Welcome to Import AI, a newsletter about AI research. Import AI runs on lattes, ramen, and feedback from readers. If you’d like to support this, please subscribe. Subscribe now A wild Russian LLM f…| Import AI
Dropping a model release as significant as Llama 4 on a weekend is plain unfair! So far the best place to learn about the new model family is this post …| Simon Willison’s Weblog
The future of generative AI could rely on smaller language models for every application an enterprise uses, models that would be both more nimble and customizable — and more secure.| Computerworld
DeepSeek's breakthroughs in lowering the entry point for creating chatbots and other generative AI tools means small players now have a more level field on which to develop new products and compete.| Computerworld
You shouldn't need a permission slip to read a webpage–whether you do it with your own eyes, or use software to help. AI is a category of general-purpose tools with myriad beneficial uses. Requiring developers to license the materials needed to create this technology threatens the development of...| Electronic Frontier Foundation
Knowledge distillation happens in many forms.| seantrott.substack.com
By default powerful ML systems will have dangerous capabilities (such as hacking) and may not do what their operators want. Frontier AI labs should design and modify their systems to be less dangerous and more controllable. In particular, labs should:| ailabwatch.org
When a dangerous model is deployed, it will pose misalignment and misuse risks. Even before dangerous models exist, deploying models on dangerous paths can accelerate and diffuse progress toward dangerous models.| ailabwatch.org
Posts about copyright written by Mike Masnick, Dark Helmet, and Leigh Beadon| Techdirt
Unlike major competitors including Google’s Gemini and OpenAI’s ChatGPT, Meta’s Llama 3 model is open source and the AI assistant it powers has previously been locked inside the company’s app ecosystem, which includes Facebook and Instagram.| Forbes
Llama 2 lead and Llama 3 post-training lead Thomas Scialom of Meta/FAIR, on the Chinchilla trap, why Synthetic Data and RLHF works, and how Llama4's focus on Agents will lead us to Open Source AGI.| www.latent.space
Thanks to our latest advances with Meta Llama 3, Meta AI is now smarter, faster and more fun than ever before.| Meta Newsroom
The open source Llama 3 AI model released by Meta today is just the start, according to the company’s chief AI scientist, Yann LeCun. He said a new, much larger version is in the works.| WIRED
AGI by 2027 is strikingly plausible. GPT-2 to GPT-4 took us from ~preschooler to ~smart high-schooler abilities in 4 years. Tracing trendlines in compute (~0.5 orders of magnitude or OOMs/year), algorithmic efficiencies (~0.5 OOMs/year), and “unhobbling” gains (from chatbot to agent), we should expect another preschooler-to-high-schooler-sized qualitative jump by 2027. Look. The models, they just| SITUATIONAL AWARENESS
Update on June 14, 2024 at 7:30am PT: We’re disappointed by the request from the Irish Data Protection Commission (DPC), our lead regulator, on behalf of| Meta
Your AI Update—April 2024 Keep up with recent news in the world of Generative AI, including new…| Technology & Digital Studies Programs
What stands behind the cost of LLMs? Do you need to pay for training an LLM and how much does it cost to host one on AWS? Read about it here| TensorOps