be a large cohort of neurons hitting their dissolution date at the same time. The whitepaper proposes a queue similar to what Ethereum uses for validator exits: if too much ICP is scheduled to dissolve on a given day, the excess gets staggered over subsequent days. This is a sensible safeguard against concentrated sell pressure from the migration itself.

Simpler and smarter maturity modulation.

The current modulation formula is, by the whitepaper’s own admission, overly complex. It uses 30-day moving averages, weekly returns, and four-week means, and it has a symmetric range of plus or minus 5%. The problem with symmetric bounds is that user behavior isn’t symmetric. People tend to disburse maturity when modulation is positive, which creates an inflationary bias.

The new system compares the 7-day moving average price to the 365-day moving average price. If the current price is below the long-term average, modulation goes negative and discourages minting. If the price is above the long-term average, modulation goes slightly positive. And here is the clever part: the bounds are now asymmetric. Negative modulation can go as low as minus 10%, but positive modulation is capped at plus 2%. This corrects the inflationary bias directly. They also add a daily speed limit of 0.3%, so if you initiate a 7-day dispersement, the modulation can only change by about 2.1% maximum during that window. More predictable for stakers.

The combined effect of all staking changes reduces voting rewards inflation from 5.88% to 3.45%, a 41% relative reduction.

The Node Provider Side: Getting Paid Fairly, Not Generously

The other major source of inflation is node provider rewards, and the whitepaper takes a hard look at the economics.

The Internet Computer currently runs 1,424 nodes, but only 701 were actually assigned to active subnets at the time of writing. That is a 49% assignment ratio. Several large Gen-1 node providers in the US and Europe have assignment ratios below 30%, meaning their nodes are sitting mostly idle and still collecting rewards. Gen-1 hardware is also fully amortized at this point. Operating costs run between $300 and $800 per month per server, while average rewards are around $1,500 per month. A reward-to-cost ratio of 2 to 5 times is hard to justify when the nodes aren’t doing much work.

The proposal is a 40% reduction in Gen-1 node rewards. The whitepaper models out a scenario where 66% of Gen-1 providers transfer to cloud engines or exit entirely. Combined with the reward cut, total Gen-1 rewards would drop by 80%. Even so, the whitepaper argues that most node providers can still cover costs and earn positive margins at the new rate.

There is also a proposal to distribute monthly payouts throughout the month rather than on a single date, which would smooth out the concentrated sell pressure that currently hits when all node providers receive rewards at once.

On the technical side, the network has 442 SEV-capable nodes, the newer Gen-2 hardware with enhanced security features. The proposal creates dedicated subnets using only SEV nodes and reduces their size from 13 nodes to 7 while keeping cycle costs the same. This would roughly double available subnet capacity without adding a single new node.

The combined effect of these node provider changes reduces node rewards inflation from 3.84% to 1.97%, a 49% relative reduction.

The Demand Side: Where the Real 70% Lives

Here is something important that gets glossed over in most Mission 70 discussions: the supply side changes alone only get to a 44% reduction. To hit 70%, the network needs demand acceleration. That means more people actually using the Internet Computer and burning cycles.

The whitepaper bets on two major catalysts.

Cloud Engines are private subnets that enterprises can spin up for their own workloads, configurable to specific security, performance, and resilience requirements. Think of them like the ISP boom of the 1990s, where independent providers built internet infrastructure and sold access to enterprises. Under the cloud engine model, node provider associations can sell this functionality directly to enterprise customers. Node providers receive 80% of the revenue, and the remaining 20% goes toward buying back and burning ICP, making cloud engines deflationary by design. Every cloud engine running on the network is burning ICP. DFINITY has reportedly already spoken to node provider associations that are interested in selling into the enterprise market.

Caffeine is DFINITY’s self-writing application platform, designed to let non-developers build fully on-chain apps. As more applications get built and deployed through Caffeine, more cycles get burned. The whitepaper describes a continuum where apps built in Caffeine can be exported into traditional development environments and vice versa, creating a seamless pipeline that drives ongoing on-chain compute usage.

To hit the full 70% target, the cycle burn rate needs to increase from roughly 0.05 XDR per second to 0.77 XDR per second. That sounds like a dramatic jump, and it is, but the whitepaper notes that the burn rate already exceeded this level for several months in 2025. The demand-side target is not entirely out of reach.

The cloud market context helps frame the ambition here. Cloud computing generated about $1 trillion in revenue in 2025 and is projected to hit $2 trillion by 2030. DFINITY is pitching cloud engines as ICP’s entry into that market, and even a small slice of that would represent enormous cycle burn.

Where Things Stand Right Now

The whitepaper dropped January 13th. The community spent weeks dissecting it. Most of the community agreed with cutting rewards, though debate remains on the degree of the cuts. That’s a pretty healthy starting point for a governance vote of this magnitude.

On February 21st, Dom confirmed the proposal is expected early next week. The NNS vote is where all of this becomes real. Not a whitepaper. Not a forum thread. An actual on-chain governance decision that will change how ICP works, how your neurons earn, and how the economics of this network function going forward.

If this passes, 2026 becomes the year ICP’s tokenomics actually changed. Supply-side measures bring inflation from 9.72% down to 5.42% by 2027. If demand acceleration hits its targets, it falls further to 2.92%. That’s a 70% reduction from baseline.

For Holders

If you’re an 8-year neuron holder, the migration caps your dissolve delay at 2 years, but you keep your age bonus and you receive a 10% reward boost flag that lasts until 2030. The system is acknowledging your commitment rather than penalizing you for it.

If you’re a newer or shorter-term staker, the lower minimum dissolve delay of 2 weeks opens up governance participation to you in a way the current 6-month barrier never did.

If you’re a node provider, the 40% Gen-1 reward cut is a real hit, and the whitepaper doesn’t pretend otherwise. But it also models out the cloud engine opportunity as a path to new revenue that is structurally deflationary for the token.

If you’re just an ICP holder trying to understand what this means for the value of your tokens: Mission 70 is a serious, detailed attempt to address the inflation problem this community has been talking about for years. The supply side numbers are not aspirational. They are concrete changes to specific parameters. The demand side is where some uncertainty remains, but the infrastructure to support that demand is actively being built.

If you made it this far, thank you for reading. And if you want more ICP articles like this one delivered straight to your inbox, consider subscribing to my Substack.

Talk to you later. BasedGiant.

Subscribe now

Except once you analyze the actual data, it doesn’t seem to support that conclusion, and a closer look at the numbers tells us something different.

BobbyO posted two videos that when taken together, make one of the more compelling on-chain arguments I’ve seen for why the “VCs dumped” narrative is probably wrong, and why the actual “forces” suppressing the ICP price have almost NOTHING to do with insider selling. Prepare to be mind blown, so let me walk you through the whole thing.

Key Concepts

Long Liquidation: when a leveraged long position is forcibly closed by an exchange because the collateral can no longer cover the loss.

OTC (Over-the-Counter): large asset trades conducted directly between two parties, bypassing public exchanges entirely.

Open Interest: the total value of outstanding derivative contracts on an asset.

Subscribe now

The Rumor

It’s not hard to understand how people got here. ICP launched in May 2021 at prices that briefly touched $700 on some exchanges, and has been dumping ever since. Andreessen Horowitz, which co-led a $102 million raise for DFINITY back in 2018 alongside Polychain Capital, hasn’t publicly mentioned ICP in years. Chris Dixon’s last notable public moment with ICP was a 2020 sit-down with Dominic Williams. Polychain has been equally silent throughout the years.

When something you invested in goes down 99.5% from its all-time high and the investors who were loudly behind it go completely quiet, the obvious interpretation is that they cut their losses and left. Now here’s the thing about “obvious” interpretations: they aren’t always the correct ones.

Math Nobody Actually Ran

Here’s where Bobby’s argument gets interesting, so let’s look at the actual numbers.

ICP launched with a total supply of roughly 470 million tokens, and today that number sits around 550 million. An increase of about 80 million ICP over nearly four years. That number breaks down into roughly 60 million in voting rewards, and 20 million in node provider rewards.

Now here’s the number that matters most: the total ICP sitting on centralized exchanges right now is around 60 to 62 million, starting from zero at launch.

The amount of ICP on exchanges is entirely accounted for by inflation alone, which is the minting of new tokens through governance and node rewards. You don’t need a single VC dumping a single token to explain what we see on exchanges today, because the math already covers it without them.

Bobby frames this as a simple question: “If a16z, Polychain, and other VCs offloaded tens of millions of ICP, where is it and why doesn’t it show up anywhere?” For those tokens to be on exchanges, they should have been there on top of the already minted 80 million ICP through normal inflation.

Another thing worth mentioning: not all of that 60 million on exchanges is actually liquid. A percentage (perhaps 20 to 30 million of it), is held by institutions using centralized exchanges as custodians. Large players use platforms like Coinbase Prime, Kraken, and Binance to hold assets in custody rather than to trade them actively, which means that supply shows up in the exchange balance but is illiquid. The liquid ICP may be closer to 30 to 40 million or so.

Subscribe now

Three Groups of Sellers

To understand the supply, we need to break down who’s actually been selling. There are three distinct groups and they each have their own logic:

1. The First Group: Seed investors. People who got into ICP at around $0.03 per token. These holders are sitting on returns of around 100x even at today’s low prices, which means selling voting rewards is essentially painless for them. They can redeem and liquidate their staking rewards and never touch a single token of their original position.

2. The Second Group: Node providers. They receive ICP as compensation for running the network infrastructure and are incentivized to sell a portion or even all of it to cover their operating costs. Running a node costs real money in hardware, bandwidth, electricity, and operations, so this isn’t “dumping”, it is a cost that’s baked directly into how the network functions.

3. The Third Group: Early VCs. This is the group everyone is worried about, like a16z and Polychain who came in at average prices somewhere in the $3 to $4 range. These are the potential “dumpers” in this narrative, but here’s how Bobby thinks through the numbers.

If any percentage of their position was liquid in the first 8 to 10 months after launch, when ICP was trading between $30 and $100, they could have sold 5-10% of their holdings at a 10 to 20x profit, recovered most or all of their initial investment, and still be sitting on 90% of their original position today.

That kind of move would be completely normal, and it would also be nearly invisible in the data, because the exchange inflow spike in that early period is entirely consistent with partial profits being taken at elevated prices, after which the trend flattens out and aligns with normal inflation.

The OTC Escape Hatch

There’s one more scenario worth addressing directly, which is what happens if the large VCs really did exit most of their position. Bobby points out that even in that case, the impact would be invisible, because a transaction of that size would never happen on an open exchange. It would be an OTC deal negotiated directly between institutions and completely off the books of any public market.

OTC trades don’t move the spot price and don’t show up as exchange inflows, so all they would mean is that one large holder was replaced by another. New money stepped in, old money stepped out, and the underlying holder base of ICP remained concentrated in institutional hands either way.

I’ll make an argument that an OTC deal would be more bullish than the original VCs simply holding on. Firms like a16z and Polychain got into ICP in 2018 for one reason: they expected a financial return, and that’s entirely what venture capital is designed to do. But if another institution quietly bought those tokens on the OTC market, the motivation behind that purchase could be different.

It could be a builder looking to deploy infrastructure on ICP. It could be an institution that has studied the technology closely and sees something the broader market is still missing. It could be a sovereign entity or an enterprise with a concrete use case already in mind. The original profit driven money may have been replaced by conviction money, and conviction money doesn’t sell at the first sign of a recovery.

This would be in check with the latest news about the Swiss and Pakistani subnet.

This Is Why They Are Not Talking About It (Part 1)

This bit is going to be longer, so bear with me, because it’s very important information, so I’ll split it into two parts to be easier to follow. This is the part of Bobby’s argument I find most compelling, because it doesn’t require any “conspiracy” and it just requires you to think critically about incentives.

Consider a16z’s portfolio. They have hundreds of investments across crypto and tech, and even if they still hold a significant position in ICP, it almost certainly represents less than 5% of their total assets under management. The other 95% is made up of projects that genuinely need a16z’s platform, voice, and endorsement to succeed. Many of those projects are in DIRECT COMPETITION with ICP for developer mindshare, capital, and narrative momentum.

If a16z publicly promotes ICP the way Chris Dixon used to, they are implicitly telling the market that ICP architecture is superior for most use cases. This would create a real problem for the other L1s and L2s they’ve backed. It creates a comparison that disadvantages their own book, and since ICP doesn’t need a16z to succeed given that it has been signing sovereign nation partnerships without a single VC shilling it publicly, there’s no real upside to speaking up while the downside is meaningful for them. The incentives clearly point towards VCs staying quiet.

This Is Why They Are Not Talking About It (Part 2)

To understand just how much is at stake in that portfolio conflict, consider what happened earlier this month. In a post on X, Vitalik Buterin said that the original rollup-centric roadmap for Ethereum’s L2 ecosystem “no longer makes sense”, and that a new path is needed. He pointed to two problems: L2s have moved far too slowly toward meaningful decentralization, and Ethereum’s base layer is now scaling on its own with gas limits expected to rise significantly through 2026, which removes the original justification for L2s existing in the first place.

He also wrote that the ecosystem doesn’t need more “copypasta EVM chains” and that chains connected to Ethereum through a multisig bridge aren’t actually scaling Ethereum in any meaningful sense.

This matters to the ICP conversation for a specific reason: VCs have poured billions of dollars into L2 projects over the past few years, and retail investors have committed tens of billions more on the theory that one of these networks would become the dominant scaling layer for the Ethereum ecosystem.

Vitalik just told the world that most of those bets were built on a premise that no longer holds. The uncomfortable truth in his post, though he doesn’t say it directly, is something the ICP community has understood for years: the vast majority of so-called “Ethereum killers” and L2 networks are essentially EVM clones that offer faster or cheaper transactions as their primary “innovation”.

ICP is a different kind of project, and the capabilities Vitalik says L2s should now be building towards, things like privacy, identity, AI, non-financial applications, and truly sovereign infrastructure, are capabilities ICP already has delivered on. The VCs who have been backing L2 ecosystems have every incentive to keep the spotlight off ICP for as long as possible.

It’s also worth noting what Chris Dixon said publicly in 2018 when the a16z crypto fund launched. He stated on the record that the firm had never sold any of their crypto investments and didn’t plan to any time soon, and that the fund was specifically structured to hold positions for ten years or more. Bobby doesn’t even cite this statement in his video, but it fits his argument precisely and adds context that the community has largely overlooked.

The Real Price Suppressor

Here’s where Bobby’s second video becomes essential, because you’re still left with an obvious question: why has the price been grinding down for four years?

Bobby’s answer is the derivatives market.

The long-to-short liquidation ratio for ICP has sat at approximately 10:1 for as long as he’s been tracking it, while the market average across major assets is closer to 1:1. To understand why this matters, you need to understand what a long liquidation actually is.

When a leveraged long position is forcibly closed, it generates a sell order on the spot market. Imagine that a $1,000 position opened at 100x leverage becomes a $100,000 sell order when it gets liquidated, 99% of which is backed by nothing real.

When longs are being liquidated at 10 times the rate of shorts, that imbalance creates a bunch of artificial sell pressure on the spot price. In 3 to 6 month, you can have more net buying pressure on the spot market than selling, and the price can still go down, because cascading long liquidations overwhelm the buying activity entirely. This is mental.

What makes the ICP data particularly notable is that this 10:1 ratio holds consistently across multiple exchanges, which is something you wouldn’t expect to see if this were just random overleveraged retail behavior. You’d expect somewhat different trading patterns between Binance and Coinbase given different user bases and regional demographics, but the ratio stays skewed the same way across platforms.

Conclusion

The supply math suggests the “VCs dumped” narrative doesn’t hold up, since the 80 million ICP minted since launch through voting and node rewards more than covers everything we see on exchanges today. If large early investors did sell significant positions, the most likely venue was OTC, making those trades invisible to the market and meaning either they stayed in or were replaced by equally (or even more) sophisticated buyers.

The silence from a16z and Polychain is evidence of rational portfolio management by firms whose incentives don’t align with publicly promoting an asset that represents a small fraction of their book and doesn’t need their help to succeed in the first place. And with Vitalik himself now calling out the L2 ecosystem for building glorified EVM clones rather than genuine innovation, the contrast with what ICP has been building quietly in the background is becoming harder to ignore.

The price suppression that’s been grinding holders for four years points toward the derivatives market, specifically an abnormal and persistent long-to-short liquidation imbalance that can, mathematically, push prices down even when spot buying outweighs spot selling.

None of this means the thesis plays out tomorrow, but it does change the question you should be asking. Based on the available data, the VCs probably didn’t leave, and even if they did, someone equally large stepped in quietly to take their place.

The more interesting question is what happens when they decide it’s time to speak up.

If you made it this far, thank you for reading. And if you want more ICP articles like this one delivered straight to your inbox, consider subscribing to my Substack.

Talk to you later. BasedGiant.

Subscribe now

Sources + Further Reading

BobbyO, “MAKING SENSE OF THE LAST 5 YEARS | ICP BY THE NUMBERS”

BobbyO, “EXPLAINING WHY THE CRYPTO MARKET IS FAKE”

Vitalik Buterin: “You are not scaling Ethereum”, CoinDesk

Vitalik Buterin blasts “copypasta” L2 chains, CoinDesk

TechCrunch, DFINITY raises $102M from a16z and Polychain

a16z DFINITY investment announcement, a16z.com

Chris Dixon on a16z crypto fund structure, PYMNTS (2018)

That someone was Kamal Youssefi, President of the Board of The Hashgraph Association. The Hashgraph Association is a Swiss-based nonprofit that drives the global adoption of Hedera technology. And what he said in the next two minutes was quietly one of the more significant partnership announcements in either ecosystem’s history up to that point.

“We finally found three use cases where we are going to collaborate with our partners at DFINITY. And I believe that combination is going to be unbeatable.”

He didn’t say what the three use cases were exactly, but more on that in a moment.

Key Concepts

Hashgraph: Hedera uses a directed acyclic graph, which allows transactions to be recorded and ordered in a more parallel, efficient way.

Canisters: Smart contracts on ICP that can store data, and serve web content directly. More powerful than typical smart contracts.

Chain Fusion: ICP technology for natively integrating with other blockchains without using bridges.

Subscribe now

What Is Hedera?

Before we get into the partnership, I’ll give a quick explanation on Hedera for anyone who focuses primarily on ICP:

Hedera Hashgraph is not a traditional blockchain. Instead of arranging transactions in a linear chain of blocks, it uses a structure called “directed acyclic graph”. The result is a network capable of processing over 10K transactions per second with finality in under 3 seconds, very low fees, and a carbon footprint so small that the network is actually carbon negative.

What makes Hedera stand out in the enterprise space isn’t just its tech though, but its governance too. The Hedera Council is a governing body made up of major global organizations such as Google, IBM, Dell, Deutsche Telekom, Ubisoft, and as of February 13th this year, FedEx. Council members actually run network nodes and vote on protocol decisions. It’s the kind of institutional credibility that makes regulators and enterprise teams far less nervous than they would be with a permissionless chain run by anonymous validators.

Hedera’s weaknesses, historically, have been on the smart contract and developer tooling side. Its EVM compatible smart contracts work fine, but they don’t offer anything close to what the ICP canister model delivers when it comes to on-chain computation, storage, and AI integration. That’s precisely where ICP comes in.

The Announcement

Kamal Youssefi and Dominic Williams officially announced the collaboration on January 25, 2025. But the details were sparse, and that seemed on purpose. This partnership is aimed at advancing on-chain finance, sustainability, and consumer engagement.

Those three things align with what Youssefi said on stage.

In his own words: Hedera brings “the sustainability aspect when it comes to energy consumption when it comes to a lower carbon footprint... and another aspect which is needed for institutional adoption, which is governance, combined with the great and powerful smart contract structure that the guys at ICP managed to develop over the last few years, including powerful AI components.”

What he described is a logical division of labor. Hedera brings the institutional trust layer, the enterprise governance, and the energy efficiency narrative (which also applies for ICP), while ICP brings the compute layer: canisters, on-chain AI, sovereign infrastructure.

The three use cases are worth thinking through, even speculatively.

• On-chain finance: Hedera’s secure and efficient consensus system complements ICP’s powerful smart contract technology. Hedera previously enabled the tokenization of bonds and real estate, demonstrating practical financial use cases. Combining that track record with ICP’s advanced smart contracts could unlock more complex and secure financial solutions.

• Sustainability: Although most of the time this is a buzzword, environmental concerns have always somewhat slowed blockchain adoption. Hedera offers a solution with its minimal carbon footprint. When combined with ICP’s powerful data processing and decentralized AI, this partnership could push sustainability standards forward.

• Consumer engagement: Hedera’s quick, low-cost transactions combined with ICP’s powerful and fast smart contracts can enable applications that users genuinely enjoy. The ICP canister model could serve the front-end, while Hedera could handle the high-throughput transaction settlement layer.

The framing from Youssefi was deliberate: “Competition is good, collaboration is better.” Neither side in this partnership is trying to absorb the other. They’re trying to build something neither could build alone.

This proves the point I have been making for years: “The future is multichain, and ICP has more chances to succeed the more chains it integrates. Tribalism is stupid, and we should all focus on collaboration.”

Subscribe now

The Partnership Has Been Quietly Moving Forward

Here’s where things get interesting if you’ve been paying attention.

Following last year’s announcement, the ICP community was a bit confused. A DFINITY forum thread from January 2025 titled “Partnership with HBAR?” had people asking whether anyone could find a link to an actual formal announcement. The answer was basically “no”.

But silence doesn’t mean nothing is happening. Fast forward to last month, and someone on X asked Pierre Samaties, DFINITY’s Chief Business Officer, for an update on the Hedera collaboration.

His reply was: “Ongoing, friends from Hedera Hashgraph are joining the event as well. We will provide an update regarding the collaboration, once it’s ready to announce.”

The event he was referencing was World Computer Day, Davos 2026, which took place January 20th this year. Hedera’s team was indeed present at Davos during that week.

So the partnership is ongoing. Both sides are at the same events. The Hedera team was specifically cited as attending a DFINITY event. And Pierre Samaties explicitly said an update is coming when it’s ready.

FedEx Joining Hedera

Now here’s the piece that makes this a genuinely compelling story right now.

On February 13, 2026, FedEx announced joining the Hedera Council. FedEx joins to help advance trusted digital infrastructure that can support the evolving lifecycle of global shipments and make supply chains smarter for everyone.

Let that sink in: FedEx does $90 billion in annual revenue, it has over 500,000 employees, and it’s a global logistics network spanning 220+ countries.

“The digital transformation of global supply chains is inevitable,” said Vishal Talwar, vice president of FedEx. “As supply chains become increasingly digital-native, trusted data must be shared and verified across many parties without increasing risk or centralizing control. Hedera provides a neutral, enterprise-grade trust layer that enables verification at global scale.”

As a Council member, FedEx will operate a node on the Hedera network and hold equal voting rights alongside other members, participating directly in the governance of Hedera software and services.

This is significant for the ICP/Hedera partnership for a reason that might not be immediately obvious: it validates Hedera’s exact value proposition in the partnership. Hedera brings institutional credibility and enterprise governance. FedEx joining the Council is the clearest possible demonstration that this credibility is real and expanding. The kind of organizations that ICP needs to reach, Fortune 500 companies with compliance requirements and legal teams who ask hard questions, are exactly the kind of organizations that are comfortable sitting at Hedera’s table.

How Deep Does This Go?

Now let me put on the speculation hat for a moment. Everything in this section is me thinking out loud, not reporting facts. But I think it’s worth exploring because the logic actually holds up.

What if DFINITY joined the Hedera Council?

The Council currently has 31 members out of a maximum of 39, which means there are open seats. Members come from technology, finance, telecom, energy, gaming, and academia. DFINITY would fit right in. It’s a Swiss foundation building decentralized enterprise infrastructure. That’s exactly the profile Hedera recruits for. And unlike most Council members who are just users of distributed ledger technology, DFINITY would bring something unique to the table: deep expertise in building one.

Council members run nodes on the Hedera network and get equal voting rights on protocol decisions. For DFINITY, joining would mean a direct seat at the governance table of one of the most (if not the most) institutionally connected networks in crypto. It would also mean direct relationships with Google, IBM, Dell, Deutsche Telekom, FedEx, and every other Council member.

For Hedera, bringing DFINITY onto the Council would signal something powerful to the market. It would be the first time a major Layer 1 protocol joined another Layer 1’s governance structure. That alone would generate serious attention. But more practically, it would give Hedera direct access to the team building the most advanced smart contract and on-chain AI infrastructure in the space. The exact capabilities Hedera lacks are the ones DFINITY has spent years developing.

Now let’s go one step further. What if Hedera ran a subnet on ICP?

This is also speculative, but hear me out. ICP’s architecture is built around subnets, independent groups of node machines that host and run canisters. We just saw the first TEE subnet go live for confidential computing, the Swiss national subnet launched at Davos, and a new Pakistani subnet announced just a few days ago. The system is designed to be modular and expandable.

There’s nothing in ICP’s architecture that would prevent a Hedera specific subnet from existing. A subnet where Hedera’s consensus data is natively accessible to ICP canisters. Where smart contracts on ICP could read and write to the Hedera network without going through bridges or oracles.

The benefits of that would be significant for both sides. Hedera developers would gain access to ICP full compute stack: on-chain storage, web serving, AI, autonomous canisters. ICP developers would gain access to Hedera’s transaction layer: fast, cheap, enterprise-trusted settlement with built-in institutional governance. The two networks would stop being separate ecosystems and start being complementary layers of the same stack.

For HBAR holders, this would mean their network becomes the preferred settlement layer for one of the most computationally powerful blockchains in existence. For ICP holders, it would mean direct integration with a network that has Google, FedEx, and IBM sitting on its governance council.

Is any of this confirmed? No. But is it technically feasible and strategically logical? Absolutely.

Conclusion

Here’s where things stand. ICP and Hedera announced a partnership in 2025. Three use cases were hinted: DeFi, sustainability, and consumer engagement. Technical specifics were held back. A few weeks ago, Pierre Samaties has confirmed the collaboration is ongoing. Hedera’s team attended World Computer Day at Davos 2026. FedEx just joined the Hedera Council on February 13th, validating Hedera’s institutional track record in concrete terms. And an update, per Pierre, is coming when it’s ready.

To all my ICP friends, keep this in mind: The people involved are credible, the use cases are logical, and the institutional credibility behind Hedera is very real. When the technical announcement finally drops, it’ll land in a much more favorable environment than it would have a year ago, because both ecosystems are stronger now.

If you made it this far, thank you for reading. And if you want more ICP articles like this one delivered straight to your inbox, consider subscribing to my Substack.

Talk to you later. BasedGiant.

Subscribe now

Sources + Further Reading

Fireside Chat, Kimbal Musk & Dominic Williams

HBAR Announces Major Partnership with ICP

The Hashgraph Association and DFINITY Partnership

Pierre Samaties on X, Partnership Update

FedEx Joins Hedera Council, FedEx Newsroom

FedEx Joins Hedera Council, Hedera Blog

Hedera at Davos 2026, Hedera Blog

Partnership with HBAR?, DFINITY Developer Forum

Key Concepts

- TEE: Think of it as a vault built directly into the CPU that even the machine’s owner can’t crack open.

- SEV-SNP: AMD’s specific implementation of TEE technology.

- Subnet: On ICP, a subnet is a group of node machines that work together to host and run smart contracts.

Subscribe now

DFINITY Proposes Another Industry Game Changer

You might be surprised about this, but in blockchain, encrypting data is the easy part. The hard part is PROCESSING that data while keeping it private. Traditional blockchains handle this by encrypting data while it’s not being used, but the second you need to actually do something with that data, for example run a calculation, or execute a smart contract, you have to decrypt it. And the instant you decrypt it, whoever is running that computation can see everything.

That’s why TEE (Trusted Execution Environment) changes everything. With TEE technology, specifically AMD’s SEV-SNP (Secure Encrypted Virtualization - Secure Nested Paging), the hardware itself creates an encrypted bubble where computation happens. The data gets decrypted inside this secure vault, processed, the results come out, but at no point can anyone outside that vault see what’s happening inside, not even the person who owns the physical machine.

ICP has been testing this technology for months. As DFINITY researcher rbirkner explained in the forum announcement, individual TEE nodes have been successfully running in all roles: unassigned, replica, and API boundary nodes. Now, with Proposal 140407, those tested nodes are coming together to form a complete subnet dedicated to confidential computation.

Less Nodes, Better Security

This subnet only uses seven nodes instead of the usual thirteen that is usual for ICP. That might sound like less security, but it’s actually quite the opposite. Because TEE nodes provide hardware-backed security guarantees, fewer nodes are needed to achieve the same or even better security level. The hardware is doing heavy lifting that would normally require more machines and more consensus rounds.

These seven nodes are spread across the world: Australia, Costa Rica, Czech Republic, Spain, Hong Kong, India, and the United States. Right now, this subnet is operating in “authorized-only” mode. No deployments are permitted yet. DFINITY is gathering operational experience, making sure everything runs smoothly before opening it up. Once they’ve confirmed stability, the plan is to either open this subnet for general use or create additional TEE subnets based on what they learn.

How SEV-SNP Actually Works

Let me break down the technology powering this subnet. AMD’s SEV-SNP isn’t new. Major cloud providers like AWS, Google Cloud, and Azure already use it for confidential computing workloads. But using it in a decentralized blockchain context is novel.

SEV-SNP encrypts the entire memory space of a virtual machine using keys generated by AMD’s secure processor. Even if someone compromises the hypervisor (the software that manages virtual machines), they can’t read the encrypted memory. Even if they have physical access to the machine, they still can’t extract the data.

But encryption alone isn’t enough. SEV-SNP also provides memory integrity protection. This means if a virtual machine reads a memory page, it’s guaranteed to only get the last value that was written there. No substitutions, tampering, or replay attacks. This protection extends to preventing the hypervisor from remapping or swapping memory pages between different VMs, closing off an entire class of side-channel attacks.

The system also includes something called cryptographic attestation. Before you trust a TEE with your sensitive data, you can verify that it’s actually running in a genuine secure environment with the correct, unmodified code. The TEE generates an attestation report signed by AMD’s hardware, proving its integrity. In other words, you’re getting a cryptographic proof from the chip itself.

Subscribe now

ICP’s Privacy Approach (Two Steps)

What makes the Internet Computer’s privacy compelling is that TEEs aren’t the only tool in the box. ICP already launched vetKeys in July 2025, providing cryptographic privacy through verifiable encrypted threshold key derivation. I know this is very technical, but bear with me.

vetKeys enable identity-based encryption where you can encrypt data for a specific user (identified by their principal, Internet Identity, or even an Ethereum address) without that user needing to have interacted with your application before. The keys are derived on-demand by subnet nodes using threshold cryptography. No single node ever has access to a complete private key. This enables use cases like end-to-end encrypted messaging, private password managers, and MEV-proof DeFi where transaction details stay hidden until they’re finalized on-chain.

Now with TEE subnets, ICP offers a complementary approach: hardware-backed confidential computation. Where vetKeys provide cryptographic privacy that’s verifiable through mathematics, TEEs provide confidential execution environments verified through the hardware itself.

Think of it this way: vetKeys are like having an unbreakable encryption algorithm, while TEEs are like having a locked room where you can safely use that algorithm without anyone watching over your shoulder. They are different tools, but both are essential for different use cases.

Real World Use Cases

Privacy on public blockchains has always been the dealbreaker for enterprise adoption. CZ actually said this exact thing in a recent post. For organizations dealing with regulations like GDPR, public blockchains have always been a no-go.

TEE subnets change that, because now you can build applications that require genuine confidentiality. Let me give you some examples:

• Healthcare and Medical Records: Hospitals could share patient data for research or coordinated care while keeping individual records completely private. The computation happens inside the TEE, results are shared, but no one outside the secure vault ever sees the sensitive medical information.

• Private DeFi and Trading: One of the issues with DeFi are MEV attacks, where bots can see pending transactions and front-run them for profit. With confidential computation, trade details can be encrypted until they’re finalized.

• AI and Machine Learning: Training AI models requires massive datasets, often containing sensitive information. TEE subnets could enable collaborative AI training where multiple parties contribute proprietary data, the model trains on the combined dataset inside the secure vault, and no party ever sees the others’ raw data.

• Enterprise Applications: Any business process involving confidential information: payroll, contracts, supply chain data with competitive implications, all of these become feasible on a public blockchain when you can guarantee that the computation remains private.

• Confidential (Real Life) Governance: Private voting where votes are encrypted inside a TEE, and only the final results published. Participants can verify the process was fair through attestation without anyone knowing how individuals voted.

All of these are real life applications that are absolutely mindblowing if you think about them being implemented in our societies. I will dedicate a whole article to expand on these further and give you guys some concrete insights on how this can change the world for the better.

TEE vs Zero-Knowledge

It’s worth knowing how TEE privacy compares to other approaches, particularly zero-knowledge proofs, which have gotten a lot of attention in the crypto space, especially with the Ethereum community.

Zero-knowledge proofs let you prove something is true without revealing why it’s true. They’re powerful for specific use cases like proving you have enough funds for a transaction without revealing your balance, or proving you’re over 21 without revealing your birthdate. But ZK-proofs come with tradeoffs. They’re computationally expensive to generate, they often require custom circuits designed for each specific use case, and they’re not suited for general-purpose computation with large datasets.

TEEs take a different approach. Instead of mathematically proving computation happened correctly without revealing the data, TEEs create a protected environment where computation happens normally but remains hidden. This makes them more flexible for complex, general-purpose applications.

The performance difference is significant. For something like a decentralized exchange handling thousands of transactions with complex order matching logic, building that with zero-knowledge proofs would be extremely challenging and computationally expensive. With TEEs, you can run that same logic at near-native speeds inside the secure vault. That said, TEEs and ZK-proofs aren’t competitors, but complementary.

The Internet Computer providing both TEE-based hardware privacy and vetKeys-based cryptographic privacy gives developers options to choose the right tool for their specific needs.

Conclusion

If you’ve been holding ICP since the early days (shoutout to my fellow 8-year gang members), you’ve watched this protocol evolve from a fascinating technical experiment into something that’s starting to solve real problems that have blocked blockchain adoption.

Privacy has always been the missing piece of the puzzle. TEE subnets remove that barrier. Combined with vetKeys, Chain Fusion, and ICP’s existing capabilities for running full-stack applications entirely on-chain, the Internet Computer offers something genuinely different from other blockchains: a platform where you can build privacy focused, enterprise-grade applications without sacrificing decentralization.

Every application running on ICP burns ICP to create cycles. More applications, especially enterprise applications with serious usage, means more burn. More burn means downward pressure on supply. Confidential computing opens up entire categories of applications that couldn’t exist on public blockchains before, expanding the potential market for ICP services.

If you made it this far, thank you for reading. And if you want more extensive ICP articles such as this one delivered straight to your inbox, consider subscribing to my Substack.

Talk to you later. BasedGiant.

Subscribe now

Sources + Further Reading

NNS Proposal 140407 - The governance proposal creating the first TEE subnet

DFINITY Forum Announcement - Official announcement with technical details and decentralization metrics

Bjoern Tackmann on X - DFINITY researcher’s announcement about confidential smart contracts

What are vetKeys? - Official documentation on ICP’s cryptographic privacy technology

Onchain Privacy in Action: A Guide to vetKeys Use Cases - DFINITY’s overview of privacy use cases

What is a Trusted Execution Environment (TEE)? - General background on TEE technology

AMD SEV-SNP Overview - Amazon’s announcement of SEV-SNP support showing enterprise adoption