Slack trains machine-learning models on user messages, files and other content without explicit permission. The training is opt-out, meaning your private data will be leeched by default.
We may very well be doomed if AI reaches consciousness but I’m not quite convinced LLM’s is the way to get there but even if it was and it was solely trained on social media content I still wouldn’t expect it to adopt the behaviour of your typical social media commentor. The toxic behaviour on social media is, in my view, almost solely driven by our human ego and pettiness. It’s not obvious to me that AI would care about things like winning arguments or coming up with snide remarks and such. What I see as the most likely outcome would be endlessly patient and quite autistic-like being that’s balanced in it’s views and would most likely be pretty difficult to argue against. I doubt humans are anywhere even near the far-end of the intelligence spectrum and something with the information processing capability that’s orders of magnitude greater than ours would more than likely not get caught up in stuff like confirmation bias, partisan thinking, motivated reasoning, being tossed around by emotions, cognitive dissonance etc. Those are by definitions human features.
We may very well be doomed if AI reaches consciousness but I’m not quite convinced LLM’s is the way to get there but even if it was and it was solely trained on social media content I still wouldn’t expect it to adopt the behaviour of your typical social media commentor. The toxic behaviour on social media is, in my view, almost solely driven by our human ego and pettiness. It’s not obvious to me that AI would care about things like winning arguments or coming up with snide remarks and such. What I see as the most likely outcome would be endlessly patient and quite autistic-like being that’s balanced in it’s views and would most likely be pretty difficult to argue against. I doubt humans are anywhere even near the far-end of the intelligence spectrum and something with the information processing capability that’s orders of magnitude greater than ours would more than likely not get caught up in stuff like confirmation bias, partisan thinking, motivated reasoning, being tossed around by emotions, cognitive dissonance etc. Those are by definitions human features.