"CS undergrad should primarily focus on teaching how LLMs work because LLMs are the endgame of CS" face the fucking wall 
Timeline
Post
Remote status
Context
3PL theory post Curry-Howard Isomorphism has a better case for being the "endgame" of CS (stupid concept) than LLMs but nearly nobody learns that in undergrad anymore
@scathach PL theory is kinda wank, I think algorithms and theory of computation matter more
@scathach the functional result of PL theory is that every language is now trying to become rust
Replies
1
bonifartius 𒂼𒄄 
@genmaicha @scathach most disciplines tend to end up in some weird dead end corners for prolonged times because publishing about other ideas is career instagib. string theory and big bang are obviously flawed and still we only see some reversal after at least half a century. climate science is largely soothsaying using made up models. computer science isn't different imo. i still don't see (haven't looked for a while) research about parallel algorithms, everyone seems to be doing the lamest "operations on matrices" markov chain on crack stuff possible. where are the cool parallelized graph algorithms etc.?