OpenAI has launched GPT-6 Astra, a new AI model capable of operating computers and completing full workflows independently, marking a significant shift from previous AI tools. This model reverses the user-AI relationship, with users now setting goals for Astra to execute, rather than directly interacting with it for answers. OpenAI's safety team reportedly flagged GPT-6 Astra as a critical cyber risk due to its advanced capabilities, with the most dangerous functions being restricted. AI
IMPACT This model's ability to autonomously operate computers and complete workflows could fundamentally change job descriptions and the nature of work.
RANK_REASON Frontier-lab model release with system card. [lever_c_demoted from frontier_release: ic=1 ai=1.0]
- AGI
- Amazon Q
- GPT-4
- GPT-6 Astra
- OpenAI
- Safety Teams: An Approach To Engage Students in Laboratory Safety
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →