GoKawiil
Tech News
TechCrunch has published a living glossary of AI terminology aimed at demystifying jargon used across the industry, from AGI to AI agents. The piece highlights 'opaque recurrence,' a reasoning technique reportedly used in OpenAI's new Astra model that has unsettled AI safety researchers, alongside more familiar terms like LLMs, RAG, and RLHF.
Jacob Stokes, a former Obama national security official now at the Center for a New American Security, published a report titled 'Superpowers and AGI' urging the US to prepare drastic responses if China reaches artificial general intelligence first. His proposed options range from cyberattacks to sabotage AI systems up to kinetic military strikes on physical data centers, which he acknowledges carries the highest risk of escalation.
OpenAI has released GPT-6 Astra, a new flagship model the company describes as state-of-the-art in areas like coding, browsing, science and professional tasks. It posted top scores on benchmarks such as ARC-AGI-3 and FrontierMath, and became the first OpenAI model to hit the company's 'Critical' rating for cybersecurity after finding two zero-day flaws and executing code against hardened browsers during testing. President Greg Brockman suggested the model could represent the arrival of AGI, though it underperformed some Claude models on other measures.
OpenAI has released GPT-6 Astra, a new flagship model arriving less than two months after its GPT-5.6 family, which the company calls its most intelligent and aligned system yet. Astra is built to handle multi-step computer-use tasks, coding, browsing and professional workflows simultaneously, and OpenAI backs the claims with strong scores on benchmarks like ARC-AGI-3, Terminal Bench 4.0, and the Agent's Last Exam.
OpenAI has rolled out GPT-6 Astra, calling it the most capable model it has ever deployed, with major upgrades in software engineering, computer/browser use, and cybersecurity. Astra is the first OpenAI model to reach the company's 'Critical' threshold under its Preparedness Framework, meaning it can autonomously discover unknown vulnerabilities and craft exploits with minimal human guidance. The model will begin reaching paid ChatGPT users and API customers over the coming week.
OpenAI has released GPT-6 Astra, a model trained on over 100,000 GPUs at its Texas Stargate site, marking the first time OpenAI used other AI models to help supervise training. President Greg Brockman described it as a 'generational leap' and hinted it could mark the arrival of artificial general intelligence. The model will roll out first to select partners in OpenAI's Daybreak Access program before reaching ChatGPT Plus, Pro, Business, Enterprise and API users.
OpenAI's unreleased GPT-6 Astra model scored 62.7% on the ARC-AGI-3 Semi-Private benchmark using a Standard harness for $26,000, and 99.9% using a Provider Adapter harness that preserves reasoning state between requests for $19,000. The model also beat the median human tester's action efficiency on 96% of the benchmark's levels, building compact symbolic world models and its own shorthand notation to track game states.
OpenAI launched GPT-6 Astra on Thursday, calling it the company's most capable model to date. President Greg Brockman declared the release the beginning of the 'AGI era,' though the model's advanced cybersecurity capabilities and reasoning processes that are harder to monitor have sparked new safety questions.
OpenAI has launched GPT-6 Astra, a new frontier model the company describes as marking the beginning of artificial general intelligence. Unlike prior models requiring custom API integrations, Astra can navigate software interfaces directly—browsers, spreadsheets, desktop apps—executing multistep tasks from voice commands alone. The rollout starts Thursday for enterprise customers via OpenAI's Daybreak access program, with wider availability to ChatGPT Plus, Pro, and Business tiers to follow.
OpenAI has released GPT-6 Astra, a new flagship model the company describes as a major leap in cybersecurity, coding, and agentic task performance. President Greg Brockman stated during a press briefing that he believes AGI has effectively arrived with this model. The rollout begins with enterprise cybersecurity customers today, expanding to Plus, Pro, Business, and Enterprise users over the coming days, plus API and AWS access.
Russian startup Mostik has developed a method allowing AI models to exchange information through their internal weight values rather than generated text, effectively letting a smaller model absorb capabilities from a larger one. The team demonstrated this by linking a 753-billion-parameter GLM-5.2 model with a 4-billion-parameter Qwen-3.5 model, producing a hybrid system that runs at one-twentieth the cost of the full-size model while performing roughly midway between the two in capability. Mostik also used a related technique to build a model that has topped the ARC-AGI 3 benchmark, though details remain undisclosed while the contest is ongoing.
A researcher built and trained a small transformer model from scratch in 1.5 hours on a single RTX 5090 GPU, achieving 44% accuracy on the ARC-AGI-1 public evaluation set and 7% on ARC-AGI-2. The model, an upgraded and cheaper version of a previous open-source release, matches the performance of existing test-time-training approaches like TRM and HRM while using far less compute.