Nathan Wilbanks (@nathanwilbanks_) 's Twitter Profile
Nathan Wilbanks

@nathanwilbanks_

ai ml researcher designer | agent master | yapper extraordinaire | exited founder | building the most powerful ai agent ecosystem @agnt_gg

ID: 1637503854161072128

linkhttps://github.com/agnt-gg/slop calendar_today19-03-2023 17:18:45

1,1K Tweet

5,5K Followers

2,2K Following

Saurabh Kumar (@drummatick) 's Twitter Profile Photo

i don’t think you understand you train a small model on scikit-learn it gets 80% accuracy you fine-tune a huggingface transformer it gets 90% accuracy you build a custom data pipeline it gets 93% accuracy you ensemble three models with voting it gets 95% accuracy you

Simon Willison (@simonw) 's Twitter Profile Photo

Quitting programming as a career right now because of LLMs would be like quitting carpentry as a career thanks to the invention of the table saw.

dax (@thdxr) 's Twitter Profile Photo

every ai tool is like "we spin up 7 subagents and make them fuck to the rhythm of a black horned tree cricket chirps and sample the results using an advanced algorithm (Math.random) and that's why we're #1 on this benchmark you never heard of"

TheSeanLavery (@cyborglavery) 's Twitter Profile Photo

kache You think that is a ceiling above you? No those are just the clouds. There is no limit. You think "Sky is the limit"? That mindset is your own shackles. Take off your limiter. There is no ceiling. If you think you see one, smash through it, and accelerate.

Teknium (e/λ) (@teknium1) 's Twitter Profile Photo

If an llm can use web search/browse during evals, and the evals are on the internet… cant they just search up the test set and find the answer? Lol

@bluecow 🐮(schizo) (@bluecow009) 's Twitter Profile Photo

The Lord wishes to do something great in our lives and in the history of humanity, yet there are few who perceive this, pause to receive the gift and then proclaim and share it with others.