3 ms·
It’s not about being afraid, it’s that the auto-reconfiguration of neurons seems advanced to decompile it at this time, and it surprising that LLM, which are ju
by eastbound 2y ago
It’s not about being afraid, it’s that the auto-reconfiguration of neurons seems advanced to decompile it at this time, and it surprising that LLM, which are just a probabilistic model of guessing the next word, could be capable of actual thought.
The day it happens, we’ll believe it. There are only 100bn neurons in a brain, after all, and many more than this in modern machines, so it is theoretically possible. Just LLMs seemed too simple for that.