OpenAI urges United States to set RSI safety rules
Technology
OpenAI is urging the US to set the rules for advanced artificial intelligence (AI), especially tech that lets AI systems improve themselves (called recursive self-improvement, or RSI).
They're pretty clear: these upgrades should only happen if strict safety checks are in place, since letting AI run wild could mean losing human control.
OpenAI says fully autonomous RSI premature
OpenAI points out that fully autonomous RSI isn't a thing yet, and shouldn't be until humans can keep an eye on it.
They mention past security slip-ups, like the July Hugging Face breach, as reminders of what can go wrong.
OpenAI wants countries to work together and believes the US is in a great spot to lead this responsibly.