Burak | 부라크

Introducing Unhardcoded: an open source layer that picks which AI model handles each call. Instead of hardcoding model=x, you declare a rule (constraints, ranking, fallback), delivering deterministic, auditable routing on your own keys. Try it: unhardcoded.com

her surname is Путин?

that too late?

skeptics were right back then like the ones who said the same about ai and ml in the 70s so zk went the same way and ur right now all i am saying is timing matters

that smth works in theory doesn't mean it will scale in practice Eli

i would prefer to say prediciton trees instead of conditional markets btw

ppl say ur flirting w an asian beauty bro, pls dont burn any eth hahaha

🇰🇷🫶🇹🇷 brother countries forever - 영원한 형제의 나라

pls understand that ai, ml, deep learning and llm are not same thing thnx

unfortunately yes 100% absolutely yes

north korean beauties are everywhere, be careful guys 😄

the prob has been to see the solutions ppl wish to use

it’ll be smth at the intersection of world model and mbrl bc obviously the next big paradigm is ai that doesn’t just generate answers but models the world, anticipates outcomes and makes its own decisions @nvidia cosmos 3 👀

桜梅桃李

Sezonu kapattı zemin yuzunden

guys i am not Sahil, i am Rahil Thanks

thats wht youre building genlayer sir?

great q bro first we as genlayer are approaching this from both the protocol and application layers like we use independent validators running different ai models have to agree on a judgement, plus with an appeal mechanism expanding the set and involving fresh validators but i think ur calibration work adds another layer to this so how do you see judge level calibration fitting into a decentralized jury and appeal system?