Feryal Behbahani at https://arxiv.org/search/cs?searchtype=author&query=Behbahani%2C+F Edan Meyer @ejmejm1 I totally forgot to post my recent video on AdA: https://youtu.be/BkWLCrLapQo Check it out if you want to see how @FeryalMP & colleagues successfully scale RL and pave the way towards future RL foundation models. It’s some great work! Human-Timescale Adaptation in an Open-Ended Task Space at https://arxiv.org/abs/2301.07608
Read More »

Richard: We talked before a bit about teaching AIs to use software, rather than trying to write it. There are a few ten million humans out of the 5 Billion using the Internet who try to write software, scripts and automation. Forcing billions of humans to become programmers, just to use computers is not an
Read More »

Google DeepMind @GoogleDeepMind Generating videos from text is challenging due to: 📈 High computational cost 📉 Limited quantities of high quality text-video data 📹 Variable videos lengths Here’s how the team addressed these issues when developing Phenaki: https://dpmd.ai/3O5U6vl Replying to @GoogleDeepMind Feed the hungry, train the jobless, comfort the helpless and lonely ones, watch out
Read More »

Richard: Can you help explain the Madelung energy and the heat of formation of ionic crystals? Bard: Sure. The Madelung energy is the electrostatic energy of interaction between the ions in an ionic crystal. It is named after Erwin Madelung, who first calculated it in 1919. The Madelung energy can be calculated using a series
Read More »

Is there a progress report for Bard that show what is happening for the whole Bard project? I would like to see how it is going. I do not have time to create a summary for myself by aggregating stuff from the web. And I would hope that Google is keeping track of all the
Read More »

@LambdaAPI site:Stanford.edu shows 3.91 Million entry points on Google search today. What would it cost to “AI-index” that so searchers ask questions in any language? If the whole process is cost effective and makes Stanford accessible it can replace “search and click” @LambdaAPI site:Stanford.edu shows 3.91 Million entry points on Google search today. What would
Read More »