3 ms·
I remain unconvinced they're (the whole LLM/"Attention Is All You Need" industry) even barking up the right tree to build anything usefully-close to "AGI".
by alabastervlog 1y ago
I remain unconvinced they're (the whole LLM/"Attention Is All You Need" industry) even barking up the right tree to build anything usefully-close to "AGI".
- wilg 1y agoits certainly currently useful and also generally intelligent, right?
- alabastervlog 1y agoYes, and "that question is in not-even-wrong territory", respectively.
- wilg 1y agoa question, especially this one, can't really be "not even wrong"
- floydnoel 1y agouseful, sure. like a calculator for text. intelligent? no, and neither is a calculator
- dgs_sgd 1y agoIt's definitely useful. It can answer specialized PhD level questions correctly, yet cannot perform tasks that an average 10 year old could. I don't consider that generally intelligent.
- wilg 1y agobut it can perform all kinds of tasks. generally intelligent isn't universally intelligent, is it?
- btown 1y agoThe idea that any situation or sensory input can be broken down into a sequence of tokens, and that action choice can be characterized by predicting a subsequent sequence of tokens in the same space, may well bear fruit. But I think that a lot of people also buy into the idea that "text and image data from the web, and from historical chats, is the right/only way to generate the data set required," and it's a dangerous trap to fall into.
- rzz3 1y agoWhat do you think the correct, or at least a more likely tree is?