This episode is a broad Moonshots-style recap centered on Google I/O 2026, with a second major segment on Cerebras’ record IPO and a shorter news item on Andrej Karpathy joining Anthropic. The speakers argue Google has reasserted itself in AI by leaning into its full stack—TPUs, search, Gemini, and consumer distribution—while also acknowledging the demos were often fast-follow or incremental rather than frontier-leading. The Cerebras interview frames wafer-scale inference as a major infrastructure bet that is finally being rewarded by the market.
Watch on YouTube ›Get the market thesis, key claims, assets, contradictions, and follow-up questions from any financial video — then unlock a version personalized to your portfolio, watchlist, and favorite speakers.
This is an x-long, highly structured market-and-tech discussion built around three pillars: Google’s AI product launches, the Andrej Karpathy/Anthropic news, and a deep interview with Cerebras CEO Andrew Feldman after a massive IPO. The core thesis on Google is that the company was widely written off, but has now re-entered the AI race by leveraging its installed base, its TPUs, and the ability to distribute AI across Search, Gemini, Workspace, Android, and shopping. The hosts repeatedly emphasize scale: quadrillions of tokens processed, hundreds of billions in capex, and over a billion users across multiple products. …
Near term, the actionable setup is around Google proving adoption for its new AI surfaces while the market watches whether its fast, cheap models actually improve search and consumer engagement. The immediate risk is that the demos impress more than the benchmarks, leaving the stock and narrative dependent on distribution rather than frontier leadership.
Over the next few months, the base case is that Google keeps integrating AI into search, shopping, and Workspace, gradually converting its user base even if rivals remain ahead on raw model quality. The setup strengthens if usage metrics continue rising and weakens if users keep defaulting to OpenAI or Anthropic for serious tasks.
Structurally, the episode points to a world where the durable winners are the firms that combine compute, models, and distribution into an operating layer for digital life. Over time, trust, provenance, and hardware supply chains may matter as much as model quality, making this as much an infrastructure regime as a software one.
Cerebras will 10x its market cap this year following its record IPO.
A speaker predicts Cerebras's market cap will increase tenfold, following the IPO closing up 68% at a $95 billion market cap.
We are moving from the information age to the verification age, and trust is becoming infrastructure.
Speaker states a paradigm shift where trust/verification replaces information access as the key value layer.
If Google hadn't leaned in at every layer of the AI stack, it would have been toast — its original search and ads model would have been cooked.
The speaker argues that Google's survival depended on vertically integrating AI from chips to applications, else its core business would have been disrupted.
How did Google's scale of AI usage and infrastructure growth change since last year, and what does that imply?
Sundar says token processing jumped from 480 trillion to 3.2 quadrillion per month, Gemini app usage more than doubled to over 900 million monthly active users, and capex is expected to rise to about 6x the 2022 level. The implication is that Google is now operating AI across a massive full-stack infrastructure with its own TPU fleet and energy-efficient chips.
Why were Google's AI growth numbers and capex expansion considered inevitable or predictable?
Alex says the outcome was inevitable because Google had to lean into AI at every layer of the stack or risk losing its core search-and-ads business. He argues the company is now full-stack in AI, from chips and data centers to products, and that Gemini's growth is partly helped by replacing Assistant.
What was Larry Page asking for early on, and how does that contrast with Google's current AI posture?
Alex recalls Larry Page asking for advice on how to get Google to spend $100 million on AI during a period when the company was not yet seriously leaning in. He uses that memory to show how far Google has come from underinvesting in AI to making it the central focus of the company.
Unlock the full claims, asset map, scores, related transcripts, follow-up questions, and AI chat — shaped around your portfolio, watchlist, favorite speakers, and risks.
