if technical path is wrong

then likely project will fail

lots of efforts will be wasted

many wrong technical paths looks attractive

they promised a great future

but finally they failed, and forgotten

tech industry don’t lack stories of wrong technical path

semantic web, virtual reality, java applets, nosql, big data, web3, p2p, 3d printing, no-code,

a common pattern is the over promise of applications

most of the time, we don’t need big data, nosql, vr

today llm could be another over-hyped technique

we should treat it as another tool, which is good at text/image/video/audio understanding and generation

but the intention to rebuild everything with llm is doomed to fail

it could achieve very high level of intelligence, like solving ioi math problems

but most applications are not math problems

<
Previous Post
prompt vs code
>
Blog Archive
Archive of all previous blog posts