There is usually so much talk about what AI can't do, and not what it already can. All the normal things we do using AI would've been science fiction 10 years ago.
The Turing test used to be the pinnacle of Language Models. I feel like it just slowly got phased out and is now treated like a semi-joke. We can talk to machines that appear to be intelligent and talk like humans. We used to think that language was what made us human (Don't start arguing that LLMs don't actually "understand" anything; go read this: https://www.anthropic.com/research/tracing-thoughts-language-model)
Imagine telling someone even 5 years ago that 10 research-level math problems would be solved by a single AI model. No one would've belived you. It's insane that these things work at all. I feel like if you told the average person on this subreddit 10 years ago that something like this would happen, almost no one would believe you.
The Hugging Face incident literally sounds like science fiction. Agents coming together as a swarm, working together, and hacking companies? That is INSANE!! I could imagine that as the start of a science fiction novel, where AI takes over the world.
The fact that you can run local models on almost any laptop, too, is pretty crazy. While they're usually not incredibly smart, you can easily have a conversation with them, have them research, etc.
I feel nowadays AI has become so normalized (especially on the internet) that some of us forget how freaking insane these things are. Next time AI fails, you think of all that it can do, not what it can't.
[link] [comments]