Corey Holmer

1️⃣ Bring a new dimension to your ideas with Dream Lab.

New AI research from Meta – CoTracker3 Simpler and Better Point Tracking by Pseudo-Labelling Real Videos. More details ➡️ go.fb.me/xiyc63 Demo on @huggingface ➡️ go.fb.me/yzuqd0  Building on our previous work on CoTracker, this new model demonstrates impressive tracking results where points can be tracked for a long time even when they're occluded or leave the field of view. CoTracker3 achieves state-of-the-art, outperforming all recent point tracking approaches on standard benchmarks — often by a substantial margin. We've released the research paper, code and a demo on Hugging Face — along with models available under an A-NC license to support further research in this space.

We're releasing a few more #NotebookLM features today – we've been so inspired by all the ways this community has been using the tool, and we're excited to see where you take it next ✨

🆕 Starting this week, you can use document tabs to organize content in a single #GoogleDoc instead of linking to multiple files. Never lose track of your documents again. → goo.gle/4gUajkR

AI is 100% out of control. 1. Let's be friends

🆕 Gemini in #GoogleSheets can now generate charts & graphs to visualize data. You can also ask Gemini questions to glean insights & understand important patterns in your data. This is currently rolling out for Workspace Labs & Alpha users → goo.gle/3BE1o72

Image generation with Imagen 3 is now available to all Gemini users around the world. Imagen 3 is our highest quality image generation model yet and brings an even higher degree of photorealism, better instruction following, and fewer distracting artifacts than ever before.

As part of Meta Movie Gen, we trained a 13B parameter audio generation model that can take a video + optional text prompts to generate high quality audio — including ambient sound, foley & instrumental background music — all synced to the video. Details ➡️ go.fb.me/vgkswc

We built and launched a viral AI product in 2 months (at Google!). Here's how:

Our little NotebookLM team has had a crazy few months and is proving that small, nimble teams not only exist within Google, but can move fast and have significant impact. Our newest feature “Audio Overviews” has taken over the internet the past few days. The team has been sprinting - we went from idea to prototype in weeks, then launched publicly in under 2 months.  It’s not perfect (yet!), but that’s the point. Here are a few takeaways:  1. It's about building products *with* our users, not just *for* them. We’re not waiting to launch, we’re shipping early and iterating. Tech is evolving faster than users can possibly know what they want. We're often anticipating needs for V1 then working alongside them to improve. Example: Very quickly we saw a massive user desire for in-line citations so our team quickly pivoted to build and launch. (Join our discord!)

2. Built-in not bolted-on: We’re building net-new, AI native products. This isn’t just AI for the sake of “AI”, we’re working to bridge the gap between state-of-the-art research and human problems. Audio Overviews are great because they sound amazing, yes -- but useful because they are (1) source grounded and (2) an easy one-click way to study and digest your own information. 

 3. Meetings are spent building, not just talking about building. We’re collaborating, deeply. I am the only UX designer on the team so we have to make every moment count. @raiza_abubakar, @stevenbjohnson and I are constantly strategizing and iterating together.  NotebookLM represents a new era of product development within Labs at Google. We're putting user feedback and community engagement at the heart of everything we do. We’re building quickly and have a lot more coming soon.. The audio below is made with the NotebookLM Audio Overview feature using only my LinkedIn post! If you haven’t tried it, I highly recommend it. We want to make it better for you! It’s been awesome to see so much investigation into the product by @karpathy and @emollick.

Introducing canvas—your coding surface in ChatGPT. ✏️ Edit code inline 🐛 Review code and fix bugs 💬 Add logs and comments 🚢 Port to different languages We’ll be adding more to canvas over time. ChatGPT Plus and Team users can try the beta starting today.

Llama 3.2 features 11B & 90B models, our first multimodal Llama models with support for vision tasks. These models can take in both image and text prompts to deeply understand and reason on inputs.

⌘ + ⇧ + P — coming soon. Pre-order now: pplx.ai/mac

Unlock deeper insights with NotebookLM: Now analyze YouTube videos & audio files alongside your docs. Plus, easily share your Audio Overview with a new sharing option! Learn more: blog.google/technology/ai/…

Starting today, you can upload YouTube video URLs and audio files directly to NotebookLM, in addition to Google Docs, PDFs, text files, Google Slides and web pages. We’re also adding new features to make Audio Overviews even more useful. Learn more → goo.gle/3BmllPF

Today, we’re announcing that the Gemini app will be available to #GoogleWorkspace customers with Business, Enterprise, and Frontline plans. Keep your organization's data secure and compliant – all while delivering higher quality work with gen AI. → goo.gle/3ZBQngy

Gemini 1.5 Pro in Gemini Advanced is about to become more capable - stay tuned for updates!

Gemini for #GoogleWorkspace is now certified according to industry security and privacy standards. With Gemini, you’re getting an enterprise-grade AI-powered assistant that can help you meet your compliance requirements today. → goo.gle/3ZBQngy

We’re honored to announce that MagicSchool received a 93% privacy rating, placing at the top of AI tools for schools in Common Sense Media’s independent privacy evaluations. We share in @CommonSenseEd @CommonSense mission to create coordinated effort to protect child and student privacy, and build in safety and security from the start.  We still have much work to do and will continue to invest resources in safety and privacy as we grow.

🌟 New Tracers and Graphs in Tinkercad Sim Lab! 🌟 The new features give students an iterative, hands-on interaction with Sim Lab simulations to see the real-time motion and paths of shapes. tinkercad.com/blog/tracesand… #TeachWithTinkercad #TeachWithAutodesk #educationmatters

You can now explore up to 10 different voices with Gemini Live *and* change up your selection at any time. Take a listen here or in the app and let us know your favorite(s) in the replies.