All-In: How AI Power Chooses Sides in US Politics

An episode of Dan's AI Intel

The four investors of All-In turned a podcast into a live map of how AI is choosing political sides — and the fault line cuts across party, not down it.

Published · By Dan Walter

Executive summary

Every Tuesday, four venture capitalists sit down to talk, and a growing share of Washington listens. The temptation is to treat the All-In podcast as a set of loud opinions. It is something more useful than that: a live map of how the people building artificial intelligence are choosing political sides. And the most important thing the map shows is that the sides do not run down party lines. They run along a single axis — build fast with a light regulatory touch, or slow down for federal safety rules — and that axis cuts straight across both parties.

Once you see AI politics that way, the year makes sense. A former White House czar who still runs a venture fund; a chip maker who hands the government a cut of its China sales; a nine-figure super PAC hunting safety-minded candidates in both parties; a lab whose staff give to Democrats while one of its leaders becomes a top Trump donor. None of it is tribal. It is a fight over who writes the rules for the most consequential technology of the decade — and the All-In table is one of the best seats to watch it from, precisely because the people at that table are also playing.

The room behind the podcast

Start with the door, because the room behind it is the real subject. All-In is a weekly roundtable of four Silicon Valley investors who call themselves "besties" — Chamath Palihapitiya, Jason Calacanis, David Sacks and David Friedberg — riffing on markets, technology and politics with the swagger of men who have made their money and stopped worrying about the invite list. It has become one of the most influential business shows in the country, and its live franchise, the All-In Summit, now draws the industry's principals rather than its commentators.

That matters because AI is no longer just a technology story; it is a power story. Who gets compute, who gets contracts, and whose warnings get heard are now political questions. All-In sits exactly where private capital, frontier labs and government meet — which is why watching it closely tells you more about AI's near future than any single lab announcement.

A podcast that became a policy on-ramp

The clearest proof came in September 2026, when Nvidia chief executive Jensen Huang was interviewed on stage at the All-In Summit and President Trump phoned in live to agree with him, calling fears of an AI takeover "a hoax." A sitting president using a tech podcast's stage to dismiss AI-safety concerns is not a cameo. It is a signal about whose framing has won inside the administration.

The on-ramp runs through David Sacks. Trump named him White House "AI and Crypto Czar" in December 2024; Sacks served as a part-time special government employee and, by his own account in March 2026, used up the 130-day limit that role carries and stepped back to an outside advisory seat. He is therefore a former czar, not a current one — but a former czar who helped build the administration's original AI doctrine around a simple premise: move fast, clear regulatory barriers, and treat China, not catastrophic risk, as the threat that matters. That doctrine became the 2025 AI Action Plan, unveiled at a Washington event co-hosted by All-In with a guest list that included Vice President JD Vance and Huang. Vance himself had sat for an All-In interview during the 2024 campaign. The podcast did not just cover the agenda; it helped stage it.

The fault line runs across party, not down it

Here is the central point, and it is the one a partisan read gets wrong. The real division in AI politics is not Republican versus Democrat. It is accelerate versus regulate.

On one side sits a broad "build fast, light touch" coalition: much of Silicon Valley's investor class, Nvidia's commercial interest in shipping chips, and a Trump administration that has embraced deregulation as national strategy. On the other sits a smaller pro-safety-rules camp — and its most prominent voice is not a politician but Anthropic's chief executive, Dario Amodei. In a June 2025 New York Times op-ed, Amodei opposed a proposed ten-year federal moratorium that would have barred states from regulating AI, calling it "far too blunt an instrument," and Anthropic instead backed targeted transparency laws such as California's SB 53 and New York's RAISE Act. A frontier lab lobbying for rules, against a deregulatory administration and much of its own industry, is the tell: this axis does not map onto party.

Exhibit — AI's fault line is accelerate-vs-regulate — and it cuts across party. Party predicts almost nothing here; position on regulation predicts everything. Source: NYT op-ed (Jun 2025); Axios; QZ; FEC-based reporting. Compiled by Dan's AI Intel.

Follow the money

Money makes the map concrete, and it flows toward acceleration. Take Nvidia. After the administration restricted its H20 chip exports to China in April 2025, it reversed course by July — and, in an arrangement Trump described as haggled down from 20 percent, agreed that Nvidia would hand the U.S. government 15 percent of those China sales, alongside Huang's pledge of some $500 billion in U.S. AI-server investment. Closeness to power and commercial interest here are the same thing.

The pattern repeats in campaign finance. A pro-AI super PAC called Leading the Future — backed by the venture firm Andreessen Horowitz and OpenAI president Greg Brockman among others — has assembled a war chest reported north of $100 million to support light-touch candidates and target those who want to slow AI down, in both parties. And it exposes the sharpest nuance of all. Inside the labs, the rank and file leans one way while a leader funds the other: OpenAI's employees gave to Democrats through 2026, yet Brockman and his wife personally became among the largest donors of the cycle to Trump's own super PAC, a reported gift on the order of $25 million (per federal filings). Employee sentiment, leadership money and corporate strategy are three different things, and collapsing them into "OpenAI is Democratic" or "OpenAI is MAGA" misses the real story.

Exhibit — The accelerationist side is spending at a scale that shapes who writes AI law. Source: NPR (Aug 2025); Built In; Transformer; Yahoo Finance/SFGate on FEC filings. Compiled by Dan's AI Intel.

The hosts are invested in the story they tell

The most direct connection is the one closest to the microphone. The All-In hosts are not journalists who happen to invest; they are working investors who happen to broadcast. Sacks runs Craft Ventures; Palihapitiya runs Social Capital; Calacanis runs the LAUNCH fund; Friedberg builds companies through The Production Board. They discuss, and often champion, the AI companies, chips and policies in which their firms hold positions.

Sacks is the documented flashpoint. Reporting in late 2025 found that he and Craft Ventures held stakes tied to more than 400 tech firms with AI exposure, and that he operated under two ethics waivers — one for crypto in March, one for AI in June — that let him shape federal policy while retaining private interests. He did divest some holdings, including in Musk's xAI, and Democratic lawmakers opened an ethics inquiry; one government-ethics scholar called the waivers "sham" documents. Whatever one concludes, the structure is the point: the people explaining AI policy to millions have money riding on the outcome, and they are shaping it too.

Musk is the proof the map keeps moving

If you still suspect these are permanent tribes, look at Elon Musk. In a single year he went from Trump's most prominent backer, to the de facto head of the Department of Government Efficiency, to an ex-official who left in May 2025, to a combatant in an ugly public feud that summer, to a partial reconciliation by year's end that Vice President Vance reportedly helped broker — and back to campaigning for Trump-aligned candidates into the 2026 midterms. All the while his companies — xAI and its Grok model, Tesla, SpaceX — remained deeply entangled with the state as contractor, regulator's subject and rival.

That volatility is the evidence. These alignments are contingent, transactional and personal, not ideological. And that is exactly why they matter: because they are unstable, the question of who holds the president's ear on AI this quarter genuinely determines who gets compute, who wins contracts, and whose safety concerns reach the room. We traced the same accelerate-versus-caution split among Musk, Amodei and Palantir's Alex Karp in our episode on why AI's own safety warnings fuel the race (episode 29, from mid-2026).

The verdict

Take the All-In worldview seriously on its own terms first, because it is a coherent one. The hosts are techno-optimists: they believe AI is the growth engine of the century, that heavy-handed regulation would hand the future to China, that "doomerism" is often a mix of fear and rent-seeking, and that a free-market, builder-first culture is what made American technology dominant. On plenty of this they have a real case, and they argue it in the open.

That is the genuine value of the show, and it is rare. Nowhere else can a listener hear the people actually shaping AI and its policy think out loud, at length, unscripted. For understanding how power sees this technology, it is close to essential listening.

The honest problem is the blur. When a host holds — or recently held — government power and a portfolio of stakes in what the show promotes, journalism, lobbying and self-interest run together, and the audience is rarely told where one ends and the next begins. Add the echo-chamber tendency of four friends who mostly agree, and you get something that informs and advertises at the same time. The fair verdict is not that All-In is propaganda. It is that it is the most candid available window into AI's ruling class — and that you should watch it exactly the way its hosts watch a pitch: fascinated, and counting who stands to gain.

Bottom line

The All-In table is worth a serious listener's time, because it maps a fight that will shape AI law for a decade — accelerate versus regulate, cutting across party. Just remember that the cartographers own property on the map.

Sources

A note on sourcing: this report was built from current web reporting accessed via search summaries rather than opened primary pages; figures are stated at the tier the reporting supports, and single-sourced numbers that could not be corroborated were dropped.

Transcript

Alex: Every Tuesday, four venture capitalists sit down to talk, and a growing slice of Washington leans in to listen.

Sam: And here's the part that reorganizes everything — the sides they're mapping don't run down party lines. They cut straight across them.

Alex: The people building AI are choosing political teams right now. And the four guys with the microphone aren't just calling the game. They're on the field. Welcome back to Dan's AI Intel — the show that takes the one AI question that actually matters that week and digs past the hype and the fear to what's really going on underneath.

Sam: I'm Sam, and as always I've got Alex here to keep me honest. And today we are getting into the All-In podcast — you know the one, the four Silicon Valley besties riffing about markets and tech and politics.

Alex: Right, but we're not here to review the show. We're here to read it as a map. Because if you want to know where AI power is flowing in America right now — who gets the chips, who gets the contracts, whose safety warnings actually reach the President — that table is one of the best seats in the house.

Sam: So here's the trip. We'll walk into the room behind the mics and figure out why a business podcast became a literal on-ramp into federal policy. We'll follow the money, and I promise you there's one company where the staff and the boss are funding opposite sides — it's wild. And then we'll land on the honest question: how should you actually listen to four guys who are talking their own book?

Alex: There's a turn in here that flipped how I think about the whole year, and it's got nothing to do with left versus right. But we'll earn it.

Sam: Before we dive in — if you've been enjoying the show, do the one-tap thing and follow us wherever you're listening. It's free, and it means the next one just shows up. Okay. The room.

Alex: So picture it. Four investors who literally call themselves "besties" — Chamath Palihapitiya, Jason Calacanis, David Sacks and David Friedberg. They sit down every week and riff with the swagger of guys who already made their money and stopped worrying about who's on the invite list.

Sam: And it's not a fringe thing anymore.

Alex: Not even close. It's become one of the most influential business shows in the country. And their live event, the All-In Summit — it used to draw the commentators. Now it draws the principals. The people who actually run the companies.

Sam: Okay, so why does a podcast full of rich guys with opinions matter to us, specifically, on an AI show?

Alex: Because AI stopped being just a technology story. It's a power story now. Who gets compute, who lands the government contracts, whose fears get taken seriously in the room — those are political questions now. And All-In sits right on the seam where private capital, the frontier labs, and the government all touch. Watch it closely and you learn more about where AI's heading than from any single lab's launch event.

Sam: Alright, give me the moment where this stopped being theory.

Alex: September this year. All-In Summit. Jensen Huang's on stage — that's the leather-jacket guy who runs Nvidia, whose chips this entire boom is basically fighting over. He's being interviewed, and mid-conversation, President Trump phones in. Live. To agree with him.

Sam: Wait, the President just... called into the podcast?

Alex: Called in, and used the stage to wave off the whole idea of an AI takeover as — his word — "a hoax."

Sam: See, that's not a cameo. A sitting president dialing into a tech podcast to personally dismiss the safety worries — that's a signal about whose version of the story is winning inside the White House.

Alex: Exactly right. And the on-ramp for that has a name: David Sacks. Think of him as the PayPal-era investor Trump tapped to be his "AI and Crypto Czar" back in December 2024.

Sam: Czar. So he ran AI policy for the administration.

Alex: He helped build its original doctrine — move fast, clear the regulatory barriers, and treat China, not some runaway catastrophe, as the real threat. That became the 2025 AI Action Plan. And it got unveiled at a Washington event co-hosted by All-In, with JD Vance and Huang on the guest list. Vance had even sat for an All-In interview during the campaign.

Sam: Hold on, so the podcast didn't just cover the policy —

Alex: It helped stage it. That's the on-ramp. And one correction I want to nail down, because people get it wrong: Sacks isn't the czar anymore. By his own account this past March, he burned through the 130-day limit that role carries and stepped back to an outside advisory seat. So — a former czar. But a former czar who wrote the playbook everyone's still running.

Sam: Okay, this is where I'd normally reach for the red-team, blue-team framing. Tell me why that's the trap.

Alex: Because it's the thing a partisan read gets completely wrong. The real division in AI politics isn't Republican versus Democrat. It's accelerate versus regulate.

Sam: Draw me the two camps.

Alex: On one side, a big "build fast, light touch" coalition — a lot of Silicon Valley's investor class, Nvidia's commercial interest in shipping as many chips as possible, and a Trump administration that's made deregulation the actual national strategy. On the other side, a smaller pro-safety-rules camp. And here's the tell — the loudest voice in that camp isn't a politician at all.

Sam: Who is it?

Alex: Dario Amodei. The guy who quit OpenAI over safety and went and started Anthropic.

Sam: Remind me — that's the lab that pitches itself as the careful one, right?

Alex: That's them. And Amodei goes and writes a New York Times op-ed against a proposed ten-year federal moratorium — a rule that would've banned individual states from regulating AI at all. He calls it "far too blunt an instrument." And instead Anthropic goes and backs the targeted stuff — California's SB 53, New York's RAISE Act. Transparency laws.

Sam: So let me make sure I've got the shape of this. You've got a frontier AI lab actively lobbying for rules —

Alex: — against a deregulatory administration, and against most of its own industry. That's the whole point. When the company building the tech is begging for guardrails while the government waves them off, you know for a fact the fight isn't about party. Party predicts almost nothing here. Where you stand on regulation predicts everything.

Sam: Alright, you keep saying follow the money. Let's actually do it.

Alex: Best single example is Nvidia. So back in April, the administration restricts Nvidia's H20 chips — the ones tuned for the China market. Export controls, national security, the usual.

Sam: Clamping down. Makes sense.

Alex: Right, except by July they reverse it. And the deal that comes out is genuinely something you have to say out loud to believe. Nvidia agrees to hand the U.S. government fifteen percent of its China chip sales.

Sam: The company pays the government a cut? Like a franchise fee on its own exports?

Alex: Trump described it as haggled down — he'd wanted twenty. And alongside it, Huang pledges something like five hundred billion dollars of AI-server investment inside the U.S.

Sam: Okay so when we say closeness to power and commercial interest are the same thing here —

Alex: They're not even two things anymore. That's it, in one deal. And then the same pattern shows up in campaign money. There's a pro-AI super PAC — it's called Leading the Future — and it's stacked. Andreessen Horowitz behind it — that's the big VC firm that went all-in for Trump in '24 — plus OpenAI's president, Greg Brockman, among the backers.

Sam: How big a war chest are we talking?

Alex: Reported north of a hundred million dollars. And the mission is naked: back the light-touch candidates, target anyone who wants to slow AI down — in both parties.

Sam: Both parties. There's that cross-cutting thing again.

Alex: And this is the nuance I teased at the top — the one that broke the tribal read for me completely. Look inside OpenAI. Through 2026, the rank-and-file employees are giving to Democrats.

Sam: Okay, so OpenAI leans blue. Filed.

Alex: Except — Brockman, the president, and his wife personally become among the single largest donors of the cycle to Trump's own super PAC. A gift reported on the order of twenty-five million dollars.

Sam: Wait, wait. The staff are funding one side and the president of the company is writing a twenty-five-million-dollar check to the other?

Alex: Same company. And that's why "OpenAI is Democratic" or "OpenAI is MAGA" are both just wrong. Employee sentiment, leadership money, and the corporate strategy are three completely different animals. Collapse them into one and you miss the actual story.

Sam: So let's go back to the four guys at the table, because this is the part that nags at me. They're explaining all of this to millions of people.

Alex: And here's the thing to hold onto: they're not journalists who happen to invest. They're working investors who happen to broadcast. Sacks runs Craft Ventures. Palihapitiya runs Social Capital. Calacanis has the LAUNCH fund. Friedberg builds companies through the Production Board. They're discussing — and a lot of the time championing — the exact AI companies, chips and policies their own firms have money in.

Sam: And Sacks is the sharpest version of that, I'm guessing.

Alex: Sacks is the documented flashpoint. Reporting late last year found he and Craft held stakes tied to more than four hundred tech firms with AI exposure. Four hundred. And he was operating under two ethics waivers — one for crypto in March, one for AI in June — that let him shape federal policy while keeping those private interests.

Sam: That's the fox-designing-the-henhouse energy people worry about.

Alex: He did divest some of it — including his stake in Musk's xAI. Democratic lawmakers opened an ethics inquiry. One government-ethics scholar flat-out called the waivers "sham" documents. And look — you can land wherever you want on his intentions. The structure is the point. The people explaining AI policy to millions have money riding on the outcome. And they're helping shape that outcome too.

Sam: Okay, but here's my pushback. Isn't this just... normal tribes? Rich people back their side, everyone knows the teams?

Alex: That's exactly the intuition I want to break. So look at Elon Musk, just over the last twelve months. He starts as Trump's most prominent backer. Then he's the de facto head of DOGE, the government-efficiency thing. Then he's an ex-official — he leaves in May. Then it curdles into a genuinely ugly public feud over the summer. Then a partial reconciliation by year's end, which Vance reportedly helped broker. And then he's back out campaigning for Trump-aligned candidates into the midterms.

Sam: That's not a tribe. That's a weather system.

Alex: And the whole time his companies — xAI and Grok, Tesla, SpaceX — stay tangled up with the government every possible way: as a contractor, as something the regulators are watching, and as a rival. So that volatility isn't noise. It's the evidence. These alignments are contingent, transactional, personal — not some fixed ideology.

Sam: And if they're that unstable, then the question of who's got the President's ear this quarter genuinely decides who gets the compute and the contracts.

Alex: You just said the quiet part. And it's the same accelerate-versus-caution split we pulled apart with Musk, Amodei and Palantir's Karp back in episode 29, "Why AI's Safety Warnings Fuel the Race" — that was mid-summer. If this thread grabs you, that one goes deep on the people who warn loudest and build fastest.

Sam: And honestly, for the Amodei side of it, our kill-switch episode — number 23 — is the companion piece. The safety-credible CEO asking the government to hit the brakes on his own tech. So bottom line me. Is All-In good or bad?

Alex: Neither, and I want to take their worldview seriously first, because it's coherent. These guys are techno-optimists. They genuinely believe AI is the growth engine of the century, that heavy-handed regulation would just hand the future to China, that a lot of "doomerism" is a mix of real fear and rent-seeking, and that the free-market, builder-first culture is what made American tech dominant in the first place. And on a lot of that? They've got a real case. And they argue it out in the open.

Sam: Which is rare, actually.

Alex: It's genuinely rare. Nowhere else can you hear the people actually shaping this technology and its policy think out loud, at length, unscripted. For understanding how power sees AI, it's close to essential.

Sam: So where's the catch?

Alex: The catch is the blur. When a host holds — or just recently held — government power and a portfolio of stakes in the exact stuff the show is promoting, then journalism and lobbying and self-interest all start running together, and you the listener are almost never told where one ends and the next begins. Add four friends who mostly already agree, and you get something that informs you and advertises to you at the same time.

Sam: So it's not propaganda.

Alex: No. The fair verdict is that it's the most candid window we've got into AI's ruling class. You just want to watch it the exact way its hosts watch a pitch — fascinated, and quietly counting who stands to gain.

Sam: I love that. The table's worth your time, because it's mapping a fight that'll shape AI law for a decade — accelerate versus regulate, cutting clean across party. Just don't forget the cartographers own property on the map.

Alex: And that's the discipline, honestly. You can think the show is close to essential and still keep one eye on the ledger the whole time. Both things are true at once.

Sam: So let's bring it home. Three things to walk out with. One — the real AI fault line is accelerate versus regulate, and it ignores party completely. Two — the money is pouring toward acceleration, from Nvidia's fifteen-percent deal to a hundred-million-dollar super PAC to a twenty-five-million-dollar check that split one company down the middle. And three — the guys narrating it all are invested in the story they're telling.

Alex: And I hope you came away seeing a bit more clearly where this is actually heading. It's a fast, tangled, genuinely high-stakes picture with a brutally short shelf life on what's true this month — and that's exactly what makes it worth following closely.

Sam: One honest note on how this show gets made: it's AI-generated. Dan builds a custom stack of AI tools to research, verify and illustrate the questions worth understanding — mostly to learn them himself — and then publishes it for anyone who wants to follow along. Deeply researched, fact-checked, and always worth a second look.

Alex: And before you go, one genuinely useful thing you can do: follow the show. Whatever app you're in right now, there's a follow or a plus button — it's one tap, it's free, and it does two things. You'll get every new episode the second it drops, and honestly, for a small independent show like this one, a follow is the single biggest lever there is for helping it reach other people trying to make sense of all this. So if today was worth your time — go hit follow.

Sam: And one last ask before we wrap. If there's something in here you'd push back on, or a thread you want us to pull harder on next time, tell us — the address is podcast@connectiveshift.com. We read every single message, and it genuinely shapes what we dig into next. So what should we go deep on? Let us know.

Alex: Thanks so much for listening. We'll see you in the next one.