“see the theory of constraints is, only improvements at the bottleneck will make a difference. and if you're putting more burden before and/or after the bottleneck, you actually make the whole thing worse, if it's a complex system https://t.co/8Vlk9Pxj2l”

“"It's autocomplete" is not a helpful analogy to understand LLMs. A LLM is more like a database that lets query information in natural language. You can query both knowledge, and "patterns" (associative programs seen in the training data, that can be applied to new inputs).”