For the record: I was never a skeptic.
I always said that human level AI was achievable within decades.
I also said it was harder than what most people thought & was not "just around the corner."
I said we weren't going to get to it with RL.
I said SSL, world models, and planning
RESEARCH
-
Achieving Human-Level AI: SSL, World Models, and Planning Approaches
By
–
-
Moravec’s Paradox: Why Robot Demos Bore Humans
By
–
As someone working on robotics in the depths of Moravec's paradox, one issue is that demos of skills that humans find effortless are highly boring
-
Python’s Role in Data Science and Machine Learning Infrastructure
By
–
Because for most of the Data Science, ML, and AI use cases almost all of the important stuff happens behind the scenes in highly optimized C, C++, Rust, and CUDA libraries. And these use cases is where Python has by now entrenched itself.
-
Vector Embeddings vs Word Token Embeddings Distinction
By
–
I’m talking about vectors that are the output of vector-encoding models, commonly called “embeddings”. not word/token vectors, also called embeddings, which you’re thinking of
-
Digiday releases latest weekly AI briefing update
By
–
It's another Monday, which means our latest @Digiday weekly AI Briefing is live.
-
Project-Based AI Training: Centrale Méditerranée’s Curriculum
By
–
[#Article] How to train tomorrow's AI professionals through projects? Discover Centrale Méditerranée school's curriculum that offers genuine engagement with companies' AI projects
-
FAIR: A Decade of Innovations in Artificial Intelligence
By
–
[#Article] FAIR, a decade of #innovations and new breakthroughs: Ego-Exo4D, Audiobox and Seamless Communication https://actuia.com/actualite/fair-une-decennie-dinnovations-et-de-nouvelles-avancees-ego-exo4d-audiobox-et-seamless-communication/
… #artificialintelligence #AI -
Embeddings Model Output vs Token Embeddings Clarification
By
–
great question; when I tweet about this, I will always be talking about the output of an embeddings model; if I mean token embeddings I will specify
-
Volume 207 Teaching Machine Learning and Artificial Intelligence Workshop Proceedings Released
By
–
Volume 207 Proceedings of the Third Teaching Machine Learning and Artificial Intelligence Workshop https://
proceedings.mlr.press/v207/ Is now available on PMLR -
AlphaGo’s Imitation Learning and Reinforcement Learning Recipe
By
–
Curiously, AlphaGo was using imitation learning (cake, yummy ) + RL refinement (cherry, yes plz ). A recipe that should sound familiar to many today.