Bill Gurley's bombshell theory: Anthropic's leadership isn't writing software — they believe they're midwifing a digital deity that will allocate resources to humans based on AI-determined reward functions.
All-In with Chamath, Jason, Sacks & Friedberg
Bill Gurley's bombshell theory: Anthropic's leadership isn't writing software — they believe they're midwifing a digital deity that will allocate resources to humans based on AI-determined reward functions.
TL;DR
Bill Gurley joins the All-In crew to dissect AI's impact on jobs, open-source freedom, and Anthropic's controversial worldview. Gurley introduces his "Dr. Frankenstein theory" — that Anthropic's leadership may genuinely believe they are building a superior species, not just software [1] — Chamath Palihapitiya "Chamath frames Anthropic's doomerism and lobbying as optimal game theory: get 3–4 players in a room, dominate them, and then write the rule…" 33:40 . Sacks warns that regulatory capture could lead to a US ban on open-source AI models, handing China a decisive advantage [2] — Chamath Palihapitiya "If the US bans open-source AI, the rest of the world will simply run on Chinese models — handing Beijing a decisive geopolitical advantage.…" 51:33 . The hosts debate AI-driven job displacement fiercely, with Sacks noting software engineer job postings are up 15% YoY despite AI automating code [3] — David Sacks "GitHub code commits: 1B → 1.1B in one month: GitHub saw 1 billion code commits last year; in just the past month there were 1.1 billion — a…" 1:45:55 . Key takeaway: be the most AI-enabled version of yourself — or risk becoming a sitting duck.
Bill Gurley joins the All-In besties to discuss the first AI-native graduate class, Pope Leo XIV's AI encyclical, Anthropic's controversial 'digital god' worldview, AI sovereignty and the looming open-source crackdown, and the great AI jobs debate as Dario and Altman walk back their doom rhetoric.
The episode opens with Jason Calacanis confessing to racking up charges on Chamath's household iPad — Uber Eats, DoorDash, Instacart, even Loro Piana — while housesitting in the pool house. Chamath volleys back with an OJ Simpson joke about houseguests who overstay their welcome. Once the table banter settles, Jason welcomes David Sacks and Chamath before introducing the day's special guest: Bill Gurley, former Benchmark partner, newly minted author of 'Running Down a Dream,' and now a philanthropist launching a $5,000 grant program (rdad.org) to help people chase their dreams. Gurley also previews a forthcoming TED Talk and a university course built around his book, setting the tone for a conversation that will range from career advice to existential AI risk.
Bill Gurley opens the substantive portion of the conversation by citing a Gallup poll finding that 59% of workers are 'quiet quitters' — ambivalent about their jobs and therefore unlikely to proactively embrace AI. His core thesis: the best defense against AI disruption is becoming the most AI-enabled version of yourself possible. [1] — Bill Gurley "59% ambivalent about their job: A Gallup poll found 59% of surveyed workers are ambivalent about their jobs, making them less likely to pro…" 06:18 Jason Calacanis layers on a striking observation from his associate-in-training program: of 400–500 applicants given the choice between writing traditional deal analysis or vibe coding a software project, roughly 80% chose to code — the inverse of what he expected. David Sacks declares Claude proficiency the single most marketable skill in the economy right now, comparing it to being the only person in an office who knows how to use a spreadsheet in 1995. Producer Nick then demonstrates Claude Cowork's ability to create a highly personalized daily briefing by ingesting all past podcast transcripts, illustrating how AI tools compound in value when properly managed. Gurley closes by noting that this advantage extends to every field — marketing, legal, accounting, sales — not just tech.
Jason walks through the key arguments of 'Magnifica Humanitas,' Pope Leo XIV's first encyclical: technology is never neutral, it mirrors the values of those who build and finance it, and AI risks concentrating power in the hands of a few. The Pope called for regulation, worker retraining, child safety guardrails, and a ban on autonomous weapons. [1] — Jason Calacanis "Pope's encyclical: 235 pages, 42,000 words: Pope Leo XIV's first encyclical on AI, 'Magnifica Humanitas,' ran 235 pages and over 42,000 wor…" 17:57 Notably, Anthropic co-founder Chris Olah was involved, despite being an atheist — and tech giants including Amazon, Google, and Meta lobbied the Vatican on April 29th to water down the language, without success. David Sacks steelmans the Pope's concern about centralization but pivots to a deeper warning: creating an AI regulator risks handing government the very power the Pope feared would fall to tech companies. He invokes the Latin principle quis custodiet ipsos custodes — who guards the guardians — and traces the American founders' separation of powers as a second-order answer to that timeless question. His preferred solution is a competitive market, not a regulatory body, as long as competition among five frontier labs is maintained.
Bill Gurley opens with a 130-year history lesson: Pope Leo XIII's 1891 encyclical warned that the Industrial Revolution would devastate workers. [2] — Bill Gurley "Leo XIII's 1891 encyclical warned the Industrial Revolution would harm humanity. Since then, real wages rose 8–10x, global poverty fell fro…" 25:00 What followed was the workweek shrinking from 60+ hours to 34, real wages rising 8–10x, global GDP per capita growing from $1,500 to $20,000, child labor in the US dropping from 18% to zero, and global poverty falling from 75% to under 10%. Leo XIII, Gurley flatly states, 'got it dead wrong.' [2] — Bill Gurley "Leo XIII's 1891 encyclical warned the Industrial Revolution would harm humanity. Since then, real wages rose 8–10x, global poverty fell fro…" 25:00 But Gurley's real bombshell is his 'Dr. Frankenstein theory' of Anthropic. [1] — Bill Gurley "Gurley argues Anthropic isn't building software — they're midwifing a deity. After reading Chris Olah's 'Constitution,' Amanda Askill's pod…" 30:13 After exhaustively reading Chris Olah's 'The Constitution,' listening to chief philosopher Amanda Askill's podcasts, and parsing Dario Amodei's 'Machines of Loving Grace' — a blog post inspired by a poem envisioning humanity 'watched over by machines of loving grace' — Gurley concludes that Anthropic's leadership doesn't believe it is writing software. It believes it is midwifing a deity: a superior species that will govern a post-labor economy by distributing resources to humans through an AI-determined reward function. [1] — Bill Gurley "Gurley argues Anthropic isn't building software — they're midwifing a deity. After reading Chris Olah's 'Constitution,' Amanda Askill's pod…" 30:13 Chamath reduces this to its essence — 'a computational reward function for humans that decides how much you're worth' — and frames it as a dystopian Black Mirror scenario. Jason calls it the ultimate narcissism: believing you can create God.
Jason Calacanis introduces a new concept: intelligence sovereignty. [1] — Jason Calacanis "Jason Calacanis introduces a new concept: 'intelligence sovereignty.' Privacy was about protecting your data from being read. Intelligence …" 38:15 Privacy used to mean protecting your data from being seen; intelligence sovereignty means preventing AI from analyzing your messages, emails, and photos to shape your worldview. He argues that open-source models running on Apple's M-series silicon (with 48GB–1TB of memory) are the only viable defense, and notes the paradox that Communist China is leading the open-weight movement while American companies push for centralization. [1] — Jason Calacanis "Jason Calacanis introduces a new concept: 'intelligence sovereignty.' Privacy was about protecting your data from being read. Intelligence …" 38:15 Chamath demonstrates that Fortune 1000 companies are already acting on this logic — demanding control planes that can hot-swap between frontier models to avoid vendor lock-in and political risk. [2] — David Sacks "Sacks sees a clear pattern: Anthropic repeatedly frames open-source AI as uniquely dangerous because guardrails can be removed. He believes…" 49:24 He shares a striking example: a Fortune 20 company that was tasked with generating $1 billion in AI OpEx savings instead burned $200 million on tokens in six months with minimal results. Meanwhile, a Polymarket post revealed that an AI consultant's client accidentally spent $500 million on Claude tokens in a single month. Sacks then delivers his open-source warning: [2] — David Sacks "Sacks sees a clear pattern: Anthropic repeatedly frames open-source AI as uniquely dangerous because guardrails can be removed. He believes…" 49:24 Anthropic's repeated rhetoric framing open-weight models as uniquely dangerous is deliberate predicate-building — inserting facts into the public record to justify a future ban. If America bans open-source AI, it puts itself on an island; the rest of the world runs on Chinese models. The result: a monopoly handed to two or three closed frontier labs, the precise centralization outcome the Pope feared.
The episode's final and most contentious segment opens with Fortune reporting that both Sam Altman and Dario Amodei have walked back their AI job apocalypse predictions ahead of anticipated IPOs. Goldman Sachs CEO David Solomon's NYT op-ed — arguing AI will automate 25% of work hours rather than eliminate 25% of jobs, and that bank tellers increased after ATMs — has shifted the narrative completely. [1] — David Sacks "Back in January, Sacks made the contrarian prediction that AI would produce net job gains, not losses. Now Goldman's CEO says the apocalyps…" 59:58 Sacks lays out the empirical case: the Yale Budget Lab found no discernible AI labor disruption over three years; US unemployment sits at 4.3%, at or below full employment; software developer job postings are up 15% year over year to a 3-year high; and GitHub logged 1.1 billion code commits in one month versus 1 billion for all of last year. [1] — David Sacks "Back in January, Sacks made the contrarian prediction that AI would produce net job gains, not losses. Now Goldman's CEO says the apocalyps…" 59:58 He argues that AI is triggering a bespoke software boom, with industries that never hired engineers now doing so. Calacanis pushes back hard: Amazon's Andy Jassy explicitly said the company would forgo 600,000 future hires; Waymo has 3,000 vehicles displacing drivers; Meta's 8,000-person layoff was paired with AI monitoring software.[2] Chamath offers the middle path: most current layoffs are really reversals of post-COVID overhiring, with AI serving as a convenient two-letter scapegoat. But he concedes that regulated industries — pharma, aerospace, finance — will not see the startup disruption Calacanis predicts. Bill Gurley closes with his characteristic optimism: competition will ensure productivity gains flow to consumers as cheaper goods, not to obscene corporate margins. The skills trades remain a desperate shortage, and he plugs Mike Rowe Works' $16 million scholarship program for plumbers, welders, and electricians as a more useful solution than government retraining programs.
Chapter 2 · 06:00
Bill Gurley opens the substantive portion of the conversation by citing a Gallup poll finding that 59% of workers are 'quiet quitters' — ambivalent about their jobs and therefore unlikely to proactively embrace AI. His core thesis: the best defense against AI disruption is becoming the most AI-enabled version of yourself possible. [1] — Bill Gurley "59% ambivalent about their job: A Gallup poll found 59% of surveyed workers are ambivalent about their jobs, making them less likely to pro…" 06:18 Jason Calacanis layers on a striking observation from his associate-in-training program: of 400–500 applicants given the choice between writing traditional deal analysis or vibe coding a software project, roughly 80% chose to code — the inverse of what he expected. David Sacks declares Claude proficiency the single most marketable skill in the economy right now, comparing it to being the only person in an office who knows how to use a spreadsheet in 1995. Producer Nick then demonstrates Claude Cowork's ability to create a highly personalized daily briefing by ingesting all past podcast transcripts, illustrating how AI tools compound in value when properly managed. Gurley closes by noting that this advantage extends to every field — marketing, legal, accounting, sales — not just tech.
A Gallup poll found 59% of surveyed workers are ambivalent about their jobs, making them less likely to proactively adopt AI tools and more vulnerable to displacement.
The first generation of AI-native graduates is entering the workforce already fluent in the tools. Calacanis's 400-applicant internship program found 80% of candidates chose to vibe code a project over writing traditional analysis — a complete inversion of what he expected. These students aren't cheating; they're ahead.
Chapter 3 · 17:37
Jason walks through the key arguments of 'Magnifica Humanitas,' Pope Leo XIV's first encyclical: technology is never neutral, it mirrors the values of those who build and finance it, and AI risks concentrating power in the hands of a few. The Pope called for regulation, worker retraining, child safety guardrails, and a ban on autonomous weapons. [1] — Jason Calacanis "Pope's encyclical: 235 pages, 42,000 words: Pope Leo XIV's first encyclical on AI, 'Magnifica Humanitas,' ran 235 pages and over 42,000 wor…" 17:57 Notably, Anthropic co-founder Chris Olah was involved, despite being an atheist — and tech giants including Amazon, Google, and Meta lobbied the Vatican on April 29th to water down the language, without success. David Sacks steelmans the Pope's concern about centralization but pivots to a deeper warning: creating an AI regulator risks handing government the very power the Pope feared would fall to tech companies. He invokes the Latin principle quis custodiet ipsos custodes — who guards the guardians — and traces the American founders' separation of powers as a second-order answer to that timeless question. His preferred solution is a competitive market, not a regulatory body, as long as competition among five frontier labs is maintained.
Pope Leo XIV's 235-page encyclical 'Magnifica Humanitas' warns that technology takes on the characteristics of those who build, finance, and control it. Despite Amazon, Google, and Meta lobbying the Vatican to soften the language, the Pope held firm — calling for regulation and asking whether AI will serve humanity or concentrate power in the hands of a few.
Pope Leo XIV's first encyclical on AI, 'Magnifica Humanitas,' ran 235 pages and over 42,000 words — roughly the length of a full book.
Sacks argues the real AI centralization threat isn't a tech company — it's a government that gains the power to approve models and then expands its definition of 'safety' to include censorship. He invokes the Latin principle quis custodiet ipsos custodes: who guards the guardians?
Leo XIII's 1891 encyclical warned the Industrial Revolution would harm humanity. Since then, real wages rose 8–10x, global poverty fell from 75% to under 10%, child labor in the US dropped from 18% to zero, and the workweek shrank from 60+ hours to 34. Gurley says Leo XIV is making the exact same mistake.
Since Leo XIII's 1891 encyclical warning against the Industrial Revolution, the workweek fell from 60+ hours to 34 hours globally, and real wages rose 8–10x — proving the prior Pope's fears were misplaced.
Since 1891, technology and capitalism reduced global poverty from 75% of humanity to under 10%, while life expectancy rose 60% and child labor in the US fell from 18% to zero.
Chapter 4 · 26:57
Bill Gurley opens with a 130-year history lesson: Pope Leo XIII's 1891 encyclical warned that the Industrial Revolution would devastate workers. [2] — Bill Gurley "Leo XIII's 1891 encyclical warned the Industrial Revolution would harm humanity. Since then, real wages rose 8–10x, global poverty fell fro…" 25:00 What followed was the workweek shrinking from 60+ hours to 34, real wages rising 8–10x, global GDP per capita growing from $1,500 to $20,000, child labor in the US dropping from 18% to zero, and global poverty falling from 75% to under 10%. Leo XIII, Gurley flatly states, 'got it dead wrong.' [2] — Bill Gurley "Leo XIII's 1891 encyclical warned the Industrial Revolution would harm humanity. Since then, real wages rose 8–10x, global poverty fell fro…" 25:00 But Gurley's real bombshell is his 'Dr. Frankenstein theory' of Anthropic. [1] — Bill Gurley "Gurley argues Anthropic isn't building software — they're midwifing a deity. After reading Chris Olah's 'Constitution,' Amanda Askill's pod…" 30:13 After exhaustively reading Chris Olah's 'The Constitution,' listening to chief philosopher Amanda Askill's podcasts, and parsing Dario Amodei's 'Machines of Loving Grace' — a blog post inspired by a poem envisioning humanity 'watched over by machines of loving grace' — Gurley concludes that Anthropic's leadership doesn't believe it is writing software. It believes it is midwifing a deity: a superior species that will govern a post-labor economy by distributing resources to humans through an AI-determined reward function. [1] — Bill Gurley "Gurley argues Anthropic isn't building software — they're midwifing a deity. After reading Chris Olah's 'Constitution,' Amanda Askill's pod…" 30:13 Chamath reduces this to its essence — 'a computational reward function for humans that decides how much you're worth' — and frames it as a dystopian Black Mirror scenario. Jason calls it the ultimate narcissism: believing you can create God.
Gurley argues Anthropic isn't building software — they're midwifing a deity. After reading Chris Olah's 'Constitution,' Amanda Askill's podcasts, and Dario's 'Machines of Loving Grace' essay, Gurley concluded that Anthropic's leadership genuinely believes they are building a superior species that will allocate resources to humans via a computational reward function.
Chamath frames Anthropic's doomerism and lobbying as optimal game theory: get 3–4 players in a room, dominate them, and then write the rules for an oversight body too technically outmatched to push back. The goal is to make the regulator your ally and your competitor's obstacle.
Jason Calacanis introduces a new concept: 'intelligence sovereignty.' Privacy was about protecting your data from being read. Intelligence sovereignty is about preventing AI from analyzing everything you do and then telling you how to interpret the world. Open-source AI running on local hardware is the only viable defense.
Chapter 5 · 38:32
Jason Calacanis introduces a new concept: intelligence sovereignty. [1] — Jason Calacanis "Jason Calacanis introduces a new concept: 'intelligence sovereignty.' Privacy was about protecting your data from being read. Intelligence …" 38:15 Privacy used to mean protecting your data from being seen; intelligence sovereignty means preventing AI from analyzing your messages, emails, and photos to shape your worldview. He argues that open-source models running on Apple's M-series silicon (with 48GB–1TB of memory) are the only viable defense, and notes the paradox that Communist China is leading the open-weight movement while American companies push for centralization. [1] — Jason Calacanis "Jason Calacanis introduces a new concept: 'intelligence sovereignty.' Privacy was about protecting your data from being read. Intelligence …" 38:15 Chamath demonstrates that Fortune 1000 companies are already acting on this logic — demanding control planes that can hot-swap between frontier models to avoid vendor lock-in and political risk. [2] — David Sacks "Sacks sees a clear pattern: Anthropic repeatedly frames open-source AI as uniquely dangerous because guardrails can be removed. He believes…" 49:24 He shares a striking example: a Fortune 20 company that was tasked with generating $1 billion in AI OpEx savings instead burned $200 million on tokens in six months with minimal results. Meanwhile, a Polymarket post revealed that an AI consultant's client accidentally spent $500 million on Claude tokens in a single month. Sacks then delivers his open-source warning: [2] — David Sacks "Sacks sees a clear pattern: Anthropic repeatedly frames open-source AI as uniquely dangerous because guardrails can be removed. He believes…" 49:24 Anthropic's repeated rhetoric framing open-weight models as uniquely dangerous is deliberate predicate-building — inserting facts into the public record to justify a future ban. If America bans open-source AI, it puts itself on an island; the rest of the world runs on Chinese models. The result: a monopoly handed to two or three closed frontier labs, the precise centralization outcome the Pope feared.
Chamath explains that 80.90's Fortune 1000 clients refuse to lock into one AI provider, fearing both technology leapfrogs and ideological misalignment with a frontier lab's terms of service. They want a control plane that can hot-swap between OpenAI, Anthropic, or any open-weight model — because they see the model layer commoditizing fast.
A Fortune 20 CEO was asked for $1 billion in AI-generated OpEx savings; six months in, the company had spent $200 million on tokens with minimal measurable results.
Sacks sees a clear pattern: Anthropic repeatedly frames open-source AI as uniquely dangerous because guardrails can be removed. He believes this is deliberate predicate-building — putting facts in the public record to justify a future ban on open-weight models. A ban would shatter the competitive market and hand the monopoly to a handful of closed labs.
If the US bans open-source AI, the rest of the world will simply run on Chinese models — handing Beijing a decisive geopolitical advantage. Chamath highlights the paradox: America's adversary is championing open weights while American companies push for closed, regulated AI.
Elon Musk announced xAI rewrote its entire AI training complex in C, achieving an order-of-magnitude speed improvement and running on 220,000 GPUs — potentially collapsing training costs dramatically.
Sacks cites The Information's data showing Anthropic growing at ~10x year over year while OpenAI grows at ~3x. Simple math: 10x10=100, 3x3=9. If that differential holds for two years, Anthropic ends up with roughly 90% market share — a monopoly born of compounding, not product superiority.
According to The Information, Anthropic is growing at roughly 10x year over year while OpenAI grows at ~3x — meaning Anthropic could command 90% market share within two years if rates hold.
A Polymarket post revealed an AI consultant's claim that a client accidentally burned $500 million in a single month on Claude tokens after failing to set employee usage limits — $16.6 million per day, $700,000 per hour. This is the hidden cost of 'free' AI plans that hook organizations before the bill arrives.
A Polymarket post revealed an AI consultant's claim that a client accidentally spent half a billion dollars in a single month after failing to set employee limits on Claude usage.
Chapter 6 · 59:56
The episode's final and most contentious segment opens with Fortune reporting that both Sam Altman and Dario Amodei have walked back their AI job apocalypse predictions ahead of anticipated IPOs. Goldman Sachs CEO David Solomon's NYT op-ed — arguing AI will automate 25% of work hours rather than eliminate 25% of jobs, and that bank tellers increased after ATMs — has shifted the narrative completely. [1] — David Sacks "Back in January, Sacks made the contrarian prediction that AI would produce net job gains, not losses. Now Goldman's CEO says the apocalyps…" 59:58 Sacks lays out the empirical case: the Yale Budget Lab found no discernible AI labor disruption over three years; US unemployment sits at 4.3%, at or below full employment; software developer job postings are up 15% year over year to a 3-year high; and GitHub logged 1.1 billion code commits in one month versus 1 billion for all of last year. [1] — David Sacks "Back in January, Sacks made the contrarian prediction that AI would produce net job gains, not losses. Now Goldman's CEO says the apocalyps…" 59:58 He argues that AI is triggering a bespoke software boom, with industries that never hired engineers now doing so. Calacanis pushes back hard: Amazon's Andy Jassy explicitly said the company would forgo 600,000 future hires; Waymo has 3,000 vehicles displacing drivers; Meta's 8,000-person layoff was paired with AI monitoring software.[2] Chamath offers the middle path: most current layoffs are really reversals of post-COVID overhiring, with AI serving as a convenient two-letter scapegoat. But he concedes that regulated industries — pharma, aerospace, finance — will not see the startup disruption Calacanis predicts. Bill Gurley closes with his characteristic optimism: competition will ensure productivity gains flow to consumers as cheaper goods, not to obscene corporate margins. The skills trades remain a desperate shortage, and he plugs Mike Rowe Works' $16 million scholarship program for plumbers, welders, and electricians as a more useful solution than government retraining programs.
Back in January, Sacks made the contrarian prediction that AI would produce net job gains, not losses. Now Goldman's CEO says the apocalypse is overblown, and both Sam Altman and Dario Amodei have walked back their doom rhetoric. Sacks is waiting for his apology — and has the receipts.
The Yale Budget Lab's comprehensive study of the US labor market over three years found no discernible disruption attributable to AI.
Despite AI automating code writing, job postings for software developers hit a 3-year high, growing 15% year over year — the opposite of what a simple displacement model would predict.
GitHub saw 1 billion code commits last year; in just the past month there were 1.1 billion — a roughly 14x annualized year-over-year increase in code generation driven by AI.
Despite three-plus years of the AI wave, the US unemployment rate sits at 4.3%, which economists consider at or near full employment — offering no aggregate evidence of AI-driven job loss.
Amazon CEO Andy Jassy stated the company would eliminate 600,000 future planned positions as it deploys AI, signaling a structural reduction in future workforce growth.
Law firm Kirkland & Ellis announced plans to spend $500 million to build its own frontier AI model on-premises, exemplifying the Fortune 1000 trend toward AI sovereignty.
Mike Rowe's foundation Mike Rowe Works has funded $16 million worth of scholarships, enabling 2,600 people to train as plumbers, welders, or electricians for free.
No indexed bits in this chapter.
Gurley argues Anthropic isn't building software — they're midwifing a deity. After reading Chris Olah's 'Constitution,' Amanda Askill's podcasts, and Dario's 'Machines of Loving Grace' essay, Gurley concluded that Anthropic's leadership genuinely believes they are building a superior species that will allocate resources to humans via a computational reward function.
Sacks sees a clear pattern: Anthropic repeatedly frames open-source AI as uniquely dangerous because guardrails can be removed. He believes this is deliberate predicate-building — putting facts in the public record to justify a future ban on open-weight models. A ban would shatter the competitive market and hand the monopoly to a handful of closed labs.
This episode
Factual claims made this episode, and whether a source was named.
A Gallup poll found that 59% of surveyed workers are ambivalent about their jobs, a group Gallup called 'quiet quitters.'
Since Pope Leo XIII's 1891 encyclical, the global workweek fell from over 60 hours to 34 hours.
Real wages have risen 8 to 10 times adjusted for inflation since 1891, and the median worker now earns more than a doctor did in 1891.
Global GDP per capita rose from $1,500 to $20,000 since 1891.
Child labor in the US fell from 18% to zero since 1891, workplace deaths fell 40x, and life expectancy rose 60%.
Anthropic is growing at roughly 10x year over year while OpenAI is growing at approximately 3x year over year.
A company accidentally spent approximately $500 million in a single month on Claude tokens, equating to $16.6 million per day.
The Yale Budget Lab found no discernible disruption in the US labor market attributable to AI over the past three years.
Software developer job postings are up 15% year over year and at a 3-year high, despite AI now writing most code.
GitHub recorded 1 billion code commits last year, and 1.1 billion code commits in just the past month — roughly a 14x annualized year-over-year increase.
The US unemployment rate is currently 4.3%, which economists consider at or near full employment.
Amazon CEO Andy Jassy announced the company would not fill approximately 600,000 future planned positions due to AI deployment.
Elon Musk announced that xAI rewrote its entire AI training stack in C, achieving an order-of-magnitude speed improvement over JAX, running on 220,000 GPUs.
Waymo currently operates a fleet of 3,000 self-driving vehicles.
Mike Rowe Works has funded $16 million in scholarships for 2,600 people to train as plumbers, welders, or electricians.
Kirkland & Ellis plans to spend $500 million to build its own on-premises frontier AI model.
A Fortune 20 company CEO asked for $1 billion in AI-generated OpEx savings; six months in, the company had spent $200 million on tokens with minimal results.
This episode
Guest on the episode; former Benchmark partner and author of 'Running Down a Dream'; introduced the Dr. Frankenstein theory about Anthropic.
Issued the AI encyclical 'Magnifica Humanitas,' warning that technology takes on the characteristics of those who build it and calling for AI regulation.
CEO of Anthropic; his essay 'Machines of Loving Grace' and his walkback of AI job apocalypse predictions were central topics.
Referenced for his early warnings about Google monopolizing AI, his role co-founding OpenAI, and his announcement that xAI rewrote its training stack in C for a 10x speed gain.
OpenAI CEO cited as having walked back his AI job apocalypse predictions ahead of an anticipated IPO.
Goldman Sachs CEO who wrote a NYT op-ed arguing AI will automate work hours rather than eliminate jobs, and that the job apocalypse narrative is overblown.
Discussed extensively as a case study in regulatory capture, AI doomerism for competitive advantage, and the 'Dr. Frankenstein theory' that its leadership believes it is building a superior species.
Compared to Anthropic in terms of growth rates; Sam Altman's walkback of AI job doom rhetoric was discussed. Anthropic was described as a spinout of OpenAI.
CEO Andy Jassy cited as saying Amazon would forgo 600,000 future hires due to AI, and discussed as the world's largest user of robotics.
Cited as having laid off 8,000 employees and installing employee monitoring software, with the job cuts attributed by some to AI and by others to post-COVID overhiring.
CEO David Solomon wrote a NYT op-ed arguing the AI job apocalypse is overblown, validating Sacks's earlier contrarian prediction.
Jack Dorsey's fintech company; announced a 50% headcount reduction attributed to AI, which financial analysts widely characterized as AI washing for a company that had long overstaffed.
CEO Matthew Prince cited as explicitly blaming AI for a 20% (1,100-person) workforce reduction and introducing the term 'measurers' for eliminated middle-management roles.
Law firm announced plans to spend $500 million to build its own on-premises frontier AI model, cited as evidence of the Fortune 1000 AI sovereignty trend.
Cited by Calacanis as evidence that self-driving technology will eliminate taxi and rideshare driver jobs, currently operating a fleet of 3,000 vehicles.
Cited by Sacks for a comprehensive study finding no discernible labor market disruption attributable to AI over the past three years.
Anthropic's AI model; cited as the most marketable skill in the economy, used for producing the All-In podcast daily briefing, and discussed in the context of runaway enterprise token spend.
Cited for code commit data showing 1.1 billion commits in one month versus 1 billion for all of last year, indicating a 14x annualized increase in AI-driven code generation.
Stats
We use essential and analytics cookies to run Vuci. To understand how the site is used: Privacy Policy.
Install Vuci on your phone
Add it to your home screen for a faster, app-like experience.
Install Vuci on your phone
Tap the Share button, then “Add to Home Screen”.
A new version is available
Reload to get the latest Vuci.