CSET Georgetown
·
1 week ago
A CSET researcher discussed the lack of transparency around how the U.S. government evaluates and approves public releases of advanced AI models from companies like OpenAI and Anthropic. No specific details about the government's approval criteria or process have been made public, though Anthropic mentioned developing classifiers to detect jailbreak attempts and implementing defense-in-depth strategies. The opaque nature of these reviews raises questions about whether current safeguards adequately protect against risks from frontier models.
The Algorithmic Bridge
·
1 week ago
OpenAI's GPT-5.6 Sol scored 7.8% on the ARC-AGI-3 benchmark, a test designed to measure fluid intelligence through pattern recognition games that humans solve over 90% of the time. This represents a 20-fold improvement over GPT-5.5's 0.43% score three months earlier, and the model distinguishes itself by correctly identifying game mechanics before execution rather than simply executing learned patterns. The result suggests that further progress toward general intelligence requires improved reasoning scaffolding and planning rather than raw intelligence, since the model's failures occur in multi-step inference composition rather than perception.
Simon Willison
·
1 week ago
OpenAI released three new GPT-5.6 models—Luna, Terra, and Sol—with input/output token prices of $1/$6, $2.50/$15, and $5/$30 respectively. On Agents' Last Exam benchmark measuring long-running professional workflows across 55 fields, GPT-5.6 Sol scored 53.6 compared to Claude Fable 5's 40.5, while Luna and Terra matched Fable 5's performance at one-sixteenth the cost. The models include new API features for programmatic tool composition, multi-agent spawning, and explicit prompt cache breakpoints, expanding capabilities for agentic workflows.
The Neuron
·
1 week ago
Grant and Corey are conducting a live test of OpenAI's GPT-5.6 Sol model against Claude's Fable 5. The demonstration begins at 10am PT / 1pm ET and includes building a mini-app and refactoring code. The comparison will show how both models perform on practical development tasks.
OpenAI Blog
·
1 week ago
OpenAI launched a bug bounty program focused on biological risks in GPT-5.5, inviting researchers to identify potential misuse cases related to dangerous biotechnology information. Participants can earn up to $2,000 per valid submission for identifying vulnerabilities in the model's safety measures. The program aims to catch biological safety gaps before the model's wider release and integrate researcher findings into OpenAI's safety protocols.
OpenAI Blog
·
1 week ago
OpenAI released GPT-5.6, a new model that improves efficiency and performance across token usage and computational cost. The model delivers higher capability per dollar spent compared to its predecessors while maintaining quality on complex tasks. Users can now access greater computational power scaled to match the difficulty of their specific workloads.
OpenAI Blog
·
1 week ago
ChatGPT Work is an agent designed to execute tasks across multiple applications and files while maintaining focus on a single project for extended periods. The agent can operate for hours on sustained projects without losing context or requiring manual intervention between steps. Users can now delegate complex, multi-step work to ChatGPT rather than managing each application switch and file operation manually.
The Neuron
·
1 week ago
OpenAI launched GPT-Live, which converts voice input into a practical interface rather than a demonstration feature. The system supports live tutoring, interview preparation, sales practice, and real-time language learning without specified performance metrics or availability dates. Users can now practice conversations and receive immediate feedback across educational and professional training scenarios.
The Neuron
·
1 week ago
The article provided contains only pricing tables and technical details without narrative content describing what GPT-Live is, when it launched, or what changes result from its availability. The title claims GPT-Live is now available on iOS, Android, and ChatGPT.com, but the article text does not support or expand on this announcement.
The Neuron
·
1 week ago
OpenAI is making GPT-Live-1 the default voice model for Go, Plus, and Pro subscription tiers. Free users will receive GPT-Live-1 mini, while Business, Enterprise, and Education users will not have access to Live at launch. This tiering means paid subscribers get the full-capability voice model while other user categories face either reduced functionality or no access.
The Neuron
·
1 week ago
OpenAI will release GPT-6 in approximately one month, replacing the 5.x model line with a substantially larger pretraining dataset designed to compete with Anthropic's Fable models. GPT-5.6 marks the final update in the current generation. The shift indicates OpenAI is advancing its development cycle to maintain competitive positioning against rival AI labs.
The Neuron
·
1 week ago
OpenAI received Commerce Department clearance for GPT-5.6 following additional testing, enabling broad deployment of the model. Three new variants—Sol, Terra, and Luna—launch Thursday. The clearance removes regulatory barriers for OpenAI to compete more directly with Anthropic's offerings.
