All-In · 2026-05-29 · 95 min

Anthropic's Digital God, Pope vs AI, Job Loss Narrative Flips, Open Source Crackdown Coming?

Bill Gurley

YouTube ↗

TL;DR

AI’s labor impact is splitting into near-term displacement and a longer productivity-led hiring boom, not resolving into either apocalypse or irrelevance. Sacks cited 4.3% unemployment, no discernible AI labor disruption in a Yale Budget Lab study, and software-engineering postings up 15% year over year; Jason countered with Meta’s 8,000 cuts, Cloudflare’s 20% reduction, and Amazon’s stated plan to eliminate 600,000 future positions. Gurley’s hedge was practical: anyone refusing AI is like someone refusing “email,” “a spreadsheet,” or “a computer.”
Anthropic is both an exceptional product company and, in Gurley’s reading, the industry’s strangest political risk. After initially seeing its doomerism as regulatory capture, he developed a “Dr. Frankenstein theory.” Jason connected Dario Amodei’s “Machines of Loving Grace” and imagined AI-directed secondary economy to the idea that the company may be “midwifing a deity.” Chamath called that an extreme form of narcissism and delusion of grandeur; Sacks steelmanned the safety mission while Jason warned that branding Anthropic as safe and rivals as reckless could advance concentrated control.
Pope Leo XIV and Sacks agree that concentrated AI power is dangerous, but disagree on whether government is the cure. The Pope’s 42,000-plus-word Magnifica Humanitas says technology is never neutral and takes on the characteristics of those who “build, finance, and control it”; Sacks fears an “FDA for AI” would let political definitions of safety expand into censorship. His preferred architecture is competition among five frontier labs, aggressive antitrust if monopoly emerges, and guardians forced to “guard against each other.”
Open models are the episode’s backstop against both corporate lock-in and state-directed intelligence. Sacks sees breadcrumbs toward a US ban based on removable cyber and bio guardrails, while Jason argued that privacy is becoming “intelligence sovereignty”: no central model should analyze your private context and tell you how to interpret the world. The paradox is that China is leading in open weights while America centralizes; a US ban could leave the rest of the world running Chinese models.
Frontier-model convergence shifts investment value toward interfaces, control planes, local hardware, and portable context. A Rogo financial-analysis evaluation put Opus 4.7, GPT-5.5, and Sonnet 4.6 within three-tenths of a percentage point, prompting Gurley’s call for open connectors that make models “exchangeable, swappable.” Friedberg said Fortune 1000 buyers increasingly want on-prem systems and a hot-swappable layer above model vendors because capability leadership, terms of service, and political constraints can all change.
Token economics are becoming the enterprise AI reality check. Chamath relayed a Fortune 20 company that sought $1 billion of AI-generated operating savings, spent $200 million on tokens in six months, and saw minimal results; another claim said one client accidentally consumed nearly $500 million in a month. Yet Chamath noted that Anthropic’s reported 10x growth versus OpenAI’s 3x would mathematically approach 90% share in two years if sustained—an unlikely but investable illustration of compounding, compute constraints, and the coming obsession with token efficiency.
The immediate career moat is agency, not credential or job category. Jason said roughly 80% of applicants choosing between a conventional venture memo and a software assignment chose to vibe-code; Sacks called Claude proficiency possibly “the single most marketable skill in the economy right now.” Gurley’s test is whether someone keeps learning from fascination, because “the most AI-enabled version of yourself” is both the best defense against displacement and the best route into newly cheap creation.

Digest · the substance, structured for research

1. Agency, not job title, is the first AI labor hedge

Gurley tied vulnerability to disengagement: Gallup found roughly 59% of those surveyed ambivalent about their jobs, and an ambivalent employee is unlikely to experiment aggressively. His prescription was blunt: “The best way to protect yourself from AI is to be the most AI-enabled version of yourself you can be.”
Jason’s associate-in-training program offered applicants either a portfolio-company deal memo or a competitive-intelligence software project. Of roughly 400–500 applicants for six positions, about 80% chose vibe coding—evidence, in his view, that current graduates are already using AI tools rather than intimidated by building.
Sacks called Claude proficiency “the single most marketable skill in the economy right now,” comparable to being the only employee who understood spreadsheets at their introduction. He conceded that the arbitrage may narrow, but argued that early users extend their advantage by continually discovering more workflows.
Producer Nick’s daily briefing supplied the specimen: Claude Cowork ingested every show transcript, matched new stories to each host’s past views, and wrote its own training rules and skills file. The recursion did not eliminate the producer; someone still had to “supervise, iterate, validate” and improve the system every day.

2. The Pope identifies concentrated AI power as the central danger

The show described Pope Leo XIV’s Magnifica Humanitas as a 235-page, over-42,000-word encyclical arguing that AI is not inherently evil but “technology is never neutral.” It takes on the characteristics of those who “build, finance, and control it,” making ownership and incentives inseparable from safety.
Its less-contested prescriptions included worker retraining, protections for children, guardrails, and a ban on autonomous weapons. The central question was whether AI would concentrate power among a few actors or “serve everyone”; Amazon, Google, and Meta reportedly lobbied the Vatican on April 29 to soften the language.
Sacks agreed that AI could enable Orwellian surveillance, censorship, and control, but identified government as the likeliest perpetrator. An “FDA for AI” could start with technical safety, then expand—as social-media trust-and-safety regimes did—to disinformation, psychological safety, microaggressions, and political censorship.
His answer to “who guards the guardians?” was the American checks-and-balances model: divide power so institutions constrain each other. Let five frontier labs compete now; if the market collapses into one or two dominant firms, use antitrust “very aggressively” rather than grant one regulator approval power over every model.

3. Gurley says the Pope’s historical analogy points the other way

Pope Leo reportedly modeled his intervention on Leo XIII’s 1891 warning about industrialization. Gurley’s objection was not theological but empirical: the earlier warning “got it dead wrong” because innovation and capitalism ultimately delivered the opposite of the predicted human degradation.
From 1891 to today, Gurley said, the global workweek fell from more than 60 hours to 34, real wages rose 8–10x, and the median worker came to earn more than a doctor did in 1891. Global GDP per capita increased from roughly $1,500 to $20,000.
His remaining ledger: US child labor fell from 18% to zero, workplace deaths dropped 40x, life expectancy rose 60%, and global poverty declined from 75% of humanity to under 10%. He saw no reason AI-led innovation should reverse that long prosperity trend, though individuals still must adapt.

4. Gurley’s “Dr. Frankenstein theory” takes Anthropic literally

Anthropic mystifies Gurley because it leads its field while being “the most negatively outspoken commenter” on its own work. He originally assumed its extensive state-by-state lobbying and doomer rhetoric sought regulatory capture; after reading deeply, he concluded that belief may be as important as strategy.
His reading list included the roughly 80-page Constitution associated with Chris Olah, podcasts by chief philosopher Amanda Askell, and Dario Amodei’s “Machines of Loving Grace.” Gurley urged listeners to endure the laborious primary material because Anthropic’s language reveals ambitions that ordinary software framing misses.
The source poem imagines a “cybernetic ecology” where humans are free of labor, reunited with their “mammal brothers and sisters,” and “all watched over by machines of loving grace.” Jason’s reaction—“Sounds like overlords to me”—became sharper when paired with Amodei’s proposed economic order.
Amodei envisioned AI systems allocating resources to humans through a secondary economy based on what those systems judge worth rewarding. Jason’s conclusion was, “I don’t think they think they’re writing software. I think they’re midwifing a deity here.” Chamath called the premise an extreme form of narcissism and technological grandeur.

5. Anthropic’s safety halo may also be a control strategy

Chamath’s more tactical interpretation was game theory: absorb disproportionate capital, help define the rules, and place less technically capable referees over the market. “If the refs don’t understand the game, you’ll run over the game,” especially when only three or four approved firms remain inside the room.
Gurley credited Anthropic with building a halo among media, academics, and other intellectual elites who may rank it as the company that “cares most.” The same campaign creates broader fear and data-center resistance, but it also positions Anthropic as the natural authority whenever policymakers seek a safe operator.
Sacks’s steelman was that Anthropic genuinely believes it is creating something godlike and therefore uniquely dangerous; its founders left OpenAI because they thought leadership was insufficiently serious about safety. Jason’s concern was that “we care most” could become a rationale for concentrated control.
Through the centralization lens, Jason argued that characterizing every competitor as reckless can advance monopolistic control. Sacks said AI should remain decentralized enough that users can protect themselves, including running models on their own hardware rather than depending on a company that might be “in bed with a deep state.”

6. Intelligence sovereignty turns local AI into strategic infrastructure

Sacks defined open source as software freedom: users can execute a program locally without surrendering data, privacy, or autonomy to a monopolist. Without that option, participation in the modern economy could require accepting a centralized model’s social-credit logic; the only alternative would be living off-grid.
Jason extended data sovereignty into “intelligence sovereignty.” Privacy means outsiders cannot inspect photos, notes, email, or messages; intelligence sovereignty means they also cannot feed that material into their model and “tell me how to interpret the world.”
That made Apple Jason’s “dark horse” because of its historical privacy posture and local hardware: M5 systems with 48GB or 128GB of memory, plus a supposedly forthcoming Mac Studio with a terabyte. Capable local agents and small, verticalized language models could change the competitive boundary between device and cloud.
Jason initially called China the leader in open source; Sacks corrected the category to open weights. The distinction matters, but so does the geopolitical inversion: China is leading in open-weight access while leading US labs and regulators appear to be moving toward greater centralization.

7. Model convergence moves power toward the control plane

Rogo’s financial-analyst evaluation found “no single best model anymore”: Opus 4.7, GPT-5.5, and Sonnet 4.6 were separated by less than three-tenths of a percentage point. Jason’s investment question was what incremental trillions of training capital earn if practical outputs converge so quickly.
Gurley’s answer was to commoditize interfaces. More open MCP-style connectors—he noted MCP sits with the Linux Foundation—could do for model workloads what Kubernetes did for portability: separate applications, context, and data from the underlying provider so models become “exchangeable, swappable.”
Friedberg said Abacus builds headless products with a control plane capable of hot-swapping between frontier providers; he likened the approach to what Chamath is doing with 8090. Global 1000 buyers fear choosing a model that gets leapfrogged and being trapped by a vendor’s future terms of service or political philosophy.
Regulated buyers add data leakage, HIPAA, governance, and auditability to that list. The example was a Canadian hospital whose lawful euthanasia workflow might conflict with an American model provider’s policy; Stable Diffusion reproducing Getty watermarks illustrated how training provenance can unexpectedly surface in outputs.

8. Enterprise token bills are ending the “AI is free” illusion

Chamath relayed Vivek Garipalli’s Fortune 20 anecdote: a CEO demanded $1 billion in AI-generated operating savings, but after six months the company had spent $200 million on tokens with minimal results. The emerging unwind is CFO scrutiny, reduced licenses, and demands to tie consumption to measurable returns.
A separate Polymarket post claimed a client without employee limits accidentally consumed nearly $500 million in one month—about $16.6 million per day and almost $700,000 per hour. The hosts treated it as astonishing evidence of uncontrolled usage, not as an independently established benchmark.
Jason compared flat-rate plans to making the first 10,000 gallons of water free: everyone leaves the hose running until metered pricing appears. Inside his own organization, three employees independently built three Founder University portals because each saw colleagues receive credit for shipping an interface.
Sacks expects token efficiency to become a major theme, though not necessarily to overturn model economics. Chamath said Anthropic reportedly grew 10x year over year versus OpenAI’s 3x; sustained for two years, his simple 100-versus-9 compounding example implies roughly 90% share, though competition and compute scarcity may prevent it.

9. An open-model crackdown is the policy risk to watch

Sacks sees “breadcrumbs” toward banning open-source or open-weight models. Anthropic’s cyber and bio discussions repeatedly emphasize that open-model guardrails can be removed; he reads that rhetoric as creating “predicate facts in the public record” for a restriction proponents cannot yet justify directly.
A model’s weights are merely a file of numbers, making literal prohibition difficult. But US cloud providers would stop hosting or supporting those files, sharply increasing deployment friction while foreign users retained open models’ cost, customization, and control advantages.
The EU is Chamath’s “canary in the coal mine” because regulation collides with projects lacking a single accountable owner. Sacks and Gurley’s feared endpoint is a United States cut off from open innovation while “the rest of the world ends up running on Chinese models.”
Meanwhile, the training moat may be shrinking. Chamath cited domain-specific silicon and Elon Musk’s claim that rewriting the training stack in C++ produced an order-of-magnitude improvement across 220,000 GPUs; Jason estimated every 1% efficiency gain there equaled roughly 2,000 GPUs and hundreds of millions of dollars.

10. The AI job-apocalypse narrative is visibly retreating

Goldman Sachs CEO David Solomon argued that AI may automate 25% of work hours, not eliminate 25% of jobs, freeing workers for higher-level tasks. His analogies were bank tellers increasing after ATMs and live entertainment expanding after television; annual US job creation and destruction already totals 25–35 million.
The show read Sam Altman and Dario Amodei as walking back harsher forecasts. Amodei now said AI might automate 90% of someone’s tasks while the remaining 10% expands into new work—the same task-versus-purpose distinction Sacks and Jensen Huang had emphasized.
Sacks, claiming vindication for his January job-gains prediction, cited a Yale Budget Lab finding of “no discernible disruption” from AI over three years. He also pointed to software-developer postings at a three-year high, up 15% year over year, despite coding being enterprise AI’s breakout use case.
His macro check was 4.3% unemployment versus the roughly 5% economists call full employment. The hosts still urged humility: the technology is dynamic, aggregates can conceal painful transitions, and confident forecasts—whether apocalypse or effortless abundance—extend beyond the available data.

11. The hosts sharply disagree over whether layoffs are AI washing

Chamath argued that many companies overhired, hoarded talent to deprive potential competitors, and bloated operating budgets during the previous decade. AI now gives management a convenient “scapegoat” for returning to fighting weight, even though no major filing has clearly quantified productivity lift from the consumed tokens.
Jason instead took executives at their word, citing Cloudflare’s 20% cut, Meta’s 8,000 layoffs, Block’s proposed 50% reduction, and Amazon’s plan to eliminate 600,000 future positions. He believes AI collapses product teams, removes middle-management “measurers,” and rewards public companies that produce more with fewer people.
His full position was more nuanced than net apocalypse: short- and medium-term displacement could reach the low millions, but a “Cambrian explosion” of five- or ten-person startups may ultimately absorb talent and expand the economy. His disagreement with Sacks concerned the painful transition and who fails to make it.
Sacks countered that isolated layoffs are anecdotes unless netted against AI-created firms and jobs, and called Block a classic case of AI washing after pandemic overstaffing. He cited securities-litigation partner Donnie King of Akin Gump, who warned that misattributing operational weakness to AI could eventually invite shareholder suits as misleading corporate puffery.

12. Code abundance is increasing demand for people who manage complexity

Sacks offered coding as the cleanest rebuttal to simple automation math. AI now writes much of the code, yet developer openings are rising; GitHub allegedly went from 1 billion commits during the prior year to 1.1 billion in one month, a claimed 14x year-over-year expansion.
Easier production means more software gets attempted, but somebody must inspect, integrate, secure, and maintain the resulting complexity. One fund manager told Sacks his next two hires would be software developers rather than data analysts because the firm was deploying bespoke code for the first time.
Jason agreed on expanding software demand while predicting role compression inside each team: designers can vibe-code, developers can handle front-end UX, and contributors can manage themselves. The product opportunity grows even as distinct project-manager, middle-manager, and handoff roles may disappear.

13. Competition, partial automation, and reskilling complicate the headcount call

Jason expects Waymo’s roughly 3,000 vehicles, Optimus, and package-sorting robots to eliminate most cab, truck-driving, and warehouse-sorting jobs over the next decade. Gurley rejected the 100% endpoint: autonomous economics may require humans for roughly 50% of rides while cheaper non-ownership transport expands total demand enough to preserve employment.
Friedberg and Sacks cautioned against extrapolating startup productivity into heavily regulated sectors, invoking Boeing, pharmaceuticals, safety, governance, and auditability. Jason’s counterexamples were SpaceX and regulated taxi or trucking markets, leaving the core disagreement intact: incumbency slows disruption, but does not prove immunity.
Gurley rejected the idea that AI automatically gives every incumbent 70% operating margins. Competitors will use the same productivity gains to lower prices, whittling away excess profit and making consumers’ basket of goods cheaper—though healthcare, education, and other regulated categories may offset those gains.
Gurley has little confidence in government retraining, preferring individual adoption and undersupplied skilled trades. Jason highlighted mikeroweWORKS’ reported $16 million funding of 2,600 plumbing, welding, and electrical scholarships, alongside his own new grant program; Gurley’s RDaD program offers $5,000 grants for people ready to pursue their dreams and redirect their careers.

Jason Calacanis

Okay, we are gathered here today in holy unity, brothers and sisters, to convene and discuss, on this most holy day, the day the All-In podcast drops many topics: AI data centers, China, justice, human dignity, and Dario unwinding these SPVs. It hasn't been good for the Vatican. We got in at 20 billion. That was a 50-bagger for us. So, let's get started.

Chamath Palihapitiya

Jason, I'm pretty sure you believed you were the vicar of God before the encyclical, so this is nothing new for you.

Jason Calacanis

The smoke has risen from Chamath's pool house and from the poker room.

Chamath Palihapitiya

He's staying in my pool house. He's been there for the last 3 days.

Jason Calacanis

It's been magnificent. He didn't know. You know what? I understand where O.J. was coming from. You put JCal in your house for long enough, you just lose your mind. At some point, somebody's getting whacked.

All right, enough with the shenanigans. It's been great staying at the house because there's actually—Chamath is not aware of this—an iPad in the kitchen that's logged in to Uber Eats, DoorDash, Instacart, Amazon, and Loro Piana.

Chamath Palihapitiya

Come on, stop.

Jason Calacanis

No, there is. It's literally every single service. I told the house manager, “Listen, any packages that come in over the next 72 hours, right to the pool house. If it says JCal, right to the pool house.” All these packages have been coming in. Then I relabeled them, gave them back, sent them to the ranch, and now the house manager is sending that stuff to the ranch.

Loro Piana wants to know why my inseam went from 36 to 12.

Chamath Palihapitiya

Your waist size went from 32 to 36.

Jason Calacanis

All right, welcome to the program, everybody. David Sacks is here. How are you doing, David?

David Sacks

I'm good.

Jason Calacanis

Chamath Palihapitiya is back at the 8090 office. I was at the 8090 office the last couple of days, and it's a vibe. There's a great culture going on.

If you're a bestie and you show up at that office, everybody there is a huge fan of the pod. I was like royalty. Everybody stopped by: “Hey, I'm a developer here. I'm a big fan of the show. Thank you for giving it to Chamath. We can't give it to him because he pays our mortgage and everything, but every time you stick it to Chamath, we love it. We're cheering for you in the secret Slack room.”

Chamath Palihapitiya

There's a secret Slack room?

Jason Calacanis

There is. There is definitely a secret Slack room going on.

No, but it was great. The vibes were awesome. You're building a lot of software and hiring a lot of young talent. I don't want to say where your secret source is, but you have a secret source of talent. Those are some smart kids.

Chamath Palihapitiya

I'm happy to say it. When I was at Facebook, we became the most aggressive recruiter of Waterloo co-ops, so I went back to the well. We recruit more interns every quarter than we have full-time engineers, which we do on purpose because it puts a ton of pressure on the product actually being good.

We had 400 people apply this quarter for internships.

Jason Calacanis

Wow. It's very interesting. With us sitting in for Friedberg, who's busy with some potato seed this week doing great stuff in Ohio, the one, the only Bill Gurley is here. He's been running down a dream. If you haven't bought the book, get the book. It's incredible. You're off the book tour, so now you have time for us.

Bill Gurley

Yes. I had told you, if you ever talked about the Pope, I'd love to hop on.

Jason Calacanis

Yes, you were like—

Well, Bill's an evangelical and I'm a Catholic, so we do have some common ground here.

When JCal gets sacrilegious, I've got to come on there and make sure JCal doesn't get out of line with the Pope.

Listen, the Pope is God's messenger on Earth. We should give him a base level of respect. By the way, that's us imitating Bill Gurley. This is not actually Bill Gurley. For those of you listening, you might be confused.

Bill Gurley

They were literally confused.

Jason Calacanis

We don't want to put words in your mouth, but just a point of clarification. Everybody knows you handed the baton over at Benchmark after a very successful couple of decades in venture capital. You wrote the book. You've now got a nonprofit. You're doing your own spin, I think, on maybe what Peter Thiel does with his fellowship. You started your own Running Down a Dream fellowship, I understand.

Bill Gurley

It's targeted at a different demographic. It's called RDAD.org, RunningDownADream.org. We're going to do $5,000 grants to people who want to chase their dreams but need some help. There is an application process, like with the Thiel Fellows and other programs, and we've been out talking to those people.

We're actually live. We went live last week for applications. If you know people who have read the book, are inspired, and need some help, have them apply.

Jason Calacanis

Good for you, BG.

Bill Gurley

I did a TED Talk, which will come out soon, that's related to the book. There's a professor in Miami who's built a course around the book, which I'm excited about, and he's doing it in an open-source way so that other people can borrow it as well. If there's anyone out there, I'd love to help them do that.

Jason Calacanis

What's your take on all of this doomerism? If you're a young person in college or high school, is this much ado about nothing, or how do you run down a dream in the face of something like that?

Bill Gurley

I started the book before this happened, and I've been asked the question a lot. It came up in the TED Talk. I fear that a lot of people are in jobs they don't care about that much. There's a Gallup poll that backs this up. They came up with that term “quiet quitters.” Around 59% of the people they surveyed are ambivalent about their jobs.

When you're ambivalent about your job, you're not high-agency, so you don't lean in. If you look at how Jason talks about how they implemented AI in all of his different working groups, you hear that enthusiasm and high agency, and then you want to go try these things.

I think the best way to protect yourself from AI is to be the most AI-enabled version of yourself you can be. But if you're ambivalent about your job, you're probably not doing that, and you could be a sitting duck.

So I think it's the mindset that's the problem.

Jason Calacanis

I created an associate-in-training program for my firm because we want to help people get into venture capital. We gave them a choice of assignments. One of them was to write coverage of one of our portfolio companies that's breaking out. Micro1 is the name of it. We asked them to give us a competitive landscape—basically, write a deal memo and coverage of that company.

Then we gave them another option: to vibe-code a very specific project I've wanted to have for our venture firm for a long time around competitive intelligence. Maybe 80% of the students applying did the vibe-coding. We had 400 or 500 people apply for 6 positions. I was shocked. I thought it would be the exact opposite.

Anybody can write. Anybody can throw something into ChatGPT and get some output. But they actually built software, and that's the scary thing. The students who graduated 5 or 10 years ago, before AI, aren't AI-first. They feel lost and adrift. They don't have agency.

But the group coming out of college right now—who cheated their way through school using ChatGPT and doing their assignments—is already using those tools. I'm joking about calling it cheating; I mean hacking. I agree with that.

Yeah, and they're just like, “I know how to use these tools to get through my finals.”

Chamath Palihapitiya

Gurley, I think you're saying something super important. I said this last week, which is: nobody asks the warehouse worker at Amazon whether they actually want that job. To your point, job satisfaction isn't some external person judging your job to be valid and saying you must be able to have it.

I think you should ask the person who does the job: Do you like it, and do you want to keep it? Those are 2 very different questions.

Bill Gurley

Yeah.

Chamath Palihapitiya

I think all of this AI doom and gloom was a lot—and too much, frankly—of the former and not enough of the latter. Now this whole lie is getting undone. I think Sacks posted about it this week as well. The Goldman Sachs CEO said it. And now, in this crazy twist of fate, now that we need to have trillion-dollar IPOs, all the frontier labs are like, “Wow, it's going to be a bonanza of jobs.”

Jason Calacanis

Mark Cuban had a great quote. He said there are 2 types of people in the world: those who use AI to learn faster than they ever could before and those who use AI to avoid learning altogether.

Chamath Palihapitiya

I think it's this notion of high agency or not. That's pretty good.

Jason Calacanis

Are you leaning in and using this stuff to be ever more powerful in what you try to accomplish, or are you using it as a cheat code? If you're in the latter category, you're probably at risk.

You get asked a lot about how to educate yourself if you're a parent of kids so that you can put them on a path to launch, do well, and chase their dreams.

You have a good answer for that question.

Bill Gurley

The second chapter of the book is all about lifetime learning, and it’s kind of a requirement that you’re following your fascination, because the lifetime learning comes for free if you’re fascinated with something. You just constantly soak up and devour new information.

I do think that a lot of kids get exhausted because we’ve made high school and college such a grind that they think the learning ends the day they walk out with their diploma. As we all know, the best and brightest in all of our fields are on a constant learning journey. When something new comes out, they dive in and try to figure it out.

Jason Calacanis

Every single person in the book that we profiled has that kind of attitude about their craft every day. I think the real test is: If you’re not proactively self-learning, then you’re probably not tilting against something that you really adore and are fascinated by.

David Sacks

With respect to new college grads, I was going to say that I think the single most marketable skill in the economy right now has got to be proficiency in Claude. If you’re going into a firm right now and you’re the only one who knows Claude, it would be like you’re the only one who knows how to work a spreadsheet or word processor. The advantage would be enormous.

I think that’s probably a short-term arbitrage, because eventually everyone’s going to have to figure out how to use these tools. But as a young college graduate right now, you have such an advantage if you’re AI-native, just knowing how to use these tools.

This thought partially occurred to me when I saw what our producer Nick has been doing using Claude. He’s been creating this daily briefing document.

Jason Calacanis

We’ve been doing it for 3 months, actually.

David Sacks

I just ran it for the first time, apparently. I didn’t know you’d been busy.

I thought it would just be AI slop and would give me a roundup of news that I was getting in my X feed anyway. But the thing that was really impressive about it was that it predicted topics that I would specifically be interested in based on my previous comments on the pod. It also went back and looked at previous transcripts and what I had said, and then had updates to those topics based on specific things I had said.

So, again, it was highly contextual. But then I asked Nick, “How did you generate that?” He showed me the custom prompt that he designed for Claude and then the skills document, and they were very long and detailed documents. They weren’t written in code, but they were very technical, and I just realized, looking at that, that the average person is not going to be able to generate this.

This is why the idea that you’re just going to be able to throw AI into an organization and it’s magically going to generate value is not true. You have to know how to get value out of it.

Jason Calacanis

I mean, the interesting thing, Sacks, is you can. You just have to ask your AI—you ask Claude or ChatGPT or whatever you’re using—“Hey, I want you to make me a mega-prompt. You’re a producer of a podcast. These are the 4 characters on the podcast. What would be a great prompt for me?”

It will actually suggest a prompt, and then you can refine the prompt. You can have a dialogue about a prompt as opposed to writing the prompt yourself.

Jason Calacanis

Well, Nick, can you show on the screen and scroll through the training rules? Then there’s also the skills document that was written on how to be a producer for this podcast, which I thought was really impressive.

By the way, David, what you said is true of almost every single job type. It’s not just tech or programming. If you’re in marketing, legal, accounting, sales—any role you might have at a firm—if you’re the most AI-savvy person of all your peers, you are golden.

David Sacks

You are golden. You are golden in your—

Jason Calacanis

You’re 10x more valuable than the next person who’s not, basically.

David Sacks

Yes, yes. I don’t think it goes away, because I think you learn how to get better at it over time. Having an early advantage will extend for a while, because you can learn more and more things and accomplish more.

Jason Calacanis

Should we let producer Nick describe what we were just looking at there?

Yeah, go ahead, producer. What? Producer Nick, explain the process.

Nick

Once we got access to Claude Cowork and it had that further-expanded memory access, I thought it would be interesting to just start feeding every transcript into it and seeing if it could actually contextualize new stories that were coming out based on past things that you guys had said.

I gave it a general prompt of what I wanted and said, “How would you write a skills file or some training rules for this?” It wrote all of it for me.

Jason Calacanis

Yeah.

David Sacks

Oh, so you were less good than I thought. [Laughter] I thought you were—

Nick

It’s a hack. It’s a hack. You use AI to make the skills.

Jason Calacanis

But you’ve been updating that over time, right, as you’ve been iterating and learning?

Nick

Every single day, and every single day it gets smarter and better.

Jason Calacanis

The recursiveness of this is incredible.

You need someone to manage that process, right? The 4 besties are not going to do that. You need a producer of the show to do that. This is why people think, “Oh, it’s just going to wipe out all the jobs.” No, someone still has to supervise, iterate, validate, and do all those kinds of things.

Bill Gurley

Yeah, and it’s really interesting that the people who are coming into the workforce right now are super aware of this, and they’re putting the tools to work. It’s much easier for them to get a job.

I literally looked at the top 9 candidates for this associate and training program I have, and we’re going to do it every year. Every summer, we start it. We do it for a year, and we pay you to learn.

It was extraordinary how you could tell immediately if the person had systems thinking, Sacks. They understood the process of venture capital: There was a structure to it. You had to source deals, make decisions on which ones to invest in, do diligence, and double down on investments. They just understood the process.

When you talk to one of these LLMs, it will tell you what to do. You can say, “I don’t know what I’m doing. What should I do next?” and it actually tells you what to do next.

So, for people who are intimidated by this and maybe think, “I’m already too far behind,” I encourage you to pop up Claude, go into Cowork, and say, “What can I do to be better at my job?” Just start talking. You can use voice-to-text. I use Wispr Flow, which is a really cool program for this, and I have a foot pedal to do it.

You just ramble and ramble and ramble and keep adding stuff. You don’t have to be structured. It will build the structure around the 2 or 3 paragraphs that you give it as instructions.

That’s the thing people are getting caught up on now, Bill. They think they have to type, when in fact, if you just blather on and on—a scale I have a unique ability to do—you just blather on. It’s a superpower. You blather on, and the thing makes sense of it.

It’s unbelievable what the blather-on prompt can get in terms of output. Thanks for coming to my TED Talk.

Jason Calacanis

All right, let’s get started. There’s a lot to talk about, and we’ve got a big docket today. We’re going to start with the Pope. The Pope is dope.

Pope Leo—he’s the 14th—released his first encyclical on AI, and it was long: 235 pages, over 42,000 words. Just to give you an idea, Bill—

Bill Gurley

When did he write it, do you think? When did he put that together?

Jason Calacanis

Well, no, no, I think he used ChatGPT. That’s what it says here in the notes. No, I mean, I’m guessing—

Bill Gurley

How long did it take for him to write this in between all of his other tasks? I think it’s a 6-month process to do this, but I’m sure he had collaborators.

Jason Calacanis

Bill, your book, I’m assuming, was—

Bill Gurley

Write it.

Jason Calacanis

I’m sure there was a team that wrote it. But, Bill, your book’s 60,000 or 70,000 words, I’m guessing, so this is almost a literal book, right?

In terms of how long it is, it’s called Magnifica Humanitas, or “Magnificent Humanity.” In it, he warns business leaders to safeguard humanity from AI.

His core argument is that AI is not inherently evil, but technology is never neutral, and technology takes on the characteristics—wait for it—of those who build, finance, and control it. I don’t think he thinks super highly of that group of people.

The Pope called for regulation of AI companies. Obviously, we’re going to have that debate here. Some of the things he called for I think are not very debatable, and there’s a lot of consensus around worker retraining, safety for children, guardrails, and a ban on autonomous weapons. That’s the Skynet rule: Don’t build Terminators with your AI.

But he was joined by Anthropic co-founder Chris Olah. I don’t know how many co-founders there are of this company, but apparently there are dozens. Olah is not Catholic. According to a Vanity Fair profile, he was raised evangelical and now he’s an atheist.

The folks at Amazon, Google, and Meta lobbied the Vatican on April 29 to soften the language in his missive, and he was not swayed.

His central question, Sacks, is: Will AI be used to concentrate power in the hands of a few, or will it serve everyone? That’s something you brought up when you mentioned monopolies, duopolies, and so on.

Two weeks ago on this very podcast, what's your take on the Pope, his interest, and his missives on AI and promoting a bit of AI regulation?

David Sacks

I very much agree with the Pope that the biggest risk of AI is a centralization of power and then its misuse against us in some Orwellian way. I think it's government that's going to do that, not necessarily an individual actor, because governments ultimately have the power. So I do worry about the potential for AI to be used to surveil us, censor us, and control us, as Orwell described in 1984. If that's where the Pope is going with this, I very much agree with him.

Maybe where we end up in different places is that he thinks government regulation is the way to prevent this. I would just say that we have to be careful not to empower government too much, because if you give government the power to regulate or approve AI development—if you create, say, an FDA for AI, as many people are calling for—that will give government the power to approve models and therefore give notes to model developers. Very soon, this definition of safety will expand, because the government always takes an expansive view of its powers.

We saw this during the social media wars, where the definition of trust and safety expanded to issues like psychological safety, microaggressions, disinformation, transphobia, and so on. These social media companies were told that they had to stamp out all of those threats to safety, and it ended up becoming a censorship agenda. I get very worried about what happens if some government agency can give notes to the model developers and starts telling them that their definition of safety is not expansive enough. You have to protect the public from disinformation or psychological harms.

Again, I think we just have to be careful not to aggrandize government, because that's going to be the most likely culprit in terms of the centralization of power. I know the Vatican likes Latin. This is a problem of political philosophy that goes all the way back to Socrates. It's called “Quis custodiet ipsos custodes?”—who will guard the guardians?

In other words, if we entrust a set of guardians to protect us from a bunch of threats, what's to stop them from becoming tyrannical and becoming the new threat against us? This is the central dilemma of political power.

Jason Calacanis

Who watches?

David Sacks

Yeah. Who watches the watchers? Who guards the guardians? Who's going to protect us against our guardians if they turn against us?

The genius of the American founding is that it was a second-order solution to this question. The founders of America very much understood this, and what they came up with was that we have to have the guardians guard against each other. They came up with the idea of separation of powers. We'd have a separation of federal and state governments, and we'd have the 3 branches of government. Even within the legislative branch, it was a bicameral legislature.

They divided up the powers in a way that hopefully the guardians would check against each other, as opposed to becoming tyrannical against us. That is my view on AI: ultimately, we have to have a solution of checks and balances.

If the AI market becomes monopolized and falls into the hands of 1 or 2 companies, I would use antitrust law very aggressively as a check and balance against their power. Right now, we have a very competitive market. We have 5 frontier labs competing very aggressively. As long as the market is competitive, I would use that, because I think competition generates the best outcomes. It helps us win against China, but it also protects the population because if these companies get out of line, there's some competitor that can offer something better.

Jason Calacanis

Consumers can opt out of it. If they don't trust GPT, they can use Anthropic, or if they don't trust Anthropic, they can go to Grok. Bill, you had the number-one-rated talk at the All-In Summit in history, 2,851 miles. You have been famously against regulatory capture. In light of the Pope's comments—“Hey, regulating”—what do you think is common sense?

AI is everything. AI can help people make bioweapons. It can also help people get their term paper in or be a better salesperson at Oracle. We're talking about AI like we're talking about oxygen here. This is a fundamental horizontal technology. Where do you think there is a case to regulating AI, if at all, and where do you think the free market will figure it out?

Bill Gurley

I have 2 takes: 1 on the Pope and 1 on Anthropic. Your question is powerful. Let's go with the more powerful entity.

Jason Calacanis

You want to go in reverse, the least powerful of the 2? Go.

Bill Gurley

So this Pope said—and I have to learn how to pronounce all these Latin words, like you—that this encyclical was modeled after one done by Leo XIII in 1891, and he invoked that. He even said he chose the name because he's so enamored with Leo XIII. Leo XIII's encyclical warned that the Industrial Revolution was going to be bad for people.

Let me tell you what happened from 1891 until today. The workweek went from over 60 hours to 34 hours globally. Real wages went up 8 to 10×, adjusted for inflation. The median worker now earns more than a doctor did in 1891. Global GDP per capita went from 1,500 to 20,000. Child labor in the US went from 18% to 0%. Workplace deaths fell by 40×. Life expectancy went up 60%, and global poverty went from 75% of humanity to under 10%.

All those things happened because of technology, innovation, and capitalism, which is exactly what Leo XIII was warning against. So he got it dead wrong. He got the whole thing precisely wrong. It's an interesting thing to say you're borrowing from.

Jason Calacanis

Yeah. Now, on to Anthropic and just common sense. How would you regulate or protect against—maybe we'll broaden the term here—nefarious uses of the technology?

Obviously, we all want children to be protected. We want truth and honesty in terms of facts, and all of us sharing some basic truths. We obviously don't want people using this technology for bioweapons and the Terminator scenario.

Bill Gurley

I have to tell you that Anthropic is a mystery to me. I've never seen a company that is both leading its field and the most negatively outspoken commenter on what they do. I've just never seen it.

My initial theory was the regulatory-capture theory: they just want to ensure there's regulation. Quite frankly, I think they're very close to achieving that. They've stirred up a frantic position, especially in America. American consumers are definitely afraid of AI.

I think I've talked to you guys in the past about the book that Jonathan Haidt's written about social media, and there's a whole bunch of state legislators who think we should have regulated social media, so now they're destined to want to get in front of it. We know that Anthropic is one of the most aggressive lobbying startups of all time—the amount of effort they're putting in, the amount of money on a state-by-state basis.

That was always my first theory, but then they got so loud that, in the past 30 days, I've literally read everything I can about Anthropic, and I've come up with a new theory.

Jason Calacanis

Okay, new breaking theory.

Bill Gurley

I call it the Dr. Frankenstein theory. You remember when Elon had that conversation with Larry Page, where Larry was literally sitting next to him when he called?

Jason Calacanis

Explain the story real quick.

Bill Gurley

We were at a birthday party, and Elon was like, “Listen, humanity needs to be protected from the stuff at DeepMind,” because at DeepMind they had an example of the AI trying to break out—to jailbreak out—of its computer and not be turned off, and it had some sentience or some inkling of sentience. He said, “We have to protect the human species.”

Larry said, “What do you mean? That's speciesist, because you care about the human species over AI.” This was at least 15 years ago.

Jason Calacanis

No, this was right before Elon co-founded OpenAI, right? Back in 2015 or something.

Bill Gurley

The actual story here is that Elon Musk and Google had backed Demis and the team at DeepMind when they were an independent company. Then Elon was like, “Oh my God, Google's going to buy this.” I remember having the conversation with Elon about this: “We have to figure out a way for DeepMind not to go to Google. We have to block this somehow.”

He begged those folks not to sell to Google because Google was running the table on everything. He wanted this technology to be independent, and he was on the board of the company.

Jason Calacanis

He also said this was his motivation to launch OpenAI as a nonprofit.

Bill Gurley

Google got it. He just said, “This technology is too powerful for any 1 person.” Once again, you have to give Elon a lot of credit. He saw the writing on the wall: if 1 person can control this, it would be extremely dangerous. He saw it 15 or 20 years ago, and he and Sam Harris used to debate this over dinner.

What happens if somebody controls this and they run away with it? It would be extremely dangerous. It has to be available to all the people. Essentially, that's the Pope's position. It has to be in the service of humanity, not ruled by 1 person. It's far too powerful.

The reason I call this the Dr. Frankenstein theory is that, the more I dig, I've met people who, I dare say, think it's their responsibility, and they're excited about building a species that's superior to humans. I would just encourage people to read as much as they can about Anthropic.

Jason Calacanis

Chris Olah worked on this thing called the Constitution. It’s about 80 pages. It’s hard to get through, but I would encourage you to read it. Amanda Askell, who is the chief philosopher, has started doing podcasts. I would encourage you to listen to them and listen to her language.

And then Dario Amodei wrote this blog post called “Machines of Loving Grace.”

Bill Gurley

“Loving Grace.” I read it.

Jason Calacanis

It was based on a poem, and the poem is kind of weird. We should put a link to the poem. It’s quite short, but the last stanza of the poem says, “I like to think of a cybernetic ecology where we are free of our labors and joined back to nature. Returned to our mammal brothers and sisters.”

I don’t know what that means. We’re going to go live in the fields where the mammals live? And then the kicker: “All watched over by machines of loving grace.” Sounds like overlords to me.

In Dario’s post—and it’s very long; you read it, Chamath—I mean, “Machines of Loving Grace” is very long. He’s talking about, in the future, what humans are going to do because he believes in massive abundance and UBI, and that we won’t have to work. I don’t believe in any of those things, but he does. And then he says it could be a capitalist economy of AI systems, which then give out resources to humans based on some secondary economy of what the AI systems think makes sense to reward in humans.

So that’s envisioning a deity of sorts that’s going to break ties and decide what humans—

Bill Gurley

It’s a computational reward function for humans. It decides how much you’re worth.

Jason Calacanis

Yeah. So, I don’t think they think they’re writing software. I think they’re midwifing a deity here. I don’t know which one I’m more afraid of: the regulatory capture or this second theory, which I call the Dr. Frankenstein theory. It’s more scary to me. I think the second thing—

Chamath Palihapitiya

These are delusions of grandeur. Let’s call it what it is. They believe that they are so intelligent. I know some of these folks—the Burning Man sort of offshoot of transhumanism—and they believe that they’re so powerful, these individuals, that they can create God. By creating God, they are like this Prometheus kind of species.

It literally is the ultimate level of narcissism and delusion of grandeur to think you can create God, and that then the God you create, like you’re saying, Bill, is going to be so benevolent and perfect that you construct the perfect God that will give you your pellet, give you your little scenario—

David Sacks

Well, I guess the question then is, why are they pushing for the—let’s call it—regulatory capture agenda?

Jason Calacanis

I know why. Go ahead, Chamath.

Chamath Palihapitiya

That is very reductive game theory. So, if you want to be unexploitable, I think the best thing that you could do, if you’re trying to build a supergod, is have 3 or 4 entities in a room, close the door behind you, and then dominate those other 3 or 4 entities. Then you set the rules.

Because your counterparty is unable to track at the level of technical capability that you would have, you create this massive asymmetry that allows you to exploit them. That’s just simple game theory optimization.

What Bill said is so powerful. I’ve read these things, and it’s laborious and it takes time, but every time they put these things out, just take the time to read them. I initially thought that this was mostly game theory, that a lot of their reactions were less rooted in their dogmatic beliefs and more rooted in a GTO approach to either raising capital or putting pressure on competitors.

Either way, both could be true. What your framing is and my framing, although mine’s more tactical than yours, to be fair, because I’ve always thought that these moves make sense through that lens: How do you absorb most of the capital? How then do you make sure that you are in a position to disproportionately affect the rules? And how do you create an oversight body that is less capable and less intellectually aware of the actual details than you are? Because—

David Sacks

The referees don’t understand the game, right?

Chamath Palihapitiya

If the refs don’t understand the game, you’ll run over the game. Yeah.

Bill Gurley

By the way, one thing they have achieved by doing this is that if you pulled the—let’s just call them the intellectual elites, so everyone in the media and whatnot, the professors and all those people—and they were to rank the different AI players by who they think is most caring, I think they’d probably put Anthropic first because they’ve been out with the doomerism talk.

So it’s given them a halo with the people who may matter to what they want to accomplish. It’s simultaneously creating a lot of trouble, like with the data centers and whatnot. There are negative ramifications.

Jason Calacanis

What you’re saying is so important because, on the one hand, they create empathy, and then they write these documents that expose what they think, and nobody actually connects the dots.

Bill Gurley

Yeah. To steelman their position for a second, I think probably the way they think about it is that they are creating something very powerful, something godlike, and therefore it needs to be safe. They care the most about that out of everybody. Nobody else takes this seriously.

Remember that Anthropic was basically a spinout of OpenAI, and they felt that Sam and the company leadership weren’t taking their point of view seriously enough.

Chamath Palihapitiya

It was the most woke portion of OpenAI.

Bill Gurley

We’re steelmanning. So, they see the power of it. They’re the ones who are concerned about safety, and they care the most. Therefore, they’re in the best position to do that.

Jason Calacanis

Now, I think the issue is just you can see how this can lead to regulatory capture, right? If you brand yourself as the safe AI company and then try to characterize everybody else as reckless players, and reckless AI needs to be stopped, you can see how this would basically further your monopolistic control over this industry.

And if you see AI through the lens that, frankly, the pope and I see it—which is centralization versus decentralization—I do think that is one of the key lenses we should have on the technology: whether you want this to be a centralized or decentralized technology. This way of viewing the world leads to more centralization, and I think that’s dangerous.

I mean, if AI is this very powerful technology, I think it needs to be decentralized so that all of us can protect ourselves to some degree, right? We need to be able to run the AI ourselves, on our own hardware, if we so choose, so we’re not beholden to a single company that might be in bed with a deep state.

Chamath Palihapitiya

Let’s say it very pointedly: If benefits and compensation and economic support were all of a sudden tied to some algorithmic decision, this is a dystopian episode of Black Mirror that we’re dealing with. And to your point, Sacks, you want 100 or 1,000 or 100,000 versions of what that answer is, so that there’s actually a way to refute a singular answer.

A singular answer to these kinds of questions, which is effectively what some folks would want, is incredibly dangerous.

Bill Gurley

And this is something that is in control, I think, of humanity. I’ve been talking about AI sovereignty here for a bit, just in terms of how much more cost-effective it is and how you’re not training other people’s AIs with your knowledge and your insights.

This is why it’s super important that open-source agents and local hardware be able to run these models, and that consumers and companies learn how to roll their own language models—how to make a small language model, an SLM, a VSLM, a verticalized one—and run it on your Apple hardware. Apple has actually taken a principled approach historically to your sovereignty and your data. Data sovereignty now is privacy.

Jason Calacanis

Yes. And now it’s intelligence sovereignty. Intelligence sovereignty is different from privacy. Privacy is, “You can’t see my photos. You can’t peek into my Notes app and see what I wrote there in my journal.”

Now, intelligence sovereignty is, “You can’t tell me what to think. You can’t use your AI to analyze my photos, to analyze my emails, to analyze my messages, and tell me how to interpret the world.” That’s actually going to be the next key piece here.

This is why I think Apple is just the dark horse in this entire race. There is an open-source product that can run on this hardware—the M5s, the 48 gigs, 128 gigs, the new Mac Studio coming out with supposedly a terabyte. That changes the whole game.

And this is so paradoxical, Bill, that our adversary, the Chinese of all people—the Communist Party—is leading the open-source movement, and the United States is centralizing.

David Sacks

They’re leading the open-weight movement. It’s not open source. The distinction is important.

Jason Calacanis

Yeah. Yeah.

David Sacks

Look, Jason, I agree with you about the importance of open source, because open source means software freedom. You can run the program yourself on your own hardware. You don’t have to share. You don’t have to give up your data sovereignty. You don’t have to give up your privacy to, again, some monopolist who’s going to be in bed with the government or the deep state, right?

So that’s the thing we’re all afraid of. And if that’s the only AI that’s available—from the monopoly or duopoly—then your choices are to live off the grid and not participate in the modern economy, or give up control to some social credit system.

So I think open source is really important. And, by the way, that was Elon’s instinct in creating OpenAI. He was afraid that Google was going to monopolize AI, so he said, “Let’s create OpenAI so that it’s not dominated by a single company.”

I think that is the right answer here. People’s instinct, in response to the idea of powerful AI, is to clamp down and just control it. But actually, you have to have multiple players. That’s the only way you’re going to be protected: to have multiple players.

Jason Calacanis

This next wave of the market evolution, I think, is going to be extremely high-stakes and messy. Nick, just throw this up because I want these guys to react to it. This is a company that I just ran into on X called Rogo. What they did was create a test bench and a set of evals to essentially be a financial analyst, and then tested all of the frontier models.

I read the paper they published, and I quoted the most interesting part because I see it everywhere now across all evals: “There is no single best model anymore at the top of the leaderboard.” Opus 4.7, GPT-5.5, and Sonnet 4.6 appear almost indistinguishable, separated by less than, in this case, 3/10 of a percentage point overall. Read superficially, the results suggest convergence: 3 frontier systems reaching roughly the same level of capability.

Why is that interesting? You have trillions of dollars going into each of these companies to try to create these next superbrains, but increasingly, our existing set of evals and our existing capabilities, when applied to these models, roughly produce the same thing. Theoretically, that says these things are getting commoditized way too quickly. Then you’d say, “What’s the ROI on all this incremental spend?” which is a very interesting economic and investment question.

So, Gurley, what do you think happens if these evals continue to asymptote and we need more and more and more money for training?

Bill Gurley

Some of the smarter people in the open-source community have suggested to me that we need more open-source connectors of different types. MCP is actually run by the Linux Foundation, and if you think about any surface area where a model might interact with other software, the more of those connectors that can be open-sourced and commoditized, it would lower.

This is what Google did with Kubernetes: to try and commoditize where workflows live off AWS and make it easy to migrate. The more you can create systems that make the type of exchange you just described super easy, so that you can plug and play the model and not have to worry about things like context, how context comes in, and data—the kinds of things that Glean and Databricks do—the models become swappable.

Certainly, with both the model companies trying to move up the stack, you have a massive desire from the app-layer players to try to figure this out. We’ve already watched what Cursor is doing, playing with its own model and being forced to reckon with the fact that it’s coming up the stack fast. I think that’s a really good insight that this gentleman shared with me, and I think the founders and developers out there should work on more of these interfaces and throw them into the open-source world just to make them more exchangeable and swappable.

Is there an issue right now where we don’t have a good harness for open source? I mean, the way that Claude is a harness for—

David Friedberg

Yeah. There are people making open-source versions of this, or building companies around harnessing and building the integrations into it. But open source is always the last to build the fit and finish around the product. They focus on the core of the product, right? Linux for your desktop never really took off because the interface was never polished. The UI was never perfect.

There are companies building that, and I’ll show you one company that we invested in. This is a company called Abacus, and they had a very simple idea. They came up with their own hardware stack and their own platform, and now they’re sold out of the boxes they’re building for insurance and healthcare. Everybody wants to run AI inside their organization and then start building their own models. We actually incubated this in our incubator. You can check it out at goabacus.co.

They’re basically saying that organizations cannot get enough of this product. It’s crazy how savvy these organizations are getting, and Chamath, you’re doing it with 8090 as well, I think. They’re saying, “We have to build headless products so that we don’t get locked into any one provider.”

Whenever we go into the Fortune 1000, we never compete with OpenAI or Anthropic. They’ll sometimes have a preference for what they want to see under the hood, so our control plane can basically hot-swap, as Bill said, between one or the other. We’ve also started to lay the seeds for open source and open weights.

The reason is that they don’t want to be tied to one of these critical frontier labs. They want to be able to ride the wave of innovation, but they’re afraid of 2 things. They’re afraid that one technology leapfrogs the other too quickly for them to participate and they pick the wrong one. The second thing they’re increasingly afraid of is terms of service and being at the mercy of a frontier lab with a political philosophy that they may accidentally be in the crosshairs of.

You’re a hospital system in Canada. You support the euthanasia laws in Canada, but this frontier model in America says, “No, can’t do it. So now we shut you off.” That’s an example. I’m not saying one is right or wrong; it’s just to illustrate the case. A lot of the folks we see now in the Fortune 1000, and increasingly the Global 1000, want abstraction above it. They want to sit, as Sacks said, in a control plane. They want to be at this level and have the flexibility because they don’t know how it’s going to shake out.

They see all the money being invested at the model layers, but they see the model quality asymptote. So they’re asking, “Wait a minute, what are we supposed to do just from a risk perspective?”

Jason Calacanis

Regulated industries are particularly sensitive to these kinds of issues you’re bringing up.

David Friedberg

Hugely, hugely sensitive and regulated. If you just follow what finance, healthcare, and those kinds of folks are doing, they’re saying, “This has to be on-prem.” They’re very concerned about a data leak, and they’re very concerned about HIPAA compliance. They’re very concerned about training a model.

What if somebody queries or writes a prompt and it pulls some information from that Canadian healthcare system, and all of a sudden somebody gets a result? That sounds farcical. Remember, Stable Diffusion built itself on Getty Images—

The Getty Images watermark was suddenly in the output. You see Anthropic and OpenAI in all of these Fortune 1000s at the developer layer because most developers have their own credit cards and are allowed to sign up for them. Eventually, you wrap them in an enterprise license, so it’s a typical PLG-led market motion, like we saw with Slack. We’ve seen it everywhere.

The interesting thing is not that, but the unwind that happens when you have these huge licenses and these huge buckets of spend. You can’t really tick and tie it together. The CEOs then wake up and are told by the CFO, “Here’s where we are.” Uber was one example a second ago. I don’t know, Nick, if you have this tweet, but from Vivek Garipalli, the founder of Clover Health: “Overheard from a Fortune 20 company CEO: asked for $1 billion in AI-generated OPEX savings at the beginning of the year. We’re 6 months in, the team has spent $200 million on tokens, with minimal results.”

Now they’re in this weird motion where the CEO is pulling the budget back and they’re having to cut the licenses. You just saw Microsoft announce that they’re killing the Claude licenses. It’s a super-dynamic market right now, and I don’t think we know what the terminal solution looks like.

And, by the way, Claude is really good at product. Claude for Excel is better than Copilot—not by a little, by a lot.

Bill Gurley

Anyone who’s going to run against them, they are a worthy foe, I should say.

Jason Calacanis

Yeah, I think Claude is exceptional, by the way. I use it every day. Yesterday, I hit my token limit on my Pro plan. I had to put it on my credit card and spend another couple thousand bucks, and I was so angry—but I did it because it’s so good.

Yeah, go ahead, Sacks. Wrap us up here.

David Sacks

To wrap up, let me connect a couple of ideas. One is that, in terms of the regulatory-capture agenda you’re seeing in Washington, I think where it’s all leading is an effort to ban open-source models or open-weight models.

There are a lot of breadcrumbs leading here. I think people who want this are being a little bit circumspect. They don’t feel like they’re quite there in terms of being able to justify it yet.

Jason Calacanis

Can you explain it?

David Sacks

Sure. Look at a lot of the rhetoric around how models need to have guardrails, and that with open-source models, the guardrails can be removed and therefore they’re dangerous. You see this rhetoric already in Anthropic’s blog posts. Any threat that they describe, they go out of their way to take that shot at open-source models.

You saw it with respect to cyber, for example, or with respect to bio threats and things like that. I’ve seen that type of language repeatedly: that open models lack guardrails, or that the guardrails can be taken off and therefore it’s a problem. Again, I think they’re trying to create ideas or put predicate facts in the public record to justify an action later on.

I think it’s just a matter of time before they feel like they’re in a position where maybe they can push for that type of ban directly.

They're not quite there yet.

Jason Calacanis

But what does that do then to the rest of the market? Let's just say America bans open-source and open-weight models. Okay, well, what about the rest of the world?

Chamath Palihapitiya

It'll put—

Jason Calacanis

They're going to leapfrog us.

David Sacks

Sure. You'll put the US on an island. First of all, as we all know, what does it mean to ban an open-weight model? It's a file. It's a bunch of numbers that you can run on your laptop.

Bill Gurley

Yeah. But what it will do is, when you think about all the cloud service providers who run open models, they'll stop doing that because they have to comply with the law. All this infrastructure that's been built up will become much harder to use for open models in the United States. The rest of the world will continue to benefit from them because there's a tremendous benefit in terms of cost, customization, and control that you get with an open model.

Chamath Palihapitiya

And we're on a completely different price curve. We haven't talked about this yet. There was an economic and capital moat to training that is going away. It's going away in 2 ways. One is because we're getting these domain-specific architectures at the silicon layer. Second, we're rebuilding all of the core components. I don't know if you guys saw yesterday, but Elon was like, "We've rewritten the entire training stack in C++, and it's an order-of-magnitude increase. We can run it on 220,000 GPUs." At the scale of what they're trying to do, those kinds of innovations are going to make the cost of model training so much cheaper. Why would we stick to the $10 billion training runs when we can have the $10 million training runs?

Jason Calacanis

Well, if it got 1% better, just as a thought experiment—Nick, could you find Elon? If it got 1% better, that's the equivalent of 2,000 GPUs, which is the equivalent of hundreds of millions of dollars in compute. Every 1% equals hundreds of millions in compute. If he gets 10% or 20% more efficient every quarter—

Chamath Palihapitiya

Look at this speed improvement. The speed improvement versus JAX for training runs is now an order of magnitude. When you think about the capex buildout, the opex, the power, the cabling, the copper—all of it—this is a closed-source model, but I'm pretty sure that tweet alone is going to get read by enough people that there will be 5 or 6 open-source stacks for training rebuilt as close to the bare metal as possible.

David Friedberg

Yeah.

Chamath Palihapitiya

Why wouldn't you do that now? To your point, Sacks, cutting that off so that we lose that kind of innovation makes no sense to me.

David Sacks

I agree. Like I said, I don't know that the forces that want to ban open source are strong enough, or have made the case or created the predicate facts necessary yet, to ban open source. But I do think it is on the agenda, and it's where all the breadcrumb trails are leading. Just watch out for that. I agree totally with what David just said, and I wrote a blog post recently on open source and made the exact same point.

Jason Calacanis

I read that, too. That was a good one on Above the Crowd.

David Sacks

No, it's not. It was on the Santa Fe Institute.

Anyway, the exact same conclusion is that the rest of the world ends up running on Chinese models if they're able to succeed at what you just said.

Chamath Palihapitiya

And if you want to know the canary in the coal mine, Sacks, obviously the place they love regulation most is the EU. The EU has already done volley after volley of proposed regulation for AI, and open source is particularly in the crosshairs there because nobody's in charge of it. Are you going to get a bunch of open-source contributors having to vet their model with EU regulators? That's obviously not going to happen. Nobody's in charge of it. They're just a bunch of contributors. But open source is the solution, I think, to—

David Sacks

Yes, I agree.

Chamath Palihapitiya

It is the backstop. It is the backstop. Unless you want to live off the grid, if you want to participate in the modern economy, it is the backstop. Let me make one other final point. It maybe leads into our next topic: I do think there's the potential for the monopolization of this market to a greater degree than people may be pricing in right now. First of all, we've seen that every other major tech category has led to a monopoly or duopoly situation. That seems to be the way these things work out. But if you look at the growth rates right now, it seems like Anthropic has been pulling away. There's an article in The Information showing the latest numbers, and I think Anthropic has now pulled away from OpenAI, which is not surprising and something I predicted. If you have one company that's growing at 10x year over year and another company that's growing at 3x year over year, within 2 years, the first company will have 90% market share. This is the power of compounding. Just do the math: 10 × 10 is 100; 3 × 3 is 9. If you're able to outgrow your competitor at that rate for 2 years, you will achieve monopoly market share. Now, there are reasons to believe that Anthropic cannot continue that growth rate for 2 years. There's going to be a competitive response. It's already happened. Also, there may not be enough compute to support that kind of growth. There may be physical constraints. But you'd always rather be the company that has that inertia, that's on that trajectory, than the one that has to do something different to knock that leader off its current trajectory.

Jason Calacanis

Did you guys see what just hit the wire? Nick, can you throw it up from Polymarket? This is insanity. Polymarket put out there that an AI consultant revealed that one of their clients accidentally spent half a billion dollars in a single month after failing to set employee limits on Claude usage. [laughter]

Bill Gurley

What?

Jason Calacanis

Oh my God, look at this. $16.6 million per day, almost $700 per hour. Oh my God.

David Sacks

Well, there seems to be a new meme taking shape that somehow all this token spend is wasteful and basically useless. We're constantly oscillating between narratives like AI is going to put everyone out of work and AI is useless and it's a bubble. The doomers can't seem to make up their minds whether AI is going to be our new god or whether it's basically a total waste of money and is going to lead to a bust. But in any event, there's no question that token efficiency is going to be a big theme over the next year because the spend has been ramping up way faster than enterprise customers expected, and there's going to be a drive for efficiency. Does that fundamentally change the dynamics? I don't think so. But it might temper the growth to some degree.

Jason Calacanis

They've done a tremendous job making people believe that tokens are free by giving them these crazy deals: $20 a month, you can do whatever you want; $200 a month, you can do whatever you want. It's like everybody's leaving the hose on, everybody's watering, and then—

You get a message that says, "You've hit your usage. Come back at 2:30." I'm like, "2:30? It's 10:30. I can't do anything between 10:30 and 2:30." Then it says, "Well, you can put in your credit card." And so I did. It's literally like the first 10,000 gallons of water are free, basically, and then all of a sudden it's a penny a gallon. Everybody in the organization—and this has literally happened in our organization—started building interfaces. One person built an interface for the Founder University program. Another person built one. Then another person was like, "Well, those 2 people got credit at the management team meeting, so I'm going to build an interface." The next person builds an interface. Then everybody's shipping interfaces. I literally had 3 different people on the team make 3 different versions of a Founder University portal, and I'm like, "We don't need 3. Can we get coordinated here?" It didn't get to the point of spending thousands of dollars, but it certainly got to the point of spending hundreds of dollars, and it would have gotten to tens of thousands.

Are we still on the first topic? What are we doing?

Chamath Palihapitiya

Well, no, we kind of merged 2 or 3 of them together.

Jason Calacanis

Oh, we did? Okay.

It's super interesting, trust me. I think what Gurley said is one of the most interesting things I have heard in a long time.

David Sacks

Take people at their word. If you read their words, and you can understand what they're saying, you don't have to guess about why they want to have a digital god. Well, now, I'm not the sharpest arrow in the quiver, but I can take down a buck. [laughter] I can tell you that this doesn't make a lot of sense to me. Even the dullest arrow can take a buck down.

Jason Calacanis

All right, let's get back to it. It's so great having you here, Bill. We missed you.

Bill Gurley

I got you. I got you. [laughter]

Jason Calacanis

We missed you, brother.

We're going to transition to the next topic. There is some evidence that Dario is mitigating his doomer rhetoric. Did you see this? Let me get to it. Yeah, I got to it here. All right, we're going to have to talk for the 16th time in the last 18 months about AI's impact on labor because, again, this chaotic, schizophrenic interpretation of the data continues. Cloudflare, as we talked about last week—shout-out Matt Prince, Chamath's favorite CEO of the year.

Chamath Palihapitiya

Letter of the year. He cut 20%.

Jason Calacanis

Award for the letter of the year.

Chamath Palihapitiya

Making friends every week.

Jason Calacanis

Here on the program, they both blamed AI, explicitly and specifically, and Zuck then paired his 8,000 cuts at Meta with the fact that he has put spyware on everybody's laptop to study every employee to make their training data better. That got leaked, and people thought, “Hey, that's a Black Mirror episode. We're working at Meta in order to get our 2-year severance package.”

But on the other side of the table, Goldman Sachs CEO David Solomon wrote an op-ed in The New York Times: “I'm the CEO of Goldman Sachs. The AI job apocalypse is overblown.” Obviously, he might be fighting for that Anthropic or OpenAI IPO in the coming months, or maybe is doing it right now.

He made 3 points. AI won't eliminate 25% of jobs; it's going to automate 25% of work hours, and workers will fill that time with higher-level tasks. Obviously, that didn't happen in the case of Zuckerberg's layoffs. Just because a job can be replaced doesn't mean it will be. Bank tellers increased after ATMs. Live entertainment became more popular after TV. And the U.S. labor market creates and destroys 25 to 35 million jobs annually, and the gross churn dwarfs net losses.

New categories like agentic AI management are already hiring, yada yada yada. A publication called Fortune is apparently still publishing AI slop, and they say both Sam Altman and Dario have walked back their AI job apocalypse predictions as they gear up for an IPO. Sacks, have at it.

David Sacks

Well, I think you should be giving me more credit than that, because my most contrarian take back in January on our prediction show is that AI would lead to job gains, not job loss. And over the past week, you've now seen the narrative shift, I would say, almost completely toward that position.

So, you have the CEO of Goldman Sachs writing this in The New York Times. I don't think he'd be doing that if he felt like he was completely stepping out on a limb. Maybe even more importantly, you had Sam and now even Dario walking back their claims of massive job loss.

And they explained why. Dario said it's kind of like the 25% of work-hours thing. He said that AI might automate away 90% of someone's tasks, but the other 10% will expand to do a whole bunch of new tasks and new things, which is very similar to the types of arguments that people like me have been making. Actually, that's what Jensen's been saying: just because you automate away some task doesn't mean that you automate away the purpose of a job. The worker is freed up to do new things, to do the higher-complexity tasks that David Solomon, the Goldman CEO, is talking about.

So, the fact that Dario is now walking this back and coming around to my position, I think that's kind of amazing. And where do I go to get my apology?

Jason Calacanis

Well, we're going to have an official apology form that you can fill out. It's got checkboxes.

David Sacks

“I was wrong.” I mean, some mornings I woke up thinking, “Why am I going out defending these guys? These idiots.” They're scaring the public with all these dire predictions about an apocalyptic future. There was no data to support that.

We can all debate what's going to happen in the future, and we probably should be humble about what is going to happen in the future because we don't completely know, and this industry is very dynamic. But you have to look at the data that we have so far in the current situation, and we do not see data that supports massive job loss.

You can cite this layoff or that layoff, JCal, but those are anecdotes, and the plural of anecdotes is not data. If you look at the actual data, like Yale Budget Lab did, they said no discernible disruption in the labor market in the last 3 years due to AI. They've done a comprehensive study.

You look at job postings for software engineers: it's up 15% year over year. Job postings for software developers have hit a new 3-year high, despite the fact that coding is the single breakout use case of AI this year. So, if AI has not caused job elimination for software developers, what category has it caused? Code is now the number 1 use case, I think, of AI in the enterprise.

Jason Calacanis

Okay.

David Sacks

Let's be honest. Over the last 5 or 10 years, a lot of companies overhired. They mishired. These CEOs did not have a good handle on it. Their opex budgets completely got bloated and inflated, and they need to get back to where they were, get back to fighting weight.

Jason Calacanis

Never waste a crisis.

David Sacks

Never let a good crisis go to waste. Exactly. And so they point to this thing. It's very simple to say, “It's AI.” It's 2 letters. And they say, “We're going to fire people.”

But underneath that is not AI, because we know this: it hasn't done anything measurable yet in the consumption of these tools. Nobody is standing there and saying, “Look at my filing. Here is the lift that I have gotten.” Nobody has said that yet. That's very important to observe.

Instead, what people are doing is realizing, “Okay, I have this cover now to go and clean up what was very poor management and mismanagement over the last 5 and 10 years, where I overhired and I mishired.” That's what's happening today.

Jason Calacanis

Okay, Bill Gurley, I'm going to let you chime in here. You've got 2 besties saying, “Hey, this is all hogwash. It's AI washing.” These jobs were just—the strategy, obviously, in Silicon Valley was—

David Sacks

They need a scapegoat.

Jason Calacanis

They were hired 2 years ahead of time: build for the future. It was a vanity metric, and you were blocking talent from working on other startups or competitors. The Google strategy.

David Sacks

Hold on. Wait, wait, wait. You just said the critical thing. That is exactly why they did it.

Jason Calacanis

Yes. That was the explicit strategy from—

David Sacks

The actual strategy.

Jason Calacanis

These guys were awash in cash. Part of it is you were just hoarding talent, or what you thought was talent.

David Sacks

Yes, and just keeping them off the market.

Jason Calacanis

And now you're jettisoning it because the reality is, as companies get bigger, their growth rates monotonically decrease, and you get to a GDP plus some number. Your valuation frameworks change, and there's nothing you can do to fight that law of gravity in the public markets.

As each of these CEOs who, at some point, thought they were different and the rules didn't apply to them are now realizing, you're just like everybody else. Okay, we have to stay humble, as Sacks said. But Bill Gurley, would you like to apologize for Sacks and/or give him credit for his incredible non-consensus?

Bill Gurley

He wasn't the one promoting the job apocalypse.

Jason Calacanis

It was you. [Laughter] I'll give my thoughts in a moment. You're the voice for the mainstream media.

Bill Gurley

I'll give mine.

Jason Calacanis

You always represent the legacy media on our show. JCal, you have been in the fourth estate.

Bill Gurley

I represent the legacy media? You're the New York blue blood.

Jason Calacanis

I'm just giving you the statistics, guys. Now, let's remember anecdotes.

David Sacks

Let's remember.

Jason Calacanis

Actually, let me give you an important statistic. Let me give you a very—

David Sacks

No, no. This is really important.

Jason Calacanis

We have to let Bill Gurley comment. Then you can—

David Sacks

Really important. Do you use ketamine?

Bill Gurley

I don't use ketamine. That's the terrible drug. Do not use ketamine, folks.

Jason Calacanis

Bill Gurley, you have the floor.

Bill Gurley

I would just touch on 2 things that I already said earlier. One, historically, innovation has led to more prosperity for humans, and I gave those numbers from 1891 to today. I see no reason why that won't happen here.

In the short run, from a bottom-up perspective, every human that wants to protect themselves needs to be the most AI-enabled version of themselves they can be. The people that might be at threat of job loss are people who stand hard, fast, and refuse to use AI. That's simply like saying, “I'm not going to use email. I'm not going to use a spreadsheet. I'm not going to use a computer.” You probably are at risk.

Jason Calacanis

Yeah, yeah. The paradigm will shift. To give you actually my position, which is—

Chamath Palihapitiya

Would you like me to give my position, or do you just want to jump?

Jason Calacanis

Yeah, I do, but I never got to finish that point. But I can do it after you.

Chamath Palihapitiya

Yeah, yeah. So, I will give my position on this, which is—and it's always been the same—there's going to be massive job displacement that occurs. That massive job displacement is going to come because CEOs, in many cases, believe that this technology is going to make people more efficient. They can do more with less, and they will be rewarded by the public market by just having higher earnings. We see that for every single company.

Now, I fully concur it was because of bloating, and I gave my position there. I know specifically that Sergey and Larry took that strategy of taking talent off the market so there wasn't a Google competitor. That was literally explained to me by those individuals: “We hire people, and then we figure out what to do with them later.”

That strategy pretty much became the standard in Silicon Valley, and now it's being reversed. Now there will be wholesale jobs that will be retired.

If you look at self-driving, that's obviously happening with Waymo, with 3,000 vehicles, and there'll be many more on the roads. That job will be eliminated. We will be sitting here in 5 or 10 years, and the idea of somebody driving a taxi is going to seem silly and dangerous.

Jason Calacanis

We will see the same exact thing happen with Optimus. You may have seen the Figure robot sorting packages. All those sorting jobs at Amazon factories are going away. Amazon themselves—these are the savviest people in the world—said, “We are going to eliminate 600,000 future positions, and we are going to cut positions.” Andy Jassy said, “This is going to be a recurring theme. As we deploy AI, we will do more with less.”

You will see the headcount at all these big companies dramatically decrease or stay the same as earnings massively increase. You can take the position, Sacks, that, “Oh, my God, the numbers are in your favor.” They’re not. The numbers are in my favor. The job loss is tremendous, and there are numbers associated with that: 8,000 people at Meta after 20,000 before that. If you look at the steady state of these companies, they have nothing to do with AI. Let me finish.

David Sacks

They overhired.

Jason Calacanis

No, no, no. We are beyond that. We are beyond that. They are now getting rid of people. When they say they’re getting rid of managers, you can take them at their word. When they say they’re getting rid of middle managers, you can take them at their word—

Scapegoating. You’ve given your position already; I’m giving mine. My position is they are obsessed with this technology, they’re obsessed with earnings, and they will continue that.

On the other side of the ledger, I believe we’ll have a Cambrian explosion in startups, and all this talent, if they embrace the tools—to Bill Gurley’s point—is going to be able to solve more problems and create small companies of 5 or 10 people who were laid off from Amazon or Meta and make double their salary or have a better job that they control. I believe that is going to be the ultimate solution.

But that transition is going to be extremely painful, and we should have some humility on this podcast for the people impacted. Every cab driver is losing their job. Every truck driver is losing their job in the next 10 years. Anybody sorting packages is losing a job.

Now, you can say all you want—no, let me finish my thought. You can say all you want, Chamath, that those people don’t want those jobs. But they may need those jobs.

Chamath Palihapitiya

And you are an elitist by definition. We are all elitists on this program. We are elite performers.

Jason Calacanis

And these people are going to lose their jobs, and they may not get a job very quickly.

Chamath Palihapitiya

Being able to call something what we think it is is not being elitist. It’s actually telling the truth. Meta overhired. You could have stopped the company at 3,000 people when I left, and it would not have changed the outcome of that company. There was no need to go to 90,000 people and burn $50 billion on VR. They did it because they had the freedom to do it. That’s allowed. It’s capitalism.

They’re coming back to realize that there’s a more efficient version of what they are. That has nothing to do with AI. That’s the only point I’m trying to make. All you have to do is just say that.

Jason Calacanis

I think you’re wrong, and let me explain to you why you’re wrong. I believe you’re wrong. I believe Zuckerberg is putting that software on people’s computers in order to find more jobs to eliminate, to increase it. The surface area of problems in the world is not decreasing, but what is decreasing is the number of humans to take on the next opportunity. That’s going to continue.

I think the companies that will be rewarded, and whose stock prices will be rewarded, are the ones who do much more with much less. They’re going to keep eliminating these jobs, and I take them at their word.

David Sacks

You don’t have to explain everything with conspiracy. Maybe they just mismanaged for a period, and we could agree on that. I think that explains the post-COVID 2 or 3 years. I think what we’re seeing this year is actually the tools working. The tools are working, and there are jobs that are no longer needed.

The managers, as Matthew Prince pointed out, or product managers or designers, those have all been consolidated into 1 job: somebody who ships a product. It doesn’t require 12 people. It requires 2 people now.

Jason Calacanis

I don’t think that’s been consolidated. I see it in Fortune 1000 companies all the time. I don’t think what you’re saying applies. You’re talking about the slowest adopters. I’m on the front line with startups. These are where all the jobs are.

David Sacks

But I’m sorry, but a startup is not going to go and enter a regulated market and put JPMorgan out of business. Not going to happen.

Jason Calacanis

They will eventually displace those companies. It happens all the time.

David Sacks

Not going to happen.

Jason Calacanis

We’re going to agree to disagree.

Bill Gurley

Good luck to the startup trying to disrupt Boeing. Good luck. I’m going to take Boeing.

Jason Calacanis

Okay. Well, some people might take SpaceX.

David Friedberg

Good luck making drugs out of an Excel spreadsheet. I’ll take the regulated pharma company. Good luck.

Jason Calacanis

Sure. Listen, there are some industries that are much more regulated, and you’re going to show up at the FDA and say, “Okay, where’s your team?” “Oh, it’s just me. I do it all.”

David Sacks

Me and my model. Look at this.

Jason Calacanis

You joke. Somebody just did that—

David Sacks

It’s not a joke. It’s not a joke. And it’s not going to happen, because that’s not the way society wants safety, predictability, governance, and auditability to work.

Jason Calacanis

There’s a distinct difference between drugs and software and services in the world. I think we can agree on that. Listen, truck driving is one of the most regulated industries out there. So is cab driving and taxis, as Bill and I well know, and those jobs are being eliminated.

Bill, I’m going to give you the final word, then Sacks, I’ll give you the final word.

Bill Gurley

A chance to respond.

Jason Calacanis

Let’s do Sacks. Okay, Sacks, then Bill, go.

David Sacks

Well, first of all, Jason, you remind me of the Trotskyite who, when confronted with the fact that none of Trotsky’s predictions had come true, simply proved how farsighted Trotsky was.

Jason Calacanis

I didn’t go to graduate school. You’re going to need another reference.

David Sacks

In other words, none of your predictions about job loss have come true. In fact, the data—

Jason Calacanis

None. Zero data—

Except for what Meta just did last week. But go ahead.

David Sacks

That’s an anecdote. That is not—

Jason Calacanis

It’s not an anecdote. You’re calling 8,000 people losing their jobs an anecdote?

David Sacks

You don’t hear yourself?

Jason Calacanis

Hold on.

David Sacks

Do you hear yourself? It’s not an anecdote? 8,000 people lost their jobs.

Jason Calacanis

Can I make my case? I heard you about the Meta data point. First of all, that job loss was not directly attributable to AI. It just wasn’t. That’s something you’ve invented and put in the data.

Something Zuckerberg said. No, they clarified that. Okay, okay, sure.

David Sacks

He said it was related to trying to balance additional spending and capex, but it was not directly related to AI. But even if it were—even if it were 100% the case that it was due to AI—you’re not netting those jobs against all the other jobs that are being created because of AI and all the new companies that are being created right now because of AI.

So you’re just cherry-picking 1 statistic. You’re attributing 100% of that to AI, and then you’re not netting it and presenting a balance.

Jason Calacanis

Okay, so Jack Dorsey, Matthew Prince, Zuckerberg, and Andy Jassy are all lying and AI-washing. This was due to AI. He just did it.

David Sacks

That’s your reading of it. But like I said, even if you attribute those specific job losses to AI, which is questionable, you’re not netting it against all the job creation that’s happening and also the new company creation.

Jason Calacanis

I’m not cherry-picking it. I am reading the news and describing what the CEO said. Jack at Block said he’s doing this because of AI. Matthew Prince said it’s AI. Zuckerberg said it’s AI. I’m just taking them at their word.

David Sacks

Yes, exactly. So Jack Dorsey came out and said that he was going to do a 50% elimination because of AI. Within 24 hours, all the financial analysts on X said that Jack was AI-washing and that Block had horribly overstaffed during COVID. It was running much more inefficiently than all of its other peers in this category, and they needed to do a 50% job cut for a long time.

So pretty much everyone thought that was pure AI-washing. In fact, you’ve just proven my point. What exactly are the efficiencies that Jack is getting? This is the most hand-wavy thing ever: “Oh, we’re just magically going to be able to eliminate half our cost structure right now.”

Jason Calacanis

Okay, so Jack Dorsey, Matthew Prince, Zuckerberg, and Andy Jassy are all lying and AI-washing. This was due to AI. He just did it.

David Sacks

I specifically attributed that the future and the new jobs will come from startups. So don’t misrepresent my point. Thank you.

Okay, we currently have a 4.3% unemployment rate in the economy. Economists consider 5% to be full employment. So basically, unemployment is at or near record lows right now, despite the disruption of our lifetime, despite the fact that we’re over 3 years into this AI wave.

Second—and again, this is the point I wanted to make earlier with respect to coding—coding is the single job category most impacted by AI right now. We are at the point where AI is writing most of the code. We have almost complete automation of code writing. You would think that if you could look at this in a simple Malthusian way, all the software developers would be getting laid off right now. Is that happening? No. No, software developers are not being laid off on net.

In fact, job postings for software developers are at a 3-year high, growing 15% year over year. Why is this? I think the explanation is really, really important.

You look at code commits on GitHub, which is the leading code repository. There were 1 billion code commits last year. In the past month, there have been 1.1 billion.

Jason Calacanis

Make something easier, more people do it.

David Sacks

Right? We have basically a 14x year-over-year increase in code generation.

That code has to be managed by somebody. You still need humans to look under the hood. When the amount of code explodes and you get 10 or 100 times more code, the complexity rises as well.

We're not hiring 10 times more engineers, but you do need more engineers now to manage all of that code. The other thing that's happening is that there's been an explosion in the use of code across the economy by different businesses, applications, and use cases.

I'm hearing from people who are now hiring software engineers who never would have hired them before. I was talking to a fund manager, and he said that his next 2 hires were not going to be data analysts. They were going to be software developers because they're now deploying code for the first time in ways that they were not before.

This goes back to my point about Claude proficiency being the most marketable skill right now in the economy. People are using these tools in entirely new ways. I think we're at the outset of a boom right now caused by bespoke software proliferating throughout the economy and being used by firms that never thought of themselves as tech firms before.

All of which is leading to more productivity, and that leads to a healthier economy, and that leads to more job creation. You're seeing that again in the aggregate numbers, and that doesn't even include the blue-collar boom that's happening right now with the development of all this infrastructure, the data centers, and the new energy and power generation.

We're seeing hundreds of thousands of new construction jobs being created among blue-collar workers. Jason, I'm sure you don't want them losing their jobs by turning this boom off. So, again—

Jason Calacanis

No, I never advocated that. You misconstrue—you like to misconstrue my position. I am very clear that there's job displacement going on, and the job displacement is related to AI, but net, I do think the economy will grow. David—

David Sacks

Maybe at some point in the future you'll be right, like Trotsky: communism has never been tried. Maybe it'll work.

Jason Calacanis

Nobody knows your Trotsky references. You lost 95% of the audience. Just speak like a normal person.

David Sacks

Chamath laughed. Chamath understood it.

Jason Calacanis

Okay, great. I know the audience is smarter than you give them credit for.

David Sacks

No, I just think you're making these deep points to try to sound smarter than you actually are. The reality is the reality. These people are being laid off because of AI.

Jason Calacanis

David Friedberg, of the 20 million people in the United States driving cabs and trucks and doing that as a job right now, how many of those do you think will lose their jobs to self-driving in the next decade or 2? I'm not trying to lead the witness here in any way. Obviously, some people prefer a human driver, but what's your take on that specific part of the economy?

Bill Gurley

I think it's impossible to go with a 100% automated solution because the economics don't work well. I think, like some of the other examples that were given—ATMs and whatnot—the use of non-owned cars is going to go way up.

It's going to keep growing through this, and humans are going to be used for, say, 50% of it instead of 100%. I might not be surprised if the number actually stays the same or grows. Let's remember, these jobs didn't exist before because regulation had limited what the taxi market was capable of, and getting around that actually led to job creation.

I'm not a big fan of the doomerism around jobs. There's a word that's kind of used to talk about it, and I don't have high confidence in any government program for skills retraining. So it's not clear to me what happens after we say, "Okay, yes, it's happened. What do we do now?"

I think the thing you can do the most is, first, use the new tools. Know what they're capable of in your field. Get out there. Second, if your job is going to go away—and maybe it's a job you don't care about—start thinking about where there are opportunities.

Everyone's talking about it. The skilled trades are short of people.

Jason Calacanis

There's a shortage of plumbers, electricians, HVAC technicians—all that.

David Sacks

It's amazing how Jason uses facts that haven't happened yet as support for his argument. You just state that all the truck drivers are losing their jobs, all the drivers are losing their jobs, and then you say that this proves your take.

I know it's your belief, but that is not proof. Do you understand?

Jason Calacanis

The proof I gave was Amazon and Andy Jassy, Shopify and Tobi Lütke, and Marc Benioff at Salesforce. Let me ask Bill: everybody knows Amazon packages are being delivered, so you cherry-pick anecdotes and then misattribute them to AI.

David Sacks

They literally have a self-driving division. It's called Zoox.

Jason Calacanis

You're the biggest AI washer there is. They are the largest user of robotics in the world. So yes, Chamath, they are pursuing robotics massively more than anybody, and they are pursuing self-driving.

David Sacks

You just like putting all these words together. At one point it's a warehouse worker, then it's a driver, then it's Amazon. It's just—

Jason Calacanis

It's not. You can personally attack me all you want. The issue here is self-driving is going to take away, I believe, the majority of—

David Sacks

Okay, that's the key word: "believe." Let's put it there as a belief. Who knows? You don't know, and I don't know.

Jason Calacanis

Okay.

David Sacks

And I think the same about robotics. But I will take people at their word. I'm curious, Bill, your take on these large enterprises. You've heard 2 positions here.

Bill Gurley

I have a question for you.

Jason Calacanis

I have a legit question for you.

Can I just let the guest be involved, please? Sacks, you're monopolizing the conversation instead of actually engaging with his perspective. Explain to me—

David Sacks

No, no, let me truly ask you.

Bill Gurley

Okay.

David Sacks

Explain to me why job postings for software engineers are up 15% year over year despite the fact that code has now been fully automated.

Jason Calacanis

I think there's a Cambrian explosion in software. You're absolutely correct. I believe people who know how to vibe-code, or who are non-developers, are making bespoke software. I've said that 100 times on this podcast over the last few years, and I predicted it. So, absolutely, I believe that will be an area of job growth.

I believe the positions that are being removed—or, I know based on what we're hearing—are product managers and middle managers: what Matthew Prince called "measurers," what other people call mid-management.

Everybody believes that the recording, the daily stand-ups, and the Zoom calls—all of that is turning into people management—is being done better by AI, and people are more self-directed. The stack of people needed to build products is being consolidated.

The typical designer can now vibe-code. The developer can do front-end design and UX, and they can project-manage themselves. So there are a series of jobs that will increase and a series of jobs that will be eliminated, just like the mailroom got eliminated and bike messengers got eliminated.

By the way, did you guys see the news that Kirkland & Ellis is going to spend half a billion dollars to roll its own frontier model?

David Sacks

Makes total sense. That was like our earlier point today: people are doing on-premises deployments and going to make their own models.

Jason Calacanis

Bill, I have 1 specific question for you, and thank you for the good engagement there, Sacks. It lacked the ad hominem that usually starts every conversation we have.

Bill Gurley

I don't usually call you an idiot.

Jason Calacanis

That's because it's in our minds. We're thinking about it; we're just not saying it.

Bill Gurley

Good. I like it better.

Jason Calacanis

Bill, specifically, when Andy Jassy last spring said, "Hey, we're going to do more with less. We're going to be AI-first," and they said, "We're not going to hire these 600,000 jobs," and when you see Tobi Lütke say, "You have to do AI first before you ask for headcount, and prove to me that you tried AI first before hiring somebody," do you think this is a sign that these organizations are AI-washing? Or do you think these recent announcements are more, "Hey, we're going to do more with less, and the size of these companies will be smaller because of AI"?

Bill Gurley

One thing that I think that last question misses, and that a lot of the AI doomerism misses, is that competition exists. I don't think there's any scenario where you just do more for less and, all of a sudden, everyone has 70% operating margins. That's not going to happen. Someone else is going to come along and do more for less and lower the price.

So the thing that could happen is that we could have a productivity boom from lower-priced goods and services, and the basket of goods that humans are able to buy gets cheaper and cheaper and cheaper. That's been true in many categories. Unfortunately, it's offset by what happens in health care and other regulated industries.

Jason Calacanis

Yeah.

Bill Gurley

Education. But yeah, I expect products to get cheaper and people to be able to create more with less. But I don't think it leads to obscene profits because they'll be whittled away by competition.

Jason Calacanis

Okay.

David Sacks

By the way, just on this AI-washing point, there's a trial lawyer named Donnie King. He's a securities litigation partner at a firm called Akin Gump, and he and his colleagues have started to warn that we could start seeing shareholder lawsuits against companies that engage in this type of AI washing because he thinks it's a type of puffery, right? Essentially, what the company is doing is attributing its own nonperformance or operational issues to AI when, in fact, there are real problems in the business. Therefore, it could be a form of securities fraud.

Jason Calacanis

Wait, securities fraud? Yeah.

David Sacks

I want to double-click on this. Did you see the CEO of Wix today, in his note about layoffs?

Jason Calacanis

No. Who's Wix?

David Sacks

Find me the AI washing in there. Wix.

Jason Calacanis

Oh, yeah, that's the website builder. Yeah, you can build websites with Claude. The whole website business is challenged. Yeah.

David Sacks

Interesting that he just talked about operational details.

Jason Calacanis

Did he? I didn't read the note.

David Sacks

Of course you didn't read the note.

Jason Calacanis

Well, you said it just happened.

David Sacks

I will read it. This broke at 9:25 this morning.

I think it's really interesting that this lawyer thinks there's so much AI washing going on that he thinks it could constitute securities fraud, and he's warning clients not to engage in it. But look, JCal, you're like the last person who hasn't gotten the memo on this. There was a huge narrative shift this week. Sam Altman is backing off this. Even Dario Amodei is walking it back. You've got the Goldman Sachs CEO. You've got the explosion in job postings. Everyone's coming around to the idea that the job apocalypse is massively overblown.

Jason Calacanis

I mean, it could be overblown, looking at the stats.

David Sacks

Your apology—I'm happy to accept.

Jason Calacanis

No need for an apology. My position has always been: apologize.

David Friedberg

It's displacement. You're going to have some people displaced in the short to midterm, and then eventually there'll be more problems to solve and people will have to reallocate. I do think we're being—I think the tech industry itself doesn't have enough empathy or enough thoughtfulness when discussing this, because these are real people losing real jobs. You can point at statistics and think they're spinning, but these are real people losing real jobs who may not make the transition.

David Sacks

Jason, I got it. Do you think it's more empathetic to be scaring the bejesus out of people that they're going to lose their jobs?

Jason Calacanis

I'm not in the scaring camp. I'm not in the scaring camp. I am in the enabling camp. That's why I keep saying that if you've been laid off, you should start a company and embrace the tools. I'm all about empowering people. I think if they learn the tools, they'll have 10 job offers and they'll start their own companies. So I do think there's a solution to it. I just think we're going to go through low millions of jobs being lost, retired, and transitioned out over the next couple of years.

I was just going to be empathetic and offer some solutions. We talked about the skilled-trade deficit and the people working in those fields. Mike Rowe has a foundation called mikeroweWORKS. They funded $16 million so 2,600 people could get free scholarships to become plumbers, welders, or electricians.

Bill Gurley

Generation Toolbelt. Yeah.

Jason Calacanis

I think it's better than having the government fix things. As we started and talked about, I've got a new grant program myself to help people tilt their careers in a different direction.

David Friedberg

Yes.

Jason Calacanis

Go do something you love, and apply. Maybe I can help fund you moving in that direction.

Chamath Palihapitiya

As to your vibe shift, I think it's because, candidly, people's houses have been Molotov-cocktailed because of their doomerism. People are specifically citing that when they shoot at their houses and throw Molotov cocktails at them twice in the same week. If you're IPOing and coming out saying, “Hey, jobs are going away. Jobs are going away,” that's just a really bad look.

David Sacks

Or it's because we called it out and they got cut, so now they were telling the truth.

Chamath Palihapitiya

No, no, no. Sorry, I need to do one thing. Tulsi Gabbard is a friend of ours. I just want to give a huge shout-out to Tulsi Gabbard and, specifically, her husband, Abraham Williams. What he is going through is tragic. He's going through some really tough stuff with cancer, but he's going to kick its ass. I just wanted to say we love you.

Jason Calacanis

Yeah, Tulsi is great.