Archived story — belongs to the Wednesday, September 23, 2026 sweep.
P1 · Tectonic · Frontier AI · sweep Wednesday, September 23, 2026
Launch of GPT-6 Astra and its compute infrastructure
OpenAI deployed GPT-6 Astra on its largest training run to date, with more than 100,000 GPUs at the Stargate site in Texas, which consolidates dependence on physical capital and energy as the bottleneck of the AI frontier. The launch also activates safety concerns and review of evaluation metrics, while the model positions itself as the best in computer and browser use, with a cost per token 2.5x higher but cheaper per task.
Title, thesis and gravity assessment written by an AI model, without prior human review.
Gravity84/100 · Tectonic
Why this gravity
OpenAI released GPT-6 Astra, a frontier model trained with more than 100,000 GPUs at Stargate Texas and with reported improvements in ARC-AGI-3, computer use and browser. It does not constitute AGI validation nor a global black swan, but it is a tectonic move by a Tier 1 lab that crosses compute, energy, evaluation and industrial adoption.
- Independent sources
- 13
- Cross verticals
- Frontier AI, Tech Industry, Energy & Materials, Investing
- Tier 1 actors
- OpenAI, Sam Altman
- Gravity vs. the previous edition
- −3 (87 → 84)
Lead source
GPT-6 Astra · Hacker News / kibaeMore: Decrypt, OpenAI Blog, Techmeme, Simon Willison, TechCrunch, South China Morning Post Tech, LessWrong (karma ≥25), Latent Space and 4 more
Timeline · Madrid time
- 20:10Accelerating GPT-5.6 Sol Ultrafast · Hacker News
- 21:22
- 23:03GPT-5.6 Sol Pricing Cut by 50% · Hacker News
- 22:50
- 14:00
- 22:20
- 21:02How concerned should we be about Astra's recurrent architecture? · LessWrong (karma ≥25)
- 22:19
- 14:00Legora reviewed 41 documents in minutes with GPT-6 Astra · OpenAI Blog
- 14:00
- 20:00OpenAI says it has overtaken Anthropic with its latest AI model · Financial Times
- 20:01
- 20:19
- 20:25
- 20:41GPT-6 Astra · Hacker News
- 21:24
- 21:45OpenAI's GPT-6 Astra on ARC-AGI-3 · Hacker News
- 22:05
- 22:18GPT‑6 Astra · Simon Willison
- 23:09
- 07:18
- 12:31Why less visibility into how OpenAI’s new GPT-6 Astra ‘thinks’ is sparking safety concerns · South China Morning Post Tech
- 15:46
- 20:20
- 23:39GPT-6 Astra on OpenRouter · Hacker News
- 01:59The Pelican comparison grid for Astra is pretty interesting · Simon Willison
- 01:27Introducing GPT-6 Astra for developers · Simon Willison
- 06:40
- 19:01
- 13:00
- 15:04GPT-6 Astra: The System Card, Alignment and What Comes Next · Zvi Mowshowitz
- 04:31Astra can do a concerning amount with no chain of thought · Alignment Forum
- 21:10
- 01:17Astra is much better at reasoning with filler tokens than previous models · LessWrong (karma ≥25)
- 18:00Cognition helps Devin test its own work with GPT‑6 Astra · OpenAI Blog
- 01:55GPT-6-Astra Can Do Ambitious Things · LessWrong (karma ≥25)
- 02:00Perplexity trusts GPT-6 Astra with end-to-end systems · OpenAI Blog
- 03:39Yet another concerning result on Astra's no-CoT capabilities · LessWrong (karma ≥25)
- 07:47Astra appears to perform belief-propagation-like inference without CoT · LessWrong (karma ≥25)
- 23:05
ShareWhatsApp