A sarcastic thread turns OpenAI’s Gemini roast into a benchmark-rankings joke

ctjlewis · x · 2026-07-22

A sarcastic thread mocks OpenAI for roasting the announcement of Gemini pre-training, reading it as a sign of weak confidence.

It jokes that dunking on a team “in 5th place on the benchmarks” suggests certainty, while Anthropic is framed as too strong to mock and Chinese models are treated as a political scapegoat in the joke. The punchline is that OpenAI’s comms strategy says more about its current standing than about Gemini.

Original post →

More from Fun

Fun channel →