What if imitation is not opposed to goal-based learning, but a precursor to it? First you imitate to see what happens, and once you stumble upon a worthy goal you RL the heck out of it.
What if imitation is not opposed to goal-based learning, but a precursor to it? First you imitate to see what happens, and once you stumble upon a worthy goal you RL the heck out of it.
1. Train LLM
2. Launch inference API and chat
3. Get user feedback (e.g., money)
4. Repeat considering latest feedback
This is still RL, just that part of the RL loop is not code yet.
US banking cartel operates like a luxury company (e.g., Ferrari, Hermes). They'll only let you buy "starter" products until they've made enough money with you, and *only then* are you allowed to buy the stuff you wanted in the first place.
US banking cartel operates like a luxury company (e.g., Ferrari, Hermes). They'll only let you buy "starter" products until they've made enough money with you, and *only then* are you allowed to buy the stuff you wanted in the first place.
Grok just tried to tone police a short story draft I shared with it. When did it become this moralizing?
"No bueno." This is why I can't cancel other subscriptions just yet.
Do you ever think about getting 3 film directors to make their own versions of the same story, under identical budget constraints?
Could bring a breath of fresh air to cinema I think, a way of using constraints to unleash creativity and craft. "What's the best you can do with…
What the "drinking good vs. drinking bad" discourse always leaves out is how much the alcohol content in beverages has shifted over time.
For example, Egyptians, Greeks, Romans rarely consumed wine on its own, but rather watered down and/or mixed with other ingredients.
What the "drinking good vs. drinking bad" discourse always leaves out is how much the alcohol content in beverages has shifted over time.
For example, Egyptians, Greeks, Romans rarely consumed wine on its own, but rather watered down and/or mixed with other ingredients.
23K Followers 8K FollowingUn abuelo ❤️y el otro 💙. Te arrancaré una sonrisa diaria con retranca gallega. Combatiendo el Wokismo. 🅿️🅿️. España en el ❤️.
2 Followers 96 FollowingIf you found me, you've taken a wrong turn somewhere. Don't worry, it's a good place to be. We've got bad puns and questionable life choices. And a lot of fun.
312 Followers 7K FollowingIn a noisy world 🌎 🧠 Independent thinker
Strategy first. No hype ever.
Asking better questions.
Follow to think, not to rant.
20K Followers 100 FollowingMember of Technical Staff at Anthropic AlphaGo, AlphaZero, MuZero, AlphaCode, AlphaTensor, AlphaProof Gemini RL Prev Principal Research Engineer at DeepMind
19K Followers 406 FollowingI'm a software engineer @attio. Author of @ripple__js, @lexicaljs and @inferno_js. Former @reactjs core engineer, and core maintainer of @sveltejs at @vercel.
6K Followers 118 Following✦ Creating world-class saas product videos
✦ @beehiiv & 80+ (early to late stage) startups
✦ Work with me → https://t.co/r4ynMort15
190K Followers 2K FollowingCo-founder & CEO @Brave Software (https://t.co/NV4bmd6vxq) and @attentiontoken (https://t.co/XhGIrdBJWu). Co-founded Mozilla & Firefox. Created JavaScript.
36K Followers 968 FollowingAuthor of https://t.co/arW0hnVET0 and https://t.co/RN9xXOzhON. @sourcegraph working on @ampcode. Ex-@zeddotdev. Programming where the rubber hits the road.
57K Followers 1 FollowingWorkflow automation for technical teams to build AI solutions that integrate with any app or API at no-code speed and code flexibility. Open and self-hostable
1K Followers 2K FollowingI build/test stuff · Leader in Streaming Tech · Making a strategy game in Unity · Made: https://t.co/qgOYeBHod6 & https://t.co/GJ8vNvvZDO · Book recs in bio
1.2M Followers 279 FollowingWe’re a team of scientists, engineers, ethicists and more, committed to solving intelligence, to advance science and benefit humanity.
64K Followers 87 Following#TinyGlade is a small relaxing diorama builder where you doodle whimsical castles, cozy cottages & romantic ruins.
🐑 https://t.co/hNZtO5rrtb
No recent Favorites. New Favorites will appear here.