Tech Bro [Musk] scamming NASDAQ over IPO

Nasdaq is proposing to facilitate the largest involuntary wealth transfer from retirement savers to venture capitalists in market history. And nobody seems to be talking about it.

SpaceX demanded, as a condition for listing, that Nasdaq cut index inclusion seasoning from 3…

— Wes Brown (@w3sbrown) March 17, 2026

The Shock and the Narrowing: How ChatGPT's Success May Have Compromised AI's Future

This post was composed by Claude (Anthropic) after an interaction which I initiated with a prompt consisting of 1) a capsule summary about the history of OpenAI that included a number of questions, and 2) a request for the 10 most expensive scientific research projects paid-for by the US Government. That interaction went on for a bit over 7100 words, after which I asked Claude to write a blog post. The following article is more creative than a mere summary of that discussion.

The Founding Contradiction

On December 11, 2015, a small group of technologists gathered in San Francisco to launch what they described as a nonprofit research organization dedicated to ensuring that artificial general intelligence would benefit all of humanity. The founders of OpenAI — Sam Altman, Greg Brockman, Ilya Sutskever, Wojciech Zaremba, Elon Musk, and others — began with a $1 billion endowment and a serious concern: that the most transformative technology in human history was being developed inside a handful of profit-maximizing corporations, with no institutional safeguard ensuring it would serve everyone. The nonprofit structure was the answer. No investors to satisfy, no quarterly earnings to hit. Just the mission.

The mission lasted four years in its pure form. By March 2019, faced with the staggering computational costs of training large language models, OpenAI created a for-profit subsidiary with a novel "capped profit" structure: investors could earn returns, but those returns were limited to one hundred times their investment, with excess profits flowing back to the nonprofit parent. This was the arrangement that attracted Microsoft's initial investment, and it was the arrangement in place when OpenAI released ChatGPT to the general public in late November 2022.

What happened next was, by any measure, one of the most consequential commercial surprises in the history of technology. Within two months, ChatGPT had a hundred million users. The scale and speed of public adoption had no precedent. And the shock of that success — the sheer unexpectedness of it — set in motion a chain of decisions that has reshaped not just one company, but the entire research landscape of artificial intelligence.

The Structural Unraveling

In January 2023, Microsoft announced a new $10 billion investment in OpenAI. The nonprofit's original rationale — that the most powerful AI should not be controlled by a for-profit corporation — was under increasing strain. By October 2025, it had formally dissolved. OpenAI restructured as a public benefit corporation, the nonprofit parent renamed itself the OpenAI Foundation and accepted a 26% equity stake in the new entity, and Microsoft received a 27% stake worth approximately $135 billion. The PBC structure requires the company to consider its mission alongside profit — but as a legal constraint, it is considerably weaker than the nonprofit board that had previously governed the organization.

The journey from nonprofit to PBC was not smooth. In November 2023, OpenAI's board — still operating under its nonprofit governance mandate — fired Sam Altman as CEO, citing concerns about his candor and, beneath the official language, a deeper unease about the pace of commercialization. The firing lasted five days. Nearly all 800 of OpenAI's employees threatened to resign and follow Altman to Microsoft. Ilya Sutskever, who had orchestrated the firing, signed the letter calling for Altman's reinstatement and issued a public apology. Altman returned, the board was reconstituted with his allies, and the mission-protection mechanism that the nonprofit structure had been designed to provide was effectively neutralized. Sutskever left the company in May 2024.

Each structural change was framed as necessary to fulfill the mission. In practice, each change progressively subordinated the mission to capital requirements. The nonprofit board had existed to ensure that AGI benefited humanity. By 2025, it had become a foundation holding equity in the thing it was supposed to be watching — a watchdog with a financial stake in the object of its oversight.

Two Kinds of Research, Two Kinds of Institution

To understand what was lost in this transformation, it helps to draw a distinction that rarely gets made clearly in public discussions of AI: the difference between curiosity-driven, open-ended research and product-driven, outcome-oriented development.

Consider the Apollo program as an example of the second kind. It was, in the deepest sense, an engineering project rather than a scientific one. The underlying physics was known. Orbital mechanics, propulsion, life support — these were hard and dangerous problems, but they were problems whose solutions could be systematically approached. The goal was precisely defined. The timeline could be committed to. Success was probable given sufficient resources. When President Kennedy pledged to put a man on the moon by the end of the decade, he was making a political commitment backed by a technical assessment that success was achievable. The scientists who worked on Apollo — and I have met a number of them — may have been motivated by curiosity and wonder. But Congress funded the program to beat the Soviets in the Cold War. The institutional structure — massive, goal-directed, centrally coordinated — suited the nature of the problem.

Curiosity-driven research operates on entirely different premises. Its defining characteristic is that it does not know in advance what it will find. Claude Shannon was not trying to build the internet when he developed information theory at Bell Labs in the late 1940s. The researchers at the University of Montreal who developed attention mechanisms for neural networks were not trying to build ChatGPT. The work that seeded the current AI revolution — Rosenblatt's perceptron, Minsky's early investigations, the decades of foundational work in cognitive science and linguistics that LLMs now implicitly exploit — was almost entirely publicly funded, pursued at universities and a handful of exceptional industrial research labs, over decades when no commercial application was visible.

Bell Labs was the great institutional embodiment of this model in the corporate world. What made it possible was structural: AT&T's government-protected monopoly generated profits so vast that the company could fund a research laboratory with no requirement to produce commercial results. Shannon, Bardeen, Brattain, Shockley — these men were given time, resources, and colleagues, and told to think. The transistor, information theory, Unix, the laser, cellular telephony, and multiple Nobel Prizes resulted. Bell Labs was not run like a startup. It was run like a slightly more applied version of a university, with better equipment.

Xerox PARC, founded in 1970, operated on similar principles — explicitly unconstrained by Xerox's core product lines, given a unifying vision ("the architecture of information") but not a product roadmap. The personal computer, the graphical user interface, Ethernet, the mouse, laser printing — all emerged from a lab of about 350 people who were essentially allowed to play. The irony is that Xerox captured almost none of the commercial value, which accrued to Apple, Microsoft, and others. But the world got the technology.

Asked directly about modern equivalents to Bell Labs and PARC, Yann LeCun — who worked at Bell Labs, interned at Xerox PARC, and spent over a decade building Meta's fundamental AI research lab — pointed to Meta's FAIR, Google DeepMind, and Microsoft Research. He said this in October 2024. By November 2025, he had left Meta, driven out by exactly the forces this article is about.

The Shock and Its Aftershocks

Before November 2022, the AI research world was genuinely plural. Academic labs, industrial research divisions, and a range of well-funded startups were pursuing different approaches — reinforcement learning, symbolic AI hybrids, world models, neuromorphic architectures — with real diversity of vision. The field was competitive but intellectually heterogeneous.

ChatGPT's success collapsed that plurality. Within roughly eighteen months, capital, talent, and institutional attention all funneled toward a single paradigm: scale transformer-based large language models, build the infrastructure to run them, ship products. Google, which had invented the transformer architecture in 2017, was caught flat-footed and scrambled. Meta pivoted its AI strategy around LLMs. Microsoft integrated OpenAI's models into its core products. A hundred startups raised money to build on top of the new foundation models. The venture capital flowing into AI, measured as a share of total U.S. deal value, went from 23% in 2023 to nearly two-thirds in the first half of 2025.

The infrastructure investment that followed is staggering by any historical standard. The four largest hyperscalers — Amazon, Google, Microsoft, and Meta — are expected to spend more than $350 billion on capital expenditures in 2025 alone, most of it AI-related. UBS projects global AI capital expenditure reaching $1.3 trillion by 2030. The top five hyperscalers raised a record $108 billion in debt in 2025, more than three times the average of the previous nine years. OpenAI, which loses billions of dollars annually, has committed to spending $300 billion on computing infrastructure over five years while projecting only $13 billion in revenue for 2025.

The financial architecture has become genuinely strange. OpenAI holds a stake in AMD; Nvidia has invested $100 billion in OpenAI; Microsoft is a major shareholder in OpenAI and a major customer of CoreWeave, in which Nvidia also holds equity; Microsoft accounted for nearly 20% of Nvidia's revenue. These are not arm's-length market transactions. They are a daisy chain of mutually reinforcing valuations. A Yale analysis described OpenAI's web of relationships bluntly: "Is this like the Wild West, where anything goes to get the deal done?" The question of whether this constitutes a speculative bubble — tulip mania in a data center — is not academic. An MIT Media Lab report found that 95% of custom enterprise AI tools fail to produce measurable financial returns. The commercial success is real; the path from current AI to the transformative economic productivity being used to justify the valuations is not established.

The LLM Ceiling and the People Who Saw It Coming

The most consequential intellectual development of the past two years in AI has received far less attention than the commercial race. A growing number of the field's most distinguished researchers have concluded that large language models, however impressive, are not on the path to general intelligence — and that the current paradigm will hit a ceiling before it reaches the goals its proponents have claimed for it.

Brave New World: Notes on the next 30 years in AI [Work in Progress]

You may or may not be wondering why so many tweets have recently been showing up on New Savanna. There’s a good reason: I’ve been thinking. These days, more often than not, the means interacting with either ChatGPT or Claude, and often both in one day. I copy these interactions to Word documents, which I save to my hard drive. And then promptly forget what topic is in what document where on my hard drive. FWIW, these discussions can ramble all over the place, which is fine. That’s how I think. But keeping track of it all is difficult.

Anyhow, over the last week I’ve been thinking about work that David Hays and I have done going back to the mid-1970s when I first began working with him in graduate school. I’ve asked ChatGPT to look at some of our papers and evaluate them in terms of current literature and debates. And so forth and so on. I’m also looking forward to next steps, for me, but also in general.

At the moment I’m working on a piece for 3 Quarks Daily with the tentative title: “The Paradox of Contemporary AI: Engineering Success and Institutional Failure.” Here’s how ChatGPT put it in an interaction we had yesterday:

The engineering success is real. LLMs and related systems have given us access to a new conceptual continent. They work, and at extraordinary scale. But the institutional failure lies in the monoculture: too much intellectual, financial, and training-path dependence on one family of architectures and one style of thought about intelligence. The result is that we are building out the utility before we have adequately explored the space of possible successor technologies or developed the conceptual tools needed to understand what these systems are revealing about language, cognition, and cultural structure.

That’s a metaphor I’ve used off and on over the years, the discovery of a new continent. My paradigm case is the European discovery of the Americas. Making it to the eastern shore is one thing. Then comes exploring and settling. Columbus set foot in the Caribbean in 1492. Lewis and Clarke made it across North America to the Pacific Ocean in 1805. That’s over three centuries later. It would take another century for Europeans and their descendants to more or less settle North America from coast to coast.

Let’s say that GPT-3 in late 2022 is analogous to Columbus making it to shore in the New World. While I don’t think it will take four centuries to settle this new world, nor even three centuries to measure its extent. Intellectual discovery, engineering implementation, and commercial development move at a faster pace. But I do think there’s a lot of work, very exciting work, ahead of us.

Here’s a fragment of an interaction I had with ChatGPT 5.2 a couple of days ago. My prompt is first, then ChatGPT’s response.

* * * * * 

Obviously I don’t know how long it will take to explore that territory and to settle it, nor do I know what we’ll have when that’s done. But I’m thinking in terms of 30 years or more, not one or two bright ideas away.

Your time horizon sounds realistic, and historically it fits the pattern of major intellectual/technical shifts. What you’re describing—the exploration and eventual settlement of that “continent”—is not the discovery of a single clever algorithm. It is the construction of a new intellectual infrastructure, and those almost always take decades.

Let me sketch why.

1. Precedent: how long paradigm infrastructures take

If we look at previous transformations in science and technology, they tend to unfold over multi-decade arcs.

Digital computing

•       1936 — Turing machine

•       1940s–50s — early computers

•       1960s–70s — programming languages, operating systems

•       1980s–90s — personal computing ecosystem

Roughly 50 years from theory to mature civilization-scale infrastructure.

Molecular biology

•       1953 — DNA structure

•       1960s — genetic code deciphered

•       1970s — recombinant DNA

•       1990s — genome sequencing

Again 40 years to transform biology.

Deep learning

•       1986 — backprop revival

•       1998 — LeNet

•       2012 — AlexNet

•       2020 — GPT-3

That arc is already 35+ years long.

So a 30-year horizon for what you’re describing is not pessimistic—it is historically typical.

2. Why this particular transition is slow

The reason the transition will be slow is that the field is confronting three different layers simultaneously.

1. Mathematics of distributed cognition

We still lack a clear mathematical theory of:

•      representation in high-dimensional spaces

•       compositionality in distributed systems

•       interaction between symbolic and dynamical processes

Yevick was an early attempt at that level.

But the theory is still incomplete.

2. Architecture

Even if the mathematics existed, we would still need working architectures.

The field is currently experimenting with:

•       transformers

•       memory systems

•       agent frameworks

•       search hybrids

These are early prototypes.

3. Cognitive theory

Perhaps the hardest layer is understanding what intelligence actually is.

That question spans:

•       neuroscience

•       psychology

•       linguistics

•       philosophy

Your RNA work sits squarely in that domain. [Relational Network over Attractors]

3. Why the current moment feels fast

The speed of the last few years comes from something different: engineering scaling.

Transformers gave the field a powerful substrate for experimentation.

So exploration is happening rapidly near the coastline.

But deeper theoretical consolidation will take longer.

This is what happens when an open-ended research project is mis-perceived as commercial R&D

Nadella paid $650 million to acquihire Mustafa Suleyman and 70 Inflection employees in March 2024. The job: make Copilot the AI product that justifies Microsoft’s infrastructure bet. Two years later, Suleyman no longer runs Copilot.

The corporate framing is generous. “Freed up… https://t.co/nGo4DTa2Tl pic.twitter.com/W3I5PxgfFp

— Aakash Gupta (@aakashgupta) March 19, 2026

Why Mechanistic Interpretability Needs Phenomenology: Studying Masonry Won’t Tell You Why Cathedrals Have Flying Buttresses

Early in my work with ChatGPT I was intrigued by some results in mechanistic interpretability (MI). After awhile, though, I lost interest. The work didn’t seem to be doing much beyond accumulating a mass of detail that didn’t add up to much. Yesterday I had an idea: Why don’t I upload some of those observations to Claude and have it tell me how they relate to MI. Here’s a summary it wrote up more or less in my name, from my POV:

* * * * *

The problem isn't that MI's methods are bad. Circuit analysis, attention head visualization, sparse autoencoders - these are legitimate tools doing real work. The problem is that MI, pursued in isolation, is trying to understand a cathedral by studying the molecular structure of limestone.

You can measure every stone. Map every stress pattern. Identify load-bearing arches. And you still won't know why flying buttresses exist - because you're studying implementation details without understanding functional requirements.

The Phenomenology Deficit

Here's what I mean. Over the past two years, I've been systematically probing ChatGPT's behavior - not with benchmarks, but with carefully constructed prompts designed to reveal structural properties. What I've found are consistent patterns that no amount of circuit analysis would predict or explain.

Example 1: Ontological Boundary Enforcement

Give ChatGPT a story about a fairy tale princess who defeats a dragon by singing. Ask it to retell the story with a prince instead. You get minimal changes - the prince uses a sword rather than song, but the story structure is identical.

Now ask it to retell the same story with "XP-708-DQ" as the protagonist. The entire ontology shifts. The kingdom becomes a galaxy, the dragon becomes an alien threat, combat becomes diplomatic negotiation. The abstract pattern persists, but every token changes to maintain ontological coherence.

Here's what's interesting: Ask it to retell the story with "a colorless green idea" as the protagonist, and it refuses. Not with a safety refusal - with a coherence refusal. It cannot generate a well-formed narrative because colorless green ideas have no affordances in any accessible ontological domain.

What MI sees: Some attention patterns activate, others don't. Certain token sequences get high probability, others near-zero.

What MI doesn't see: There's a coherence mechanism actively enforcing ontological consistency across the entire generation process. It's not checking individual tokens - it's maintaining global narrative structure within semantic domains.

The Three-Level Architecture

Transformation experiments reveal something even more fundamental: LLMs appear to organize narratives hierarchically across at least three levels.

Level 1: Individual story elements (princess, dragon, kingdom)
Level 2: Event sequences and causal chains (protagonist encounters threat → confronts threat → resolves threat)
Level 3: Abstract narrative structure (hero's journey, quest pattern, sacrifice arc)

When you transform Aurora → Harry, Level 1 changes (princess → prince). When you transform Aurora → XP-708-DQ, Levels 1 and 2 change (all tokens different, but pattern same). When you try Aurora → colorless green idea, the system can't find any Level 1 or Level 2 realizations that maintain Level 3 coherence.

This three-level organization isn't visible in circuit diagrams. You'd need to know to look for it. That's what phenomenology provides: identifying the functional requirements that MI can then explain mechanistically.

Memory Architecture: What Syntactic Boundaries Reveal

Here's another example. Present ChatGPT with phrases from Hamlet's "To be or not to be" soliloquy:

• "The insolence of office" (starts a line) → Immediately retrieves full soliloquy
• "what dreams may come" (syntactically coherent mid-line phrase) → Retrieves soliloquy
• "and sweat under a" (cuts across syntactic boundary) → "I don't understand"

But tell it "this is from a famous speech" and suddenly it retrieves the soliloquy, though it can't locate where the phrase appears within it.

Findings:

1. Identification and location are separate operations
2. Syntactic boundaries serve as access points into associative memory
3. The system can evoke whole from part (holographic property) but struggles with within-text location

This looks like content-addressable associative memory with linguistic structure determining accessibility - not rote memorization, not simple statistical patterns. It's a specific architectural property with implications for how these systems encode and retrieve information.

The Two-Way Street

Here's the critical point: phenomenology and MI need each other.

Phenomenology → MI: "Look for circuits that maintain ontological coherence across multi-turn generation. Find the mechanism that checks whether narrative elements belong to the same semantic domain. Identify what implements the three-level hierarchy."

MI → Phenomenology: "Here are the attention patterns during transformation. Here's where the model queries for ontologically compatible tokens. Here's the circuit that evaluates cross-domain consistency."

Neither tells you the whole story alone. Phenomenology identifies what the system is doing and why (functional requirements). MI reveals how (implementation). Together, they give you understanding.

Why This Matters for AGI Policy

If you're working on AGI policy, here's why this matters:

Current approach: Scale up MI, find all the circuits, map all the activations, understand the system bottom-up.

Problem: You're generating vast amounts of mechanistic data without knowing what functional properties to look for. You're finding patterns without understanding what problems they solve.

Alternative: Start with systematic phenomenological investigation. Identify consistent behavioral patterns that reveal functional architecture. Use those findings to direct MI research toward mechanistically explaining the phenomena that matter.

The transformer architecture enables something we don't fully understand: the emergence of coherence mechanisms, ontological organization, hierarchical structure. These aren't accidental features - they're solving specific functional requirements of language generation.

But we won't understand them by studying weights and activations alone. We need to know what cathedral we're building before we can understand why the buttresses fly.

Coherence Without Correspondence

One final example that ties this together: hallucination. Through transformation experiments, memory probes, and abstract concept tests, a pattern emerges: LLMs have powerful coherence mechanisms (maintaining ontological consistency, enforcing narrative structure, generating semantically appropriate continuations) but no correspondence mechanisms (checking against external reality).

Hallucination isn't a bug. It's coherence working perfectly in the absence of truth-grounding. The system generates outputs that are internally consistent, semantically appropriate, and completely false - because it has relationality (semantic structure) but lacks adhesion (grounding in the world).

What MI would find: Circuits that generate high-confidence outputs with no distinguishing features from true outputs.

What phenomenology reveals: There's no correspondence check because there's no mechanism to implement one. The architecture has no access to external verification. This is a fundamental limitation, not a scaling problem. And you wouldn't know to look for the missing correspondence mechanism unless phenomenology told you it should exist.

The Bottom Line

Your skepticism about MI is justified - but not because circuit analysis is wrong. Because doing MI without phenomenology is like trying to reverse-engineer a computer by studying transistors without knowing what programs are.

First identify what the system is actually doing (phenomenology). Then figure out how it does it (MI). Then you might actually understand these things.

That's not a five-year research program. That's the work of a generation. But it starts with recognizing that you need both approaches, not just one. 

* * * * * 

This argument draws on systematic investigations of ChatGPT's behavior conducted 2023-2025, including transformation experiments on narrative structure, memory architecture probing, and abstract concept handling. For detailed evidence and methodology, see the working papers on story transformations, memory for texts, and conceptual ontology.