Modlog

CONTENT WARNING: Some deleted posts may contain disturbing or adult material. Proceed with caution.

/c/morewrite Modlog

Filter by user

Filter by mod

Time	mod	Action
2 years ago	mod	Removed Comment Well lol I'm sure my ad-hoc suggestion for an experiment could be improved and a better definition could be found without affecting the presumed outcome. My point is that there is a lot of anger and bias and fallacies surrounding this. by LarmyOfLone ( LarmyOfLone@lemm.ee ) reason: even more boring wrong nonsense!
2 years ago	mod	Removed Comment Yeah an inner mental process that has been replicated - well to a (very) limited degree. My argument is that this is exactly what is so offensive to people. The painful realization or disillusionment that one of the things we held most special about being human, turns out is not so special after all. I never would have predicted that before GPT-3. I look at movies or stories with a character that is an artistic archetype differently now. Meat good! Silicon bad! by LarmyOfLone ( LarmyOfLone@lemm.ee ) reason: more boring nonsense
2 years ago	mod	Removed Comment Same, I'm not quite sure what the argument is. Also, it seems that a lot of the AI hate comes from how capitalism uses it and hypes it. That is not the fault of the poor innocent AI! I also don't see how you could plausibly say that AI isn't creative at all. If you'd take the least creative humans on earth and compare them to LLMs in some sort of designed experiment, I'm pretty sure the LLMs would win against some percentage of humanity. In the coming decades this percentage of "more creative than X percent of humans" will increase. THAT is a deep insult to our collective human psyche. THAT is where I suspect a lot of AI hate comes from. A lot of the arguments against AI hype are just rationalizations that are ultimately the fault of capitalism or greedy stupid people (e.g. tech bros). It also seems to me that LLMs represent creativity without general intelligence, and that fundamentally limits them. They can't prompt themselves or understand what they created. Again, for now. What this shown is that if you throw enough data and computing power at the problem, it solves it just like nature did with natural selection. For now it's only like one small portion of the human mind. Cobble together enough pieces we might be closer to AGI than we'd like. Some seem to view this thought as some sort of heresy. It would be more important to argue for more ethical control of AI outside of profit motives and dogmatic views on copyright. We should develop "artificial ethics" first before we try to extract profit from the advances. The strong emotional reactions will only be exploited so the capitalists will get to control for the worst possible outcomes. by LarmyOfLone ( LarmyOfLone@lemm.ee ) reason: boring nonsense