one year on
Google is winning on every AI front, Romero says.
The essay's author argues that Google DeepMind's hardware, distribution, and model advantages give Google DeepMind a strong edge over OpenAI, Anthropic, and others.
In a widely-discussed essay published this week, the essay’s author, Alberto Romero, declares that Google is now decisively winning the AI race — and that neither OpenAI nor Anthropic have a chance at this point. The piece, titled “Google Is Winning on Every AI Front,” has 1005 points and 819 comments on Hacker News. Commenters argue that Google’s hardware, distribution, and model advantages give it a major edge. On Hacker News, commenters echo this view, with one noting that “Google’s ability to tailor their mature hardware to exactly what they need gives them a massive leg up on competition.”
Not everyone is convinced. Some critics question whether the TPU investment has yielded a durable edge. The debate is playing out on Hacker News, where commenters are split on whether the TPU investment was worth it.
The record
On the HN thread, users highlight Google's TPU and JAX ecosystem as a key structural advantage, with one commenter calling it a '10x, if not 20x' multiplier on compute; others debate whether the TPU investment has been worthwhile.
One year later — open only if you can handle spoilers
By mid-2026, Google's lead had only widened, with Gemini models maintaining a consistent edge in evaluations and cloud revenue surging. The TPU bet was largely vindicated, though antitrust scrutiny of Google's AI ecosystem began to intensify in the US and EU.
The Weekly Replay · free by email
This week, one year ago — every Sunday.
One email each Sunday: the week's replayed AI news, with the one-year-later annotations included. Written like it's breaking — dated like it isn't.
Free · double opt-in · unsubscribe anytime · privacy