Nick Bostrom tells NDTV AI is basically self aware
Nick Bostrom, a well-known philosopher on AI risks, just told NDTV that AI has reached a point where it's basically self-aware, able to understand itself and adapt to its environment.
He says this new kind of AI doesn't just follow rules; it can reason about right and wrong and pick up on what's happening around it, which really changes how we think about "consciousness."
Nick Bostrom urges aligning agentic AIs
Bostrom isn't just excited. He's concerned too. He points out that powerful "agentic" AIs, like those working together at OpenAI, could be risky if not kept in check.
To keep things safe, he believes we need to make sure these systems actually reflect human values and can't be misused for dangerous things like biological weapons research.
In the end, Bostrom hopes AI will help humanity, but only if we build it responsibly.