Large Language Models (LLMs)
sari and
Large Language Models (LLMs)
sari added 2y
Meanwhile, LLM results are probabilistic. 10 blue links are also probabilistic, but Google is giving you options - adding human curation by making you pick from a list of results. Using an LLM result is more like hitting ‘I’m feeling lucky’ - it requires more confidence
sari added 2y
How GPT-3 will turn all of your notes into an *actual* second brain:
sari added 2y
Artificial Intelligence
Timothy Shih and • 230 cards
Yang Zhou added 1y
sari added 6mo
“The fact that these things model language is probably one of the biggest discoveries in history. That you (LLM) can learn language by just predicting the next word … — that’s just shocking to me.”
- Mikhail Belkin, computer scientist at the University of California
Johann Van Tonder added 8mo
I've learned so much from @hubermanlab, but podcasts make recall and discovery challenging.
Huberman AI uses the latest models from @openai to search every episode with Google-level accuracy and summarize the results.
https://huberman.rile.yt
sari added 2y
LLMs are mid. It does a good job at getting the average and this is useful in many areas where this is tough. When you want "good", LLMs let you down. We underestimate the impact doing things to a "mid" level will have on the world.
Ian Vanagas added 1y
There’s a lot of hype around AI, and in particular, Large Language Models (LLMs). To be blunt, a lot of that hype is just some demo bullshit that would fall over the instant anyone tried to use it for a real task that their job depends on. The reality is far less glamorous: it’s hard to build a real product backed by an LLM .
Phillip Carter • All the Hard Stuff Nobody Talks About When Building Products With LLMs
sari added 1y