Hi Gregory > I haven’t settled on a subject for my dissertation yet. One I am > considering is the nature of consciousness. The idea is that > consciousness could be an emergent property in complex systems. As we > have observed with LLMs many of the things it doesn are unexpected and > not specifically programmed. Or possibly to study the larger impact of > AI on software development. That last one is more to the point but a > bit more pedestrian than the first.
Wow. This is still so broad. I think this depends a lot on the input data and the prompt. I think actually if you wanted to implement something, then you could do way better job than Copilot does. There are many non-privacy invasive, not energy consuming things that an AI system could do with a desktop. > Stage 2, bad but recoverable with the right treatment. All the best with this. Sounded like time to step back and reduce the workload. I hope the government that is so notorious for expensive medical care can at least somewhat support you in such a situation. > A 2-3 day workshop would be a lot of fun. This could be setup as in collaboration with FSF, they probably can disseminate link to an online workshop and programme to their user lists. Someone would need to come up with a system to compile a programme with a list of talks or workshop session themes for that. Regards, Gryllida
