OpenAI's aggressive compute bet may explain why stronger models are held back

haider1 · x · 2026-08-17

haider1 argues OpenAI has invested far more aggressively in compute, which may explain why stronger models haven't shipped yet — the models are ready, waiting for serving infrastructure to catch up. He speculates that if Astra launches this month at 5–10T parameters, OpenAI could still serve it cheaper and better than Anthropic serves Mythos 5. Cited article: Jensen Huang on land, power and shell as critical resources for AI factories.

Original post →

More from AGI Musings

AGI Musings channel →