@mysticaltech@wesroth@gdb Basically LLMs are far worse than the human brain. It takes 500,000 equivalent years of learning to still be dumber than a human.
AGI might be coming, but LLMs are not it.
@David_Moscrop Did you listen to his interview with Althia Raj? It was a trainwreck when she asked about policy. I think the wheels will come off his campaign when people see him on the debate stage or in longer interviews.
@JustinLewis1977@NateSilver538 In my experience, AI is not that good for coding. It's a time saver for simple work, but it wastes time if you try to use it for harder issues.
@NateSilver538 I'm glad that AI is making coding more accessible to amateurs, but errors like this would be easily caught by an IDE.
I find AI to be bad at debugging even moderately difficult issues. It's wasted far more of my time than it's saved.
@GeorgeP98181161@JasonKPargin The overpopulation problem. People were extrapolating human population growth and figuring we wouldn't be able to feed everyone. Then it turned out fertility drops as living standards rise and now they're expecting the world population to level out or even start to drop.
@mattgurney That's surprising to me. It makes sense that a language model sucks at math. But you would think its bread and butter world be searching for information in a text file.
@ayesitisme21@TWilsonOttawa If Dems were able to steal the 2020 election, while they were not in control of the executive, how come they weren't able to steal 2016 or 2024, when they were in control?
@kateg_tweets@Chadwick100@WorldDarkWeb By repeatedly crossing the double line and stopping in the middle of the road??? The safer thing would be to let the maniac pass and leave the policing to the police.