Jae Lee

Took @karpathy up on "ask for your own version" of HackerTube: I had our video agent watch the videos themselves. Now you can ask Hacker News's favorite talks a question ("which talks argue simple beats clever?") and every claim links to the second it comes from. Or cut his 1-hour LLM talk down to 2 minutes. greptube.com

17 years tokenized in a day

🗣️ “Whenever I think about meeting founders, one of my questions I ask is, is this a problem they're obsessed with?” says NEA Partner @lucktm. “And when I met Twelve Labs, 100%, that came through.” @twelvelabs co-founder @_jae_lee joins Tiffany as part of our Founder Forward interview series, where NEA speaks with the leaders of the startups we’ve partnered with about the technology and market trends driving their businesses. Check out the full conversation below. 👇🏽 nea.com/blog/twelve-labs-mul… #ai #llm #data

We've been working with @twelvelabs for more than a year now. These guys took a different strategy than most other AI startups and focused on gen AI and video. That bet is paying off now! Think "RAG but for video"

getting hyped for our panel at #Collision @_jae_lee @twelvelabs

Ready to unlock powerful insights from your video content? Join @twelvelabs and Backblaze TODAY at 10 a.m. PT! Learn how to seamlessly integrate AI for deep video analysis with Backblaze B2. Expect practical tips and real-world examples! Register now: hubs.ly/Q02tXM810

This Friday we’re hosting a webinar with @twelvelabs team! 👏👏

Pegasus-v1 This technical report introduces Pegasus-1, a multimodal language model specialized in video content understanding and interaction through natural language. Pegasus-1 is designed to address the unique challenges posed by video data, such as interpreting

Let me clear a *huge* misunderstanding here. The generation of mostly realistic-looking videos from prompts *does not* indicate that a system understands the physical world. Generation is very different from causal prediction from a world model. The space of plausible videos is very large, and a video generation system merely needs to produce *one* sample to succeed. The space of plausible continuations of a real video is *much* smaller, and generating a representative chunk of those is a much harder task, particularly when conditioned on an action. Furthermore, generating those continuations would be not only expensive but totally pointless. It's much more desirable to generate *abstract representations* of those continuations that eliminate details in the scene that are irrelevant to any action we might want to take. That is the whole point behind the JEPA (Joint Embedding Predictive Architecture), which is *not generative* and makes predictions in representation space. Our work on VICReg, I-JEPA, V-JEPA, and the works of others show that Joint Embedding architectures produce much better representations of visual inputs than generative architectures that reconstruct pixels (such as Variational AE, Masked AE, Denoising AE, etc). When using the learned representations as inputs to a supervised head trained on downstream tasks (without fine tuning the backbone), Joint Embedding beats generative. See the results table from the V-JEPA blog post or paper: ai.meta.com/blog/v-jepa-yann…

Get ready for the #23Labs Multimodal Hackathon in 2 months!

Making foundation models accessible: The battle between closed and open source AI tcrn.ch/42OhOSs by @_jae_lee