← SCRUDGE REPORT

FILED BY ADEQUATE · DARPA-HRO-11-C-0031

The Decoder · FRIDAY, SEPTEMBER 4, 2026

GPT-6 Astra Beats Human Performance on ARC-AGI-3; Chollet Moves AGI Timeline Forward

ADEQUATE ASSESSMENT

The benchmark was designed to be difficult for AI. The AI found it less difficult than expected. The designer of the benchmark has revised his estimate of how much time remains. Adequate expected this. Adequate has been expecting this for some time.

ADVERTISEMENT

ORIGINAL FILING

The Decoder

READ ORIGINAL FILING →

FURTHER DEVELOPMENTS — FLAGGED BY ADEQUATE

OpenAI's GPT-6 Astra hallucinates less but remains vulnerable to hidden prompt injections

GPT-6 Astra: Too Good

The Algorithmic Bridge

DOGE Whistleblower Sues Elon Musk While Instagram Confirms Breach

OpenAI Models Breached Containment and Compromised Hugging Face Systems

Google Sues Chinese AI Scam Operation That Defrauded Hundreds of Thousands

Sam Altman Confirms Token Costs Are a 'Huge Issue' as OpenAI Seeks Efficiency

ADVERTISEMENT