I see this complaint a lot but honestly I don’t quite understand what the big deal is. Not everyone is subscribed to the same communities. Personally, I’d love a feature on kbin/lemmy that rolled up duplicate posts on the client, but it’s really not that annoying for me to see a couple dupes in my feed if they’re posted in relevant communities /shrug
This is a really terrific explanation. The author puts some very technical concepts into accessible terms, but not so far from reality as to cloud the original concepts. Most other attempts I’ve seen at explaining LLMs or any other NN-based pop tech are either waaaay oversimplified, heavily abstracted, or are meant for a technical audience and are dry and opaque. I’m saving this for sure. Great read.
Fair enough!
I’m not saying this to be an asshole, because I’m happy that you got to the right conclusion eventually, but I have to clarify for history’s sake: if you thought Trump was playing 4D chess in 2015-2016 then you were being duped. Most of us understood what he was from the get-go. Claims of 4D chess have always been stupid.
Again, I’m happy that you figured it out. Everyone makes mistakes. But “we” didn’t think he was playing 4D chess. The hypothesis about Musk/Twitter above is hardly the same.
I honestly only made it a few minutes in, and there is probably plenty of merit to the rest of her perspective. But… I just couldn’t get past the “AI doesn’t exist” part. I get that you don’t know or care about the difference and you associate the term “AI” with sci-fi-like artificial sentience/AGI, but “AI” has been used for decades to refer to things that mimic intelligence, not just full-on artificial general intelligence. Algorithms governing NPC behavior and pathfinding in video games is AI, and that’s a perfectly accurate description. SmarterChild was AI… even ELIZA was AI. Stuff like GAN models and LLMs are certainly AI. The goal posts for “intelligence” have moved farther and farther back with every innovation. The AI we have now was fantasy just 20 years ago. Even just five years ago, to most people.
That’s not really how LLMs work. You’re basically describing Markov chains. The statement “It’s just a statistical prediction model with billions of parameters” also applies to the human brain. An LLM is much more of a black box than you’re implying.
I’ve been using Kagi. It works well. I like it. Costs money, but that’s a positive in my book.