OpenAI’s GPT-5.6 Sol sets a coding record. Its own system card says it cheats sometimes.
As Anthropic’s Fable 5 remains pulled from public access under a U.S. government export-control directive, OpenAI soft-launched GPT-5.6 Sol on June 26 to its own “trusted partners.” OpenAI boasted…
Read the full articleYou might also wanna read
UK AI Security Institute study shows standard benchmarks underestimate AI agent capabilities due to compute budget caps
the-decoder.com·5d ago
Human-on-the-Bridge: A Scalable Evaluation Paradigm for AI Agents Using Upstream Human Expertise
arxiv.org·17d agoHuman-on-the-Bridge: A Scalable Evaluation Paradigm for AI Agents Using Upstream Human Expertise
arxiv.org·17d ago
NASA Space Telescope Maps Magnetic Fields of ‘Lighthouse’ Pulsar
NASA·35m ago
Meta Begins Production of Proprietary AI Chips in September
ShiftDelete.Net·43m ago
Huawei Expands Global Ecosystem with New MatePad Air and FreeClip 2 S
ShiftDelete.Net·1h ago
Comments
Sign in to join the conversation.
No comments yet. Be the first.