Comment on OpenAI says it’s “impossible” to create useful AI models without copyrighted material
luciole@beehaw.org 11 months agoThere’s this linguistic problem where one word is used for two different things, it becomes difficult to tell them apart. “Training” or “learning” is a very poor choice of word to describe the calibration of a neural network. The actor and action are both fundamentally different from the accepted meaning. To start with, human learning is active whereas machining learning is strictly passive: it’s something done by someone with the machine as a tool. Teachers know very well that’s not how it happens with humans.
When I compare training a neural network with how a trained to play clarinet, I fail to see any parallel. The two are about as close as a horse and a seahorse.
intensely_human@lemm.ee 11 months ago
Not sure what you mean by passive. It takes a hell of a lot of electricity to train one of these LLMs so something is happening actively.
I often interact with ChatGPT 4 as if it were a child. I guide it through different kinds of mental problems, having it take notes and evaluate its own output, because I know our conversations become part of its training data.
It feels very much like teaching a kid to me.
luciole@beehaw.org 11 months ago
I mean passive in terms of will. Computers want and do nothing. They’re machines that function according to commands.
The way you feel like teaching a child when you feed input in natural language to a LLM is known as the ELIZA effect. To quote Wikipedia: