🎙️ Sam's Highlights

Sam Lessin's best standalone moments from every episode of More or Less — one clip at a time, newest first. Updates automatically as new episodes publish.

Home · All topics · Episode archive & transcripts · YouTube · Spotify

1:22 minSAM
#175 · Nov 13, 2026

Sam: Powerful People Don't Cheat Because They Think They'll Get Away With It — They Do It Because They've Stopped Feeling Anything

Sam's sharpest, most standalone theory of the episode: extreme wealth and power numbs you, and scandal is what happens when you chase danger just to feel something again.

1:08 minSAM
#175 · Nov 13, 2026

Sam: I Set an AI Bot to Monitor Every Executive's Insider Trading Filings

Sam shows off his latest bot project — scanning daily SEC filings to back out executives' real stock-sale strategies — and calls it one of the best public-data use cases for AI he's found.

1:04 minSAM
#175 · Nov 13, 2026

Sam: I Told Claude to Order Cameras for My House and It Just... Did It

Sam describes handing Claude full autonomy to buy, ship, and self-configure a home camera and VPN network — no manual setup required — and calls the result both amazing and slightly unsettling.

1:13 minSAM
#175 · Nov 13, 2026

Sam: AI Companies Have a Secret Dial Between 'Make It Smarter' and 'Make It Dumber'

Sam's core theory of the episode, delivered in one unbroken monologue: AI labs may be quietly tuning model quality down to protect margin, the same way ad businesses dial ad load up or down to hit a quarter.

1:09 minSAM
#175 · Nov 13, 2026

Sam: The Only Honest Reason to Buy Anthropic at IPO Is Regret Minimization

Sam explains why he'll never buy a trillion-dollar asset in size, why a genuine ten-trillion-dollar Anthropic barely makes sense, and why he'd rather dream big on a hundred-million-dollar company instead.

0:48 minSAM
#175 · Nov 13, 2026

Sam: The World Isn't Just More Algorithmic — It's Genuinely Getting More Entertaining

Sam's closing-stretch riff ties the entire Jensen Huang saga back to his running theory: humans are wired for entertainment, and AI is making the world dramatically better at supplying it.

1:00 minSAM
#174 · Nov 6, 2026

Sam: Introducing "Cuck Financing" and Slow's 0.0001% Offer to Every Seed Founder

Sam explains the new hot financing trick, where the lead invests a sliver at a huge price so everyone else pays it, then announces Slow Ventures will do it for any company that asks.

1:40 minSAM
#174 · Nov 6, 2026

Sam: To Understand the Agentic Future, Read Ari Emanuel's Book on Hollywood Agents

Sam's framework for how agent businesses will work: a few trusted relationships doing whatever you ask, a brutal market with no competency moat, and startups that get customers first.

1:34 minSAM
#174 · Nov 6, 2026

Sam: I Gave My Whole Fleet of AI Agents My Notes and Made Them Compete

Sam's real personal setup: a private notes file that every agent watches, with the agents racing to finish tasks he never assigned.

0:51 minSAM
#174 · Nov 6, 2026

Sam: Humans Are Terrible at Using Software. AI Is Excellent at It.

A tight, standalone reframe of the AI-software story: the bottleneck was never writing software, it was people being bad at using it.

1:13 minSAM
#174 · Nov 6, 2026

Sam: The Fun Jobs Aren't Nuclear Missiles, and in High-Stakes Software You're Really Paying for Trust

Sam explains why fast, experimental jobs go to agents while businesses where you can't screw up, like payroll, are really selling trust and a balance sheet.

1:53 minSAM
#172 · Oct 23, 2026

Sam: AI Labs Are Sitting on a Pile of Secret 'Skeleton Keys'

Sam riffs on the OpenAI PHP vulnerability to argue labs are quietly hoarding undisclosed zero-days, then ties it to his own early-Meta experience of product teams losing touch with what regular users actually see.

1:11 minSAM
#172 · Oct 23, 2026

Sam's Hobby Horse: Make the AI Labs Carry the Insurance for Their Own Risk

A tight, pure-Sam monologue laying out why he thinks labs centralizing AI's profits while radiating its risk onto society is unsustainable, and what a real liability-insurance fix would look like.

1:18 minSAM
#172 · Oct 23, 2026

Sam: My 'Low Tech' Mazda Dealership Is Secretly More AI-Forward Than Your Tesla

While the Morins brag about their self-driving Tesla fleet, Sam tells the story of his manual Mazda Miata service visit turning out to be unexpectedly teched-out, AI-transcribed inspection video and all.

1:17 minSAM
#172 · Oct 23, 2026

Sam: The Real Reason for the Merger Timing Is a Stock-Lockup Cliff

Sam's hot-off-the-press theory, delivered live: merging SpaceX and Tesla now dilutes the share count right before a massive SpaceX stock-lockup cliff hits next spring.

1:14 minSAM
#172 · Oct 23, 2026

Sam: My Bot Wrote Our Entire AGM Deck From One Prompt -- and It Beat Ours

Sam describes giving his AI bot a single prompt to write Slow Ventures' whole annual-meeting deck from scratch, and being genuinely blown away that it beat the version his human partners wrote.

1:02 minSAM
#172 · Oct 23, 2026

Sam Torches Anyone Who Calls Nathan Fielder an 'Untrustworthy Narrator'

Sam defends the new Elizabeth Holmes film as a masterpiece about humanity and lies, then roasts online critics who missed that the narrator's unreliability is the entire point.

1:27 minSAM
#171 · Oct 16, 2026

Sam: I'd Rather Be Right Than Rich as the AI Hype Finally Meets Sobriety

Sam's unscripted admission that watching the AI super-cycle narrative crack open feels better, intellectually, than the fact that it's making him poorer.

1:00 minSAM
#171 · Oct 16, 2026

Sam: 'Disney's a Public Company' — His Litmus Test for Trusting an AI Agent

Sam explains, via an old line from his dad on a Disney World roller coaster, why he'd hand Meta his data before OpenAI or Anthropic.

1:15 minSAM
#171 · Oct 16, 2026

Sam: My Old Startup Fin Booked Mark Andreessen's Lunch at the Wrong Restaurant

Sam tells the decade-old origin story of Fin, his human-plus-AI concierge startup with Andrew Cortina, and the exact mistake it made with Marc Andreessen as a user.

2:26 minSAM
#171 · Oct 16, 2026

Sam: Why His AI Concierge Startup Fin Could Never Make the Economics Work

Sam lays out, in full, the two structural reasons an AI assistant service can't survive: unbounded downstream liability and an error rate the business can't afford to eat.

2:30 minSAM
#171 · Oct 16, 2026

Sam: 'I Think What's Inevitable Is the Death of the Web'

The episode's title line, straight from Sam, followed by his theory that every API is about to become a live, Hollywood-agent-style negotiation for access.

1:10 minSAM
#171 · Oct 16, 2026

Sam: The Meme Lords vs. the Truth Subculture — Who Actually Ends Up Running Things?

Sam's compact theory of two parallel subcultures — the nihilistic majority and the truth-seeking minority that has historically held the money and the power.

1:08 minSAM
#170 · Oct 9, 2026

Sam: Great Technology Is Not a Great Business If It Has a Perfect Replacement

Sam explains why he loves Claude but can't see a durable business: the product is excellent, but if it vanished tomorrow a perfect replacement would exist, and Silicon Valley mistakes technology for strategy.

2:00 minSAM
#170 · Oct 9, 2026

Sam: The West Always Hated Zero and Infinity, and Silicon Valley's Whole Story Is Tails

A compressed theory of Silicon Valley's narratives: from the Pythagoreans to Bitcoin and GameStop, put a zero or an infinity in the chain and the tail takes over the story.

1:17 minSAM
#170 · Oct 9, 2026

Sam: Forget P(Doom). Make the AI Labs Carry Insurance for Lab Leaks

Sam turns the religious AI-doom debate into a practical policy: lab leaks could cause Katrina-scale damage, so require reserves and insurance the way post-2008 banking did.

1:24 minSAM
#170 · Oct 9, 2026

Sam: Agent Standards Don't Matter, Either Machines Can Reach You or They Can't

Sam dismisses MCP and agent-commerce standards as irrelevant and describes the real signal: sites like Hilton's are already throwing up anti-bot walls.

0:55 minSAM
#170 · Oct 9, 2026

Sam: AI Lets Every Company Negotiate Bilateral Deals With Bots at Scale

Sam sketches how bot-to-bot business development replaces the need for standards: Hilton can't do human deals with a gajillion agents, but their agents can negotiate with each other.

0:56 minSAM
#170 · Oct 9, 2026

Sam: There Are Two Kinds of LPs, and One Is Personally Offended When GPs Get Paid

Sam's sharp taxonomy of limited partners: the Silicon Valley kind that wants its managers to win, and the finance kind that treats GP profit as an insult.

1:06 minSAM
#169 · Oct 2, 2026

Sam: DC, Silicon Valley and New York All Need AI to Work (The Triangle Trade)

Asked what "not working out" even means, Sam flips it: Washington needs the GDP growth, Silicon Valley is having a party, and New York has built a financial engine on top. Without AI, the economy looks dire.

1:03 minSAM
#169 · Oct 2, 2026

Sam: If Claude Disappeared Tomorrow, I'd Care for About One Second

Sam, a heavy Claude user, explains the baseline problem in the core Anthropic business: forward deployment can lock customers in, but users barely care which model they use.

1:20 minSAM
#169 · Oct 2, 2026

Sam: There Is No Moat in Agents, Switching Costs Are Literally Zero

Sam shows how he moved from Claude to Muse by feeding it everything, and argues the only lock-in left is BD, which would make business teams the next overpaid class.

1:19 minSAM
#169 · Oct 2, 2026

Sam: The Real Story Is the Split Between the Haves and Have-Nots

Playing peacemaker in the screens-versus-real-life fight, Sam argues everyone will use tech differently depending on their life, and the bigger issue is a world splitting into haves and have-nots.

1:15 minSAM
#169 · Oct 2, 2026

Sam: Being Disconnected Is the Ultimate Status Symbol

Sam argues the rich go offline the way they play at self-sufficiency in the woods, and that disconnection has become the new luxury, until Adrian asks why he wants Starlink on his plane.

1:59 minSAM
#169 · Oct 2, 2026

Sam: AI Could Send Us Into Middle Ages 2.0, Where Everyone Hoards Behind a Moat

With no other way to build moats, Sam argues AI could produce a closed, stratified society of black-box systems and locked-down networks, a more practical worry than the extreme doom scenarios.

0:54 minSAM
#168 · Sep 25, 2026

Sam: "I'd Rather Send My Children to the Military Than to This School"

Sam drops his real, unguarded objection to a16z's new academy — not that it's marketing, but that he doesn't trust it to teach grit, discipline, and great books.

0:38 minSAM
#168 · Sep 25, 2026

Sam: AI Labs' Liability Isn't a Cost Line, It's "You Set Off Some Nukes"

Sam's clean standalone riff on why AI company liability is categorically different from normal corporate risk, built on a childhood memory of his banker dad explaining Disney at Disney World.

0:42 minSAM
#168 · Sep 25, 2026

Sam Explains the St. Petersburg Paradox in Under a Minute

Sam walks through the 300-year-old coin-flip math puzzle he's giving a talk on next week — and why its "pay infinity to play" logic is exactly the trap the AI narrative has fallen into.

0:54 minSAM
#168 · Sep 25, 2026

Sam: "A Perfectly Efficient Economy Destroys Capitalism"

Sam's sharpest line of the episode — if agents strip out all platform loyalty and trust, every marketplace's margin gets driven to zero, which is the opposite of what capitalism is supposed to produce.

0:49 minSAM
#168 · Sep 25, 2026

Sam: "I Don't Want a Computer, I Want a Ton of Computers"

Sam reveals he spends about $2,000 a month on CPU, not GPU, because his real bottleneck with agents like Muse isn't quality, it's not having enough parallel machines running at once.

40 Sam Highlight clips so far · browse all topics