this post was submitted on 16 Jan 2025
36 points (78.1% liked)
Futurology
1891 readers
122 users here now
founded 2 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
There's so much to legitimately worry about with AI, that we often lose sight of its potential good.
The moment you become truly aware of how often ai is incorrect is the moment you are ready to use it properly.
I do hope we have started a long study for the short and long term influence of llms on critical thinking skills.
AI juste needs grounding. If you ask gpt4 to look online before answering you (giving it a database of reliable info), it will always be correct. With this technique AI can be a great teacher.
gpt4 is not capable of searching online. It's trained on a fixed database with no internet access.
OpenAI gpt4 can't?!? I am fearly certainly it can, considering I use this feature daily.
I'm pretty worried about the massive acceleration of our environmental destruction of the planet that it causes, so I hope they turn it to getting us the fuck off this rock pretty soon then.
https://www.nature.com/articles/d41586-024-00478-x
Capitalism will fix this (bizarrely). Making AI more efficent (read more profitable) is on LOTS of peoples minds right now.
Whether it's more efficent chips, better algorithms or whatever, rest assured, LOTS of effort is going into it.
Our environment will still be destroyed, but it won't be AI that does it, just boring greed
I won't hold my breath.
I'm worried about ideological capture. Since teachers are individuals, it's fairly hard to uniformly push extreme views - there's always gonna be some resistance.
That no longer exists with AI. Whoever produces it has full control over what it says, knows and "thinks".