​Welcome to another edition of Generalist Intelligence, the weekly intelligence briefing that delivers situational awareness in 20 minutes or less. To unlock the full briefing, join as a member.

This week, we examine surging memory prices, Meituan’s US-chip-free breakthrough, and a surprising data center bottleneck.

— Mario

Subscribe now

How much?!

Samsung is pushing for another memory price increase of up to 20% for next quarter, on top of around 90% in Q1 of 2026, and 50-60% in Q2. That’s three consecutive quarters of major price hikes for DRAM, the type of memory chip that go into everything from your phone to AI servers, driven by data centers buying up nearly all available supply. For the first time in years, memory makers like Samsung, SK Hynix, and Micron hold real pricing power, which will show up in their margins and, eventually, in the price of every laptop, phone, and car you buy. For AI infrastructure, it’s a growing tax on anyone trying to build or scale compute that isn’t a hyperscaler with the leverage to lock in supply years in advance.

Read more

​Welcome to another edition of Generalist Intelligence, the weekly intelligence briefing that delivers situational awareness in 20 minutes or less. To unlock the full briefing, join as a member.

This week, we distill the second and third-order effects of GLP-1s, a narrative violation in AI, and growing tension around rare earth minerals.

— Mario

Subscribe now

Welcome to GLP-World

There is a large, venomous desert lizard called a Gila monster, which eats only once or twice a year. In the early 1990s, an endocrinologist named John Eng started wondering how that was possible – just an idle curiosity, but he cobbled together a small study. He soon isolated a novel peptide in the reptile’s venom that dramatically lowered blood sugar, just like the human body’s naturally occurring GLP-1 hormone, only it lasted for hours instead of minutes. Three decades later, a synthetic version of the molecule Dr. Eng discovered in Gila monster venom is quietly upending capitalism. You know it as Ozempic, and it’s changing everything.

So, no more obesity?

There’s that, obviously, with all the knock-on effects for junk food makers (bad – research shows high-calorie, ultra-processed foods are first to go) and tailors (very good – all those alterations and newly body-confident folks buying their first really good tuxedo). But the mechanism by which GLP-1 drugs like Ozempic suppress people’s appetites looks likely to be far more powerful than cutting out mere “food noise” for those who want to snack less.

Read more

“There are a lot of people that are going out there and saying things like, ‘all jobs are going away,’ or ‘we're going to automate everything,’ or this language about the permanent underclass. And it really upsets me because not only is it inaccurate, but it's really harmful to the psychology of a lot of people that I think have a very important role to play in the future of humanity.” – Matan Grinberg

Matan Grinberg is the co-founder and CEO of Factory, an AI company valued at $1.5 billion that helps enterprises like Nvidia, Morgan Stanley, and Adobe automate software development through “Droids,” intelligent agents designed to streamline software engineering. Before Factory, Matan spent more than a decade in theoretical physics, studying string theory at Princeton and UC Berkeley. His work now centers on a different kind of complex system: how software gets built in an era of increasingly capable AI agents, open models, and shifting compute economics.

In our conversation, we explore:

• How Emmy Noether’s theorem continues to shape Matan’s approach to technology, business, and AI

• Why Matan believes there will always be more problems to solve, even as AI becomes more capable

• The resource allocation problem facing CEOs as they balance headcount, compute, and token budgets

• Why Factory is betting on model independence and Matan’s take on the SpaceX-Cursor deal

• Why Matan pushes back on conflating open models with “Chinese models” and wants a stronger open-model ecosystem

• The identity crisis that followed Matan’s decision to leave physics

• Lessons from Factory’s first few years, including learning to push back and identify gaps in his own knowledge

• Factory’s culture, values, and Matan’s partnership with co-founder Eno Reyes

Thank you to the partners who make this possible

.tech domains: An identity for builders at their core.

Brex: The intelligent finance platform.

Persona: Trusted identity verification for any use case.

Explore the episode

Timestamps

(00:00) Intro

(03:50) Noether’s theorem explained

(06:45) How the search for what’s conserved informs Matan’s work

(10:53) Why there will always be more problems to solve

(11:58) The resource allocation problem of the AI era

(15:54) Factory’s mission: bringing autonomy to software engineering

(18:28) How Factory decides what to build next

(20:10) Why Factory abstracts away model choice

(22:07) How Factory wins enterprise customers

(23:15) Matan’s take on the SpaceX-Cursor deal

(27:48) Why open-weight models matter

(29:19) Anthropic’s Fable 5 release and the debate over AI guardrails

(35:33) How Matan got into string theory

(38:21) Working with Juan Maldacena

(41:53) Startup founders vs. theoretical physicists

(46:15) Rethinking physics and redefining his identity

(51:29) Discovering AI and code generation

(52:53) The origins of Factory

(55:52) Lessons from Factory’s first few years

(59:58) Learning to push back and finding the holes in his knowledge

(1:03:17) Factory’s culture and values

(1:08:11) Matan’s predictions for the future of AI and Factory

(1:10:49) Final meditations

Follow Matan Grinberg

LinkedIn: https://www.linkedin.com/in/matan-grinberg

X: https://x.com/matanSF

Website: https://factory.ai

Resources and episode mentions

Books

• A Tale of Two Cities: https://www.amazon.com/Tale-Cities-Dover-Thrift-Editions/dp/0486406512

• The Picture of Dorian Gray: https://www.amazon.com/Picture-Dorian-Gray-Oscar-Wilde/dp/0141439572

• The Brothers Karamazov: https://www.amazon.com/Brothers-Karamazov-Fyodor-Dostoevsky/dp/1685781152

• Zero to One: Notes on Startups, or How to Build the Future: https://www.amazon.com/Zero-One-Notes-Startups-Future/dp/0804139296

• Invisible Cities: https://www.amazon.com/Invisible-Cities-Italo-Calvino/dp/0156453800

• Collected Fictions: https://www.amazon.com/Collected-Fictions-Jorge-Luis-Borges/dp/0140286802

People

• Emmy Noether: https://en.wikipedia.org/wiki/Emmy_Noether

• Juan Maldacena: https://www.ias.edu/scholars/maldacena

• William Shakespeare: https://en.wikipedia.org/wiki/William_Shakespeare

• Shaun Maguire on X: https://x.com/shaunmmaguire

• Frank Slootman on LinkedIn: https://www.linkedin.com/in/frankslootman

• Sridhar Ramaswamy on LinkedIn: https://www.linkedin.com/in/sridhar-ramaswamy

• Eno Reyes on LinkedIn: https://www.linkedin.com/in/enoreyes

Other resources

• Noether’s theorem: https://en.wikipedia.org/wiki/Noether%27s_theorem

• Factory: https://factory.ai

• Nemotron: https://developer.nvidia.com/topics/ai/nemotron

• Why One Superintelligence Is More Dangerous Than a Thousand (Vincent Weisser, CEO & Co-Founder of Prime Intellect): https://www.generalist.com/p/why-one-superintelligence-is-more

• IMDb Top 250 movies: https://www.imdb.com/chart/top/

• Harakiri: https://www.imdb.com/title/tt0056058

• La Grande Bellezza: https://www.imdb.com/title/tt6155190

• La Haine: https://www.imdb.com/title/tt0113247

• The Master: https://www.imdb.com/title/tt1560747

• Interstellar: https://www.imdb.com/title/tt0816692

• Proper Time to the Black Hole Singularity from Thermal One-Point Functions: https://www.researchgate.net/publication/345261524_Proper_time_to_the_black_hole_singularity_from_thermal_one-point_functions

Subscribe to the show

I’d love it if you’d subscribe and share the show. Your support makes all the difference as we try to bring more curious minds into the conversation.

YouTube

Spotify

Apple

Production and marketing by penname.co. For inquiries about sponsoring the podcast, email jordan@penname.co.

Friends,

​This week, we look Eastward. While the West concerns itself with which model is the smartest, Beijing published a 17-point plan to shove AI into every shop, clinic, and classroom in the country. It is one America cannot afford to ignore.

Read for our view on what Xi Jinping’s AI push means, Google’s brain drain, what satellite imagery tells us about data center build-outs, and the obscure crystal in high demand.

— Mario​

P.S. To learn more about our process and vision for Generalist Intelligence, read our introductory piece. To unlock the full briefing, join as a member.

Subscribe now

China’s latest AI push

Earlier this month, almost entirely unnoticed by the Western press, China dropped an AI policy document – the fabulously resistible Implementation Opinions of the Ministry of Commerce and 8 Other Departments on Accelerating the Development of ‘AI+Consumption’ – that we think is worth a closer look. It feeds into two things old China hands often say: “America invents, China applies” and, effectively, “China likes central planning.” And once you get past the title and the usual platitudes about how the plan is “Guided by Xi Jinping Thought on Socialism with Chinese Characteristics for a New Era,” it becomes clear that something serious is going on.

Read more

By the end of his life, Henry Luce was at home in any room. Not loved, of course. His strong opinions, a staunch conservatism and a fervent anti-communist bent among them, hardly earned universal approbation. But he appeared unfazed, unperturbed, more than a little superior. “He lived well above the tree line on Olympus,” one of his editors remarked. Fundamentally, at home.

As the founder of Time, Fortune, Life, and Sports Illustrated, Luce had met the good and great of every industry. And not merely met them, but shaped their lives. That had propelled him not just to eminence but extraordinary wealth. In 1966, Luce’s holdings generated $503 million. He’d founded the company 43 years earlier with $86,000.

He died the year after, aged 68, leaving behind the trappings of the Great Man, the consummate insider.

Whether the man ever felt that solidity, that sense of ultimate belonging, only he could have answered. The boy, the Luce of five and nine and fifteen, did not. No matter where Luce went, he did not fit, did not belong. First, he was the son of a missionary in Shandong, an American in China. Then, after he was sent to the China Inland Mission school at Chefoo, he was the American among English students. Worse yet, he was a stammerer, a tic he’d acquired the year prior and that his tutor had declared “absolutely imaginary” and an “evil habit.”

“Everything is going as usual but not very well,” Luce wrote in a letter home from Chefoo. “It sort of seems to hang on not in spells of homesickness but a hanging torture, I well sympathise with prisoners wishing to commit suicide.” He was ten years old, precociously articulate, and utterly alone. “I am getting that hatred of which I will never get over even tho I was here hundreds of years.”

In tandem with Luce’s suffering, a commensurate competitiveness emerged. In other missives, he writes of his desire to destroy his peers. “My years ambition has been accomplished, that is to lick Hayes.... For this year I have made the form record in going up places, that is 8 places.” A sense of deficiency seemed to give way to obsessive self-comparison and ferocious drive.

At fifteen, Luce was sent to boarding school in Connecticut, attending Hotchkiss. Yet again, he was a child apart, an “aid boy” among New England’s blue bloods. As a condition of his scholarship, he had to wait tables in the dining hall, serving breakfast to his paying classmates. Rather than living on campus, he was sent to a boarding house a mile away. He knew little of American life (“I did not know a single rule of football when I came here,” he said, wore shabby clothing fabricated in China, and his halting speech was devoid of the Yankee idioms. He earned the nickname “Chink.” He had gone from an American boy in China to a Chinese adolescent in America.

Luce’s response to serving the richer students was not to disdain them, but rather to aspire to join them. Instead, he looked down on the scholarship boys, assuming himself to be their superior. “These scholarship fellows here are all fundamentally fine, but they lack certain qualities — especially noticeable at table — that one misses!” Rather than find acceptance within a group, Luce chose to separate himself from it, to set his sights on ascending America’s social ladder. He achieved it.

Not every founder was ostracized quite as totally as Luce. Culture, class, geography, and even his manner of speaking conspired to separate him from his peers. But the experience of being an outsider, of being profoundly out of place, is the most common trait observed from studying the formations of great entrepreneurs. It is also among the most acute, contributing to a desire to prove one’s relative worth and breeding a sense of specialness, even if reflected through a dark mirror. Perhaps every adolescent feels some moment of difference; this is the beginning of selfhood. But not everyone experiences it quite like this.

Part I of the Saplings series, which studies the early lives of legendary founders, explores this quality, its variations, and its impact. As follows:

• A world apart. We discuss the ways in which differences in nationality, ethnicity, class, and other external factors make these children feel like outsiders. Often, this seems to be coupled with a commensurate desire to overprove one’s worth to the rejected environment.

• Familial interlopers. In some cases, children are strangers in their own homes. They’re shunned by parents and siblings, or perceived as somehow alien, which influences a sense of independence.

• Isolated by their body. Illness, disabilities, and deformities impacted some founders similarly, separating them from their peers physically and forcing them to spend more time in their own minds.

• Wired differently. Having a mind that runs on unusual tracks distances many founders from those around them. They often struggle to communicate, finding it easier to demonstrate their abilities and intelligence in other ways.

• From the other side. Negative transformative experiences, like a close familial death, force children to grow up early and contribute to a sense, in some cases, of being somehow special.

It is worth noting that neither this series in general nor this piece in particular imagines that it can explain everything about these people. To notice that so many remarkable founders were marginalized by the world around them in their early years is not to say that it is a necessity of greatness or to draw a direct correlation between exile and achievement, nor to valorize the very real, often terrible suffering these people endured. Humans are capable of metabolizing hardship into many things — helplessness, energy, cruelty, inspiration, revenge, and grace—none of which ennoble the original sin.

And, of course, every case is a matter of degree and a matter of perspective. Only the founders themselves could tell us how they perceived their remove from the world around them, and given what unreliable narrators these characters often are, you would be a fool to take them at their word. (“I’m super normal” is Jensen Huang’s self-appraisal.) This is, then, just a fragment of something, a sliver of colored glass we are putting up to the light to look at the shadows it casts.

If you are interested in the broader process behind the Saplings project, you might enjoy our piece introducing it here.

A world apart

A young Masayoshi Son, Softbank’s flamboyant impresario, considered his grandmother “repulsive.” This was the woman who raised him and loved him. The son of Korean immigrants in Japan, the Son family had little money. Each morning, the elderly woman placed a tiny Masayoshi on a two-wheeled wooden cart and pulled him through the neighborhood, through dirt and pig slop, to see what food the restaurants near the train station had discarded the night before. This was what they fed the family pigs. Son recalled the scene as an adult, describing the “slippery” ground and its “rotting smell.”

Read more

“No one is data ready. Absolutely no one is data ready.” —Yash Patil

Yash Patil is the 23-year-old founder and CEO of Applied Compute, a $1.3 billion company helping businesses train custom AI models on their own data: smaller, cheaper, and purpose-built for the work they actually do. Before founding the company, Yash dropped out of Stanford and spent two years at OpenAI working on post-training infrastructure and Codex. He left with one core conviction: every company that runs its critical workflows on someone else’s model is building on shifting sand. Applied Compute is his answer to that problem, already serving customers including DoorDash, Cognition, and Mercor.

In our conversation, we explore:

• Why “own or be owned” is becoming existential for any company that relies on frontier AI models

• What it was like inside OpenAI the weekend the board fired, and then reinstated, its CEO

• Why post-training is where competitive advantage is now being built, and what reinforcement learning with verifiable rewards actually is

• Why evals have become the new production environment, and why companies will never share them with frontier providers

• How a specialized model built for DoorDash outperformed frontier models on a narrow, high-value task

• Why cost, not capability, is now the primary driver pushing companies toward custom models

• Why Yash believes AI’s transformation of the economy will unfold over decades, and why near-term fears about mass job displacement are misplaced

Thank you to the partners who make this possible

Brex: The intelligent finance platform.

Guru: The AI source of truth for work.

Persona: Trusted identity verification for any use case.

Explore the episode

Timestamps

(00:00) Introduction

(03:50) Fable 5 and the case for owning your own models

(09:22) Why Applied Compute is betting on custom AI models

(12:30) Yash’s early influences and first projects

(17:42) His brief time building at Stanford

(19:29) Leaving Stanford for OpenAI

(25:58) Inside OpenAI during Sam Altman’s firing

(28:18) What Yash admires about Sam Altman

(29:43) Teaching models to reason

(35:39) The core insight behind Applied Compute

(39:40) How Applied Compute works with its customers

(45:55) Why model training never ends

(48:56) Why not every task needs a frontier model

(51:25) The culture and people of Applied Compute

(54:50) Applied Compute’s training infrastructure

(58:43) The coming compute crunch and other predictions

(1:03:48) Final meditations

Follow Yash Patil

X: https://x.com/ypatil125

Website: https://yashpatil.me

LinkedIn: https://www.linkedin.com/in/yash-s-patil

Resources and episode mentions

Books

• The Design of Everyday Things: https://www.amazon.com/Design-Everyday-Things-Revised-Expanded/dp/0465050654

People

• Brendan Foody on X: https://x.com/BrendanFoody

• Luke Metz’s website: https://lukemetz.com

• Barret Zoph on X: https://x.com/barret_zoph

• John Schulman on X: https://x.com/johnschulman2

• Ian Osborne on LinkedIn: https://www.linkedin.com/in/ian-osborne

Other resources

• Applied Compute: https://www.appliedcompute.com

• Claude Fable 5 and Claude Mythos 5: https://www.anthropic.com/news/claude-fable-5-mythos-5

• Goldman Sachs: https://www.goldmansachs.com/

• Sarah Chieng’s post on X: https://x.com/MilksandMatcha/status/1983587401688870972

• Science Olympiad: https://www.soinc.org

• Chai Discovery: https://www.chaidiscovery.com

• TreeHacks: https://treehacks.com

• Thiel Fellowship: https://thielfellowship.org

• Building a hill-climbing machine: Launching seven new MAI models: https://microsoft.ai/news/building-a-hillclimbing-machine-launching-seven-new-mai-models

• Kirkland & Ellis: https://www.kirkland.com

• Palantir: https://www.palantir.com

• Automating Merchant Onboarding at DoorDash: https://www.appliedcompute.com/case-studies/doordash

• Composer: https://cursor.com/blog/composer

• NVIDIA: https://www.nvidia.com

• DeployCo: https://www.deployco.co

Subscribe to the show

I’d love it if you’d subscribe and share the show. Your support makes all the difference as we try to bring more curious minds into the conversation.

YouTube

Spotify

Apple

Production and marketing by penname.co. For inquiries about sponsoring the podcast, email jordan@penname.co.

Friends,

Welcome to another edition of Generalist Intelligence, the weekly intelligence briefing designed to deliver situational awareness and help you see around corners. All in 30 minutes or less.

To learn more about our process and vision for Generalist Intelligence, read our introductory piece. To unlock the full briefing, join as a member.

This is a new initiative, so we’d love to hear what you think. Reply anytime to tell us what's working and what you'd like us to cover more deeply. Our goal is for this to become the highest signal-to-noise newsletter you receive each week, and one worth setting aside the time for.

Subscribe now

Dario has been having one of those weeks. Just days after releasing Fable — a supposedly guard-railed version of its “too powerful to be safely released” Mythos model — Anthropic got a call from the US government. Amazon had found a way to bypass some of the safety features. The fear was that bad actors (read: China) would find the same exploit and use the power for nefarious purposes. Anthropic was apparently given 90 minutes (last thing on a Friday afternoon) to comply with export controls that denied access to foreigners, even if they worked for Anthropic. The only practical option was to deny everyone, Americans included.

Whatever you think of the relevant parties (some at Generalist Towers describe themselves, unashamedly, as “Anthropic-pilled”), this sucks for the Claude-maker, is bad for America, and is revealing about the future.

What if it’s actually a good thing?

Nice try. There are people arguing that getting knocked around by Donald Trump (again) might be “good for business,” pointing to an uptick in Claude receipts after the president previously declared Anthropic a “supply-chain risk.” Others, rather gamely, suggest that being officially branded “too powerful” creates an aura that attracts customers. Both are wrong. Anthropic already declared Mythos too powerful, then spent two months building guardrails that have, one way or another, landed it in this mess. Either the guardrails aren’t good enough, which is embarrassing, or the US government (and, for some reason, Anthropic’s closely intertwined partner Amazon) has decided to try and slow them down. Both are major red flags to anyone thinking of buying from Anthropic.

I thought Anthropic wanted AI to be regulated?

Yes, there’s an irony in Dario, having made so much noise about the dangers of the AI he’s building and the need for government to constrain his almighty power, looking quite so hurt when they take him at his word. But what Anthropic actually wants is something like the Federal Aviation Administration: an arms-length regulator that keeps passengers safe without dragooning business. Not whatever this is.

What does it mean for America?

Seems bad, basically. It’s great having world-beating AI firms, but you do rather rely on the rest of the world buying their stuff to get the full benefit. The next few years look crucial in the development of AI, and for those next few years, Trump will be president. That means Anthropic, and in theory, every US model-maker, lives or dies by Caesar’s thumb. If you’re an international company buying AI tools, that’s a lot of risk.

Where else are they going to go?

Read more

Friends,

For several years, I’ve wanted to launch a weekly intelligence briefing. While there are exceptional private intelligence groups covering the financial markets and geopolitics, and many great tech publications, I nevertheless found myself wanting something a little different from what I was able to find.

Something that delivered an elegant, insightful analysis of the hidden shifts and fresh opportunities bubbling up in our industry. Something high-signal, low-fluff, and with, perhaps, a dash of wit. Something that provides a little greater situational awareness in thirty minutes or less.

But doing it right would require taste and craft, as well as real time. For an intelligence product to succeed, it must find fresh angles on known stories and the non-obvious ones just bubbling beneath the surface. That takes scrounging and scouring.

Earlier this year, we were able to unlock both. Most important was convincing one of my favorite writers, an old hand at this kind of work, to jump aboard and lead the charge. By leveraging frontier models, we’ve been able to build an extensive signal-gathering apparatus that would not have been possible to conduct without a very large team, spanning news, industry outlets, social media, academic research, government and regulatory filings, code repositories, model-adoption aggregators, fundraising data, prediction markets, website monitoring, open-source intelligence sources, and talent flows. We also monitor a select group we refer to as “super signalers” across various verticals, selected by us, who we believe are frequently ahead of the curve. Not all of these approaches are bearing fruit yet, but we expect them to improve as we build and fine-tune. Already, we’ve seen the benefits of a process that searches so expansively. Not only does it surface tidbits a human might have missed, but it spots echoes or juxtapositions that are revealed only with capacious context.

However, the story selection and writing are done by us, the humans. The system hunts ceaselessly, bringing up a furious exhaust of data, noise, and signal, all in one. It is up to the human to think, judge, analyze, and write.

Generalist Intelligence is the result of this marriage, a weekly intelligence briefing designed for tech’s most discerning professionals. It will arrive Friday mornings. Our hope is that it opens you up to perspectives you hadn’t considered and offers a look at fresh pockets of the future you might have missed. I’ve been reading internal versions of this as we tuned the format over the past couple of months, and have found myself looking forward to it every week.

Subscribe now

We’re in the very early stages of getting this right, and there’s a long way to go. We’d love to know what you think, especially what you’d like to see more of, as it comes up. Our goal is for this to become the highest signal-to-noise newsletter you receive each week, and one worth setting aside the time for.

Now, onto the briefing.

— Mario

With SpaceX blasting into the public markets and Anthropic and OpenAI straightening their ties to follow suit, the bajillion-dollar question is, obviously: are they worth it? Anyone claiming to know the answer very likely has a bridge to sell you. Nevertheless, we can’t stop thinking about it.

SpaceX feels like a special case. Public markets are being asked to price in total interstellar domination at IPO, with a good trillion of the $1.77 trillion price tag – at which the loss-making company would trade at over 90x last year’s revenue – based on a series of highly speculative goals including repeat business on Mars, data centers in orbit, and making a key contribution to the development of AI. On the other hand, it’s Elon. So, you know.

What’s really got our chins wagging at Generalist Towers this week is a more existential question. Anthropic and OpenAI haven’t filed public prospectuses, but they look likely to be priced as though they are in, effectively, a two-horse race for control of the future (ok, maybe Google too). Some of us think that’s right – that it’s only a matter of time before one of the model labs cracks recursive self-improvement, and then it’s a short hop to AGI and either Sam or Dario as World King.

But a couple of things have landed in the past week that have got us thinking seriously about the other side. So we thought it would be fun to take the contrarian stance, and make the bear case for AI.

AI, huh, what is it good for?

LLMs are incredibly good at coding. At the end of last week, Anthropic dropped an essay titled When AI builds itself, in which the Claude-maker claimed that more than 80% of the code it merges into its codebase was written by Claude, and that its engineers were shipping “8x as much code per quarter” compared to their relatively flat productivity between 2021 and 2025. (When I asked Claude about this, it raised an eyebrow at the figures, pointing out they were “self-reported by a company currently raising money on exactly this story.”) And to be fair, rather gallantly, the good folks at Anthropic point out that the 8x figure is likely an overstatement as it measures “quantity over quality.” But the direction is clear: Claude is incredible at writing code. Presumably, this is translating into giant gains for all?

Well, no.

A new study of more than 100,000 GitHub developers found that Claude Code allowed coders to create or edit almost 300% more files. But that uplift was halved to 150% by the time they got to submitting pieces of work for review, and that in turn shrunk a further 5x by the time it got to shipping. In other words, coders using the latest agentic AI tools released just 30% more software, compared to those raw-dogging it. The researchers attribute that remaining 30% to what they call “strong complementarities” between the agents and the coders, meaning the agent very much needs the coder. They also noted that while there was a marked uptick in new apps being made this way, there was “no increase” in overall demand. This doesn’t mean agents are bad at code, but it does mean they might not be nearly as good as you thought at the thing they’re best at.

So what happens to the model makers?

Read more

“I chose cultural anthropology, since it offered the greatest opportunity to write high-minded balderdash.” — Kurt Vonnegut.

***

Friends,

One way I think about The Generalist’s work is as the anthropologist of a small, strange tribe.

This is not a normal tribe. It has no set territorial home, no blood lineage, no linguistic or ethnic markers. Its people ha…

Read more

“Warfare is always just adapting to whatever the other side is doing. And the person that wins is whoever adapts faster. And at this point in time, American aerospace is uniquely poorly suited to do that.” —Bryon Hargis, Co-Founder & CEO of Castelion

Bryon Hargis is the co-founder and CEO of Castelion, a defense startup building low-cost hypersonic missiles designed to be manufactured at scale. Before founding Castelion, Bryon spent more than a decade at Johns Hopkins Applied Physics Laboratory and nearly six years at SpaceX, where he worked on national security space programs and saw firsthand how iterative engineering and manufacturing speed could reshape aerospace. Castelion’s first missile, Blackbeard, is slated for integration on the Navy’s F/A-18 Super Hornet in roughly a year.

In our conversation, we explore:

• Why Bryon believes building missiles is paradoxically essential to maintaining peace

• The game theory behind warfare and why tit-for-tat strategies require credible middle-ground responses

• How China’s 2021 hypersonic test revealed not just a capability gap but a manufacturing and cost advantage

• Why traditional aerospace processes—optimized for low risk and high cost—can’t compete with rapid iteration

• What Bryon learned in his first week at SpaceX (after 12 years in traditional aerospace)

• Why building a carrier-based, air-launched hypersonic missile as a first product was the hard but right choice

• How focusing on manufacturability and cost over maximum capability can produce more effective deterrence

• Why the person who adapts faster in warfare always wins, and how that shapes Castelion’s philosophy

Thank you to the partners who make this possible

.tech domains: An identity for builders at their core.

Ahrefs Brand Radar: Find your brand in AI results.

Persona: Trusted identity verification for any use case.

Explore the episode

Timestamps

(00:00) Intro

(04:01) Why America needs hypersonic missiles

(07:13) China’s edge in hypersonics

(12:05) The missing middle ground in deterrence

(18:05) Preventing warhead ambiguity

(19:40) How hypersonics differ from ballistic missiles

(25:05) The economics of defensive vs. offensive systems

(28:21) How SpaceX differs from traditional aerospace

(37:40) Why Bryon chose to build in defense over space

(42:42) Key factors that drove Castelion’s success

(48:28) Designing Blackbeard, Castelion’s first hypersonic missile

(1:01:06) The importance of lower costs and quicker manufacturing

(1:10:04) Book recommendations

Follow Bryon Hargis

LinkedIn: https://www.linkedin.com/in/hargsb

X: https://x.com/hargsb

Resources and episode mentions

Books

• Carrying the Fire: An Astronaut’s Journeys: https://www.amazon.com/Carrying-Fire-Astronauts-Michael-Collins/dp/0374531943

• Skunk Works: A Personal Memoir of My Years at Lockheed: https://www.amazon.com/Skunk-Works-Personal-Memoir-Lockheed/dp/0316743003

• The Hard Thing About Hard Things: Building a Business When There Are No Easy Answers―Straight Talk on the Challenges of Entrepreneurship: https://www.amazon.com/Hard-Thing-About-Things-Building/dp/0062273205

• Gravitation: https://www.amazon.com/Gravitation-Charles-W-Misner/dp/0691177791

• “Surely You’re Joking, Mr. Feynman!”: Adventures of a Curious Character: https://www.amazon.com/Surely-Youre-Joking-Mr-Feynman/dp/0393355624

• Perfectly Reasonable Deviations from the Beaten Track: https://www.amazon.com/Perfectly-Reasonable-Deviations-Beaten-Track/dp/0465023711

People

• Kip Thorne: https://en.wikipedia.org/wiki/Kip_Thorne

• Richard Feynman: https://en.wikipedia.org/wiki/Richard_Feynman

Other resources

• Castelion: https://www.castelion.com

• Chinese hypersonic weapon fired a missile over South China Sea: https://www.ft.com/content/a127f6de-f7b1-459e-b7ae-c14ed6a9198c?syn-25a6b1a6=1

• Prisoner’s dilemma: https://en.wikipedia.org/wiki/Prisoner%27s_dilemma

• APL: https://www.jhuapl.edu

• Planet Labs: https://www.planet.com

• Falcon 9: https://www.spacex.com/vehicles/falcon-9

• Veritasium: https://www.youtube.com/@veritasium

• 3Blue1Brown: https://www.youtube.com/@3blue1brown

• Making Mistakes with Greg: https://www.youtube.com/@makingmistakeswithgreg

• Super Fast Matt: https://www.youtube.com/@SuperfastMatt

• This game theory problem will change the way you see the world: https://www.youtube.com/watch?v=mScpHTIi-kM

Subscribe to the show

I’d love it if you’d subscribe and share the show. Your support makes all the difference as we try to bring more curious minds into the conversation.

YouTube

Spotify

Apple

Production and marketing by penname.co. For inquiries about sponsoring the podcast, email jordan@penname.co.

"China can do it because the labor is cheap. That's not true anymore. They are incredible at automating things and automating them in a way that makes them cost competitive." —Davide Asnaghi

Davide Asnaghi is the co-founder and CEO of Diode, a Brooklyn-based startup using AI to design and manufacture circuit boards in the United States.

Before Diode, Davide worked on Apple’s Special Projects Group and spent time in Hong Kong and Shenzhen studying Asia’s electronics manufacturing ecosystem. That experience convinced him that PCB design, despite powering everything from smartphones and satellites to medical devices and autonomous systems, remained one of the most overlooked layers of the tech stack.

Since its founding just two years ago, Diode has landed Physical Intelligence and Saronic as customers and partnered with Anthropic to help Claude become a better electrical engineer. The company’s ultimate ambition: to make hardware as nimble as software.

In our conversation, we explore:

1. Why the West outsourced PCB manufacturing to Asia in the 2000s and why bringing it back matters for American competitiveness

2. What Shenzhen’s manufacturing culture does better than Silicon Valley (and what the U.S. can learn from it)

3. How Diode’s models can one-shot much of schematic design and compress hardware timelines from months to weeks

4. The three-week YC pivot that transformed Diode from a design validation tool into a full-stack manufacturer

5. Why circuit boards are the “forgotten middle child” between silicon and software

6. How Diode partners with Anthropic to make LLMs better electrical engineers

7. What it takes to build a hardware company in 2025—and why the talent bar must stay incredibly high

8. How Italian, American, and Chinese cultures shaped Davide’s approach to entrepreneurship and manufacturing

Thank you to the partners who make this possible

.tech domains: An identity for builders at their core.

Guru: The AI source of truth for work.

Brex: The intelligent finance platform.

Explore the episode

Timestamps

(00:00) Intro

(04:15) Why Davide calls himself a copper merchant

(05:53) Diode’s mission to rebuild PCB manufacturing in the U.S.

(07:58) What success looks like

(09:00) Growing up in northern Italy and spending a year in Minnesota

(13:14) Why Italy produces fewer venture-backed founders

(15:30) Why Hong Kong accelerated Davide’s learning

(19:09) Silicon Valley vs. Shenzhen

(22:05) What Davide learned in Apple’s Special Projects Team

(24:11) Why Davide left Apple after two years

(26:54) Meeting his co-founder, Lenny

(29:32) How Davide uncovered the need for better PCB design and manufacturing

(33:23) PCB manufacturing in Asia, and Diode’s approach

(41:29) The YC pivot that changed Diode’s business

(44:39) Inside Diode’s customer journey

(48:10) Where the value is in electronics manufacturing, and Davide’s AGI thesis

(51:30) What separates a working board from a great one

(55:32) Where Diode fits in the electronics stack

(59:55) Diode’s early near-death moment and long-term vision

(1:02:30) Diode’s exceptionally high bar for hiring

(1:04:48) Where Davide gets his best ideas

(1:07:00) Final meditations

Follow Davide Asnaghi

LinkedIn: https://www.linkedin.com/in/d-asnaghi

X: https://x.com/davideasnaghi

GitHub: https://hexdae.github.io

Resources and episode mentions

Books

• Breakneck: China’s Quest to Engineer the Future: https://www.amazon.com/Breakneck-Chinas-Quest-Engineer-Future/dp/1324106034

• Chip War: The Fight for the World’s Most Critical Technology: https://www.amazon.com/Chip-War-Worlds-Critical-Technology/dp/1982172002

People

• Alex Wong on LinkedIn: https://www.linkedin.com/in/alex-wong-6b8930205

• Lenny Khazan on LinkedIn: https://www.linkedin.com/in/lennykhazan

• Brendan Eich’s website: https://brendaneich.com

Other resources

• Diode: https://www.diode.computer

• AWS: https://aws.amazon.com

• Arduino: https://www.arduino.cc

• Bending Spoons: https://www.bendingspoons.com

• Fortell (formerly Chromatic): https://www.fortell.com

• Butterfly Network: https://www.butterflynetwork.com

• Foxconn: https://www.foxconn.com

• JLCPCB: https://jlcpcb.com

• Anthropic: https://www.anthropic.com

• Quilter: https://www.quilter.ai

• KiCad: https://www.kicad.org

• Cadence: https://www.cadence.com

• Stripe: https://stripe.com

• Dogpatch: https://en.wikipedia.org/wiki/Dogpatch,_San_Francisco

Subscribe to the show

I’d love it if you’d subscribe and share the show. Your support makes all the difference as we try to bring more curious minds into the conversation.

YouTube

Spotify

Apple

Production and marketing by penname.co. For inquiries about sponsoring the podcast, email jordan@penname.co.

“AI is going to be the age of the polymath. If you have the ability to think about different systems and how they might work together, you’re going to be able to come up with outcomes that were previously impossible because those disciplines didn’t work well together.” — Cyan Banister, Co-Founder of Long Journey Ventures

Cyan Banister has built one of the most distinctive early-stage track records of the last fifteen years, with early bets on companies like Uber, SpaceX, DeepMind, Niantic, and Postmates. Today, she is co-founder and general partner at Long Journey Ventures, where she backs what she calls “magical weirdos.” Banister describes herself as a professional daydreamer, running constant thought experiments and paying close attention to signals others ignore. In this episode, she explains how that mindset translates into investing, and why many of her best opportunities have come from observation, curiosity, and a willingness to look in unlikely places.

In our conversation, we explore:

• Cyan’s philosophy of treating life as a series of experiments

• The strange, profound experiences that led her to question and ultimately move beyond her atheism

• How scanning Wi-Fi networks in a Four Seasons café led her to Flock Safety, last valued at $8.4 billion

• Long Journey Ventures’ “Biz, Tizz, and Rizz” framework for identifying exceptional founders and why the trifecta is rare

• How AI will enable the age of the polymath

• Why she believes brain-computer interfaces are closer than most people think

• Why she says Pokémon Go was “the closest we ever came to world peace”

• Why she lives part-time in a retirement community and her vision for a more connected future

Thank you to the partners who make this possible

.tech domains: An identity for builders at their core.

Brex: The intelligent finance platform.

Persona: Trusted identity verification for any use case.

Explore the episode

Timestamps

(00:00) Intro

(03:51) Never playing the game you appear to be playing

(07:18) Practicing childlike wonder as a daily discipline

(10:08) Questioning belief after her stroke

(13:30) Cyan’s metaphysical experiments

(23:24) Non-local consciousness and creativity

(27:22) Investing with extreme openness to signals

(29:05) The importance of timing in investing

(32:26) Meeting Travis Kalanick

(34:19) Finding Flock Safety through a chance encounter

(38:23) The summer of Pokémon Go (what worked and what didn’t)

(39:55) Human nature and what makes something “stick”

(42:15) Brain-computer interfaces and AI’s accelerating effect

(52:53) “Biz, Tiz, Riz:” her framework for evaluating founders

(59:20) Why Cyan lives in a retirement community part-time

(1:03:50) A unique way of finding books that speak to you

(1:08:44) Final meditations

Follow Cyan Banister:

LinkedIn: https://www.linkedin.com/in/cyanb

X: https://x.com/cyantist

Newsletter: https://uglyduckling.substack.com

Website: https://cyanbanister.com

Resources and episode mentions

Books

• Harry Potter and the Philosopher’s Stone: https://www.amazon.com/Harry-Potter-Philosophers-Stone-Rowling/dp/0747532745

• Impro: Improvisation and the Theatre: https://www.amazon.com/Impro-Improvisation-Theatre-Keith-Johnstone/dp/0878301178

• The Diamond Age: https://www.amazon.com/Diamond-Age-Neal-Stephenson/dp/0553573314

• Snow Crash: https://www.amazon.com/Snow-Crash-Neal-Stephenson/dp/0553380958

• My Life and Loves: https://www.amazon.com/My-Life-Loves-Frank-Harris/dp/9358712538

• Milton Friedman’s books on Amazon: https://www.amazon.com/Books-Milton-Friedman/s?rh=n%3A283155%2Cp_27%3AMilton%2BFriedman

• The Razor’s Edge: https://www.amazon.com/Razors-Edge-W-Somerset-Maugham/dp/1400034205

People

• Kurt Vonnegut: https://en.wikipedia.org/wiki/Kurt_Vonnegut

• Aldous Huxley: https://en.wikipedia.org/wiki/Aldous_Huxley

• J.K. Rowling’s website: https://www.jkrowling.com

• Neal Stephenson’s website: https://www.nealstephenson.com

• Travis Kalanick on X: https://x.com/travisk

• Ryan Graves: https://en.wikipedia.org/wiki/Ryan_Graves_(businessman)

• Keith Rabois on LinkedIn: https://www.linkedin.com/in/keith

• Garrett Langley on LinkedIn: https://www.linkedin.com/in/glangley

• John Luttig on LinkedIn: https://www.linkedin.com/in/luttig

• John Hanke on X: https://x.com/johnhanke

• Paul Stamets website: https://paulstamets.com

• Keith Corso on LinkedIn: https://www.linkedin.com/in/keith-corso

• Brendan Eich on LinkedIn: https://www.linkedin.com/in/brendaneich

• Arielle Zuckerberg on LinkedIn: https://www.linkedin.com/in/ariellezuckerberg

• Josh Browder on X: https://x.com/Joshuabrowder

• Aleister Crowley: https://en.wikipedia.org/wiki/Aleister_Crowley

• George Gurdjieff: https://en.wikipedia.org/wiki/George_Gurdjieff

• P.D. Ouspensky: https://en.wikipedia.org/wiki/P._D._Ouspensky

Other resources

• Josh 31: https://uglyduckling.substack.com/p/josh-31

• The Telepathy Tapes: https://thetelepathytapes.com

• The Intuitive Hour: Awaken Your Inner Voice: https://open.spotify.com/show/3Ws7WZsFuQuwgl6uiPMJq3

• Spellers: https://www.imdb.com/title/tt27558471

• Animal Telepathy & Consciousness: https://thetelepathytapes.com/podcast/the-telepathy-tapes-s2-e5

• Brain-computer interface: https://en.wikipedia.org/wiki/Brain%E2%80%93computer_interface

• Saturday Night Live: https://en.wikipedia.org/wiki/Saturday_Night_Live

• Ouija: https://en.wikipedia.org/wiki/Ouija

• Kozmo: https://en.wikipedia.org/wiki/Kozmo.com

• Modern Meditations: Keith Rabois: https://www.generalist.com/p/modern-meditations-keith-rabois

• Y Combinator: https://www.ycombinator.com

• Pokémon Go: https://www.pokemongo.com

• Ingress: https://ingress.com

• Monopoly Go: https://www.monopolygo.com

• Pikmin Bloom: https://pikminbloom.com

• Niantic: https://nianticlabs.com

• Niantic Spatial: https://www.nianticspatial.com

• This Beanie Is Designed to Read Your Thoughts: https://www.wired.com/story/this-beanie-is-designed-to-read-your-thoughts

• CTRL-Labs: https://dsvw8jmuo45j8.cloudfront.net

• Fungi Perfecti: https://fungi.com/pages/bees?srsltid=AfmBOoqnjt2xKnfY7H3bW_yLaUuVrvoogzsa8mt5BHEqN_iWze0WQEkx

• Etsy: https://www.etsy.com

• Busright: https://www.busright.com

• Rizz & Tizz (and a Little Biz): Our Framework for Magical Weirdness: https://www.longjourney.vc/news/founderrizzandtizz

• Crusoe: https://www.crusoe.ai

• Mafia: https://mafia.gg

• B Street Books: https://www.bstreetbooks.com

• Automat: https://en.wikipedia.org/wiki/Automat

• The Restaurant With No (Visible) Workers: https://www.npr.org/sections/alltechconsidered/2015/08/31/436377616/the-restaurant-with-no-visible-workers

• “The Magic Glasses” by Frank Harris: https://hermetic.com/crowley/equinox/i/i/eqi01007

• Sufism: https://en.wikipedia.org/wiki/Sufism

• Kabbalah: https://en.wikipedia.org/wiki/Kabbalah

• The Razor’s Edge (1946 film): https://en.wikipedia.org/wiki/The_Razor%27s_Edge_(1946_film)

• The Razor’s Edge (1984 film): https://en.wikipedia.org/wiki/The_Razor%27s_Edge_(1984_film)

Subscribe to the show

I’d love it if you’d subscribe and share the show. Your support makes all the difference as we try to bring more curious minds into the conversation.

YouTube

Spotify

Apple

Production and marketing by penname.co. For inquiries about sponsoring the podcast, email jordan@penname.co.

Friends,

Sometime, about three months ago, I had a realization. It may sound obvious to many of you, and absurd to others.

It is this: the Claude models released in December (and improved upon since) have driven the greatest personal capability leap in my lifetime.

The iPhone changed the way I accessed information and navigated the world. Social networks altered how I communicated and fraternized. Any number of apps and programs improved my workflows, smoothed opportunities, and accelerated my thinking. Doubtless, dozens of quieter improvements have changed the texture of my life without me realizing it — superior medicines and more efficient machines.

But as a single, discrete jump, nothing else comes close. Before Opus 4.5 — Claude’s winter upgrade — I had opened my computer’s terminal on a handful of occasions, mostly by accident. The last time I’d actually tried to use it had been a 2015 continuing education course at General Assembly that I’d bungled through with the grace of a macaque running a printing press.

Today, I spend more than 70% of my working hours in the terminal. I build software, construct systems, and conduct sprawling research assignments. Agents parse hundreds of articles, looking for tidbits aligned with my interests or writing goals, and prepare reports in my preferred style. Others streamline administrative duties, find bugs, polish edges, and ship the next feature I have in my mind.

The result is a system that increasingly feels not just like an extra employee, but closer to 20.

One of the challenges of running a business like The Generalist is that you cannot play just one role. I may get the greatest pleasure from sitting in front of an empty screen and beginning to write, but there are bills to be paid, scheduling to do, dashboards to survey, growth strategies to deploy, sponsors to consider, and emails to send. Every day is a battle between deep work and the realities of running a business. It is an unavoidable fact that every hour spent on these things is one I cannot spend on the parts I enjoy most and where I feel I have the greatest advantage.

Even outside these especially mundane tasks, there are thousands of little irritations and distractions. Try to conduct detailed research on a person or topic, and count the moments of friction. How many paywalls will you hit for publications you already purchase? How many pop-up ads or cookie advisories? What purposefully distracting news carousel will spin across the bottom of your article? What video will stuff itself in from the side? If you want to study someone’s background on LinkedIn or Twitter, how many red notifications will raise themselves like welts? How many DMs call for your attention? What new product announcement screams for you from the feed? How many greedy fingers try to grab your focus in the simple act of trying to travel from A to B, from intent to information?

The modern internet has become an attention casino, and we have grown accustomed to working in the middle of it, averting our eyes from clanking slots and spinning wheels. But what if your workspace could feel more like a quiet desk in a well-lit library? What if you could dispatch 20 virtual employees to wade through the morass of the modern web for you and deliver the results? What if you could cut back on the logistics and wrangling and give yourself the time to focus on the one thing you do not want AI to do for you? What if you could double the time you spent in deep work?

Over the past few months, I have been using Claude Code to explore these questions for myself. The result is a full-stack knowledge management system spanning multiple software products, internal systems, and a local model. (I have built several other things, but this is the crispest distillation.) It is not perfect, nor will it be for everyone. Showing these systems to friends and family over the past months, I’ve found that it often helps people better understand what is now possible with these models, and how you might use them to your benefit.

When I’ve caught up with these people days or weeks later, I’ve often found they’ve undergone a similar transformation, from dabbler to devotee. While there’s undoubtedly a risk in outsourcing too much, including the act of thinking itself, so far I haven’t found this to be true for myself or others. Rather, people seem to be having an extremely fun time, building systems that take away drudgery and interfaces explicitly shaped for how their minds work. It’s incredible how much easier and more pleasant it is to navigate an app that intuitively works the way your brain does, rather than one constructed for the general population.

If you’re curious about these tools, I would recommend setting yourself the goal of building the smallest possible thing you can think of. You’ll quickly find your ambitions grow as your comfort sitting in front of a terminal does.

With that, here’s a look at my system and how it’s changed my workflow.

The Library

There is a particular type of cognitive annoyance which one might call “thing-finding.” You have undoubtedly experienced it. It is the moment when you’re forced to stop in the middle of something and ask, “Agh, what was that thing again? That thing that person said? That thing I read? That thing I wrote? Did I save it somewhere? Is it in Google Drive? Will Finder locate it? (Nope!)”

An occupational hazard of running a media company is that there are a lot of things. There are research things and podcast things and article things and note things and PDF things and email things and interview things. And they are invariably quite hard to find. Historically, they have been spread out across my desktop, Google, Obsidian, Ulysses, The Generalist’s website, and Dropbox. In the grand scheme of the travails of the world, the callous mysteries of the universe, dark matter, and the infinite unknown, I recognize this is not a real problem. But when you are in the middle of a piece, just beginning to find that fragile rhythm that even an ambling ice cream truck can disrupt, this is among the most annoying interruptions possible.

Delphi, the all-seeing oracle, is the solution I’ve built to address this issue.

To start, I compiled every piece The Generalist has ever written, the transcripts of every podcast I have recorded, my full compendium of Obsidian notes, my Readwise highlights, a good chunk of Google Drive, and assorted files from my desktop. In total, there are more than 45,000 searchable “chunks,” with more added by the day.

The search pipeline relies on three layers, fused together: vector search via Voyage-3 embeddings, keyword search via SQLite FTS5, and a locally trained cross-encoder reranker, distilled from Cohere. I created the reranker by starting with a compact, open-source model pre-trained on Microsoft’s MS MARCO search dataset and fine-tuning it on nearly 40K query-passage pairs from our data. That taught the model what “relevant” looks like from our initial Cohere setup, and allowed Delphi to deliver high-quality results more quickly and cheaply.

If you don’t know what this means, don’t worry. It’s absolutely not necessary to get this deep into the weeds. I certainly didn’t expect to fine-tune even a tiny model of my own, but bit by bit, you start to become interested in what else you can do. The system above is, I’m sure, far from perfect, but it’s working well for me at the moment.

Now, using Delphi, I can type in a half-formed query like “remind me what philosopher Karol talked about on the podcast,” and it will recall that Karol Hausman, CEO of Physical Intelligence, shared his interest in Spinoza.

If I want to ask something that requires referencing multiple sources—for example, what CEOs have said about their hiring practices—it can handle that too, pulling in references from across my corpus.

I could have answered these questions in the past. The first one would have taken me a few minutes of pecking and tabbing; the second perhaps several hours. In all likelihood, I simply wouldn’t have bothered.

As a destination, I don’t use Delphi that often, and I still think it can be greatly improved. The UX isn’t quite as polished as I would like, and search is good, but could still be faster and smarter. But it mostly does what I want it to, and it’s reassuring to know that it’s there when I need it. As you’ll see, its power is leveraged in our other tools.

The Researchers

The most useful thing I have built is not visible, nor easily explicable. It is a concatenation of skills, tools, techniques, and preferences that allow me to gather information widely and thoroughly, without having to do the hunting and pecking myself.

Fundamentally, my research system operates through a collection of agents dedicated to snuffling out relevant information from particular mediums. One scans my internal corpus for existing work. Another reads relevant articles. A third seeks out podcast appearances.

Like Liam Neeson’s protagonist from Taken, each of these has been equipped with a “particular set of skills” that aid them in their quest. Jina and Firecrawl turn webpages into clean, readable text. An open-source tool searches YouTube and pulls crisp transcripts of interviews. Various scripts search for a subject’s published writing (a blog or personal webpage) or secondary media appearances. A headless browser allows agents to access articles on paywalled sites for which I have subscriptions. Instead of having to check The Financial Times, The Economist, and The New Yorker one by one, the agent can do it for you, as long as you’re logged in.

Crucially, it can do all of this in the background while you work on something else. Not only have you avoided the attention pitfalls and switching costs imposed by the modern web, but you’ve effectively hired a capable research assistant. I had always hoped The Generalist would grow large enough for it to make sense for me to hire such a person; now, I have been granted a dozen, each with a very strong grasp on what I am likely to find most relevant.

To ensure the research is conducted at a high standard, I’ve created a set of skill files for these agents to reference. These include dictums such as:

Read more

"Most often, we tend to conflate innovation and novelty. But if I look around, most things that have changed the fabric of our lived experience were actually not new when they did." — Viswa Colluru

For decades, drug discovery has shifted away from nature and toward biology-first approaches. Viswa Colluru believes that shift was a catastrophic mistake. His company, Enveda Biosciences, has raised over $500 million to build a “search engine for nature’s chemistry.” The mission is personal: he grew up around his father’s pharmacy in India and later lost his mother to a treatable cancer whose medicine his family couldn’t afford. Many life-changing medicines, including morphine, aspirin, and metformin, originated in nature, but there has never been a reliable, scalable way to systematically explore its chemistry. Colluru founded Enveda in 2019 with $55,000 of his own savings to change that. The company has since identified 18 drug candidates, with three now in clinical trials.

In our conversation, we explore:

• Why the pharmaceutical industry abandoned nature (and why that was a massive mistake)

• How Enveda built a system to decode unknown molecules in nature

• The deeply personal story of his mother’s battle with leukemia and how it shaped his life’s work

• Why old ideas, from immunotherapy to natural products, often hold the most latent potential

• How Enveda developed 18 drug candidates for about $1 million each instead of $10-15 million

• Enveda’s three leading drug candidates targeting eczema, obesity, and ulcerative colitis

• Why first-in-class medicines capture the vast majority of returns in pharma

• What competitive table tennis taught him about building companies

Thank you to the partners who make this possible

Brex: The intelligent finance platform.

Ahrefs Brand Radar: Find your brand in AI results.

Persona: Trusted identity verification for any use case.

Explore the episode

Timestamps

(00:00) Introduction to Viswa Colluru

(03:57) His father’s pharmacy and early exposure to Western and Ayurvedic medicine

(07:06) Early pull toward technology

(09:29) His mother’s leukemia diagnosis

(14:24) Studying Biotechnology

(16:07) Graduate school

(17:55) Studying immunotherapy when it was unfashionable

(24:23) Innovation vs. novelty

(27:24) Lessons from table tennis

(32:05) Joining Recursion

(37:10) Learning urgency and courage

(40:42) What launched Enveda

(45:40) The limits of reductionist drug discovery

(49:53) Chemistry-first approach

(52:17) Raising $225K and investing $55K personally

(56:04) Initial studies and targets

(1:04:30) Three categories of leading drugs: Eczema, obesity, ulcerative colitis

(1:13:27) Why GLP-1s are not the whole answer

(1:18:27) Enveda’s long-term vision

(1:21:31) Book recommendation

Follow Viswa Colluru

LinkedIn: https://www.linkedin.com/in/viswacolluru

X: https://x.com/viswacolluru

Resources and episode mentions

Books

• Hit Refresh: The Quest to Rediscover Microsoft’s Soul and Imagine a Better Future for Everyone: https://www.amazon.com/Hit-Refresh-Rediscover-Microsofts-Everyone-ebook/dp/B01HOT5SQA

• The World of the Cell: https://www.amazon.com/World-Cell-7th-Wayne-Becker/dp/0805393935

• String Theory: David Foster Wallace on Tennis: A Library of America Special Publication: https://www.amazon.com/String-Theory-Wallace-Library-Publication/dp/1598534807

• The Vital Question: Energy, Evolution, and the Origins of Complex Life: https://www.amazon.com/Vital-Question-Evolution-Origins-Complex/dp/0393088812

People

• Satya Nadella on X: https://x.com/satyanadella

• Pablo Lubroth on LinkedIn: https://www.linkedin.com/in/pablolubroth

• David Foster Wallace: https://en.wikipedia.org/wiki/David_Foster_Wallace

• Chris Gibson on LinkedIn: https://www.linkedin.com/in/chris-gibson-5ab66065

Other resources

• Enveda: https://enveda.com

• Chronic Myeloid Leukemia: https://my.clevelandclinic.org/health/diseases/21845-chronic-myelogenous-leukemia-cml

• Gleevec: https://www.novartis.com/us-en/content/gleevec%C2%AE-tablets

• The Moat of Low Status: https://sashachapin.substack.com/p/the-moat-of-low-status-68a

• Learn to love the Moat of Low Status: https://usefulfictions.substack.com/p/learn-to-love-the-moat-of-low-status

• Roger Federer vs Pete Sampras | Wimbledon 2001 fourth round | Full Match: https://www.youtube.com/watch?v=9uJ2IyL0cPc

• Goliath: https://en.wikipedia.org/wiki/Goliath

• Recursion: https://www.recursion.com

• Celebrating the Departure of an Outstanding Employee (Chris Gibson’s post about Viswa): https://medium.com/recursion-pharmaceuticals/celebrating-the-departure-of-an-outstanding-employee-65db2455fd1f

• Silk Road: https://en.wikipedia.org/wiki/Silk_Road

• Artemisinin: https://en.wikipedia.org/wiki/Artemisinin

• GW Pharmaceuticals: https://en.wikipedia.org/wiki/GW_Pharmaceuticals

• Cannabidiol: https://en.wikipedia.org/wiki/Cannabidiol

• CERN: https://home.cern

• GWAS: https://www.genome.gov/genetics-glossary/Genome-Wide-Association-Studies-GWAS

• Charlie Munger’s quote: https://www.goodreads.com/quotes/8690661-take-a-simple-idea-and-take-it-seriously

• Modern drug discovery using ethnobotany: A large-scale cross-cultural analysis of traditional medicine reveals common therapeutic uses: https://pubmed.ncbi.nlm.nih.gov/37701812

• Analysis of virus genomes from glacial environments reveals novel virus groups with unusual host interactions: https://pmc.ncbi.nlm.nih.gov/articles/PMC4490671/

• Rinvoq: https://www.rinvoqhcp.com

• Dupixent: https://www.dupixent.com

• Humira: https://www.humira.com

• Singulair: https://www.drugs.com/singulair.html

• Eli Lilly: https://www.lilly.com

• Keytruda: https://www.keytruda.com

• Merck: https://www.merck.com

Subscribe to the show

I’d love it if you’d subscribe and share the show. Your support makes all the difference as we try to bring more curious minds into the conversation.

YouTube

Spotify

Apple

Production and marketing by penname.co. For inquiries about sponsoring the podcast, email jordan@penname.co.

"Within 2 or 3 years, AI mathematicians will surpass human mathematicians for any specific mathematical task. I don't think it'll be a decade, like some people say." — Tudor Achim, CEO of Harmonic

Tudor Achim is the co-founder and CEO of Harmonic, a startup working to solve one of AI’s hardest problems: mathematical reasoning. In July 2024, Harmonic achieved gold-medal-level performance on International Math Olympiad problems alongside systems from OpenAI and Google DeepMind—but with a key difference: every proof Harmonic submitted was formally verified. Tudor's path to Harmonic wound through competitive piano, computational biology, and autonomous driving. He studied at Carnegie Mellon's music preparatory school, worked on machine learning at Quora, briefly pursued a PhD before dropping out, and then co-founded an autonomous driving company, Helm.ai. Harmonic's core product, Aristotle, uses reinforcement learning and the programming language Lean 4 to solve problems and verify solutions.

In our conversation, we explore:

• Why Tudor believes math is the fundamental toolkit to understand the world

• How Harmonic uses hallucinations as a feature, not a bug

• How Aristotle works and the applications beyond pure mathematics

• The reinforcement learning process that lets Harmonic generate synthetic training data and solve problems humans have never attempted

• Why Tudor believes AI could surpass human mathematicians on specific tasks within 2–3 years

• Why the future of mathematics looks more like GitHub than academic journals

• The alternating pattern between intellect leaps and data leaps throughout scientific history

• How studying piano under an extraordinary teacher taught Tudor discipline and the value of sticking with hard problems

Thank you to the partners who make this possible

Brex: The intelligent finance platform.

Guru: The AI source of truth for work.

Rippling: Stop wasting time on admin tasks, build your startup faster.

Explore the episode

Timestamps

(00:00) Intro

(03:34) From competitive piano to computer science

(06:28) The mathematical foundations of music (and why Tudor keeps them separate)

(08:24) Can AI ever create art with true intent?

(09:51) Early obsessions

(12:52) Defining intelligence

(14:49) Discovering machine learning’s potential at Quora

(17:30) Why Tudor chose computational biology for his PhD

(19:19) The decision to drop out and build Helm.ai

(22:55) The two breakthroughs that made mathematical AI possible in 2023

(25:28) The importance of Lean 4

(28:21) How Tudor and Vlad Tenev discovered they shared the same impossible dream

(32:35) Why formal verification became the core conviction

(34:21) The timeline for AI surpassing human mathematicians

(35:25) An overview of Aristotle: the world’s first always-correct mathematical agent

(38:12) Why Tudor says hallucinations are the engine of creativity

(39:30) The translation challenge from natural language to formal proof

(40:40) Reinforcement learning

(42:10) Why Aristotle is both faster and cheaper than alternatives

(43:34) Tradeoffs and use cases

(45:34) Math in AI now and what’s next

(47:38) Tying with OpenAI and DeepMind at the International Math Olympiad

(49:08) Democratizing AI and correctness

(53:13) Tudor’s 2030 thesis

(56:02) History’s alternating rhythm of thinking and measuring

(57:53) What Tudor has been wrong about

(58:52) What Tudor’s best at

(1:00:18) Final meditations

Follow Tudor Achim

LinkedIn: https://www.linkedin.com/in/tudorachim

X: https://x.com/tachim/with_replies

Resources and episode mentions

Books

• Anna Karenina: https://www.amazon.com/Anna-Karenina-Leo-Tolstoy/dp/0143035002

People

• Mineko Avery’s obituary: https://obituaries.post-gazette.com/obituary/mineko-avery-1089363974

• Stefano Ermon on LinkedIn: https://www.linkedin.com/in/ermon

• Vladislav Voroninski on LinkedIn: https://www.linkedin.com/in/vladislav-voroninski-83200213

• Leonardo de Moura’s website: https://leodemoura.github.io

• Vlad Tenev on LinkedIn: https://www.linkedin.com/in/vlad-tenev-7037591b

• Vladimir Novakovski on X: https://x.com/vnovakovski

• Terence Tao on LinkedIn: https://www.linkedin.com/in/terence-tao-6291246

• Grigori Perelman: https://en.wikipedia.org/wiki/Grigori_Perelman

Other resources

• Harmonic: https://www.harmonic.fun

• Carnegie Hall: https://www.carnegiehall.org

• Carnegie Mellon Music Preparatory School: https://www.cmu.edu/cfa/music/preparatory-summer-programs/index.html

• Why A.I. Isn’t Going to Make Art: https://www.newyorker.com/culture/the-weekend-essay/why-ai-isnt-going-to-make-art

• Quora: https://www.quora.com

• The Unreasonable Effectiveness of Mathematics in the Natural Sciences: https://webhomes.maths.ed.ac.uk/~v1ranick/papers/wigner.pdf

• Helm: https://helm.ai

• Honda Introduces Next-generation Technologies for Honda 0 Series Models at Honda 0 Tech Meeting 2024: https://global.honda/en/newsroom/news/2024/c241009eng.html

• Lean: https://lean-lang.org

• Mathlib: https://lean-lang.org/use-cases/mathlib

• Aristotle: https://aristotle.harmonic.fun

• Poincaré conjecture: https://en.wikipedia.org/wiki/Poincar%C3%A9_conjecture

• OpenAI: https://openai.com

• DeepMind: https://deepmind.google

• International Mathematical Olympiad (IMO): https://www.imo-official.org

• Go: https://en.wikipedia.org/wiki/Go_(game)

Subscribe to the show

I’d love it if you’d subscribe and share the show. Your support makes all the difference as we try to bring more curious minds into the conversation.

YouTube

Spotify

Apple

Production and marketing by penname.co. For inquiries about sponsoring the podcast, email jordan@penname.co.

Friends,

You didn’t mind speaking with Hanns Scharff. No, it was better than that — you liked it. He spoke English well, for one thing, the product of years spent in South Africa and a wife from London. But that didn’t fully explain it either. He spoke softly (unlike the others), was good for a joke or a story, and when he directed his dark, thoughtful eyes in your direction, you didn’t feel fearful, but at ease.

You could almost forget that he was the Luftwaffe’s most effective interrogator. And you, his prisoner.

While the Nazi Party’s other torturers and wheedlers relied on threats and violence, Scharff found success with a more genteel approach, taking downed pilots on long walks in the Taunus hills northwest of Frankfurt, during which he seemed to avoid discussing military matters. Only later, in some cases much later, would prisoners realize what had happened.

One American pilot recalled such a stroll. Only after he and Scharff had wandered and chatted for a while did the German mention, almost in passing, that a chemical shortage seemed to have impacted American munitions: their tracer bullets now trailed white smoke rather than their usual red.

No, no, the American told him. That wasn’t caused by a chemical shortage; it was a matter of design. American tracer bullets shifted from red to white when a pilot was running low on ammunition. It was a kind of warning system.

There it was. By purposefully saying the wrong thing, Scharff prompted the pilot to correct him, all without asking a question. A pocket had been picked without the wallet’s owner even registering a rustle. The pilot would not realize what had happened until it was much too late.

This technique, and others like it, are the topic of Confidential by John Nolan, a 1999 book that is as fascinating as it is difficult to obtain. Though prized by intelligence officials and professional “elicitors,” Confidential is no longer in print and only available to buy second-hand. To grab my copy, I spent a few hundred dollars on eBay and waited weeks for it to arrive. All of which only adds to its strange allure, as if someone decided Nolan’s work was too useful to simply leave lying around.

Confidential is a wildly entertaining and impressively insightful book. In studying it closely these last few months, I’ve also come to believe it’s an important one. Though Nolan is ostensibly writing for the professional intelligence gatherer, his conversational techniques are useful to anyone, in any context. They are liable to make you more engaging and persuasive, as well as a better conversationalist.

It is also worth knowing when someone else is using them. Why did that salesperson seem to purposefully misspeak? Was I imagining it, or did that headhunter seem to disbelieve everything I said? What is it about this person that makes me want to open up so much? For founders working in sectors of national interest, Confidential will help you protect what you know. If you are building almost anything of note, there is a good chance that someone out there — whether in a bland concrete building, a glassy office tower, or a grassy tech campus — would love to understand it better than you’d like them to.

This is the second piece in an occasional series about books that change how you see everyday interactions. The first, on Keith Johnstone’s Impro, explored the invisible power dynamics in every conversation. Confidential picks up the other side of that coin: how information actually moves between people, and what a former spy figured out about making it move faster.

If you’d like to read more work like this — pieces that dig into overlooked ideas and the people behind them — a premium membership is $22/month or $220/year. Subscribe now.

Subscribe now

Lessons from Confidential

Like a character in Atlas Shrugged, the natural question when beginning Confidential is, “Who is John Nolan?”

The first thing to say is that this is the name you would want for a spy — a vaguely heroic sound being stirred into a bowl of porridge.

There are, fittingly, few details online, so we must rely on Nolan’s own telling. For twenty-two years, he worked as a spy. From the sparse available details, it seems Nolan spent time in some of the intelligence community’s more controversial programs, a background that lends Confidential both its authority and occasional chill.

After his time working for the government, Nolan founded a corporate espionage consultancy that advised business clients and gathered intel on their behalf. (One of the only articles I can find that mentions Nolan outside of Confidential covers an espionage campaign conducted by P&G against Unilever in 2000 to obtain the “secrets of shampoo.” Nolan’s firm was ostensibly the orchestrator.)

As part of his work, Nolan’s team relied on the psychological tools he outlines to extract sensitive information — all while being perfectly explicit about who they were. Beyond his team, Nolan also trained executives to use his techniques and protect themselves against him.

Nolan’s own “call to adventure” is a memorable starting point for the book. In 1960s New Jersey, Nolan started work as a typewriter salesman for a small local outfit, competing with a rep from IBM. Despite offering a superior product, Nolan struggled to shift them. While he flailed around trying to convince companies to have him in for an appointment, his rival was having companies call him - everyone knew that if you needed a typewriter, you went to “Big Blue.”

Sitting in a coffee shop one day, Nolan watched as the IBM salesman loaded typewriters into his station wagon. “In a brief moment of clarity,” Nolan came to a realization. Why bother to hunt and scrabble for customers when he could simply follow his rival and figure out who was in the market for typewriters?

For the rest of the day, he tailed the station wagon, watching it go from office to office. The next morning, he set out and visited every company, one after another, showcasing what his product could do. That week, he sold twelve typewriters.

Over the following months, he repeated the trick, shadowing the IBM rep a few days each week. He grew cocky enough to wait outside an office building and follow him an hour later. Without meaning to, Nolan stumbled into the world of intelligence gathering and had seen what it could yield.

This is the first of Nolan’s rollicking stories, but it would be wrong to classify this as a collection of yarns. Across Confidential’s 350 pages, Nolan outlines techniques of striking psychological acuity, interleaved with lessons from the history of espionage, and detailed examples. On a given page, you’re just as likely to learn about the subterfuges Johnson & Johnson deployed to defend the Tylenol market as to analyze the brilliant sinuousness of Sherlock Holmes’s questioning style.

For this piece, we’ll focus mainly on Part I: “Eliciting the Information You Want and Need.” Though the latter two parts offer interesting details, they primarily address how organizations can collect intelligence more effectively or protect against spies.

As the title of Part I suggests, it covers the art of “elicitation.”

Even if you are familiar with this word, its place in the Nolan lexicon is particular and benefits from definition. When the author writes about elicitation techniques, he explicitly means the following:

Elicitation…is defined as that process which avoids direct questions and employs a conversational style to help reduce concerns and suspicions—both during the contact and in the days and weeks to follow—in the interest of maximizing the flow of information.

As Nolan explains, elicitation is expressly distinct from interrogation and interviewing. “Interrogation [is] obtaining what you want from someone who possibly has it, who has not admitted to having it, and who knows who you are and why you want it,” he writes. Meanwhile, “interviewing is the process of obtaining information from someone who probably has it, who has more or less admitted to having it, and who knows who you are and why you want it.” Interrogation is, by definition, adversarial, while interviewing tends not to be.

As you’ll see, elicitation is a subtler dance.

The worst way to get an answer might be a question

Read more

“I don't think Silicon Valley knows anything about networking anymore.” — Anil Varanasi

Anil Varanasi, co-founder and CEO of Meter, is building a new kind of networking company for the AI era. Alongside his brother Sunil, he has helped raise more than $250 million to challenge incumbents like Cisco with a vertically integrated approach spanning hardware, software, deployment, and ongoing operations, all delivered through a utility-style model. His view is that networking has remained largely unchanged for decades, even as it has become foundational to everything from AI workloads to real-world infrastructure. Meter’s ambition is not just to improve existing networks, but to make them autonomous over time. Before starting the company, Anil and Sunil were deeply involved in filmmaking, a background that still shapes their philosophy of building with cathedral-level craft across every layer of the stack.

Together we explore:

• The “burden of knowledge” and why progress is getting harder across fields

• Why most companies over-index on technology and ignore business model innovation

• The three ways companies create advantage: technology, delivery, and business model

• How Meter’s trade-in model borrows from the automotive industry

• Why networking should function like electricity or water—not hardware

• Lessons from Japanese vending machine logistics for infrastructure deployment

• The hidden coordination problem behind vertically integrated companies

• Why Anil believes “common knowledge” is often wrong

• How COVID forced Meter to abandon geographic constraints and scale nationally

• The case for fully autonomous networks in a world of exploding demand

Thank you to the partners who make this possible

.tech domains: An identity for builders at their core.

Granola: The app that might actually make you love meetings.

Brex: The intelligent finance platform.

Explore the episode

Timestamps

(00:00) Introduction to Anil Varanasi and Meter

(03:52) The burden of knowledge and slowing innovation

(08:18) Losing creativity vs gaining expertise

(10:25) What Meter actually does

(13:26) Early life, immigration, and upbringing

(15:47) Parental influence

(20:03) Film, storytelling, and creative influence

(22:55) Why Anil didn’t pursue filmmaking

(25:44) Parallels between company building and filmmaking

(27:00) Early programming and building

(28:05) George Mason and understanding systems

(29:59) The dynamic of working with his brother as a co-founder

(34:03) His first business and lessons learned (or lack thereof)

(35:15) Lessons from successful companies

(38:16) Japanese vending machines and logistics insight

(41:10) Scrapping 18 months of work

(42:40) Conviction and long-term company building

(46:02) COVID shock and near-death moment

(49:59) Building hardware like a cathedral

(52:25) Rethinking the networking business model

(57:06) Build vs buy and transaction costs

(59:39) Networking as infrastructure and utility

(01:01:30) The case for autonomous networks

(01:03:25) Hiring, talent, and what actually matters

(01:06:15) Big unanswered questions (sleep, science)

(01:07:28) Rethinking education

(01:09:02) Infinite games and long-term thinking

Follow Anil Varanasi

LinkedIn: https://www.linkedin.com/in/anilcv

X: https://x.com/acv

Website: https://anilv.com

Resources and episode mentions

Books

• The Great Stagnation: How America Ate All The Low-Hanging Fruit of Modern History, Got Sick, and Will (Eventually) Feel Better: https://www.amazon.com/Great-Stagnation-Low-Hanging-Eventually-eSpecial-ebook/dp/B004H0M8QS

• Finite and Infinite Games: https://www.amazon.com/Finite-Infinite-Games-James-Carse/dp/1476731713

People

• Matt Clancy’s blog: https://www.newthingsunderthesun.com

• Warren Buffett: https://en.wikipedia.org/wiki/Warren_Buffett

• Charlie Munger: https://en.wikipedia.org/wiki/Charlie_Munger

• Orson Welles: https://en.wikipedia.org/wiki/Orson_Welles

• Ronald Reagan: https://en.wikipedia.org/wiki/Ronald_Reagan

• Arnold Schwarzenegger: https://en.wikipedia.org/wiki/Arnold_Schwarzenegger

• P.C. Sreeram: https://www.imdb.com/name/nm0820269

• Satyajit Ray: https://en.wikipedia.org/wiki/Satyajit_Ray

• Mani Ratnam: https://en.wikipedia.org/wiki/Mani_Ratnam

• Elon Musk on X: https://x.com/elonmusk

• Ashlee Vance: https://www.ashleevance.com

• Dwarkesh Patel’s podcast: https://www.dwarkesh.com

Other resources

• Meter: https://www.meter.com

• Age: https://anilv.com/age

• Coefficient Giving: https://coefficientgiving.org

• George Mason University: https://www.gmu.edu

• You Are Not Late: https://medium.com/message/you-are-not-late-b3d76f963142

• The End of Asymmetric Information: https://www.cato-unbound.org/2015/04/06/alex-tabarrok-tyler-cowen/end-asymmetric-information

• Sam Altman on Trust, Persuasion, and the Future of Intelligence - Live at the Progress Conference (Ep. 259): https://conversationswithtyler.com/episodes/sam-altman-2

• P. C. Sreeram (Babu Export Company): https://babuexportcompany.com/9381/

• The Apu Trilogy: https://www.criterion.com/boxsets/1145-the-apu-trilogy?srsltid=AfmBOor2rfr6GalePfNZXS7QiljZiehumuoJXvu4TyL2szofL16qrAIp

• Nayakan: https://www.imdb.com/title/tt0093603

• Marty Supreme: https://www.imdb.com/title/tt32916440

• Ford: https://www.ford.com

• Toyota: https://www.toyota.com

• Intel: https://www.intel.com

• Adobe: https://www.adobe.com

• Cisco: https://www.cisco.com

• NVIDIA: https://www.nvidia.com

• Things we don’t understand: https://anilv.com/understand#

• Waldorf education: https://en.wikipedia.org/wiki/Waldorf_education

Subscribe to the show

I’d love it if you’d subscribe and share the show. Your support makes all the difference as we try to bring more curious minds into the conversation.

YouTube

Spotify

Apple

Production and marketing by penname.co. For inquiries about sponsoring the podcast, email jordan@penname.co.

Friends,

If Silicon Valley has any religion, it is that of the founder. Nowhere else puts as much faith in, nor grants as much latitude to, sovereign individuals attempting to build something from scratch. Only within this strip of approximately 35 miles might a broke 20-year-old in pajama pants and Adidas sliders command greater reverence than a celebrated researcher, diligent doctor, or decent executive. This is the strangeness of Silicon Valley and its genius.

The cult of the founder has enjoyed a fresh, febrile burst. Now, more than anytime in the last decade, operating in “founder mode” — the term popularized by Paul Graham’s post — is seen as synonymous with efficacy. Managers (that wretched, blighted species) are viewed not only as less productive but less legitimate, usurpers and meddlers that merely disrupt the glowing chi that stems from the central chakra of those who build.

Look across the tech landscape, however, and there is one manager that bears closer inspection: Satya Nadella. Since his appointment as Microsoft CEO in 2014, few executives boast a more impressive record. Given Microsoft’s current strengths, it is easy to forget the company Nadella inherited. Unlike Tim Cook, who stepped into an innovative organization still in the early innings of capitalizing on a new product category, Nadella stepped into a company that was culturally rotten, creatively blocked, and stuck with a sideways stock price. It is true that Ballmer had sown the seeds for a cloud computing renaissance, as we’ll discuss, but this was far from the finished article.

In the intervening 12 years, Nadella not only drove the company to a $3 trillion market cap but also oversaw an authentic internal revolution, expanded its product suite, and positioned Microsoft to keep pace in the AI era. He has done so while portraying himself as the consummate modern manager, fond of borrowing from the Buddha, and peddling the MBA-circuit bon mots of empathetic leadership and a “growth mindset.” Nadella’s own chronicle of his turnaround, Hit Refresh, is stuffed with such cheery banalities. While the great CEOs of the past and current generation are prone to fits of rage, savage dressing-downs, and impossible expectations, Nadella appears genuinely reasonable, a happy guru who would like you to work hard, sure, but don’t forget to take time for your family and maybe a restorative hobby.

How has he done this? How does a peacetime CEO win in a war zone? Can one really win at this scale without the animal intensity of Musk or Huang? Is the balmy public presentation the whole story?

To answer these questions, I have spent the past three months studying Nadella’s leadership from as many angles as possible. That includes Hit Refresh, Acquired’s two-part series on Microsoft before Nadella, a slew of podcasts and long-form articles, internal emails released in court filings, annual shareholder letters, and confidential expert interviews with former Microsoft executives.

What emerged is a nuanced portrait of how a manager built fresh power structures beneath him, constructed new mythologies, reset cultural norms, and developed founder-like authority.

This piece is part of The Generalist’s ongoing series of managerial “playbooks,” exclusively available to premium subscribers. You can find our previous editions on Elon Musk, Jeff Bezos, and Jensen Huang here.

Our mission, across all of these playbooks, is to reveal the real strategies legendary entrepreneurs use to build their businesses. These are often uncomfortable and in direct conflict with traditional managerial advice. However, if you believe progress depends on innovation, as we do, then understanding these principles, foibles included, is not only interesting but essential.

To unlock all four playbooks and everything else The Generalist has to offer, join us now for $22/month. You’ll get immediate access to our best long-form writing, company case studies, exclusive interviews, and private databases.

Subscribe now

Here is the Satya Nadella Playbook:

1. Manifest authority through mythology.

2. Borrow power from the old regime (even as you counterposition against it).

3. Remake the aristocracy beneath you.

4. Make it safe to fail.

5. Once the narrative is set, use it as cover.

6. Hone your sharpest knife.

7. If you can’t win the future, at least don’t lose it.

In each section, we’ll unpack the strategies behind these principles and outline their benefits and tradeoffs.

What to expect

• A 10,000+ word playbook of tech’s most effective non-founding CEO

• How Nadella earned founder-like authority without founding anything

• How Nadella dismantled Microsoft’s infamous stack ranking culture

• The bathroom-break decision that opened Azure to Linux

• The $2.5 billion acquisition that had nothing to do with productivity (and everything to do with distribution)

• The licensing maneuver that imposed a 400% tax on competitors’ cloud customers

• How a panicked 2019 email led to the $13 billion OpenAI bet

• Over 100 hours of research, confidential executive interviews, and court filings distilled

…and much more. To unlock the full playbook and learn how a “safe pick” turned a stagnant giant into a $3 trillion force, join our premium newsletter today.

Brought to you by Vanta

Learn How to Automate Compliance for SOC 2, ISO 27001, and More with Vanta

With customer expectations rising and compliance needs shifting, keeping up can be daunting. Vanta’s Agentic Trust Platform helps fast-moving startups and security teams get audit-ready fast and stay continuously compliant, turning compliance into a deal accelerator, not a blocker.

Join to learn how Vanta can help you:

• Automate evidence, policies, and remediation across SOC 2, ISO 27001, HIPAA, HITRUST, ISO 42001, and more

• Build real security foundations, not check-the-box fixes

• Show credibility faster with a public Trust Center and AI-powered questionnaires

• Keep engineers focused with guided workflows and developer-native automation

Learn more

Part I: The Inheritance

By definition, a non-founding CEO does not start from scratch. They enter an environment of someone else’s making and must transform it into something of their own. To understand how Satya Nadella changed Microsoft then, we must first grasp the state of the company he inherited. It was one just emerging from what became known as its “lost decade.”

A stalled stock

When Steve Ballmer stepped into the CEO role in January 2000, he was taking the reins of the most valuable company on the planet. Less than three weeks earlier, Microsoft had hit a peak valuation of $615 billion, with a stock price approaching $60.

When the crash came, Microsoft cratered, dropping below $250 billion. It was not the fall that was remarkable, but what happened after. Or rather, what didn’t happen after. In the years that followed, as other wounded tech players stabilized and then climbed, Microsoft stayed stuck, even as its underlying performance improved. During Ballmer’s reign, revenue compounded from $23 billion to $86 billion while operating income improved from $11 billion to $28 billion. And yet, the stock barely moved, flatlining at about $30 a share. Over a similar timeframe — between late 2000 and mid-2012, Apple snowballed from a $4.8 billion pipsqueak into a $541 billion behemoth. By the time Nadella’s reign began, Microsoft was firmly in its shadow.

Read more

“I would argue the biggest risk is actually locking in a very narrow monoculture for superintelligence. One superintelligence is much less safe than infinite superintelligence.” — Vincent Weisser

Much of the fear around AI centers on misalignment – the idea that powerful systems might act against human interests. Vincent Weisser worries about something different: what happens if advanced AI systems are perfectly aligned with the interests of a small group of institutions? That concern led him to co-found Prime Intellect, a startup building open infrastructure for training and deploying advanced AI models. Before Prime Intellect, Weisser helped organize Vitalik Buterin’s Zuzalu experiment and worked in decentralized science, where he helped unlock roughly $40 million in funding for unconventional research. Today, he’s applying that same open ethos to AI, working to ensure the tools that shape superintelligence remain broadly accessible rather than concentrated in the hands of a few.

In our conversation, we explore:

• Why Vincent believes multiple superintelligences are safer than one

• The intellectual influences that shaped Vincent’s thinking about intelligence and progress, including David Deutsch and Nick Bostrom

• Prime Intellect’s evolution from distributed compute infrastructure to frontier model training and reinforcement learning tools

• Why Vincent believes open and decentralized science could accelerate discovery

• The Zuzalu experiment and what it suggests about the future of scientific communities

• The role of aesthetics and craft in building technology

• Why Europe might have a cultural advantage in a post-superintelligence world

• Vincent’s predictions for the next five years of AI

Thank you to the partners who make this possible

Granola: The app that might actually make you love meetings.

Brex: The intelligent finance platform.

Rippling: Stop wasting time on admin tasks, build your startup faster.

Explore the episode

Timestamps

(00:00) Introduction to Vincent Weisser

(03:28) The book behind Prime Intellect’s name

(07:35) The case for suffering

(09:35) An overview of Prime Intellect

(13:03) Why open source models matter

(21:18) Vincent’s intellectual influences

(25:17) Early years in the startup scene

(31:48) Funding science outside traditional institutions

(41:22) The past 6 months of AI progress

(43:45) Deciding to build Prime Intellect

(46:55) Why GPUs were the right starting point

(51:39) Training models on Prime Intellect

(59:48) Why beauty matters

(1:03:48) The Zuzalu experiment

(1:06:27) Prime Intellect’s AGI Easter egg

(1:11:13) Predictions for the next five years

(1:15:09) Final meditations

Follow Vincent Weisser

LinkedIn: https://linkedin.com/in/vincentweisser

X: https://x.com/vincentweisser

Goodreads: https://www.goodreads.com/user/show/69248416-vincent-weisser

Website: https://primeintellect.ai

Resources and episode mentions

Books

• The Metamorphosis of Prime Intellect: https://www.amazon.com/Metamorphosis-Prime-Intellect-Roger-Williams/dp/1411602196

• The Beginning of Infinity: Explanations That Transform the World: https://www.amazon.com/Beginning-Infinity-Explanations-Transform-World/dp/0143121359

• Superintelligence: Paths, Dangers, Strategies: https://www.amazon.com/Superintelligence-Dangers-Strategies-Nick-Bostrom/dp/0198739834

• Steve Jobs: https://www.amazon.com/Steve-Jobs-Walter-Isaacson/dp/1982176865

• The Singularity Is Near: When Humans Transcend Biology: https://www.amazon.com/dp/0143037889

• Physics of the Future: How Science Will Shape Human Destiny and Our Daily Lives by the Year 2100: https://www.amazon.com/Physics-Future-Science-Shape-Destiny/dp/0307473333

• A Pattern Language: Towns, Buildings, Construction (Center for Environmental Structure Series): https://www.amazon.com/Pattern-Language-Buildings-Construction-Environmental/dp/0195019199

People

• George Hotz on X: https://x.com/realGeorgeHotz

• Andrej Karpathy on X: https://x.com/karpathy

• Benjamin Bratton’s website: https://bratton.info

• David Deutsch’s website: https://www.daviddeutsch.org.uk

• Albert Einstein: https://en.wikipedia.org/wiki/Albert_Einstein

• Vitalik Buterin’s website: https://vitalik.eth.limo

• Celine Halioua’s website: https://www.celinehh.com

Other resources

• Prime Intellect: https://www.primeintellect.ai

• Anthropic: https://www.anthropic.com

• OpenAI: https://openai.com

• DeepSeek: https://www.deepseek.com

• Google DeepMind: https://deepmind.google

• Are We Alone In The Universe? Sara Seager on Exoplanets, Venus, and the Hunt for Alien Life (Astrophysicist and Planetary Scientist at MIT): https://www.generalist.com/p/are-we-alone-in-the-universe-sara-seager

• Y Combinator: https://www.ycombinator.com

• Loyal: https://loyal.com

• Cursor: https://cursor.com

• Zuzalu: https://zuzalu.city

• Why I Built Zuzalu: https://www.palladiummag.com/2023/10/06/why-i-built-zuzalu

• Dyson sphere: https://en.wikipedia.org/wiki/Dyson_sphere

Subscribe to the show

I’d love it if you’d subscribe and share the show. Your support makes all the difference as we try to bring more curious minds into the conversation.

YouTube

Spotify

Apple

Production and marketing by penname.co. For inquiries about sponsoring the podcast, email jordan@penname.co.

Karol Hausman is the co-founder and CEO of Physical Intelligence, a robotics company building a general-purpose “AI brain for the physical world.” The company has raised more than $1 billion in funding to develop foundation models that allow robots to operate across many machines, environments, and tasks rather than being programmed for a single purpose. The core thesis: the same scaling dynamics that transformed language models may also unlock robotic intelligence. But only if you resist every commercial pressure pushing you toward specialization. The central challenge isn’t mechanical design. It’s intelligence: how robots learn, generalize, and interact with a physical world that is far harder to simulate than it is to describe. Before launching Physical Intelligence, Karol worked at Google Brain and Stanford University, studying robot learning alongside researchers Sergey Levine and Chelsea Finn, who later became his co-founders.

In our conversation, we explore:

• How growing up in a small town in Poland and watching Star Wars sparked Karol’s fascination with robots

• The moment a lecture from Sergey Levine convinced him to abandon his PhD research direction and pivot fully to deep learning

• Why robotics has historically lagged behind breakthroughs in language models

• The case for building a general “AI brain” for the physical world rather than a single specialized robot

• The role of real-world data in training robots, the limits of simulation, and how deployment could create a powerful data flywheel

• The return of reinforcement learning and the parallels between human learning and robot training

• The unique challenges of physical intelligence and why robots must operate with far higher reliability than language models

Thank you to the partners who make this possible

Brex: The intelligent finance platform.

Granola: The app that might actually make you love meetings.

Explore the episode

Timestamps

(00:00) Intro

(04:05) Karol’s early fascination with robots

(07:38) How Karol relates to Fei-Fei Li’s biography

(08:52) What inspired Karol to build better robots

(11:19) Philosophical influences

(15:33) Parallels between The Inner Game of Tennis and robotics

(18:21) Karol’s entry point to robotics and PhD program

(25:49) Combining robotics with LLMs: The Taylor Swift demo

(30:48) The 1970s SHRDLU AI experiment

(32:33) Founding Physical Intelligence

(35:13) How Lachy Groom got involved

(39:40) How research shapes what Physical Intelligence builds

(45:22) The importance of real-world data

(49:07) The return of reinforcement learning in robotics

(53:31) The risk of commercializing too early

(55:47) Finding the right partners for the business

(57:13) Open research questions

(1:00:00) NVIDIA’s simulation engines

(1:01:57) The surprising speed of progress

(1:04:16) Reliability in robotics

(1:07:31) Compensating for missing senses

(1:12:28) Book recommendation

Follow Karol Hausman

LinkedIn: https://www.linkedin.com/in/karolhausman

X: https://x.com/hausman_k

Resources and episode mentions

Books

• Worlds I See: https://www.amazon.com/Worlds-I-See-Fei-Fei-Li/dp/1250389895

• The Inner Game of Tennis: The Classic Guide to the Mental Side of Peak: https://www.amazon.com/Inner-Game-Tennis-Classic-Performance/dp/0679778314

• On the Move: https://www.oliversacks.com/oliver-sacks-books/on-the-move

• Why Greatness Cannot Be Planned: The Myth of the Objective: https://www.amazon.com/Why-Greatness-Cannot-Planned-Objective-ebook/dp/B00X57B4JG

People

• Fei-Fei Li on X: https://x.com/drfeifei

• Plato: https://en.wikipedia.org/wiki/Plato

• Baruch Spinoza: https://en.wikipedia.org/wiki/Baruch_Spinoza

• Sergey Levine on X: https://x.com/svlevine

• Brian Ichter on LinkedIn: https://www.linkedin.com/in/brian-ichter-26875978

• Lachy Groom on LinkedIn: https://www.linkedin.com/in/lachy-groom-b218895

• Chelsea Finn on LinkedIn: https://www.linkedin.com/in/cbfinn

• Lee Sedol: https://en.wikipedia.org/wiki/Lee_Sedol

Other resources

• Physical Intelligence: https://www.pi.website

• Karol’s post on X about Worlds I See: https://x.com/hausman_k/status/1732087549034889688

• Ontology: https://en.wikipedia.org/wiki/Ontology

• The History of AI in 7 Experiments: https://www.generalist.com/p/the-history-of-ai

• NVIDIA: https://www.nvidia.com

• AlphaGo: https://deepmind.google/research/alphago

• Proprioception: https://en.wikipedia.org/wiki/Proprioception

Subscribe to the show

I’d love it if you’d subscribe and share the show. Your support makes all the difference as we try to bring more curious minds into the conversation.

YouTube

Spotify

Apple

Production and marketing by penname.co. For inquiries about sponsoring the podcast, email jordan@penname.co.