6 ms·
models are stuck in time. eventually the world moves forward and it is trained on too much obsolete data. training cannot end for LLMs intrinsically. it's not
by whateveracct 1mo ago
models are stuck in time. eventually the world moves forward and it is trained on too much obsolete data.
training cannot end for LLMs intrinsically. it's not some fixed cost. it's an ongoing one.
- jurgenburgen 1mo ago> eventually the world moves forward and it is trained on too much obsolete data. This is why LLMs are never going to be AGI. Humans don’t become obsolete just because they age.
- saghm 1mo agoHumans do die eventually though, which is an even more severe form of obsolescense. I don't think they'll be AGI either, but that argument doesn't seem super compelling.
- spwa4 1mo agoOn the other hand you have the old saying "science advances one funeral at a time", indicating death solves at least as many problems as it causes. (the point being that old, powerful professors have more than once blocked progress in their fields for decades. And only a funeral, eventually, solves the problem ...)