AI-pilled execs are depressing. I'm not looking forward to my role transitioning to a reverse centaur and moral crumple zone.
Coworkers are going to be introduced to some new terminology fairly soon. They're pretty comfortable with convoluted acronyms but I think they're going to have a difficult time digesting TESCREAL.
This is the stupidest timeline.
We can do better than this.
Or maybe we can't. The people who could are being stopped by the people who won't.
@arclight isn't nuclear too dangerous to fuck with?
@hananc It's far safer than it's made out to be but there's a reason for that - br regulation and self-selection there is a very strong emphasis on safety. We have seen repeatedly what happens when that focus is lost. We have a very tenuous social license so it has to be this way.
Chatbots are immature commercial-grade technology with low repeatability and low accuracy and effectively zero traceability. They are nowhere near capable of performing safety-grade work and they are unlikely to by design - they say what they were trained to say and don't provide sources for their data. That's a key feature of chatbots - they can't provide sources because they are specifically designed to launder attribution. The whole chatbot industry is based on theft, the technology requires it. That's absolutely counter to engineering where sources are cited and verified. There's no place for chatbots in safety related work, nuclear or otherwise.
@arclight it's good to know that some people with some sense are still around.
I remain unconvinced that AI has a role in designing and fabricating nuclear fuel.