Speaker
Iain Thomson
Appearances over time
1 episodes
Episodes
1
Podcasts
Quotes & moments
Only 16% of people surveyed said AI was actually helpful in their products, while 61% said they didn't want AI in any products at all.
After Gmail offered two-factor authentication for 10 years, only 10% of users had actually enabled it — highlighting why security defaults matter enormously.
The US government banned Anthropic's Fable model, expecting to suppress it. Instead, Anthropic's share of paid AI business subscriptions surged to 41% in May, beating OpenAI's 39.5% for the first time. Forbidden fruit is the world's best marketing.
OpenAI's independently verified financials show a $38.5 billion loss last year. But most of it is non-cash — nonprofit-to-for-profit restructuring and stock compensation. The real question is whether the underlying business can ever generate enough revenue to justify a near-trillion-dollar valuation.
OpenAI donated to Trump's PAC, pledged 10% of its IPO to the US government, and put Sam Altman on stage at the inaugural to announce Stargate. Anthropic did none of that and got banned. OpenAI's chief global affairs officer is Chris Lehane — the political mastermind behind Airbnb's regulatory battles. Anthropic needs to play the same game.
Elon Musk predicts SpaceX will generate $1 trillion in annual revenue by 2030 and build a million-person Mars colony. Iain Thomson read the S-1 and says it makes WeWork's prospectus look conservative. Born on Mars? You'd need an exoskeleton to walk on Earth.
Waymo recalled nearly 4,000 robotaxis after 13 incidents of vehicles plowing into closed highway construction zones — 6 in Phoenix, 7 in San Francisco. This is the sixth recall for Waymo. Despite this, Waymo claims a 13x improvement in serious injury crashes versus human drivers.
The UK government is banning children under 16 from YouTube, TikTok, Snapchat, and Facebook by year's end. The only real enforcement mechanism would be mandatory government ID for all social media — a privacy nightmare. Iain Thomson, who has friends in Australia, says the equivalent Australian ban simply hasn't worked.
Meta is lobbying for the Kids Online Safety Act to shield platforms from child-harm lawsuits and shift age-verification responsibility to Apple and Google app stores. They're already facing 2,000+ lawsuits, and the LA jury ruling that Instagram is a 'defective addictive product' opened the floodgates.
Fox is spending $22 billion to acquire Roku, giving the Murdoch media empire control over smart TV hardware and software used in tens of millions of homes. Doc Rock is already noticing the propaganda creep in the Fox Sports World Cup app — and worries what Fox will do to Roku City.
Snap's Spectacles X launched at $2,195 — self-contained AR glasses with a heads-up display, running 4 hours on a charge. They're heavy, widely mocked as ugly, and potentially pivotal for Snap, whose stock is down 95% from its 2021 high. Everyone's waiting for Apple.
Midjourney — best known for AI image generation — is opening a spa in San Francisco's Union Square where you can get a full-body ultrasound scan using 40 Butterfly Network imaging chips analyzed by Midjourney's AI. It's a complete pivot and an all-chips-in bet on a new market.
A federal court forced Google to hand over data on 300+ users who searched for pipe bomb-related terms near the DNC and RNC. The case, just unsealed, reveals the emerging threat of keyword warrants. More alarming: every AI chatbot keeps logs of everything you tell it — including documents you upload.
OpenAI and Anthropic are each approaching one million square feet of leased office space in San Francisco — enough for 6,000–7,000 employees each. Anthropic has cornered three consecutive blocks of Howard Street. San Francisco may be being reshaped by AI companies the way Seattle was by Amazon and Microsoft.
Two years after Google paid roughly $2.7 billion to hire Noam Shazeer via Character AI, the inventor of the transformer architecture — the technology underlying all modern AI — has now moved to OpenAI. Nobody knows what OpenAI offered him, but it must be staggering.
Doc Rock does morning pages by hand every day, then walks while dictating to Claude, which synthesizes raw thoughts into tagged notes with priorities. For someone with ADHD, the combination of analog writing and AI synthesis has been transformative — a model for responsible AI adoption.
The official reason for banning Fable was AI-assisted vulnerability discovery. The real reason, according to TechDirt's Mike Masnick, may be that the security researcher who reviewed the research was associated with Chris Krebs — the CISA director Trump fired for saying the 2020 election was fair.
Analysis
What they talk about
- Government 67%
- Business 33%
Connections
Shows they appear on and people they share episodes with. Drag to explore.
