Over just a few months, ChatGPT went from correctly answering a simple math problem 98% of the time to just 2%, study finds. Researchers found wild fluctuations—called drift—in the technology’s abi…::ChatGPT went from answering a simple math correctly 98% of the time to just 2%, over the course of a few months.
I don’t understand why anyone even considers that. It’s a toy. A novelty, a thing you mess with when you’re bored and want to see how Hank Hill would explain the plot of Full Metal Alchemist, not something you would entrust anything significant to.
These models are black boxes right now, but presumably we could open it up and look inside to see each and every function the model is running to produce the output. If we are then able to see what it is actually doing and fix things up so we can mathematically verify what it does will be correct, I think we would be able to use it for mission critical applications. I think a more advanced LLM likes this would be great for automatically managing systems and to do science+math research.
But yeah. For right now these things are mainly just toys for SUSSY roleplays, basic customer service, and generating boiler plate code. A verifiable LLM is still probably 2-4 years away.
The problem is if you open it up, you just get trillions of numbers. We know what each function does, it takes a set of numbers between -1 and 1 that other nodes passed it, adds them up, checks if the sum is above or below a set threshold, and passes one number to the next nodes if it’s above and one if it’s below, some nodes toss in a bit of random variance to shake things up. The black box part is the fact that there are trillions of these numbers and they have no meaning individually.
