[ML News] DeepMind controls fusion | Yann LeCun's JEPA architecture | US: AI can't copyright its art

Yannic Kilcher

description

#mlnews #deepmind #fusion Updates on what's going on in the ML world! Check out w&b's alerts feature: https://wandb.me/yannic

OUTLINE: 0:00 - Intro 0:20 - Sponsor: Weights & Biases 2:35 - DeepMind uses Reinforcement Learning to control nuclear fusion 4:35 - Google responds to carbon emission estimates 8:40 - Yann LeCun proposes new architecture for world models 11:05 - Fruit fly neurons may perform multiplication 12:00 - Emojisearch App 12:30 - Ar5iv officially in arXiv labs 12:55 - Language Model Consciousness & Media Hype 16:45 - Vision models are more fair when trained on uncurated data 18:30 - CLIPasso 19:15 - NLP with Transformers Book 20:15 - Helpful Things 26:00 - US Office: AI can't copyright its art

Sponsor: Weights & Biases https://wandb.me/yannic

References: https://wandb.me/yannic

DeepMind uses RL to control nuclear fusion https://deepmind.com/blog/article/Accelerating-fusion-science-through-learned-plasma-control https://www.nature.com/articles/s41586-021-04301-9/figures/1 https://www.nature.com/articles/s41586-021-04301-9.pdf https://www.alexirpan.com/2018/02/14/rl-hard.html

Google responds to carbon emission estimates https://ai.googleblog.com/2022/02/good-news-about-carbon-footprint-of.html

Yann LeCun proposes new architecture for world models https://ai.facebook.com/blog/yann-lecun-advances-in-ai-research

Fruit fly neurons may perform multiplication https://www.nature.com/articles/s41586-022-04428-3

Emojisearch App https://twitter.com/lilianweng/status/1488791391358513153 https://www.emojisearch.app/ https://github.com/lilianweng/emoji-semantic-search/blob/main/server/app.py

Ar5iv officially in arXiv labs https://blog.arxiv.org/2022/02/21/arxiv-articles-as-responsive-web-pages/

Tech media may be only slightly conscious https://twitter.com/ilyasut/status/1491554478243258368 https://futurism.com/the-byte/openai-already-sentient https://interestingengineering.com/ai-might-be-conscious https://futurism.com/mit-researcher-conscious-ai https://www.dailymail.co.uk/sciencetech/article-10503703/Artificial-Intelligence-expert-warns-slightly-conscious-AI.html https://futurism.com/conscious-ai-backlash https://www.dailystar.co.uk/tech/news/conscious-ai-already-exist-expert-26223303

Vision models are more fair when trained on uncurated data https://arxiv.org/pdf/2202.08360.pdf

CLIPasso https://clipasso.github.io/clipasso/

NLP with Transformers Book https://www.amazon.de/dp/1098103246?linkCode=gs2&tag=oreilly200c-21