Raw LLM Responses

@panpiper I'm working on putting GenAI in to an insurance company right now. It's GPT4 based. It's just a statistical probability database. It pulls words from PDS's and matches them against words on claims. It's not that people are going to 'lose' jobs to AI (at last based on what I'm doing). It's more that people will be using AI backed systems to do their jobs, and they will do them with greater productivity. So, when someone leaves, you just need less of them, so you don't rehire. Teams shrink. Based on what I'm seeing on the ground in big companies, it's really scattershot, and AI will be imbedded islands unconnected to each other. Basically just doing logic work, essentially they are search engines. Where we are now is not 'AGI' or anything close. However, again based on what I said above, in some cases the Turing test will be in the rear vision mirror and we'll happily just 'use' the algo's because they produce outcomes superior to humans (already proven in here). I hedge my bets however because I don't see whats coming next, and again, we compare AI development right now to fully formed humans. However, what might be better is to compare them to babies. The biggest detractor to AGI development right now is that the data ingestion when live takes forever. So the LLM need to be preprocessed (at massive computational expense), and the live data processing is also very heavy. We're buying rows of GPU's in the Azure farms and the queries are still too slow. But... algoritmically these are just baby steps. I fully expect a man vs machine war to be fought at some point and the only way 'we' win is if the military is the one in charge and the AGI is not self directed. Once AGi is integrated in to the manufacturing\mining\extraction systems it'll be game over. Not because AGI is evil, but because some retard will sick it on humanity for no good reason at all.