AnnouncementsVideosCensorshipReligionFunnyConspiracyAwwwAnarchoCapitalismCryptoThemesIdeas4MatrixAskMatrixHelpTop Subs
2
AskMatrix - - (+1|-0)
Which company do your trust/distrust more Facebook/Meta or OpenAI/Microsoft(self)
3 commentsbyx0x7
I ask because I did a coding test between chatGPT and meta.ai and honestly meta.ai came out by a hair even though I'm paying chatGPT. That means I can cancel my premium and stop giving them money.
The question though is which one is more dangerous or worse to hand AI dominance to? Or which one is more likely to take AI dominance because I'd rather have two bad actors played off of each other than one.
Comment preview
[-]Winston Smith0(+0|0)
I think AI is a horrible idea and humans need to write better code and algorithms. Too many variables in the real world for AI to make any kind of reliable predictive analysis or decision-making. Let alone any kind of ethical judgment as AIs are already infused with too much PC bullshit.
parentlinkreply
[-]x0x70(+0|0)
I think human evaluation is always going to need to be a thing. Even if you make some multi-agent AI system that can get further along a problem that something else then that is its own closed system and its work is going to have to be human evaluated.
Same as any other algorithm. That's all AI is. One more algorithm. But if you ran an aircraft model into some optimizer back in the 90s or even today humans have to test and sanity check it, and that's 80% of the work in any case. So the question really is if AI can improve quality to cost on the 20% that is design and prototyping.
If AI does replace humans I really think we should blame the humans chomping at the bit to do that more than AI itself. AI will ultimately do whatever we put the most effort into making it do.
parentlinkreply
[-]Winston Smith0(+0|0)
It has its applications but we shouldn't be too reliant on it.
parentlinkreply