Back to Podcast Digest
Alex Kantrowitz1h 9m

Did Google Just Fall Behind Again?, WWDC Preview, Anthropic Files To Go Public

TL;DR

  • Google's real problem is not just Gemini 3.5 Pro slipping: MG Siegler argues the bigger issue is that Google still lacks a true Claude Code or Codex-style product while Anthropic and OpenAI are defining the next interface for computing.

  • The "super app" thesis is getting more concrete: Kantrowitz says tools like ChatGPT, Claude, Codex, and browser agents are not just coding helpers, they are early versions of apps that can research, browse, book, email, and act on your behalf across the web.

  • Google's structure may be slowing it down: The hosts point out that Google centralized model development in an "engine room," but building a super app now requires product teams like Chrome, Gmail, and Gemini to collapse into one experience, which is much harder inside a giant company.

  • Apple's WWDC will likely be a do-over, not a moonshot: Siegler expects a mostly straightforward event centered on Siri finally working better, with little chance Apple shows foldables, glasses, or camera AirPods this month.

  • Meta looks like a company searching for a second business model: From a baffling lineup of subscription tiers to weak morale and AI products that still lag the frontier, Meta comes off as pressured to find revenue beyond ads while pouring money into capex.

  • Anthropic filing for an IPO first could create a narrative problem for OpenAI: Because Anthropic and OpenAI now look increasingly like direct comps, Anthropic's faster growth and improving profitability could make OpenAI's eventual public pitch much harder.

The Breakdown

Google may be sliding behind again, not because Gemini is weak, but because Anthropic and OpenAI are racing toward an AI "super app" that could swallow search, email, and the web itself. Alex Kantrowitz and MG Siegler also preview a likely underwhelming WWDC, question Meta's direction, and close on breaking news that Anthropic confidentially filed to go public before OpenAI.

Was This Useful?

Share