:
[AI]■ STORY TIMELINE

GPT-6 ASTRA BEATS HUMANS ON ARC-AGI-3, PULLING AGI TIMELINE FORWARD

OpenAI's GPT-6 Astra achieved human-level efficiency on the ARC-AGI-3 benchmark for the first time, prompting ARC Prize chief François Chollet to accelerate his AGI forecast. However, benchmark disagreement clouds the broader picture of the model's capabilities.

9 SOURCESFIRST SEEN SEP 2, 07:45 PM► READ THE ARTICLE
Techmeme+0m

Mark Zuckerberg / @finkd: Mark Zuckerberg says Meta's Watermelon model and Muse Spark open weights are “coming soon”  —…

Techmeme+1h 25m

@artificialanlys: Muse Spark 1.3 with max reasoning, in limited preview for partners, scores 62 on the Artificial Analys…

Techmeme+22h 30m

The Deep View: OpenAI prices GPT-6 Astra at $10/1M input tokens and $50/1M output tokens, matching Anthropic's pricing f…

The Guardian — Technology+1d 14h

The CEO is the throughline in a history of billion-dollar blunders that have caused real harm, culminating in last week’…

The Verge+1d 14h

Just hours after OpenAI launched GPT-6 Astra, CEO Sam Altman was already apologizing for what he describes as a "messy r…

The Decoder+1d 15h

OpenAI's GPT-6 Astra is drawing contradictory benchmark verdicts. Epoch AI puts it out in front with 169 points, while A…

Techmeme+1d 17h

Artificial Analysis: Artificial Analysis Coding Agent Index: GPT-6 Astra scored 67, roughly equal to Claude Opus 5, Fabl…

Techmeme+1d 19h

M.G. Siegler / Spyglass: By declaring GPT-6 Astra to be AGI, OpenAI is being flippant and cementing the term's status as…

Techmeme+1d 19h

Matt Shumer / Something Big Is Happening: Review: GPT-6 Astra can adeptly use tools like Unreal Engine to build complex…