I think this is the heart of most of the angst around AI: it runs on computers, computers are precise and deterministic, therefore AI must be precise and deterministic.
But… it just doesn’t work that way. There is tons of room for improvement in safety and reliability, but expecting a multi-billion parameter neural network to have the same accuracy properties as a software calculator is always going to lead to frustration.
Complex systems have complex failure modes. There is a reason we use hammers and not CNC machine presses for nails.
Right but this is way too often glossed over in the rush to hype the new models. I see even many people that should know better failing to treat their output with appropriate skepticism.
So much money is being dumped into this stuff now there's a huge incentive to sweep the shortcomings under the rug.
Perhaps? But I'm not sure I see the value in saying that other people aren't doing a good job of setting expectations with yet other people. Presumably we around here know, right? And it's always fraught to imagine problems third hand.
But… it just doesn’t work that way. There is tons of room for improvement in safety and reliability, but expecting a multi-billion parameter neural network to have the same accuracy properties as a software calculator is always going to lead to frustration.
Complex systems have complex failure modes. There is a reason we use hammers and not CNC machine presses for nails.