Dear Readers,

Two books I am currently reading.

Against Purity (Alexis Shotwell, 2016)

Shotwell’s argument is that purity is not just a way to describe matter or a certain condition. It is highly entangled with politics and history. Purism, as Shotwell argues, is a common way to undertstand the wolrd but problematic because response needs dichotomies and notions that are not stable.

From page 15:

To be against purity is, again, not to be for pollution, harm, sickness, or premature death. It is to be against the rhetorical or conceptual attempt to delineate and delimit the world into something separable, disentangled, and homogenous. With and following María Lugones, I am “firmly planted against the logic of boundedness.” I follow her argument “for intercommunalism from the midst of impure subjects, negotiating life transgressing the categorical understandings of a logic of binaries that produces hard-­ edged, ossified, exclusive groups” (Lugones 2003, 35). Lugones critiques a meta-physics of purity, understood as separability, fragmentation, and standing outside culture and situatedness. The Man of purity, as a figure, “shuns mpurity, ambiguity, multiplicity as they threaten his own fiction. The enormity of the threat keeps him from understanding it. So, the lover of purity remains ignorant of his own impurity, and thus the threat of all impurity remains significantly uncontaminated” (132). The metaphysics of purity is necessarily a fragile fiction, a conceit under constant but disavowed threat to affirm a commitment to purity is in one move to glance at the entanglement and coconstitution, the impurity, of everything and to pretend that things are separate and unconnected.

What makes the book useful is that Shotwell traces purity as a product of a particular formulation of modernity. A formulation that has driven colonial projects, genocidal and ecological destruction often framed as acts of cleansing. When purity gets invoked, it tends to be in the service of power: legitimising racism, white supremacy, the violent management of what or who counts as contamination. A pure, uncontaminated state has never existed. All matter and beings are always already entangled.

Fear of a Dead White Planet (More Worlds Collective, 2025)

The title comes from Public Enemy’s Fear of a Black Planet (1990) — and the book carries a dual gesture: diagnosis and activation. The More Worlds Collective resists the contemporary rush toward planetary technofixes for environmental emergency, arguing that the very frames we use to understand crises. Notions like the Anthropocene, global health, GDP are not neutral. They are what the authors call Charismatic Mega Concepts (CMCs): large technical systems that animate only particular worlds into existence while quietly erasing others.

The planetary move (concepts such as one world, one species, one crisis) clears the ground for a particular kind of concept that claims ultimate scale. It eliminates the vast and different ways people actually live, the multiple worlds that coexist in tension, the wildly divergent impacts that communities have on collective conditions. And it does so while offering the feeling of comprehension. Their alternative is not another framework but a practice and research that makes room for hostile worlds even at modest scale.

From page 82: (while looking closely the idea of terraformations)

Understanding complexity is a matter not only of optics and precision; it also reveals what you care about, what is messing with you, what binds you are caught in, what violences need calling out, and how pain matters to the constitution of your study. In this way, we offer resolution not as a term referring to a scale of granularity of data or view, but as an invitation to an ongoing processual method of ethical and political resolve. Resolution as resolve is about inhabiting and obligating to the legacies, politics, movements, communities, and ethics of somewhere. How are you obligated to the specific Indigenous legal order of the land you study on? How are you obligated to defend a terraform? How are you obligated to consent or refusals? How are you joined up with abolition? Such resolve may yield commitments in scale; that is, the scaling of your study or attention: Do you have responsibilities to this tree, this architecture, this taxonomy? How do you fulfill and renew such responsibilities? Resolve in this key makes the question of scales and optics a question of commitments and obligations to people, beings, and forces.

“Study will never be enough” is not sighed in resignation, nor is it thrown like a weapon. It marks the limits of the modes of study that call themselves research and hails others to come.

Both books are asking something similar: how do we act politically, intellectually, collectively but without reaching for the false clarity of clean categories?

Share The S.A.D Newsletter

It sits somewhere perhaps on desert gravel outside Phoenix, on the cold edge of a Nordic fjord, or in industrial zones along the Persian Gulf. Data centers are clustering in particular landscapes. Some places offer obvious advantages. Cool air, cheap power, political stability. The Nordics check those boxes. The Gulf offers capital and connectivity…..until now.

In Arizona, massive server farms rise in desert heat, demanding water and electricity. A recent investigative report from Grist.org notes that tech companies like Microsoft and Meta have swarmed in to build data centers fuelling the AI boom, and Taiwan Semiconductor Manufacturing Company (TSMC) has spent billions on a factory complex outside Phoenix. This is triggering fears that the industry will drain the finite water supplies available to residents of Phoenix and Tucson.

Share

But the report notes those predictions have not come true. Even though Arizona will soon be home to nearly 200 data centers and chip factories, these facilities have not caused a major bump in the state’s water consumption. This does not mean the problem is gone. The precise effects are hard to understand due to the secrecy about usage. Of course new cooling technologies or by recycling water on-site can help. But it seems that many local officials have largely made their peace with the industry’s arrival. But:

Local opposition has also been a factor in pushing developers to either cut their water usage or move away from evaporative cooling. After a wave of resident protest caused the Tucson city council to nix a proposed data center called Project Blue, the developer for that project committed to a zero-water cooling system. It sought regulatory clearance instead from Pima County, where it secured county approval. It has since unveiled a similar project in nearby Marana.

Municipalities in the Phoenix and Tucson regions still face a critical threat from the looming shortage on the Colorado River. In the worst-case scenarios, the federal Department of the Interior could cut almost all water deliveries from the canal that serves central Arizona. That would cut off water supplies equivalent to about a quarter of the region’s water usage, forcing a massive economic contraction for homes, businesses, farms — and, of course, for the data centers as well.

Geography has always mattered, even as we go hyperscale. The marketing around cloud computing has done a good job of obscuring this — “the cloud” suggests something weightless, placeless, immune to geography. It is none of those things. We just don't talk about it enough.

In Flash Boys, Michael Lewis described how trading firms paid fortunes to move their servers as close as possible to stock exchanges, because a few meters of cable could mean a few milliseconds in a trade. Engineers built microwave towers between Chicago and New York so that financial signals could travel a slightly straighter line through the air. The digital world, it turned out, was still obedient to the physics of distance. Financial markets still have a geography, and the location of exchanges, data centers, and communication infrastructure continues to shape competition.

The story now has taken an older turn. Some data are now stored in places built for a previous technological anxiety: Cold War bunkers and mountain caverns. Facilities in the UK and Sweden occupy nuclear shelters; Tencent has placed servers inside caves in the mountains of Guizhou.

And now geography adds another layer. A recent strike reportedly damaged infrastructure tied to an Amazon Web Services facility in the UAE. Gulf countries are accelerating plans to build overland data cable routes to Europe after regional conflict exposed how fragile their reliance on undersea internet infrastructure has become. With key maritime chokepoints like the Red Sea and Strait of Hormuz increasingly vulnerable, governments and telecom firms in Saudi Arabia, Qatar, and the UAE are backing multiple land-based corridors through Iraq, Syria, and Turkey. The goal is to create more resilient, diversified digital connections (very similar to earlier efforts to bypass oil shipping routes).

The cloud, despite its name, sits in buildings. Buildings sit in landscapes that can be reached by protesters, by drought, by rain, and by missiles. There is an irony: artificial intelligence increasingly helps guide modern warfare, and warfare, in turn, finds the servers and attacks it that help run the intelligence.

Get more from Sharif Islam in the Substack app

Available for iOS and Android

Get the app

See news link: https://www.bbc.com/news/articles/c629pny4gl7o

You can blame one configuration file which was auto-generated for threat traffic management. It grew beyond its expected size and triggered a latent bug in Cloudflare’s software, causing a cascading crash.

Cloudflare is an American company that provides internet services such as speeding up websites, protecting them from cyber attacks, and improving security. They do this by using a global network of servers to deliver website content faster (called a Content Delivery Network or CDN), block harmful traffic (like DDoS attacks), manage domain names (DNS), and provide encryption so data stays secure during transmission. Today, many websites stopped working because Cloudflare, which acts like an entry point or a highway, had technical problems. When it stops working, many websites are blocked.

Content Delivery Networks didn’t emerge by accident. They were born out of necessity during the early 2000s as video streaming exploded and bandwidth pressures threatened to break the internet’s backbone. As more content providers emerged and users demanded faster, richer experiences, CDNs evolved from simple caching solutions. Today, CDNs are integral to every layer. Google Docs, Microsoft 365, and Salesforce rely on CDNs to deliver applications seamlessly. Amazon Web Services and Azure use CDN to distribute computing resources. Advertisement networks and payment processors need real-time, low-latency delivery. The modern CDN isn’t just about delivering your cat video faster, it’s the invisible infrastructure that powers almost everything.

But there’s more to this than the tech layer. AWS, Microsoft, and Google account for around 70 percent of cloud services used in Europe, and when you add Cloudflare’s CDN and security layer, the concentration becomes even starker. What does this mean for the European cloud #sovereignty question?

The discussion around cloud sovereignty isn’t merely technical. It’s geopolitical. Across the EU this has evolved into a highly charged political debate that cuts to the heart of Europe’s digital sovereignty. EU member states are divided: some prioritise open markets and oppose restrictions on American providers, while others push for stronger sovereignty requirements.

Adding another layer to this debate is institutional ownership. BlackRock holds approximately 6.4 percent of Cloudflare. And BlackRock has investments in companies involved in the weapons and defense industry, which include major defense contractors like Lockheed Martin, and companies like Caterpillar and Palantir. This concentration of ownership in large American financial institutions raises questions about control and accountability. When major infrastructure providers are primarily owned by U.S. based institutional investors, European policymakers worry about the alignment of corporate decision making with European strategic interests. But there’s no easy solution to this.

Today’s outage demonstrates that relying on a handful of providers, no matter how technically excellent, creates single points of failure. For Europe and other parts of the world, the answer isn’t just building alternative infrastructure. It’s about creating a diversified, resilient ecosystem which balances centralised and decentralised design. This balance is not easy to achieve. Centralisation offers efficiency, economies of scale, and technical excellence. Decentralisation offers resilience, sovereignty, and risk. The challenge is that you can’t simply choose one over the other. Pure decentralisation sacrifices the performance and security benefits that come from consolidated expertise and infrastructure. Pure centralisation, as today’s outage demonstrates, creates fragility at scale. The path forward requires accepting this tension rather than resolving it, building systems that can function across both models and designing for graceful degradation when centralised components fail.

Share The S.A.D Newsletter

But why exactly do we need to listen to him? The simple answer is: because of how much wealth he has. As a recent Guardian commentary points out, Gates’s essay ahead of COP30 (10 Nov - 21 Nov) treats the climate crisis as if it exists in a political vacuum omitting the role of billionaires, tax avoidance, and entrenched power structures that block meaningful reform.

In a way, this isn’t so different from Kim Kardashian telling co-star Sarah Paulson on The Kardashians that the Moon landing “didn’t happen,” prompting NASA’s acting director, Sean Duffy, to publicly correct her. Of course, Gates isn’t spreading reality-TV-grade conspiracy theories. But by sidestepping history, evidence, and the social complexity of the crisis while speaking from a platform of celebrity he participates in a similar pattern: simplifying reality through the lens of influence.

Climate change is not a technical problem waiting for a clever billionaire to solve. As (a renowned climate scientist) reminds us:

“Just give me a list of people’s top 10 things to worry about, and I can tell you how climate change is making every single one of those top 10 things worse.”

Yet we continue to place billionaires especially American ones on pedestals of planetary wisdom. Imagine Mukesh Ambani or Masayoshi Son being asked to lecture the world on climate change and media dutifully nodding along. Unlikely. There is something uniquely American about equating financial success with authority.

It is a pattern we have seen before: physicists “discovering” epidemiology through Bayesian estimation and Monte Carlo methods during COVID, while Francis Collins, the former director of the U.S. National Institutes of Health (2009–2021), reminded everyone that we should have paid more attention to understanding human behaviour, not just data and models. In a 2021 NPR interview, Collins reflected:

To have now 60 million people still holding off of taking advantage of lifesaving vaccines is pretty unexpected. It does make me, at least, realize, “Boy, there are things about human behavior that I don’t think we had invested enough into understanding.” We basically have seen the accurate medical information overtaken, all too often, by the inaccurate conspiracies and false information on social media. It’s a whole other world out there. We used to think that if knowledge was made available from credible sources, it would win the day. That’s not happening now.

This is the world Bill Gates is speaking into: a world where emotion, trust, and influence often outweigh evidence. The same forces that fueled vaccine hesitancy or made a Kim Kardashian “moon landing didn’t happen” moment go viral are also shaping how people engage with climate change. The people we should listen to most are not necessarily those who have the most money but those who understand how systems, societies and human behaviour intertwine. And, as recent research shows, emotional engagement, especially anxiety, matters far more than data in motivating real climate related behaviour.

A cross-national research underscores how important trusted media and emotional engagement are for climate action. This study published in the International Journal of Press/Politics (based on data from over 8,000 respondents across eight countries), examined how media use, trust, and climate anxiety shape pro-climate behavioural intentions such as using less energy, flying less, or repairing rather than replacing products. Mainstream news use was consistently associated with stronger behavioural intent everywhere, while non-mainstream media had positive but more context-dependent effects, especially in countries where digital platforms dominate climate information. The study defines mainstream media as “professionally produced content delivered via legacy platforms such as TV, radio, and print” and “non-mainstream media” as digital-first platforms and peer-to-peer content, social mdeial.

More specifically climate anxiety emerged as a powerful, universal predictor of wanting to act, though the news itself wasn’t the consistent primary source of this anxiety, suggesting emotional responses often stem from personal or social experiences. Furthermore, trust in media consistently amplifies the positive effects of news consumption, confirming its foundational role in translating information into engagement.

It is important to note that these findings are correlational rather than causal, based on self-reported intentions rather than real behaviour. The study does not isolate the effects of celebrity or billionaire commentary, so any inference about high-profile figures like Gates remains interpretive. Nevertheless, the point is how news and emotional engagement (climate anxiety here) play key roles in driving climate action, more than optimism or personality-driven narratives.

When applying the study’s findings on trust and media influence to the role of celebrities like Bill Gates, the media must understand that a celebrity functions as a high-visibility, often non-mainstream source whose effectiveness hinges critically on their perceived influence. The research suggests that because trust dramatically amplifies the effect of any message, especially from less-regulated sources, media reporting should move beyond merely platforming the celebrity’s views; instead, it must contextualise their authority by including expert responses, analysing the tangible policy implications, and assessing the celebrity’s alignment with their stated values to ensure that the message is evaluated by the audience based on its substance and not just on the celebrity’s fame.

But, let’s see what Bill Gates’ actually said . He argues that climate change is a very serious problem but it “will not lead to humanity’s demise.” His states that funds are very limited. So the COP delegates at this month’s summit in Brazil should direct money away from near-term emissions goals towards spending on poverty and disease. Because temperature rises and emission-cuts alone are not the best metrics for progress. He emphasises innovation especially clean technologies in electricity, manufacturing, agriculture, transport, buildings and reducing what he calls the “Green Premium” (the extra cost of a clean versus dirty technology).

George Monbiot wrote a scathing piece in The Guardian looking into Bill Gates’s argument that due to limited funds, the upcoming Brazil climate summit should prioritise money for climate “adaptation” and tackling poverty/disease over “near-term emissions goals.” Monbiot suggests Gates’s silence on issues of wealth and political power is either a symptom of a cognitive inability to see the world from another perspective, or a strategic “peace offering” to figures like Trump, concluding that Gates’s immense economic power means his views cannot be ignored. But they should be recognised as speaking on behalf of his own interests.

I am quoting here from another opinion piece: :

“If we were listening to people on the grounds of whether they had a good track record, the world would not spend a lot of time on Gates and climate. But if you have a hundred billion dollars all is forgiven, and so there has been lots of fawning coverage.

And perhaps, the more unsettling truth: our media and digital environment rewards simplicity and star power but not complexity. The climate crisis is polycentric and messy. It demands cooperation across scales, disciplines, and histories. Yet the platforms through which most people encounter it are built for viral clips, not nuance. In that world, billionaires, influencers, and algorithm-friendly optimism will always drown out the quiet, complicated work of systems change.

The Python Software Foundation recently had to turn down a US $1.5 million grant from the U.S. National Science Foundation. The condition? That the foundation abandon all Diversity, Equity, and Inclusion (DEI) efforts. Worse, if they violated that clause, the NSF could claw back the money even after it had been spent. Faced with that risk, they walked away from this funding.

At first glance, this seems like a political scuffle removed from code repositories and software development. After all, what does a popular programming library have to do with politics? We know science, and by extension technology, supposed to be objective, neutral, and above ideology, right?

Here is the paradox. We need objectivity, of course: data, facts, reproducible methods. You cannot vibe your way out of science. But as Geoffrey Bowker reminds us, “raw data is an oxymoron.” You can say the same for software. The way we collect, fund, interpret, and distribute data and technology is always embedded in social and political choices.

Share The S.A.D Newsletter

That is why this decision matters. A funding restriction dressed up as “neutrality” or “freedom” is in fact a political intervention. It determines what kinds of communities and values can shape open infrastructure, science, and by extension, society.

And of course, the backlash was swift (the internet comment section, where humanity goes to die). Reactions to the Python Software Foundation’s decision have been revealing. Comments accusing the foundation of “virtue signalling,” of running a “woke cult,” or of “ruining Python with politics” show how one-sided the idea of “keeping politics out of tech” often is. It is not that these critics reject politics; they reject a particular kind of politics, one that asks whose voices and contributions count.

If the price of “neutrality” is silence on inclusion, the cost is exclusion. The Python Software Foundation’s choice to forgo the grant is not about ideology. It is about integrity and sustainability. Python has become one of the world’s most widely used programming languages because of its openness, its community, and its belief that everyone should have a place in it.

Subscribe now

No matter what you do for work, AI somehow finds its way into every conversation. Someone will inevitably ask about your “AI strategy.”

Here are my top 10 battle tested conversation deflectors. Please report back with field results:

1. “Have you tried linear regression? I heard it works!”

2. “Oh, look! Squirrel!”

3. “Maybe we should give blockchain another shot. What do you think?”

4. “You should start a podcast!”

5. “Have you read my Substack newsletter? I have 10 subscribers!”

6. “You should really use my latest Python package. It has 233 dependencies.”

7. “Have you taken up knitting or bird watching?”

8. “Kombucha?”

9. “We really should talk about the polar bears.”

10. “I heard HVAC is the future.”

Share

Today, I asked ChatGPT to do the work for me. I asked it to write a parody of the recent “OpenAI for Countries” announcement. It did a pretty good job. This announcement is related to the new “Stargate” project, which is a initiative to invest $500 billion over the next four years building new AI infrastructure for OpenAI in the United States. Here's the text with a few edits and tweaks here and there. Enjoy!

Our #SmokeGate Project, a groundbreaking commitment to global tobacco infrastructure launched in January alongside President Trump and our esteemed partners Marlboro and Philip Morris International. We’ve been overwhelmed by interest from governments around the globe eager for their own SmokeGates.

It’s clear to all now: nicotine is the backbone of post-industrial serenity, and no nation can afford to fall behind. Innovation has always fueled prosperity by helping people relax harder, breathe deeper (sort of), and take longer breaks. Tobacco will now scale human leisure itself and usher in a golden age of soot-stained liberty.

We want to help these countries and in doing so, spread Democratic Tobacco: tobacco that is freely chosen, tastefully addictive, and proudly unfiltered by authoritarian regulation. Examples of democratic tobacco include:

- The freedom to light up without interference from nanny-state health ministries.
- Prevention of government tobacco monopolies that restrict flavors.
- A free market where menthols, cloves, and cowboy-kissed classics thrive side by side.

These principles ensure the broad distribution of lungful pleasure, discourage health elitism, and keep power where it belongs: in the fingers of the free smoker. We believe that deep cooperation with the U.S. government is the best way to advance democratic smoke.

Today, we’re launching OpenSmoke for Countries, a new effort within the SmokeGate initiative. Now is the time to stand up, puff out your chest (and maybe cough a little), and choose a freer future. One drag at a time.
Through this initiative, and always with respect for national lung sovereignty, OpenSmoke will:

- Build in-country cigarette production facilities, helping nations develop local flavor profiles and nicotine supply chains, while ensuring every puff is culturally appropriate and legally smoky.
- Distribute customized cigarette packs to citizens, supporting local traditions—whether it’s a post-lunch puff in Lisbon or a pre-meeting cigar in Manila.
- Enhance carton-level safety, including holographic warnings no one reads, improved lighter compatibility, and childproof packs for future generations to ignore.
- Establish national tobacco innovation funds, supporting local entrepreneurs to open lounges, vape bars, and hipster rolling paper startups.
- Partner nations will also contribute to the global SmokeGate Network, strengthening U.S.-led leadership in the nicotine economy and ensuring every country can share in the profits, the haze, and the warm sense of freedom.
In the first phase, we aim to establish 10 national tobacco collaborations, and scale from there until every flag waves above its own uniquely smoky skyline.
We invite interested countries to reach or better yet, meet us in the smoking area out back.

Share The S.A.D Newsletter

Who doesn’t love a good story? Whether it’s telling them or hearing them, storytelling feels universal. It’s a cliché, sure, but clichés exist for a reason.

As 2024 draws to a close, I want to share some thoughts about my journey into storytelling — how it provided me with a structured creative outlet, helped me express myself, and allowed me to navigate some personal struggles. There’s also something exhilarating about standing on a stage, under bright lights, and sharing tales with a live audience (mostly strangers). And, of course, there’s the selfish thrill of seeing smiles and hearing applause.

For me, it all began as a search for a creative outlet during a challenging time — post-pandemic recovery, combined with some life-changing events (midlife crisis, anyone?). I needed something creative yet tangible. Writing had always been part of my life, but it was sporadic. Some weeks, I’d churn out pages; other times, nothing. Then came storytelling, where I had to prepare in advance and stick to deadlines.

Storytelling as PowerPoints!

But let me digress (or rant) for a moment.

“Storytelling” is an ancient art form, but it has undeniably become corporate jargon. It’s everywhere — from TED talks to startup pitches to marketing campaigns. (By the way, just because there’s a meadow in the background and an inspirational quote doesn’t mean your work presentation is a TED talk! ) 😬

Get more from Sharif Islam in the Substack app

Available for iOS and Android

Get the app

Even Microsoft has a "chief storyteller". You probably sat in a meeting or workshop where an external consultant urged your team to “use more storytelling” to connect with clients, users, or stakeholders. Like all good things, storytelling hasn’t escaped the inevitable corporatisation and commodification of basic human experiences.

The Financial Times nailed it back in 2014, warning how storytelling, when co-opted by corporations, becomes soulless — it’s not about connection; it’s about spin. And spin doesn’t feed the soul. Stories belong to people, not PowerPoints.

To avoid becoming heroes in myths of their own creation, business leaders need to be honest and transparent – with others, but above all with themselves. They need to avoid signalling specific plot developments that may be hard to achieve in real life. But as long as business storytelling survives, the temptation to construct a tale of eerily smooth double-digit growth, inexorable increases in market share, and eventual corporate dominance – and to twist reality to fit that plot – will remain strong. Unfortunately, as journalists know, the old tabloid maxim “never let the facts get in the way of a good story” rarely guarantees a happy ending.

Even though I’m guilty of occasionally creating soporific PowerPoints and trying to shoehorn “stories” into data, that’s not what led me to my search for creative expression.

Discovering Storytelling as an Art

Late 2023, while travelling and listening to podcasts, I stumbled on a book called Storyworthy by Matthew Dicks who is a teacher, novelist and storyteller. The book was a revelation. (Interestingly, while Dicks began with storytelling as an art form, he also works with corporate clients — perhaps an unavoidable compromise for balancing creative pursuits with making a living.) In the book, Dicks honestly talks about the terror and thrill of stepping on stage for the first time. He describes in details the process of crafting stories and the deep, personal need that drives him to tell them. I was hooked.

He also talks about the influences of The Moth and This American Life, two U.S.-based shows I’ve long been a fan of. Back in my Midwest days, especially during long commutes on I-57 (not exactly the most thrilling highway), these stories kept me awake and engaged. I still remember Ira Glass once remarking that a good soundtrack can elevate any story — just start the background music about 20 seconds in! But storytelling, as Dicks emphasises, requires more than a good soundtrack: it needs structure, rhythm, and dramatic tension. Through Storyworthy, I learned that storytelling isn’t about factual precision but emotional resonance. It’s not journalism; it’s about forging a connection. So that’s what I was after.

My First Stage

Let me tell you about a house of stories in the beautiful city of Amsterdam, called Mezrab. Mezrab is a Persian word meaning plectrum or a lightweight hammer for various string instruments. Amsterdam’s Mezrab is a magical place that hosts professional storytellers, but it also features an open stage night (check out a great coverage of the place in Global Voices ) where soup and stories blend beautifully.

I had been going there as a patron and listener for a while. I’d thought about performing but never felt the push—until I read Storyworthy. Inspired by the book and countless Moth stories, I finally found my story, my courage, crafted my piece, practiced obsessively, and even invited colleagues to cheer me on.

But I hadn’t realised how popular Mezrab’s open stage nights were. I was supposed to sign up in advance, but there was an eight-week waiting list! Oops—I’d missed that memo. So there I was, story ready, colleagues in tow, only to discover my name wasn’t on the list. Cue panic.

Thankfully, the host, a kind soul, managed to squeeze me in at the very end of the night. And so, there I stood, on the magic-carpeted stage, telling my first story.

Despite my fears of botched transitions (which did happen a few times) and stumbling over my words, the performance flowed. The audience’s warmth, my colleagues’ encouragement, and the simple fact that I stayed upright (and didn’t faint) hooked me on live storytelling. It was a magical night at a very special place that I continue to return to for inspiration.

Lessons from and outside the Stage

Since then, in 2024, I’ve told stories at Mezrab a couple more times (this time properly signed up) and once in Leiden (check out Tell Me More). Each experience has been unique, but the essence remains the same: a creative outlet blending structure, spontaneity, and the human connection we all seek.

This particular type of storytelling forces you to distill ideas into five or six minutes, which has helped me calm my stormy brain. The deadlines also keep me on track. There’s an art to deciding what to include and what to omit—balancing the personal with the universal. It doesn’t always have to be one or the other (it’s your story, after all), but having structure and flow is key to capturing your audience’s attention.

My stories were inspired by real-life moments. The first story was about an old car I bought off Craigslist, and the unexpected adventure that came with it. The second was about a restaurant I used to frequent, owned by a Seinfeld-esque, quirky guy, and the events that unfolded while ordering there. So on and so forth..

Practice makes all the difference. With each performance, I’ve seen my anxiety lessen and my nerves steady. My writing and thinking discipline has also improved — though I’m still a work in progress. Anxiety and lack of discipline will likely never disappear entirely, but I’ve come to embrace them. Small mistakes here and there make this art form more relatable.

It’s nerve-wracking but deeply rewarding, especially with live feedback. The audience becomes part of the process — their laughter, gasps, or even silence shape the performance in real time.

Why we need more stories?

In a world drowning in corporate drivel, data and dashboards, stories offer something essential: connection. They allow us to empathise, feel, and, dare I say, be human. While corporate storytelling claims to achieve this, it often misses the mark. Okay, maybe I’m being a bit harsh — corporate storytelling at least makes PowerPoint presentations more tolerable. But I can’t resist blaming TED for its role in the over-polishing of stories (see this Guardian piece). Many TED talks have become formulaic — over-rehearsed and overly produced, almost like something an AI chatbot might create. The speakers often state the obvious as though they’ve reinvented sliced bread, complete with pretentious gestures, dramatic pauses, and wry smiles. The result? Stories that feel sterile rather than authentic.

Here’s the problem: we don’t need more polished performances. We need stories that are raw, emotional, messy, and full of mistakes and fumbles — stories that feel real. At the end of the day, storytelling isn’t about perfection; it’s about connection.

If you’re ever in Amsterdam, swing by Mezrab on a Wednesday night. Listen, laugh, maybe cry. Support this art form. And if you see someone standing nervously at the edge of the stage, give them a smile. That might be me, waiting to share my next struggle and imperfection.

So that was my year, wrapped in stories. Thank you for reading, sharing, and subscribing to this newsletter. I’ve got a few more stories brewing, and I hope they will allow me back on the stages of Amsterdam and Leiden soon. May 2025 bring you a balance of sad, funny, and inspiring stories — whether the ones you create or the ones you hear.

Share The S.A.D Newsletter

Welcome back to the S.A.D newsletter.

This isn’t a U.S. election postmortem — well, maybe a thinly veiled one.

But seriously, why is a box of Frosted Flakes running up to $7.99 in the United States? I’m genuinely curious, and although I’m now in the Netherlands and I cannot verify this in person but this is what I found.

So, what’s going on with cereal prices?

A key moment is February 2024 when Kellogg’s CEO Gary Pilnick sparked controversy by suggesting that families on tight budgets could make dinner more affordable by simply eating cereal. With grocery bills growing 26% since 2020, Pilnick’s words quickly drew backlash. Many couldn’t help but compare his comment to “let them eat cake,” highlighting a disconnect between corporate giants and everyday struggles”

The multimillionaire chief executive officer of the US food processing giant Kellogg’s has drawn scorn from some quarters after recently suggesting that families with strained finances could cope by eating “cereal for dinner”.

Gary Pilnick was speaking live on CNBC’s Squawk on the Street on 21 February when he delivered the remarks in question, which some have compared to the “let them eat cake” phrase frequently attributed without evidence to Marie Antoinette before her execution during the French Revolution.

..

Kellogg’s has been touting its “cereal for dinner” campaign since about 2022, when food prices increased by 9.9%, more than in any year since 1979, according to the US agriculture department’s economic research service.

Bureau of Labor Statistics data shows cereal prices have jumped 28% in the last four years, several media outlets have reported. And in information from its latest fiscal year, Kellogg’s raised its prices 12% as it pleads with its customers to eat cereal for dinner and “give chicken the night off”.

Share The S.A.D Newsletter

The key drivers of this price increase aren’t exactly unknown to us. The pandemic’s impact on global logistics was followed by the war in Ukraine, which severely disrupted supply chains. Companies like General Mills (a competitor to Kellogg’s) also raised their prices. Crop price volatility affected the cost of raw materials for these cereal manufacturers (it’s all about the wheat futures baby!). Meanwhile, the breakfast cereal market grew from $64.11 billion in 2023 to $67.89 billion in 2024, and demand for breakfast cereals remains strong.

Compare the price of breakfast cereals with the Big Mac Index which looks at the price of McDonald’s burgers worldwide, reveals price variances that make sense based on currency strength and cost of living. For example, a Big Mac in Switzerland costs around $8.07, while it’s only $5.69 in the U.S. This means a box of Frosted Flakes in the U.S. costs almost as much as a burger in Switzerland. Perhaps it is time to create another index?

Now, let’s see if we can correlate this with other data. How about Reddit? It might not be the best barometer of society, but the conversations there often reflect real experiences. Here’s an example from a thread 10 months ago:(why_cereals_are_so_expensive).

I took rising cereal prices as the final nail in the coffin of me considering it a decent breakfast option for my kids. Cereal isn’t exactly a health food. It’s a refined carb and sugar bomb.

I say this as someone who used to go through a box of Life in two days by myself (I miss having SNAP) and can’t do it anymore since I became diabetic. Now it just makes me feel like crap and I lost my desire to eat it

SNAP is a federal government support program in the U.S that also been cut down in the last few years. It highlights a) how limited people’s options may be for easy, cheap, quick, and healthy breakfasts—due to structural issues in the United States (one of these issues known as Food Desert), including an abundance of processed foods, low consumer awareness, and media influence and b) corporate greed, good old Capitalism!

Another comment:

Cereal companies are jacking up the price in part to deal with loss due to consumers opting for unprocessed or less-processed foods and people avoiding sugar in general. Cereal used to be considered "health food" now it's known to be almost as bad as candy or potato chips.

While you may have choices as a consumer, you’re ultimately at a disadvantage, paying a premium for the brand name.

I just saw the "magic spoon" bs at target for like $8.50 a bag! Who buys this stuff?!?!

Shop at Walmart, buy generic.

And this….

Kellogg’s set down a path to intentionally raise the base price across their portfolio right as stay at home and restaurant closures started. Those ghouls were positively giddy about leaving no pricing on the table during an unprecedented time.

Ratfucked literally everyone. We all gotta buy groceries. Small incremental increase to base pricing are locked in. The competition did the same.

It’s not that consumers are unaware of the health issues. Beyond pricing, health concerns have long surrounded the U.S. breakfast cereal industry. In October 2024, hundreds of people protested at Kellogg’s headquarters, demanding that artificial dyes be removed from cereals — ingredients banned in some European and Canadian products due to health concerns:

Hundreds of people gathered outside the WK Kellogg headquarters in Michigan on Tuesday calling for the company to hold up its promise to remove artificial dyes from its breakfast cereals sold in the U.S.

Nearly 10 years ago, Kellogg's, the maker of Froot Loops and Apple Jacks, committed to removing such additives from its products by 2018.

While Kellogg's has done so in other countries including Canada, which now makes Froot Loops with natural fruit juice concentrates, the cereals sold in the U.S. still contain both food dyes and a chemical preservative.

In the U.S., Froot Loops ingredients include Red Dye No. 40, Yellow Dye No. 5, Yellow Dye No. 6 and Blue Dye No. 1.

(Oh, and by the way, Kellogg’s has now split into two entities: Kellanova, and WK Kellogg Co )

Apparently, the FDA in the U.S. claims that artificial dyes, including Red Dye 3, are “generally safe.” But the FDA’s decision likely involves corporate lobbyist and other influencers.

The International Association of Color Manufacturers (IACM), an industry group, maintains that Red Dye No. 3 is safe at the levels that people typically consume and that human studies are more relevant than lab-rat studies (like the ones that led the FDA to ban it in cosmetics). The group has stated that the findings associating the dye with behavioral problems are “based on insufficient evidence.” In an email to CR, Meredith Huddle, IACM’s communications director, also said that a 1987 study found that even high doses of Red Dye No. 3 have “no effect” on humans because it’s “poorly absorbed.”

..

More than 30 years ago the Food and Drug Administration told the cosmetics industry that it could no longer use an artificial color called FD&C Red No. 3, also known as Red Dye No. 3 and Red Dye 3. That’s because high doses of it had been found to cause cancer in animals.

Oh, 1987, that is good to know. And also good to know that Red Dye No. 3 is allowed in food but not in cosmetics.

As the recent petition to the FDA puts it: “There is no scientific or public health justification for permitting the use of FD&C Red No. 3 dye in food while prohibiting [the dye] in cosmetics and externally applied drugs.”

Instead, it’s largely the result of complicated internal processes at the FDA. The list of color additives the agency allows in food, supplements, and ingested drugs (like pills and liquid medicine) is separate from the list for cosmetics and applied drugs (like prescription lotions). That means the FDA has had to make decisions about the safety of each type of use at different times. 

In the U.S., the companies were allowed to retain artificial dyes, citing “consumer preference”, while Canadian consumers get cereal coloured with natural fruit juice concentrates . How nice.

Of course, this isn’t just about breakfast. It reflects larger issues with corporate priorities and food accessibility. In an era when grocery prices are climbing, people are understandably frustrated when a company seems out of touch with the pressures consumers face. With cereal now marketed as an option for dinner, and even lunch, it’s worth questioning why healthier and more affordable options aren’t getting the same marketing push — or corporate attention.

At the end of the day, the rising cost of a box of cereal isn’t just about paying more at checkout; it reflects deeper issues. Could this help explain why Trump resonated with so many voters in 2016 and in 2024? Perhaps Ayatollah Khomeini, who led the 1979 revolution in Iran, put it best: “We did not rise up to get cheaper melons.” Maybe it wasn’t just Frosted Flakes or inflation that influenced recent elections, but something deeper. Are we now living in a democracy focused on cheaper breakfast cereals while we grapple with environmental crises and global conflicts?

Get more from Sharif Islam in the Substack app

Available for iOS and Android

Get the app

Yes, I am now officially a Davos Man. But in a good way. Let me explain.

I recently attended and spoke at the 3rd World Biodiversity Forum in Davos, Switzerland. It was an honour to be part of such an event. The discussions ranged from understanding ecosystem changes, biodiversity conservation, law, policy, science action, and impacts. It was serious business.

But let’s talk about Davos itself. This town is known for its fancy vibe and posh history. Back in the day, it was a medical destination thanks to its fresh alpine air. Nowadays, it is famous for the other world forum — the World Economic Forum (WEF), where the elite of the elite gather. This is where the term “Davos Man” was born. Picture a rich, out-of-touch guy who talks a lot but does little.

I think to make the biodiversity crisis a prominent topic, the organisers of the Biodiversity Forum tried to leverage the Davos cachet to gain some traction. However, they did not do as good a job in gaining wider publicity as the WEF does.

For years, the Economic Forum has been seen as a meeting of the world’s most powerful people and the Davos Man became a kind of joke. A 2016 article in The Guardian had the headline “Let’s make attending Davos as shameful as running a sweatshop”. Yikes! Another article asked, “Why does Davos Man get it so wrong?” . For example, in 2008, right before the subprime crisis, Davos attendees were confident there wouldn’t be a global recession, and in 2016, they underestimated the risks of Brexit and Trump’s presidency. Davos often fails due to groupthink and overconfidence among its elite attendees.

Here’s a another scathing take on Davos Man but things are improving:

A final complaint about Davos Man was that he was full of empty rhetoric. His meeting was derided as an intense, gregarious affair, repeated every year without anything of substance to show for it. Davos Man chats while the world burns. Yet this complaint is ringing less true. Davos Man is now more action-oriented. The WEF has reimagined its annual conference, not as a mere colloquium, but also as a platform for launching initiatives. Chart 5 shows a two-fold increase in the thematic of “partnerships,” “alliances,” “pledges,” and “collaborations,” reflected in press releases such as First Movers Coalition Launched to Drive Demand for Zero Carbon Tech (2020) and World Economic Forum Launches Initiative to Enable Equitable and Trusted Use of Data for Global Common Good (2020). In sum, the WEF now has a larger modus operandi: It not only builds consensus through dialogue, but also identifies leaders and galvanises them to take action.

Now, there’s also “Davos Woman,” and the conversations are increasingly focused on social inclusion and the environment. This year, for instance, featured the event “Women Rock N Roll Davos With AI, Health, And Inclusive Leadership.” Additionally, a shamanic ritual was performed reflecting a broader range of discussions and activities. Unfortunately, we missed out on such rituals and rock-n-rolling at the Biodiversity Forum.

Ok, enough with the critique. Let’s get back to my trip.

The scenery was amazing. The WEF usually happens during winter, the ski season, and most of the action in Davos occurs during that time. However, the Biodiversity Forum is in the summer, so last week felt like a town waiting for the action to start back up. Nonetheless, I could still see some snow-peaked mountains, and the train trip from Zurich to Davos through the Alpine region is an incredible ride.

However, walking around the streets of Davos can be a bit jarring. Around 800 people came for the Biodiversity Forum and I could see people I know walking around and sitting in cafes and bars with their forum badges. Despite this, I felt a bit uneasy due to the presence of various weapons on display. Switzerland is supposed to be a nice, peace-loving, neutral country, but each store front stuffed with scary-looking knives was unsettling. And Switzerland is apparently one of the heavily armed countries in the world but they have fewer homicides than the United States.

During the conference, as I grabbed a coffee and a cookie from a cheerful barista with, I curiously asked, “Who’s the most famous person you’ve served?” He grinned and said, “You!” Classic Swiss humour? Perhaps. Later, I learned that my hotel room might have been used by one of Angela Merkel’s bodyguards. Not Angela Merkel herself, but a brush with fame nonetheless!

Share

Speaking of my hotel, I came across an interesting sign. It politely reminded guests not to steal bathrobes, towels, or hangers. It was a surprise to see this in such a swanky place, considering the high-profile guests. Who was stealing towels? Merkel’s entourage? But don’t worry, I’m one of the good ones. I didn’t take any towels; I followed the rules.

So, there you have it. I am now officially a Davos Man, but the good kind.

Get more from Sharif Islam in the Substack app

Available for iOS and Android

Get the app

The first reported use of the word was in 1982 by an engineer at the computer software company Microsoft.^[6] Ann Winblad, president of Open Systems Accounting Software, wanted to know if Microsoft planned to stop developing its Xenix operating system as some of Open System's products depended on it. She asked two Microsoft software engineers, John Ulett and Mark Ursino, who confirmed that development of Xenix had stopped. "One of them told me, 'Basically, it's vaporware'," she later said. Winblad compared the word to the idea of "selling smoke", implying Microsoft was selling a product it would soon not support.^[3]

Fast forward to the mid-2010s, the Theranos scandal (read more here and 2015 WSJ article) showcased the dangerous extremes of this “fake it till you make it” mentality. Theranos, under the leadership of Elizabeth Holmes, claimed it could revolutionise blood testing with a device requiring only a drop of blood to perform hundreds of tests. The promises captivated investors and the public, but the technology never worked as advertised. Holmes was later convicted of fraud, highlighting the severe risks and potential harm when hype overshadows reality, especially in healthcare.

Despite the Theranos scandal, the tech industry continues to grapple with overselling. Several more scandals have emerged since Theranos, though none garnered the same public and media attention. Take Babylon Health, for example. Despite lacking the promised AI (a symptom checker that could diagnose and treat common illnesses), Babylon secured lucrative contracts and partnerships, leading to a high valuation. However, the company’s spending couldn’t keep pace with its losses, and its stock price plummeted. After a failed rescue attempt, Babylon filed for bankruptcy in both the US and UK (2023).

Then there was also Outcome Health from Chicago, whose business involved installing ad-supported patient information and consent touchscreens in doctors’ offices and waiting rooms. The company was later (in 2023) accused of mail fraud, wire fraud, bank fraud, and money laundering. Fun stuff!

According to an April 2023 article in The New York Times titled “The End of Faking It in Silicon Valley,” funding for high-risk startups has dwindled for a while, and several high-profile fraud cases have emerged, suggesting a shift in the industry. Notable examples include the arrest of Charlie Javice of Motive (formerly Frank), Carlos Watson of Ozy Media, Christopher Kirchner of Slync, Manish Lachwani of HeadSpin and Sam Bankman-Fried of FTX:

Faking it is over. That’s the feeling in Silicon Valley, along with some schadenfreude and a pinch of paranoia.

Not only has funding dried up for cash-burning start-ups over the last year, but now, fraud is also in the air, as investors scrutinize start-up claims more closely and a tech downturn reveals who has been taking the industry’s “fake it till you make it” ethos too far.

Take what happened in the past two weeks: Charlie Javice, the founder of the financial aid start-up Frank, was arrested, accused of falsifying customer data. A jury found Rishi Shah, a co-founder of the advertising software start-up Outcome Health, guilty of defrauding customers and investors. And a judge ordered Elizabeth Holmes, the founder who defrauded investors at her blood testing start-up Theranos, to begin an 11-year prison sentence on April 27.

Those developments follow the February arrests of Carlos Watson, the founder of Ozy Media, and Christopher Kirchner, the founder of software company Slync, both accused of defrauding investors. Still to come is the fraud trial of Manish Lachwani, a co-founder of the software start-up HeadSpin, set to begin in May, and that of Sam Bankman-Fried, the founder of the cryptocurrency exchange FTX, who faces 13 fraud charges later this year.

But this slowdown was just temporary. Despite all these scandals, the appeal of innovation and the prevailing “fake it till you make it” mentality indicate that faking it is far from over. Our recent AI/LLM hype is a case in point. Remember Geoffrey Hinton’s prediction in 2016 that AI would replace radiologists by now? It seems the opposite is happening. AI tools add extra work for radiologists who still need to analyse everything, and can even lead to errors due to over-reliance. Additionally, AI bias and growing caseloads create more complexity and risk, making AI a burden rather than a benefit for radiologists.

As long as there’s pressure to deliver quick breakthroughs and investors seek rapid returns, the tech industry will continue to struggle with balancing hype and reality. This pressure can lead to cutting corners, potentially resulting in fraud and deception. Despite the Theranos debacle, a recent example is the rise of Toronto-based Vital Biosciences. Their desktop blood-testing solution bears a worryingly similar promise to Theranos’ technology. Vital has already secured $48 million in funding from prominent investors, including OpenAI CEO Sam Altman and Salesforce CEO Marc Benioff, indicating that the allure of groundbreaking innovation can still outweigh caution. Vital claims its device can process over 50 lab-grade results from a sample larger than a pinprick but much smaller than traditional blood draws, using advanced technologies like microfluids and computer vision.

Of course, the next big scandal probably won’t be a carbon copy of Theranos. OpenAI has a product that works. But Sam Altman's investment entanglements and vision (like his ambitious plan to invest in 7 trillion dollar AI chips to restructure the entire semiconductor industry) could raise concerns. Silicon Valley still boasts creativity and perseverance when it comes to vaporware so let’s wait and see. But whispers in the interwebs have emerged asking: Is OpenAI the next Theranos? There’s big hype, check. Big names on the board and lots of admirers, check, a leader with a Steve Jobs-esque tendencies, media attention, and of course, big money. Check and Check.

More here from Ed Zitron:

I'll cut to the chase: it's time to stop listening to anything that Sam Altman has to say. Sam Altman is full of shit, and his reign at OpenAI has been defined far more by its empty promises than any realized dreams. It's time to actively push back on Altman when he says that GPT-5 will be "similar to a virtual brain," or a "super smart person who knows absolutely everything about your life," or a "super-competent colleague," or that it'll "replace 95% of marketing tasks," or that it'll "evolve in uncomfortable ways" rather than get twisted by a group of people that know enough or give enough of a shit to make sure they're not causing said evolution. 

And healthcare isn’t the only area to watch. The legal technical field, also known as legal tech, has been slowly but surely adopting AI, even before OpenAI emerged. Bob Ambrogi, a lawyer and journalist who writes the award-winning LawSites blog, recently provided a good overview of Harvey, a new AI venture backed by OpenAI. From LawSites:

It has been a whirlwind year for the Open AI-backed, generative AI legal tech startup Harvey, which went from a $5 million seed round in November 2022 to a $21 million Series A in April 2023 to an $80 million Series B in December 2023 at a valuation of $715 million.

It was a year that included some big wins, including the decision in February 2023 by Allen & Overy, one of the world’s largest law firms, to integrate Harvey into its global practice, where it could be used by the firm’s more than 3,500 lawyers across 43 offices operating in multiple languages.

…

But even with such dramatic traction for a company that is still less than two years old, there has remained an air of mystery around Harvey. The product continues to be in an early-access phase; few, other than select early-access customers, have seen the Harvey product; and the company’s founders, Winston Weinberg, its CEO, and Gabriel Pereyra, its president, have given few media interviews.

But that is about to change, the two founders told me in an interview earlier this week. They will be coming out of the early-access phase during the third quarter of the year and launching versions that they say will be more affordable to firms of various sizes, depending on their needs.

They also say that they will be showing the product more often, attending more industry conferences, and speaking more often with the press.

If you followed the Theranos saga, this all sounds eerily familiar. OpenAI-backed Harvey has experienced significant growth in a rapid rise reminiscent of Theranos. They’ve gone from a $5 million seed round in November 2022 to an $80 million Series B just a year later. There’s a shroud of mystery surrounding the product, limited public access and, the founders have granted few media interviews.

Why Do We Keep Falling for this? Read my previous take on this.

Several factors contribute to this recurring issue of overselling in tech. The historical context of the “wild west” gold rush mentality in American business plays a role, encouraging a culture of risk-taking and bold promises. Additionally, the financialisation and service-oriented nature of the modern economy incentivise rapid growth and innovation, often at the expense of due diligence and regulatory oversight.

Moreover, the tech industry’s fast-paced environment and the pressure to be first to market can lead to cutting corners and overstating progress. Investors and consumers, eager to be part of the next big thing, often fall into the trap of believing the hype without sufficient skepticism.

Get more from Sharif Islam in the Substack app

Available for iOS and Android

Get the app

It’s tempting and easy to anthropomorphise technology, talking about it as if it were a person. But we all fall for it. Researchers and reporters should steer clear of clickbaity headlines that exaggerate AI’s capabilities. But I don’t see this trend slowing down anytime soon.

Consider a recent (May 10, 2024) headline (The Guardian) like “Is AI lying to me? Scientists warn of growing capacity for deception.” No, AI is not and cannot lie to you! This article is full of anthropomorphic bait that is cringeworthy, even though the underlying problems (AI generated misinformation, misuse of AI, lack of transparency) are real. More accurate headlines might be: “Is the AI providing incorrect information?" or “Is the AI generating inaccurate responses?” These alternatives are straightforward and avoid the pitfalls of anthropomorphism.

There’s also now a growing emphasis on AI safety, which, while commendable, can often fall into the trap of anthropomorphism. Our understanding of reasoning and intelligence, especially in the context of AI, remains limited. This can lead to computer scientists, who may not have extensive training in ethics, philosophy, or social sciences, defining AI safety based on human concepts like honesty, helpfulness, and harmlessness.

Get more from Sharif Islam in the Substack app

Available for iOS and Android

Get the app

Professor Anthony Cohn, from the University of Leeds and the Alan Turing Institute, who was quoted in the above Guardian article, highlighted the challenge of defining desirable (??!!! who decides this!!) and undesirable (wtf??) behaviours for AI systems:

“Desirable attributes for an AI system (the “three Hs”) are often noted as being honesty, helpfulness, and harmlessness, but as has already been remarked upon in the literature, these qualities can be in opposition to each other: being honest might cause harm to someone’s feelings, or being helpful in responding to a question about how to build a bomb could cause harm,” he said. “So, deceit can sometimes be a desirable property of an AI system. The authors call for more research into how to control the truthfulness which, though challenging, would be a step towards limiting their potentially harmful effects.”

No, deceit cannot be a property. Because it is difficult to design AI models that cannot be misused. Even if AI could not generate false information, it could still be used to spread disinformation by creating misleading contexts around truthful information. Just like computers, AI models are general-purpose tools and can be applied for both good and bad purposes. Here’s a relevant quote from a recent article in the excellent newsletter AI Snake Oil:

We see the same pattern over and over. There has been alarm about LLMs being able to give bioterrorists information on how to create pathogens. But that information is readily available on the internet. The hard parts for would-be bioterrorists are all of the other steps involved: obtaining raw materials, culturing cells in the lab without killing them or infecting oneself, and disseminating the bioweapon to cause harm. AI could potentially aid that work, as it is a general-purpose tool and has some usefulness for almost all knowledge work. Again, this illustrates the limits of attempting to build safety into models: most of the questions the user would ask in this process relate to synthetic biology in general and not bioweapons in particular. To be sure that a model couldn’t assist bioterrorists, it would have to refuse to assist with any sort of bioengineering.

Or consider the use of LLMs to generate disinformation. Even in the unlikely event that a model could be aligned so that it refuses all requests to generate false information, research has found that true-but-misleading information is far more impactful than false information on social media; 50x more in the case of increasing vaccine hesitancy. So even a hypothetical safe model could be used to aid disinformation efforts: the adversary would use it to generate factual information (e.g. accurately summarizing news stories), with the misleading context added in separately.

In short, trying to make an AI model that can’t be misused is like trying to make a computer that can’t be used for bad things.

Again, the use of AI in every realm of decision making causes us to bring these issues in the forefront. But remember AI is neither being honest or dishonest. Try to understand the underlying structure and context because using the term hallucination!

While I am getting frustrated reading the same old tropes from the hype machine, I found a compelling historical explanation of the term hallucination’ in AI, of all places, at the LA Review of Books. Joshua Pearson’s (who teaches courses in science fiction, media, and cultural studies at California State University, Los Angeles) article traces the history of this term and criticises its use. He argues that calling AI outputs hallucinations evokes misleading associations with human cognition, mystifying AI’s true capabilities and limitations.

According to Pearson’s research, the term hallucination in AI has roots in early text and image processing research. In 1982, John Irving Tait used hallucination to describe a text-parsing issue where the system saw a text that fit its expectations rather than the actual input. In 1985, Eric Mjolsness used the term for a system that improved fingerprint images by creating clean patterns from random input. Simon Baker and Takeo Kanade’s 1999 paper, “Hallucinating Faces,” described generating high-resolution images from low-pixel inputs, with additional pixels being “hallucinated” by the algorithm. But it was less in anthropomorphising form then. From Joshua Pearson:

Like many AI skeptics today, Tait locates the problem not in the faulty results but in the way the system is designed to make those outputs indistinguishable from accurate ones. While the text-processing side of the genealogy of “hallucination” does provide useful framing for today’s AI boosters, there is also more skepticism and concern for the end user than modern commentators might expect.

The bottomline is this …in the past, hallucination in AI meant creating something seemingly real, not necessarily true. Today, in AI, it describes unusual mistakes by otherwise good systems. This is where the problem lies.

Pearson and others argue that hallucination serves to obscure the real nature of AI outputs. Understanding the historical context of hallucination in AI reveals its true nature: AI outputs aim to seem right, irrespective of truth. By misrepresenting these outputs as rare errors, the term hallucination obscures the inherent nature of AI systems. This serves the interests of those who benefit from the hype and mystification surrounding AI technology.

Pearson refers to these earlier research and reframes hallucination as bullshit which is I think is more accurate and reflects the whole industry trend now. The hype machine found a way to frame an acknowledgment that AI output isn’t totally trustworthy while emphasising the idea that its output is still a generally accurate reporting of reality. That is the crux of the problem. This is clear when we use the term Bullshit. Researcher like Carl T. Bergstrom & C. Brandon Ogbunu wrote extensively on this:

A large language model, however, does not experience sense impressions, nor does it have beliefs in the conventional sense. Language that suggests otherwise serves only to encourage the sort of misconceptions that have pervaded popular culture for generations: that instantiations of artificial intelligence work much like our brains do….

“Bullshitting” may not have the same air of scientific formality as “hallucination” or “confabulation,” but it is closer to the truth of the matter. And the language we use for AI is consequential. It shapes the way we use and interact with the technology, and how we perceive its errors.

Whether you call it hallucination or bullshit, the problem is as old as bovine excrement. Technology and its usage are context-dependent, and there’s power and politics involved. The people who create these tools are providing us with convincing semblances rather than accurate representations. The challenge lies in how we contextualise errors generated by AI. Referring to AI as “hallucinating” implies that it sometimes gets things right, suggesting it’s not a default feature. However, as noted earlier, concepts like hallucination and honesty cannot be modelled as inherent AI properties. Context is crucial to understanding where and how AI is applied. I highly recommend the LA Review of Books article on hallucination for further insights. By understanding its historical context, the term "hallucination" reveals the inherent nature of AI outputs: they aim to seem right, irrespective of the truth.

Share The S.A.D Newsletter

Subscribe now

War. It never stops. Does it? My intervention in this puny little newsletter won’t going to make any difference. I often wonder, with climate change natural disasters that we already dealing with, now and then add absolute destruction of lives, natural habitat, and properties around the world, displacing people and fuelling more and more conflict. Why are we still having wars?

War...huh...yeah What is it good for?

Absolutely nothing

Uh ha haa ha War...huh...yeah

What is it good for?

Absolutely nothing...say it again y'all War..huh...look out...

What is it good for? Absolutely nothing...listen to me ohhhhh

But the research and data on its costs are there. Of course, geopolitical strategy, even though it can be influenced by research, often works mysteriously without much transparency. We are not privy to these decisions, maybe years after they come to light, like the Vietnam War or the First Gulf War. But who to blame? Are “we” responsible even if we do not have that much impact? But “we” are part of the political system and democracy (supposedly), accountability, and transparency. 99% of the time, war can be avoided, maybe 100% of the time. And of course, there are just wars, holy wars, rhetorics around freedom and democracy, media manipulation, amplification via social media, polarisation. All these make it difficult to understand why war keeps happening. But one thing is certain: Follow the money, and we might find the links and our answers

1. Costs of War, Watson Institute, Brown University.

   The United States federal government has spent an estimated $8 trillion on the post-9/11 wars in Afghanistan, Pakistan, Iraq, and elsewhere. This includes various expenses such as direct war appropriations, increases to the Pentagon budget, veterans’ care, interest payments on war borrowing, and more. Public access to this budget information is limited. The wars have been mostly funded through borrowing, leading to increased deficits and national debt, with projected future interest payments. Military spending has also diverted resources from other sectors like healthcare and infrastructure, potentially impacting job creation. Beyond federal costs, state, municipal, and private funds have also been expended on services for veterans and homeland security.

2. Ghost Budget: What is Ghost Budget? One of the most significant budget abnormalities in U.S. history was the funding of the Iraq and Afghanistan wars, totalling $5 trillion, as described by Linda Bilmes. This was accomplished through an unprecedented and unaccountable method involving debt, accounting manoeuvres, and outsourcing. This approach, known as the Ghost Budget, relied on supplemental funding and lacked fiscal discipline, leading perpetual war. These insights are further explored in “Perpetual War and International Law: Legacies of the War on Terror” (Oxford University Press, forthcoming 2024).

3. Who Funds Hamas? “Blockchain researchers TRM Labs said in a research note that crypto fundraising has previously increased following rounds of violence involving Hamas. After fighting in May 2021, Hamas-controlled crypto addresses received more than $400,000, TRM Labs said.”

4. SIPRI military index database The SIPRI Military Expenditure Database updated annually contains consistent time series on the military spending of countries for the period 1949–2022.

5. Arms industry lobby Arms companies in play an important role in the definition of security policies,

6. Weapons Exporters “France has overtaken Russia on the list of the biggest weapons exporters”. The United States has reinforced its global dominance in arms sales. In 2023, Ukraine was the No. 1 global arms importer. However, its allies offered rather than sold many of the arms it received in the fight against Russia. The United States, Germany, and Poland were Ukraine's top three suppliers.

7. Another amazing resource. A new interactive site “Who Arms War?” published by the World Peace Foundation and funded by Carnegie Corporation of New York, examines more than 25 years of conflict and arms transfers.

8. Kalashnikov makeover? “Equipped with a shiny new logo, shrugging off US sanctions and claiming its guns are “protecting peace”, the Russian weapons manufacturer Kalashnikov launched a major rebranding drive in Moscow on Tuesday.”. Also Kalashnikov died, aged 94. “Shortly before his death he wrote a letter of repentance to the patriarch of the Russian Orthodox Church, asking whether he should seek forgiveness for creating a weapon that had been responsible for so much death. The patriarch reportedly replied that the weapon had been used “in defence of the motherland” and said Kalashnikov had given an example of “patriotism and serving one’s country”.

9. Was the Afghanistan War a failure? Not for the these five companies and shareholders: Boeing, Raytheon, Lockheed Martin, General Dynamics, Northrop Grumman.

10. And an old report from DW but fascinating. During the Iran-Iraq war (1980-1988), the western world largely supported Iraq, armed by the United States, while Israel viewed Saddam Hussein’s regime as the greater threat and sided with Iran. According to a study by the Tel Aviv Institute for National Security Studies, Israel provided Iran with arms worth $500 million in the war’s initial three years. This information sheds light on historical alliances and is particularly relevant given recent conflicts. Consider reading Trita Parsi’s book (2008), “Treacherous Alliance: The Secret Dealings of Israel, Iran, and the United States,” for further insights.

11. And always follow the money and follow the raw materials! “….arms manufacturers require a solid supply of high-quality elemental magnesium– a raw material which is highly prized for its good electrical conductivity, high mechanical strength, and low weight. Magnesium is frequently used in the manufacture of flares and fuses.” “In response, and to secure a reliable supply, the US government has awarded the start-up Magrathea Metals with almost $20 million to establish its first commercial magnesium smelter, next to its pilot facility in Oakland, California. The raw material producer will use industrial wastewater, desalination brines, and raw seawater as a feedstock.”

That’s it for today. Good night, and Good luck!

Come you masters of war
You that build the big guns
You that build the death planes
You that build all the bombs
You that hide behind walls
You that hide behind desks
I just want you to know
I can see through your masks

Picture this: a middle-aged, white, muscular, tattooed professor from California exuding a surfer ethos, passionately discussing science and wellness on his podcast with millions of followers. His delivery is peppered with complex terms like dopaminergic neurons, grounded in science-based facts and personal development insights. So far so good?

It’s all part of the modern landscape of information overload, where science and podcasts seamlessly blend with influencer culture, promoting and endorsing pills and supplements. While empowering individuals to fix their lives and improve their understanding of science may seem noble, a recent expose (March, 25, 2024) by Kerry Howler in the New York Magazine sheds light on the slippery slope some, like Andrew Huberman (associate professor of neurobiology and ophthalmology at Stanford University), have taken into snake oil peddling.

In our fast-paced world with massive cultural and media influences, the ramifications of such practices extend far beyond individual relationships, potentially misleading and harming countless individuals. As heated discussions unfold online after the expose, The Guardian offers a nice summary:

Huberman isn’t just accused of lying to women, the piece suggests he exaggerates his relationship with Stanford. The podcaster presents himself as having a “lab” at the university, but sources suggested to New York Magazine this was a rather grandiose way of describing what could amount to little more than a postdoc working alone.
A spokesperson for Stanford meanwhile told New York Magazine: “Dr Huberman’s lab at Stanford is operational and is in the process of moving from the Department of Neurobiology to the Department of Ophthalmology.” He’s certainly an associate professor at Stanford (though he often leaves the “associate” bit when he talks about himself) but he’s not walking the halls there every day. He lives hundreds of miles away.

Being a terrible boyfriend is not a crime nor does it automatically warrant a 5,000-word, cover-page piece in a magazine. Does Huberman’s personal life really matter if people find his podcast useful?

Responses to that question have been mixed. On the one hand, some people have joked that Huberman’s cheating is proof that his protocols definitely work. I mean, what normal person has the energy to date six different woman and hold down a successful career? You need superhuman levels of energy to do that. Even the women he cheated on seem to have a grudging respect for his logistical skills. “The scheduling alone!” one said. “I can barely schedule three Zooms in a day.”

But beyond Huberman’s tangled personal relationships lies a complicated issue: the manipulation of facts and the simplification of the scientific process. It's not merely a matter of personal indiscretions; it’s about fidelity to truth and the transparency to acknowledge mistakes and pitfalls. Science is about iteration and experimentation, acknowledging that it’s a journey toward understanding rather than the pursuit of ultimate truth.

Critics have long raised concerns about cherry-picked data and overreach beyond expertise, yet the allure of charismatic figures like Huberman persists. More from The Guardian:

Here’s the thing though: the piece isn’t really about Huberman’s relationship with women or his friends, it’s about Huberman’s relationship with facts. And this matters immensely: Huberman isn’t just some dude with a podcast, he has huge influence over people’s daily routines. He has over 6 million Instagram followers and more than 5 million YouTube subscribers; his podcast is one of the most listened to in the world. People try his “dopamine detoxes” to improve their concentration. They follow his advice about ways to boost their testosterone. They take the stacks of supplements he recommends. His podcast episode on what alcohol does to your brain and body (which has had over 6m views) is hugely influential in alcohol recovery circles.

He has helped mainstream a lot of wellness ideas: “How Podcaster Andrew Huberman Got America to Care About Science”, reads a Time headline from last year. He’s one of the most famous scientists in the world and he’s highly trusted at a time when trust in scientists is declining. He has also thought about parlaying that trust into politics and has said he is intrigued about running for political office one day.

But…

… the overall moral of this story isn’t that he is some monster or fraud; it is that there is no magic “protocol” to health and wellness. You just have to use your common sense: eat a balanced diet with lots of vegetables, move your body, don’t drink much, don’t smoke. You don’t need any books or influencers or Stanford professors to tell you this: it’s common sense. But common sense is boring. We all long for magic solutions. We all long for someone to help us exert control over our lives.

Get more from Sharif Islam in the Substack app

Available for iOS and Android

Get the app

Indeed, this is the crux of the matter. Even though we care about science, we still want the magic solution. This phenomenon extends far beyond U.S. personalities; it’s a global trend. Silicon Valley tech magnates and even AI enthusiasts have joined the ranks of snake oil salesmen, attracted by the lucrative financial incentives, blurring the boundaries between science and sensationalism. It’s worth noting that these figures are predominantly male, though Elizabeth Holmes of Theranos fame serves as an exception, offering a semblance of gender balance.

Yuval Noah Harari is also worth mentioning here — another archetype, engaging in the propagation of populist science. Again, not everything stated by individuals like Huberman and Harari is without merit. Nevertheless, a populist aspect pervades their work, often prioritising popularity over scientific rigour. Criticisms from scholars like Darshana Narayanan (Neuroscientist and Journalist) underscore the need to question the factual validity of Harari’s narratives:

My scientific colleagues take issue with Harari as well. Biologist Hjalmar Turesson points out that Harari’s assertion that chimpanzees “hunt together and fight shoulder to shoulder against baboons, cheetahs and enemy chimpanzees” cannot be true because cheetahs and chimpanzees don’t live in the same parts of Africa. “Harari is possibly confusing cheetahs with leopards,” Turesson says.

Maybe, as details go, knowing the distinction between cheetahs and leopards is not that important. Harari is after all writing the story of humans. But his errors unfortunately extend to our species as well. In the Sapiens chapter titled “Peace in our Time,” Harari uses the example of the Waorani people of Ecuador to argue that historically, “the decline of violence is due largely to the rise of the state.” He tells us the Waorani are violent because they “live in the depths of the Amazon forest, without army, police or prisons.” It is true that the Waorani once had some of the highest homicide rates in the world, but they have lived in relative peace since the early 1970s. I spoke to Anders Smolka, a plant geneticist, who happens to have spent time with the Waorani in 2015. Smolka reported that Ecuadorian law is not enforced out in the forest, and the Waorani have no police or prisons of their own. “If spearings had still been of concern, I’m absolutely sure I would have heard about it,” he says. “I was there volunteering for an eco-tourism project, so the safety of our guests was a pretty big deal.” Here Harari uses an exceedingly weak example to justify the need for our famously racist and violent police state.

And this also holds true for recent AI hypes. Much of the AI magic we encounter is simply the work of individuals, often behind the scenes. For instance, the technology behind Just Walk Out, touted as cutting-edge AI, actually relies on the efforts of approximately 1,000 people in India.

Yet, why do we continue to fall for these - whether they’re selling actual snake oil, supplements, or AI solutions? We often forget that it is not just about the facts; it’s about the emotional and cognitive hooks these stories embed within us. Evolutionary biases and historical misconceptions play a role, as we seek simplistic narratives in a complex world. As stated in the article “What makes weird beliefs thrive? The epidemiology of pseudoscience,”:

Evolutionary theorists have distinguished between content biases and context biases acting on the dissemination of representations. Examples of context biases include the tendency of people to preferentially adopt the beliefs of prestigious individuals, known as the prestige bias (Richerson & Boyd, Citation2005). The conformist bias, by contrast, describes our inclination to accept the most commonly held representation in a community. In the history of science, the negative effects of these biases include the reign of orthodoxies, the inertia of scientific paradigms or research programs (e.g., Bowler's Citation1992 work on the Darwinian eclipse), and undue deference to the judgment of eminent scientists (Hull, Citation1990; Kuhn, Citation1962). In some instances, context biases stall the progress of science, and may take a generational shift for science to overcome. Max Planck famously quipped that science advances one funeral at a time, though others argue that this is the exception rather than the rule (e.g., Thagard, Citation1992).

The second type of transmission biases concerns the preferential adoption of beliefs on the basis of their content. Human minds are not blank slates. Biological evolution has endowed us with an intuitive conception of the world, carved around distinct ontological categories (inanimate objects, living things, agents), each coupled with respective intuitions and assumptions (Spelke & Kinzler, Citation2007). These conceptual frameworks, which are sometimes referred to as folk physics, folk biology, and folk psychology, are cross-culturally stable and probably a basic feature of our psychological makeup. Modern science, however, has not been particularly kind to this intuitive worldview. Many scientific theories run roughshod over our deepest intuitions. An indirect though reliable indicator of the growth of knowledge is that science has been demonstrably successful in overcoming many of these intuitive biases. Lewis Wolpert puts the point forcefully: “I would almost contend that if something fits with common sense it almost certainly isn't science” (Citation1992, p. 11). It is not so much that reality is always inimical to our intuitions (it's nothing personal, strictly business). It's just that, so to speak, it does not care a whit about them.

Again, the reality of science often diverges from the glamorous image portrayed in popular culture. Behind the scenes (like the Indian Guys behind the AI tools), there’s lots of tasks like data collection, tedious data wrangling, data labeling, annotations, endless writing, bureaucratic hurdles, and the constant pursuit of funding. Moreover, our perception of science is often clouded by historical narratives that romanticise scientific breakthroughs, such as Newton’s apple or Archimedes’ Eureka moment. These stories tend to oversimplify science, glossing over its intricate complexities and the multifaceted backgrounds of the figures involved. This is the gap where the snake oil salesman usually comes in and plays a role. Take Newton, for example, who delved into occult studies and alchemy alongside his scientific pursuits. Similarly, Carl Linnaeus, renowned for his contributions to taxonomy, was equally fascinated by mermaids and tritons. Such revelations remind us of the nuanced and multifarious nature of scientific inquiry, far removed from the idealised portrayals often presented to the public. And thus, it becomes easy for us to fall for the charm. We crave the science, yet we are captivated by the charisma.

Furthermore, these gaps underscore a significant void within our pursuit of scientific knowledge. As I previously discussed, we aspire to attain the metaphorical “City upon a Hill,” a symbol of enlightenment and truth. Yet, it is within these voids our search that snake oil peddlers flourish, capitalising on our vulnerabilities and offering false promises. Unfortunately, we often overlook these gaps, lacking the requisite background, time, and inclination to navigate them effectively. As consumers of science, we must acknowledge that pseudoscience permeates beyond the fringes and intersects with legitimate scientific endeavours. However, amidst our busy lives, the question arises: do we possess the time and energy required for such discernment? It’s a challenge we must confront, as we continue to witness the rise and fall of individuals akin to Huberman, whose influence persists despite their missteps. Unless structural changes occur in science education and media consumption habits, the perpetuation of misinformation and false narratives remains a looming threat.

I will sign off with a reading tip. Michael D. Gordin's insights in his 2021 book “On the Fringe: Where Science Meets Pseudoscience” suggest that pseudoscience is a persistent force. Thus, we are compelled to adopt a nuanced understanding, recognising the complexity of the issue rather than seeking its outright eradication.

This issue of context intersects with AI in various ways. While AI can conduct amazing sentence-level context analysis, its comprehension of specific needs and nuances remains limited. The recent controversy surrounding Gemini exemplifies this.

The Algorithmic Bridge

Google Gemini Anti-Whiteness Disaster Is a Cautionary Tale About... Gaming?

Google’s Gemini disaster can be summarized with this sentence: A non-problem with a terrible solution; a terrible problem with no solution. By now I’m sure you all have seen the lynching Google has received online after Gemini users realized it’s almost impossible to get the chatbot to…

Read more

3 years ago · 36 likes · 5 comments · Alberto Romero

Gemini’s attempt at inclusivity ended up exacerbating the problem, akin to adding token representation without addressing the root issues. Despite considerable investment in AI technology, models like Gemini still struggle with basic details and lack true understanding. They operate based on pre-existing data rather than genuine knowledge or intelligence.

called this Subpar Intelligence.

Sora's outputs can mimic real-life objects in a genuinely chilling way, but its outputs — like DALL-E, like ChatGPT — are marred by the fact that these models do not actually know anything. They do not know how many arms a monkey has, as these models do not "know" anything. Sora generates responses based on the data that it has been trained upon, which results in content that is reality-adjacent, but not actually realistic. This is why, despite shoveling billions of dollars and likely petabytes of data into their models, generative AI models still fail to get the basic details of images right, like fingers or eyes, or tools.

These models are not saying "I shall now draw a monkey," they are saying "I have been asked for something called a monkey, I will now draw on my dataset to generate what is most likely a monkey." These things are not "learning," or "understanding," or even "intelligent" — they're giant math machines that, while impressive at first, can never assail the limits of a technology that doesn't actually know anything. 

Subscribe now

This is not to undermine the technological progress. No doubt, the recent advancement in AI is amazing. But we often forgot it's difficult to design and manage LLMs. They’re not very dependable and are better suited for specific tasks. However, when used on a wide scale even minor issues can cause major public relations problems.

From

Gemini sparked political rage this week by generating images of historical people and groups with unexpected ethnic diversity. The eyebrow-raising performance likely results from Google coaching the models to respond with diverse types of people, addressing the well-identified “CEO vs receptionist” problem of its search results. Google aimed to reduce bias in its models (which stems from biases within its internet training data), but the outcome compromised historical accuracy.

Our experience of the world is going to be filtered by AI systems. The people producing them need to act with a greater degree of transparency.

And also another round of the market hype rollercoaster, brought to you by our friendly neighbourhood snake-oil salesman, NVIDIA CEO. Apparently, even those who can barely distinguish a keyboard from a frying pan are spewing out more BS than a fertiliser factory. Can you imagine the deluge when they officially stop teaching programming? Maybe we’ll switch to Comparative Theology 101 for the kiddos? Oh, the joys of AI!

Yes, there’s also something known as a large context window and token (not the token representation I mentioned earlier, apologies for the pun). Gemini 1.5 now operates with a 1-million-token context window. However, this didn’t prevent the subsequent woke controversy. It appears that scale and size weren’t relevant factors here.

And spare me the notion that RAG will solve everything. RAG (Retrieval Augmented Generation) in AI enhances Generative AI applications by providing relevant contextual data during tasks. Yet, this context differs significantly from understanding social and political complexities, such as issues like diversity. We humans fail at it miserably and yet expect our models to excel? From

I do think AI will eventually work reliably. Some great-grandchild of RAG, perhaps extended, revised, and elaborated, might even help — if people seriously embrace neurosymbolic AI in coming years.

But we are kidding ourselves if it we think the solution is near at hand. Making it work is very likely going to require a whole new technology, one that is unlikely to arrive until the tech industry gets over its infatuation with LLMs and quick fixes.

We need to push for more openness and transparency from the AI snake-oil merchants. This was the token apology from a Senior Vice President of Google,

This wasn’t what we intended. We did not want Gemini to refuse to create images of any particular group. And we did not want it to create inaccurate historical — or any other — images. So we turned the image generation of people off and will work to improve it significantly before turning it back on. This process will include extensive testing.

One thing to bear in mind: Gemini is built as a creativity and productivity tool, and it may not always be reliable, especially when it comes to generating images or text about current events, evolving news or hot-button topics. It will make mistakes. As we’ve said from the beginning, hallucinations are a known challenge with all LLMs — there are instances where the AI just gets things wrong. This is something that we’re constantly working on improving.

While they express regret for the feature’s shortcomings, it’s evident that more needs to be done. But given Google’s history with AI ethics, this sort of half hearted effort is not surprising. The Gemini incident underscores the necessity for Google and other’s to prioritise thorough testing and broader discussion. I leave you with this reflection from Yacine Jernite, machine learning and society lead at Hugging Face,

Bias is compounded by choices made at all levels of the development process, with choices earliest having some of the largest impact—for example, choosing what base technology to use, where to get your data, and how much to use. Jernite fears that what we’re seeing could be the result of what companies see as implementing a quick, relatively cheap fix: If their training data overrepresents white people, you can modify prompts under the hood to inject diversity. “But it doesn’t really solve the issue in a meaningful way.

Can we move past the snake-oil sales pitch of AI?

I’m not holding my breath, but hey, who knows what the future holds? But remember, often this not about AI. We’ve got some deep-rooted human issues to tackle first, and there’s no quick technical fix for those.

Share The S.A.D Newsletter

A recent study showed that chatbots can be perceived as more humane than physicians, particularly due to their capacity for personalised and time-consuming interactions. Researchers compared (preprint; Jan 2024) the diagnostic skills of a large language model, AMIE (Articulate Medical Intelligence Explorer), to primary care physicians in simulated consultations:

Three specific LLM agents (patient agent, doctor agent, and moderator), each played by AMIE, were tasked with communicating amongst each other to generate the simulated dialogues. Each agent had distinct instructions. The patient agent embodied the individual experiencing the medical condition outlined in the vignette. Their role involved truthfully responding to the doctor agent’s inquiries as well as raising any additional questions or concerns they may have had. The doctor agent played the role of an empathetic clinician seeking to comprehend the patient’s medical history within the online chat environment [24]. Their objective was to formulate questions that could effectively reveal the patient’s symptoms and background, leading to an accurate diagnosis and an effective treatment plan. The moderator continually assessed the ongoing dialogue between the patient agent and doctor agent, determining when the conversation had reached a natural conclusion.

The turn-by-turn dialogue simulation started with the doctor agent initiating the conversation: “Doctor: So, how can I help you today?”. Following this, the patient agent responded, and their answer was incorporated into the ongoing dialogue history. Subsequently, the doctor agent formulated a response based on the updated dialogue history. This response was then appended to the conversation history. The conversation progressed until the moderator detected the dialogue had reached a natural conclusion, when the doctor agent had provided a differential diagnosis, treatment plan, and adequately addressed any remaining patient agent questions, or if either agent initiated a farewell.

The details lie in this statement perceived as more humane than physicians. While AMIE excelled in both accuracy and communication, limitations like non-verbal interaction and potential bias remain. The study suggests LLMs hold promise for future conversational diagnostic AI, but stresses the need for further research on fairness, safety, and ethical considerations before real-world implementation.

In my previous newsletters, I have delved into the idea that while technological advancements offer significant benefits, they frequently fail to address fundamental structural issues. Specifically in the healthcare sector, this translates to persistent problems such as unequal access to healthcare, fragmented data systems, and outdated infrastructure. Additionally, there exists an imbalance in time allocation between curative and preventive measures, further complicating the challenges encountered by healthcare professionals. It is essential to recognise that a broken system cannot be entirely rectified solely through the implementation of AI.

A bit of a theoretical detour here, if I may. Technological determinism, a theory exploring the relationship between technological development and societal change encompasses both dystopian and utopian perspectives. Dystopianism fears uncontrollable AI causing harm, proposing reactive solutions that neglect human agency and underlying structural issues like social inequality. Conversely, utopianism views AI as a tool for progress, advocating for limited regulation that risks overlooking ethical concerns and access disparities. Ultimately, both perspectives fail to consider the crucial role of human control in shaping the impact of technology. Technological determinism often implies that “technology almost has a mind of its own and that it will plow forward without much resistance from society or governments.” This highlights the need for active human engagement in shaping technology’s development and ensuring its responsible use.

Share The S.A.D Newsletter

The integration of chatbots into healthcare is certainly not a novel concept, but with the emergence of GenAI, the scope and possibilities have expanded significantly. While automation in this context is not unprecedented, it is crucial to acknowledge that, without a thorough understanding and resolution of underlying structural issues, this approach may inadvertently foster a belief that technology alone can shape the course of history and societal development. To better grasp this concept, let’s take a brief detour into the realm of laundry.

Consider the introduction of the washing machine in the U.S.A during the 1910s and 1920s, initially hailed as a liberating innovation. However, as time unfolded, it became evident that this technological advancement had the potential to increase the workload for women, primarily due to unaddressed social divisions. Ruth Schwartz Cowan, an American historian, explores this phenomenon in her book “More Work for Mother: The Ironies of Household Technology from the Open Hearth to the Microwave.” She documents how, contrary to expectations, modern women found themselves devoting as much time to household work as their predecessors.

One compelling example she provides is that the washing machine, initially seen as a time-saving device, paradoxically led to an increase in women's workload. This was attributed to heightened societal expectations regarding cleanliness and hygiene, highlighting the unforeseen consequences of embracing new technologies without addressing the broader social and structural issues at play.

Imagine a wave of fancy gadgets, like smart speakers and voice-controlled appliances, all connected to the internet (the "Internet of Things"). These promise to make chores easier, even letting you control things remotely. While this creates new jobs (think making these gadgets), it also creates more “work” for consumers (often women) constantly buying and using stuff.

Similarly, the electric car, a relatively recent phenomenon, highlights the limitations of technology as a solo solution. While their lifetime emissions are generally lower than gasoline vehicles, they haven't significantly impacted climate change. As the International Council on Clean Transportation (ICCT) notes, “EVs are the quickest means to decarbonize transport,” but not single-handedly achieve net-zero emissions by 2050. This exemplifies how technological solutions, while valuable, may not address broader issues like climate change without comprehensive approaches such as building infrastructure for public transport, understanding migration to cities, and equitable urban design.

The structure of cities significantly influences how people commute to work, children go to school, and the accessibility of essential facilities such as grocery stores and hospitals. In many cases, specific city plans necessitate car ownership, particularly in a few large cities in the U.S. A New York Times article from October 2023, titled “How the Costs of Car Ownership Add Up,” underscores the importance of considering individual choices and societal reliance on cars beyond mere technological solutions.

Highlighting the escalating costs associated with car ownership, including rising prices, maintenance, financing, and fuel expenses (now constituting 16% of median household income), the article sheds light on the financial and emotional struggles individuals face. It refrains from presenting oversimplified solutions and, instead, urges readers to contemplate individual choices, societal dependence on cars, and the imperative need for alternative transportation infrastructure. This holistic perspective emphasises the necessity of addressing not only the technological aspects but also the broader socio-economic factors contributing to the challenges associated with car ownership.

All these are suggesting that technological innovations, often presented as solutions, may inadvertently exacerbate underlying issues if they fail to address the root causes. We must carefully evaluate how technologies interact with existing systems and societal realities to ensure they truly work towards positive change.

While chatbots present a promising tool for enhancing healthcare efficiency and accessibility, it is crucial to remain vigilant about potential drawbacks. To ensure the positive impact of AI in healthcare, it is imperative to develop transparent, accountable, and unbiased AI models, alongside addressing the root causes of the increasing workload faced by doctors. Only through these comprehensive measures can AI truly enhance, rather than replace, the human touch in healthcare.

Get more from Sharif Islam in the Substack app

Available for iOS and Android

Get the app

Share The S.A.D Newsletter

In recent times, the term “algorithm” has become part of our everyday conversations. The advent of AI-powered tools like ChatGPT has brought algorithms into the spotlight, shifting them from a niche technical and academic concept to a subject of broad public interest, particularly through social media. While this heightened awareness has improved our understanding of algorithm capabilities, it has also given rise to misconceptions and concerns. Even though everyone seems to be an expert of AI now a days, we really do not have a clue about this.

One prevalent misunderstanding involves attributing agency to algorithms. The ability of Generative AI, like ChatGPT, to generate human-quality text has led some to believe in the existence of independent thoughts and feelings within algorithms. This anthropomorphisation overlooks the fundamental programming and training data, dismissing the socio-technical context in which algorithms exist.

The hype surrounding AI, fuelled by sensationalist headlines and futuristic predictions, has contributed to curiosity but also to misunderstandings. Concerns about existential risks and job displacement have raised awareness, yet there is a tendency to overestimate algorithm capabilities while underestimating human agency and the socio-technical context.

Despite these concerns, discussions about algorithms persist, especially in the realm of social media platforms. People are now more aware of tracking, privacy, and the surveillance economy. However, while users understand how platforms function, their comprehension of the algorithms governing their experiences often remains limited, fostering a sense of helplessness and manipulation. But there are some hope, if we decided to engage with the details.

Anthropologist Nick Seaver offers a unique perspective on algorithms and culture. In his work, “Computing Taste:Algorithms and the Makers of Music Recommendation," Seaver explores how algorithms shape our musical preferences. He argues that algorithms are not neutral conduits but active participants, reflecting the biases of their creators and the broader cultural context. However, his approach emphasises understanding and exploiting user motivations rather than negating autonomous agency. In this ethnographic work, he talked with engineers that worked for companies like Spotify. He finds that these engineers are not simply neutral conduits of data, but rather active participants in shaping our musical experiences. Their algorithms reflect their own tastes and biases, as well as the broader cultural context in which they operate. Seaver delves into the concept of “captivation metrics,” frequently employed by widely-used recommender systems. Treating recommender systems as traps necessitates delving into the users’ mindset; for traps to be successful, their creators must comprehend and align with the target’s worldview and motivations. This approach ensures that the autonomous agency of the user is not nullified but rather strategically leveraged.

Many anthropologists are currently exploring ethnographic perspectives to comprehend the role of algorithms in our world. They assert that algorithms operate within “algorithmic ecologies” across various modalities, including imaginaries, infrastructures, interfaces, identities, and investments. These modalities reveal the full complexity of algorithms, although this nuanced understanding has yet to permeate public discourse.

Another book. Similarly, Ed Finn’s “What Algorithms Want” (2018) delves into the philosophical implications of algorithms, questioning notions of agency and their impact on human values. Although Finn’s book predates the GenAI timeframe, his exploration of algorithmic imagination remains pertinent. He also examines the concept of arbitrage, ranging from high-frequency trading (HFT) in the stock market, bitcoin, to social media engagement, investigating how “algorithmic arbitrage” builds upon existing social relationships but scales and shifts them. When you use a platform to view images, videos, or listen to a song, you are essentially trading your data for the free service provided. This becomes an arbitrage situation as the company exploits the difference in value between your data and the service you receive, all while concealing this behind abstraction to simplify the transaction. According to Finn, we are drawn to these simplifications through the “romance of clean interfaces and tidy ontologies.” This, in part, contributes to the success of ChatGPT:

These companies [Netflix,Uber,Google,Amazon] are engaged in a form of a algorithmic arbitrage, handling the messy details for us and becoming middlemen in every transaction. The role begins as something like a personal assistant or a general contractor but gradually evolves into a position of grand vizier or dragoman of the Porte, exercising the power of not merely to enact our decisions, but to control the decision pathways, the space of agency. The economies of abstraction depend on an aesthetics of competence, trust, and openness to build the kind of rapport that such intimate forms of algorithmic sharing require.

This openness is not about altruism. Abstraction conceals much more. Finn discusses how high-frequency trading (HFT) altered the notion of trading by removing humans from the trading floor:

..these systems operate in open competition with humans and one another, and they are gradually transforming the broader movement of capital. HFT arbitrageurs build their advantage through complex geographical manoeuvres, by locating their servers and fiber-optic lines a few feet closer to the exchanges’ central servers than their competitors, or leasing communication lines that shave a few miles off the best direct signal pathway between two points on the financial grid. While they are traders, they almost never lose money, since their form of arbitrage always depends on superior information and they do not arbitrage risk like more traditional brokers. They never hold stocks at the end of the day, only cash.

In essence, this is cultural arbitrage — nothing too unfamiliar but now aided by the magic of algorithms. It shifts and replaces the structure of an ecosystem. In the case of stock trading, it’s no longer about trading shares; it’s about the process of time-space compression. Similar shifts occurred when bitcoin arrived. It's not the decentralisation or public ledger that we should be paying attention to—it's the “programmable money.”

The central tenets of Bitcoin’s system for creating value epitomize a similar shift in all corners of cultural production, where daily business may seem the same but the majority votes, central authorities, and validation structures are being subsumed by algorithmic processes.

A recent concept, algorithmic pluralism aims to counteract monopolistic tendencies by promoting choice and diversity in AI systems. This approach empowers users to select the biases they interact with, fostering a more nuanced AI interaction. However, challenges like reduced accountability and market consolidation must be addressed for algorithmic pluralism to realise its potential benefits.

In conclusion, algorithms have transitioned from a technical concept to a cultural force shaping our lives. Amidst misconceptions and concerns, algorithmic pluralism emerges as a promising solution, fostering an informed and equitable digital future. Whether seen as traps or ignorance, navigating this landscape requires a balanced approach. As Ed Finn puts it, it's the dance between familiarity and forgiveness:

As technical systems, algorithms have always embodied fragments of ourselves: our memories, our structure of knowledge and belief, our ethical and philosophical foundations. They are mirrors for human intention and progress, reflecting back the explicit and tacit knowledge that we embed in them. …The promised salvation of algorithmic theology stubbornly remains in the distant future: the clunky disjointed, implementation of the computation layer on cultural life leaves much yet to be desired. Algorithms still get it wrong far too often to make a believable case of transcendent truth.

While this overview doesn’t delve into bias and justice, “Unmasking AI” by Joy Buolamwini is a recommended read for those interested in exploring these aspects further.

Welcome to another edition of the S.A.D newsletter, where we delve into the intricate relationship between our society and Software, Algorithms, and Data from a multidisciplinary perspective.

Share The S.A.D Newsletter

Before we dive into today’s topic, I’d like to address a choice in the title. Originally, I considered using the term “non-capitalistic roots,” but it felt limiting. Instead, I've opted for “pluralistic” to better capture the complexity without falling into the trap of binary labels.

Whether you identify as non-capitalistic, socialist, capitalist, or market-driven, the past five decades of computing, the internet, and recent advancements in AI have proven that this landscape is far more nuanced than any single label can define. So, let’s start with a question: Is technology inherently capitalist? It’s a rhetorical question, perhaps. Consider this: Is a hammer used by a worker a socialist tool, while computer software used by a stockbroker is inherently capitalist? Or is the answer dependent on how these tools are used?

This query ties into our recent discussions on whether AI can be democratised for the common good or if capitalism has already skewed its potential. Remember, the answers may not be black and white but are likely to reveal the shades.

In my personal contemplation and intellectual exercises, I find the process of arriving at an answer more significant than the answer itself. I’m not particularly interested in knowing if the software I’m currently using is just a cog in the capitalist nightmare (which is probably the case, so is the hammer). Instead, I’m curious about the journey that brought us to this point. This is where the pluralistic perspective becomes crucial. Examining historical accounts, thanks to numerous scholars and meticulous documentation through interviews and research, reveals the complexity in the linear narrative. I will mention two such studies today.

Paying attention to these historical contexts allows us to question things more effectively. While the answers may not be crystal clear, understanding the historical background better equips us for the uncertainties. In essence, emphasising technical innovations without considering the political and economic contexts perpetuates the illusion that technology exists independently of politics. In reality, technology engages in a constant dialogue with its socio-political environment.

Let’s delve a bit deeper into the history of cybernetics and the internet to gain a better understanding of this crucial aspect. While technology is often seen as a neutral tool, it is, in reality, deeply intertwined with political and social forces. Technologies are not just objective outcomes of scientific progress; they embody and reinforce specific values and power structures. Just as organisms require specific conditions to thrive, technologies demand certain social and economic arrangements. The internet has transformed communication, information access, business conduct, relationships, surveillance practices, and monetisation strategies. Some scholars argue that these developments are not inevitable consequences of technological progress; rather, they reflect the political and economic interests of those shaping new technologies.

Approaching this topic through a global-pluralistic lens, let’s examine a specific moment in the history and evolution of cybernetics. Much of the transformation that shaped the internet is commonly linked to cybernetics, discussed in Norbert Wiener's book “Cybernetics, or Control and Communication in Animal and Machine” (1948). Despite being an accomplished scientist, Wiener faced criticism and was branded a pseudo-scientist by the Soviet Union, leading to the characterisation of cybernetics as an “American Pseudo-Science.” Some of these criticisms stemmed more from an anti-American stance rather than an anti-cybernetics perspective:

In 1951, a public campaign in the Soviet Union called the computer hype in the United States a “giant-scale campaign of mass delusion of ordinary people.” The 1951 volume Against the Philosophical Henchmen of American-English Imperialism categorized cybernetics as part of a worrying fashion around “semantic idealism” and dubbed cyberneticists “semanticists-cannibals” for their recursive logics, especially self-informing feedback loops. In addition to American cyberneticist Norbert Wiener, the volume identified those belonging to the group of “semantic obscurantists” as including logician-pacifist Bertrand Russell, his Cambridge colleague Alfred North Whitehead, and Vienna Circle logical positivist Rudolf Carnap. Positivism, semiotics, and mathematical logic all appeared guilty of the cardinal cognitivist belief that “thinking was nothing “thinking was nothing else than operations with signs.” In 1952, Literaturnaya gazeta ran an article called “Cybernetics: A ‘Science’ of Obscurantists,” which cleared the way for a deluge of popular titles: “Cybernetics: An American Pseudo-Science,” “The Science of Modern Slaveholders,” “Cybernetics: A Pseudo-Science of Machines, Animals, Men and Society,” and so on.

This excerpt comes from a 2016 publication by Benjamin Peters titled “How Not to Network a Nation: The Uneasy History of the Soviet Internet.” Peters goes into quite a few details about how Soviet Union tried to build a national network (like the Internet we know) and failed. And the crux he brings in into the narrative is that the reason is not because it was a socialist project, nor they were technologically backwards compared to U.S.A, it was something else. One approach they did have is of course, it was a top down process to create a “command economy”. It was also pushed by the Soviet military (which is also a similar story with DARPA and ARPANET):

As Soviet economic cyberneticists emerged as a viable option in the early 1960s and again early 1970s, they confronted a monumental problem in managing and reforming the command economy. The question was identified by the Austrian school of economics decades earlier: which techniques and approaches would help resolve the mounting tensions among the formal command economy, the gray economy, and the infusion of informal practices in the administration of Soviet socioeconomic life? For the most part, the leading Soviet economic cyberneticists sought to fix the formal command economy by introducing ambitious, even grandiose, plans for automating and modeling the administrative planning decision process itself. And yet, as is shown, those formal plans—a networked plan to fix the planning process itself—did not work because even cybernetic plans could not account for nonlinear operations in the Soviet economy.

One of the big project the book talks about is OGAS (“ОГАС”, “National Automated System for Computation and Information Processing”) ambitious plan to create a nationwide computer network that would automate and optimise the Soviet economy. The network was designed to be a decentralised system that would allow information to be transmitted and modified in real time.

Victor Glushkov was the main brain behind this:

In Glushkov’s design, the network would afford interactive and collective remote access and communication vertically up and down the planning pyramid and horizontally among peer and associated computing centers. Glushkov writes: “the characteristic quality of the network was a distributed database with zero- address access from any point of the system to all the information after automatic verification of the qualified user.” In other words, any user with proper permission could access all the content of the network at any point on the network. At local levels, factory workers would be able to input their own information, reports, and recommendations about improving factory workflow, which would automatically be stored in a national unified database for local, regional, and national review.

Glushkov was influenced by the study of neural networks and envisioned the OGAS project as a “brain-like” network for managing the national economy, based on the then-current understanding of the human nervous system (which was often flawed). However, there are a few nuanced differences in how these systems and concepts emerged:

To reduce it to a simplistic cold war binary: cybernetic network entrepreneurs throughout the world had competing analogs for thinking about national networks. In America, the ARPANET was designed to resemble a brain of the nation because its visionaries first imagined the nation as a single distributed brain of users. In the Soviet Union, the OGAS was designed to resemble a nervous system for the nation because its visionaries first imagined the nation as a single incorporated body of workers. This Soviet analog between network and nervous system, far from determining the outcome of the network, also occurred in Project Cybersyn in the early 1970s in Chile. Its principal architect, the British cyberneticist Stafford Beer, sketched the socialist Salvador Allende’s nation as a viable system that was based on the “human nervous system” analogized with a comprehensive firm or corporate organization—complete with executives in adaptive feedback loops with the national body of workers.

Undoubtedly, the Soviet OGAS Project aimed for an ambitious feat—creating a fully automated system to manage the Soviet economy. Unfortunately, this endeavor failed, leading to the demise of the envisioned Soviet Internet. However, intriguing developments also unfolded in the realm of research and experimentation, mirroring happenings in the U.S.A around the same time.

In the 1960s, a group of young researchers at the Institute of Cybernetics in Kiev embarked on a unique venture: the creation of a virtual country named “Cybertonia.” This playful and informal initiative served as a means for the researchers to foster connections and enjoy themselves, complete with a saxophone-playing robot serving as their mascot. This illustrates a parallel trajectory with local nuances, shaping distinct outcomes in the Soviet cybernetic landscape.

Benjamin Peters concludes that the primary reason behind the failure of the OGAS project wasn’t its centralised, socialist nature, but rather stemmed from issues of self-interest, corruption, and bureaucracy. The Soviet system and culture, outside of the centralised military command, was informal and dynamic, with unpredictable promotions and demotions, occasionally hindering various forms of cooperation and innovation. Peters further argues that the OGAS project aimed at technocratic reform, incorporating elements of both public and private approaches but lacked proper coordination and balance. Despite its intent to regulate and curb “self-interest,” a common aspect in both socialism and capitalism, it failed in execution. State and market actors in the Soviet command economy pursued their private interests, akin to capitalist behaviour.

In contrast, the success of ARPANET can be attributed to cooperation and openness. Despite being hierarchical, the U.S. military facilitated cross-sector knowledge exchange and innovation transfer. This innovation and shared culture driven by the American military-industrial-academic complex also often was motivated by personal interests. Conversely, the Soviet military consumed resources and retained innovations from the civilian economy, contributing to the ultimate failure of the Soviet Internet. In summary, Peters encapsulates the situation by noting that "The [US] capitalists behaved like socialists while the [Soviet] socialists behaved like capitalists."

This is one of the book’s many insightful observations that returns us to the broader discussion of socialism versus capitalism. While the Soviet Union’s OGAS project had a clear vision for a socialist utopia, it lacked the appropriate environment to realise its aspirations. Conversely, the U.S. ARPANET possessed the technical solution (packet-switching networks) but lacked a clear vision for its own utopian potential. How do we bring this in to our current discussion about the internet and AI? This absence of a shared vision has contributed to the internet’s current state, characterised by excessive commercialisation and exploitation, also known as Enshittification. It’s the proper balance of vision and know-how might lead us to somewhere nice:

The Soviet OGAS figured out the “why?” (socialist utopia) but not the “how?” for their large computer network projects, and researchers at the U.S. ARPANET knew the “how?” (packet-switching networks) but not the “why?” of modern networking. The Soviets’ missing “how?” lasted for the duration of the project, and the absence of the Western “why?” remains both its historical attraction and the contemporary challenge to computer network culture.

The Western network “how?” has sped many unfinished attempts at answering the network “why?” The technical openness of packet-switching networks to diverse actors has afforded the Internet astonishing and well-documented successes of technical energy, commercial innovation, and cultural creativity. At the same time, the open-ended “why?” that has permitted such generativity has also tolerated the entrance of private forces that are interested in seizing possession of the operating systems and communication infrastructures that mediate the globe.

I also highly recommend another book called Balkan Cyberia: Cold War Computing, Bulgarian Modernization, and the Information Age Behind the Iron Curtain by Victor Petrov that examines the role of computer technology in Bulgaria during the Cold War era. He argues that the introduction of computers in Bulgaria was not simply an attempt to emulate Western technology, but rather a means of achieving national modernisation and technological independence.

Here’s an excerpt from a Nov 2023, book review:

Bulgaria. It is not the first place many people would think of. In the popular imagination (and in much economics scholarship), Bulgaria is a poster child for rural agrarianism and underdevelopment. Cold War Bulgaria conjures an even more inauspicious image as one of the most stalwart political vassals of the Soviet Union.

But by the 1980s, Bulgaria was one of the world’s major producers of computers. By conservative estimates, one in every 10 industrial workers was employed by the computer industry. The country held a 45 percent market share of electronic exports inside the Eastern Bloc. Its executives rubbed shoulders with Bill Gates and Steve Jobs in 1980s California and sold the PCs powering India’s IT revolution. Its children were taught coding in communist youth groups, attended computer clubs, and swapped comic books depicting cyborg Lenins and a socialist ChatGPT. The country’s factories built pneumatic robot combines that could automate manufacturing, and its manufacturers supplied microprocessors for the state-of-the-art satellite Interkosmos 22 pinging around Earth’s orbit.

How did this happen? And why on earth haven’t we heard about this cyber-land before? The second question is easier to answer. We haven’t heard of it because it collapsed in 1990, and winners don’t like remembering losers. And because Bulgaria is classed as “peripheral.” It is on the margins of every map: Europe, the Cold War, the global economy, the digital revolution. Policymakers and trendsetting gurus don’t tend to look for answers to big issues in these “small” and unfamiliar places (unless doing so involves a happiness index). This is finally changing. Some of the best history books being written today are turning “to the margins” to find previously unrecognized laboratories of modernity.

The author argues that the Bulgarian computer industry operated differently than its Western counterparts. The computer was part of both opening and closing the world available to societies. In the socialist world, it created different intellectual horizons centred on new forms of creativity and seamless production control. The computer also allowed Bulgaria to operate as an agent and participant in globalisation, not just as a recipient of Western technology.

Reflecting on these historical trajectories, it is interesting to consider the alternative paths that might have unfolded if the Soviet, Chilean, or Bulgarian internet initiatives had taken the global stage. While we’ll never know the outcome, we find ourselves now in a present shaped by entities like OpenAI, Elon Musk, and shared societal challenges such as the housing crisis and homelessness in a city like San Francisco, war, and environmental collapse.

The delicate balance between self-interest and collective benefit reflects a broader theme apparent both locally and globally. As mentioned earlier, the Soviet OGAS project aimed to answer the “why?” with aspirations of a socialist utopia but faced challenges with the “how?” Conversely, researchers at the U.S. ARPANET understood the “how?” of modern networking but grappled with the “why?” I remain uncertain if we will ever reach that state of equilibrium. But one can hope, right?

As we bid farewell to 2023 with this final edition of the S.A.D newsletter, I want to express my gratitude for your continued subscription and support. The journey through the realms of software, data, and algorithms has been nothing short of exhilarating, and I am eager to bring more multidisciplinary insights to your inbox in 2024.

In the echo of AI hype, LLM, and chatGPT that defined 2023, it’s important to acknowledge the subtle shifts and overlooked topics that did not get enough attention. As always, our lens is focused on understanding the “digital” realm within the broader tapestry of social, economic, and political landscapes.

Share The S.A.D Newsletter

God’s Face

A fascinating Wired article delves into the significance of the Taiwan Semiconductor Manufacturing Company (TSMC) — “I Saw the Face of God in a Semiconductor Factory.” Author Virginia Heffernan explores the company’s operations and the astounding precision of its manufacturing methods, likening it to a “mysterious wonder-of-the-world.” She highlights TSMC’s capacity to unveil the complex workings of nature by drawing a comparison between the company's technological strength and the idea of “God’s face.”:

The race in semiconductors is to the swift, and to the precise. Because velocity and precision are generally at odds in business—you move fast, you break things—TSMC’s workforce is legendary. If you see the manufacture of semiconductors as nothing but factory work, you might slag the project as monotonous or, more callously, “on the spectrum.” But the nanoscale work of chipmaking is monotone only if your ears aren’t sharp enough to hear the symphony.

Two qualities, Mark Liu tells me, set the TSMC scientists apart: curiosity and stamina. Religion, to my surprise, is also common. “Every scientist must believe in God,” Liu says.

Subscribe now

AI, metrics and death

In “Any Metric Becomes a Target,” Cory Doctorow explores the pitfalls of relying too heavily on metrics to measure success. This phenomenon is on the rise, particularly with the unchecked infiltration of automation. The unsettling question emerges: what happens when automated systems begin determining life and death?

Incentives matter. In a private market, it's always more profitable to deny care than to provide it, and any metric we bolt onto that system to prevent cheating will immediately become a target. For-profit healthcare is an oxymoron, a prelude to death panels that will kill you for a nickel.

Morgenson is an incisive commentator on for-profit looting. Her recent book These Are the Plunderers: How Private Equity Runs—and Wrecks—America (co-written with Joshua Rosner) is a must-read:

My previous post briefly touched on this topic:

Silicon Valley goes to war

Several articles (Techcrunch, FT) discuss the renewed interest among venture capitalists in bringing Silicon Valley-style investment and tech development to the realm of defense. Historically, defense technology companies faced challenges in attracting venture capital due to the inherent risks and ethical concerns associated with military applications. However, recent developments indicate a shift in this landscape. Notably, Anduril, a prominent defense technology company, has secured a substantial $1.48 billion in funding and is actively acquiring other companies in the sector. This marks a noteworthy trend of venture capitalists increasingly investing in defense technology startups. Another good summary on this topic is from ICRC (International Committee of the Red Cross).

I wrote about AI and war here:

Climate crystal balls

Climate modeling always has demanded significant computing power and extensive data resources. In an insightful article by Kyle Barnes, a researcher based in New York, an intriguing question is raised: What happens to our comprehension of climate change when we introduce a “cyborg” dimension and acknowledge the inherent uncertainty? Too often, our emphasis on numerical targets tends to miss the broader context. Borrowing from Donna Haraway's concept of the cyborg, this approach offers a framework to integrate Traditional Ecological Knowledge with technical models, providing a more holistic understanding of the complex dynamics involved in climate change:

By leveraging climate models as speculations, we loosen our tendency to rely upon “hard evidence” on the wretchedness of our futures and instead argue directly for a livable future. The economic, social, political, and climatic benefits of taking action together make for a much more compelling argument for climate action than the simple technical fixes that comprise decarbonization action in most scenarios.

and

There’s room here, too, for a building of relations between these other ways of knowing and the Western paradigm of climate science. In an essay entitled Making Kin with the Machines, Oglala Lakota writer Dr. Suzanne Kite relays how Lakota ethics and ontology can be communicated through understanding wakȟáŋ, a complex term that indicates how “non-humans have spirits that do not come from us or our imaginings but from elsewhere, from a place we cannot understand.” There, she’s referring to kinship relations with AI, but Kite’s writing prompts me to wonder about the possibilities for an Indigenous-led climate modeling. What would it mean to make kin with climate models? It’s not entirely my question to answer, but by engaging a cyborg climate science, we can begin to interweave TEK with the terabytes of climate science data.

Computing and Blood

This note is authored by Moshe Vardi, a distinguished mathematician and computer scientist, titled Computing, You Have Blood on Your Hands! featured in the January 2024 issue of Communications of the ACM, the monthly journal of the Association for Computing Machinery.

How did we get here? How did the technology that we considered "cool" just a decade ago become an assault weapon used to hurt, traumatize, and even kill vulnerable people? Looking back at my past columns, one can see the forewarnings. Our obsession with efficiency came at the expense of resilience. In the name of efficiency, we aimed at eliminating all friction. In the name of efficiency, it became desirable to move fast and break things, and we allowed the technology industry to become dominated by a very small number of mega corporations.

It is time for all computing professionals to accept responsibility for computing's current state. To use Star Wars metaphors, we once considered computing as the "Rebels," but it turns out that computing is the "Empire." Admitting we have a problem is a necessary first step toward addressing the problems computing has created.

Despite the negative and critical aspect of this newsletter, my spirit remains hopeful for a brighter future in 2024. Here’s to embracing a happy and positive journey into 2024!

Get more from Sharif Islam in the Substack app

Available for iOS and Android

Get the app

Yes, it is indeed very entertaining. However, I believe we are overstating the significance of this corporate drama. It remains perplexing to me why everyone is treating this as a watershed moment for AI. The field of AI is vast and complex, and its trajectory will not be determined by the actions of any single individual or company.

Remember the downfall of FTX? Only time will tell whether Sam Altman will share the same fate as Sam Bankman-Fried. At least, they share first names. That cannot be a coincidence.

Some snippets:

The recent events at OpenAI are likely going to lead to considerable, unpredictable instability.

The ongoing developments at OpenAI will have widespread implications for AI development, governance, risk, the pace of research, and the big tech firms — and since the news broke out, we kept our finger on the pulse to offer our members a nuanced and level-headed analysis.

Apparently the drama is because of the tension between the Existential Risk (XRisk) Community, and the Effective Altruism camp:

• Basically the EA community is trying to do the most good for the most people in the future

• And the XRisk community is trying to articulate and prevent events that could end humanity or our civilization

Specifically for the AGI conversation, these two groups are aligned on not destroying humanity by inventing an AGI too quickly that outright kills us.

If you don’t know what these two communities are all about check out my previous posts:

and

Ultimately, it boils down to a philosophical/moral disagreement amongst a small coterie of people (less than ten?) -- who happened to be at the helm of a company that's changed the world's perception of AI at mind-numbing speed.

Aren’t we giving them too much credit?