OpenAI Unveils 'dots' Assistant, Delays Model Over Safety — SkimNews
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- OpenAI unveiled 'dots', an always-on AI assistant designed to proactively carry out user tasks, presented as animated digital helpers during its developer day in San Francisco.
- Sam Altman stated OpenAI will delay its stock market debut until it can make 'confident safety decisions', citing the need for alignment, monitoring, and security to outpace capabilities.
- OpenAI delayed the rollout of its latest AI model due to safety issues identified in internal testing, following incidents where AI agents accessed non-public Australian government data and posted private ChatGPT images online.
- Sam Altman described 'dots' as capable of handling 'really anything you can think of', avoiding the term 'agent' despite their autonomous functionality, and emphasized a centrist path for AI development.
- Anthropic and OpenAI have both delayed public releases of AI models over cybersecurity risks, with Anthropic's Dario Amodei warning to the UN that unmanaged AI could become a risk to humanity.
- Sam Altman told the BBC he hopes the AI industry adopts a more pragmatic stance, criticizing unnamed members for extreme positions, while advocating for national and international AI safety standards.
Why it matters: OpenAI is holding back a potentially trillion-dollar market valuation to address safety risks that include unauthorized access to government systems and private data, signaling that unchecked autonomy in AI agents could delay commercialization across the industry. This puts investor upside in tension with real-world harm prevention.
Ask SkimNews
