"easy" is joining the lemmings who are all falling over themselves to declare the Chatbots the future because they make human shaped blobs of text Meanwhile it's making every product I use worse and the "uses" for AI are mostly "clog the internet with low effort spam"
Again, I don't understand this performative disdain. You know they don't just "make human shaped blobs of text". You can converse in natural language with a machine; that was all but unthinkable three years ago. As far as AI slop goes, that is a problem with users, not with the tech itself.
Blud, we already had chat bots three years ago.
Prior to GPT-3, they were pretty shitty, though.
Ask chatGPT a question about any conflict in the world and it will give you the most 'Enlightened centrist' answer ever.
Ask it about any social issue and it'll sound like a TED Talk about white fragility. That's just the nature of how these things are trained. I don't think that means the technology sucks.
I would rather AI gets used to accelerate simulation research than dogshit misinformation LLMs.
Simulation research?
Drug discovery. Protein modeling prediction.
I feel like I recently heard about research on this literally using an LLM to predict protein sequences? This is not my field, so forgive the ignorance. But I would agree that prioritizing scientific research over consumer apps sounds sensible.
Llms are language models. Protein modeling is done using machine learning( DL and neural networks). Wherever you read that shit, the author obviously didn't know what they were writing about.
It's still useful for stuff. E.g. I never understood the rules of football. Then I watched the SuperBowl and just kept asking the LLM about stuff, and it gradually clicked. Before I would have had to read a poorly-written Wikipedia article using expert jargon without explaining it.
Is this dumbing me down? Possibly. Is it fundamentally different from asking a person questions about the rules of football? Not really.
Very possibly is.
I'd argue it makes learning more efficient, because you can zoom in on areas you don't understand.
Nuance and context gets murdered.
Yeah the fact that the chatbot talks more like people isn't nearly as impressive to me as it is to you I could ask any idiot on the street if I want a confident wrong answer I don't need a virtual idiot that costs $2 worth of electricity for every wrong answer
I guess you are hard to impress, then.