Your curated collection of saved posts and media
🎉 It's been a big week on the Veo front. Here are the highlights: —You can now generate vertical format videos in @FlowbyGoogle and via the Gemini API. —We reduced prices in the Gemini API by ~50%, so now developers can build with Veo 3 at $0.40/second and Veo 3 Fast at $0.15/second. To make sure you're getting the most out of Veo 3, here’s an easy prompt guide you can use for optimal video outputs ⬇️
The question is whether we’ll develop AI governance thoughtfully today, or reactively tomorrow. SB 53 offers a solid path toward the former. We encourage California to take it. Read more: https://t.co/n7azVO0uEe
Claude can now create and edit files. Turn conversations into Excel spreadsheets, documents, PowerPoint slide decks, and PDFs directly. https://t.co/F3yOiNMasU
New on the Anthropic Engineering blog: writing effective tools for LLM agents. AI agents are only as powerful as the tools we give them. So how do we make those tools more effective? We share our best tips for developers: https://t.co/N1kFYrTtax
Their ongoing testing of models like Claude Opus 4 and 4.1 has helped us find vulnerabilities and build strong safeguards before deployment. Read more: https://t.co/H0hoj3y2Ln
New from the Anthropic Economic Index: the first comprehensive analysis of how AI is used in every US state and country we serve. We've produced a detailed report, and you can explore our data yourself on our new interactive website. https://t.co/YI2qXIQzJO
In our report, we find that people delegate complete tasks to Claude 39% of the time—up from 27% just eight months ago. Meanwhile, DC and Utah lead the US in Claude use per capita. Read more: https://t.co/vlBnWuJer9
Listen to the OpenAI Podcast on— Spotify https://t.co/hLcRdGrfUX Apple https://t.co/0AdZ1ZteOV YouTube https://t.co/J0XZAq1uu4
We have updated ChatGPT's personalization page: personality configuration, custom instructions, and memories are now all in one place. Going live over the next couple of days. https://t.co/CAeBpP5j12
Typically, as models become smarter, their problems become easier to address—for example, smarter models hallucinate less and follow instructions more reliably. However, AI scheming is different. As we train models to get smarter and follow directions, they may either better internalize human goals or just get better at hiding their existing true goals. The core of anti-scheming research is to distinguish between these two, which requires understanding the reasoning behind a model's behavior.
In this new research with @apolloaievals, we found behaviors consistent with scheming in controlled tests across frontier models, including OpenAI o3 and o4-mini, Gemini-2.5-pro, and Claude Opus-4. We can significantly reduce scheming by training models to reason explicitly, using an extension to the Model Spec that prohibits scheming. That method is called deliberative alignment. With this technique, we can reduce covert actions by 30x for o3. However, situational awareness complicates results. Model spec: https://t.co/MiBFr9aNLz
Frontier models can recognize when they are being tested, and their tendency to scheme is influenced by this situational awareness. We demonstrated counterfactually that situational awareness in their chain-of-thought affects scheming rates: the more situationally aware a model is, the less it schemes, and vice versa. Moreover, both RL training and anti-scheming training increase levels of situational awareness.
We’ve heard your feedback that GPT-5 Thinking can sometimes take longer than you’d like. Now Plus, Pro, and Business users can set the pace to match the moment. Select GPT-5 with Thinking in ChatGPT on web to toggle thinking time in the message composer. - Plus, Pro, Business have Standard (new default, balancing speed and intelligence) & Extended (previous default for Plus) - Pro also has Light (snappiest) & Heavy (deeper) Your thinking time selection will remain in place for future chats on web until you change it.
@huggingface @CerebrasSystems Models provided through this API are currently only available to users on individual GitHub Copilot plans. Learn more in our release notes! https://t.co/PI8G8Hw2hF
This Wednesday, join us for a special @code livestream, straight from Microsoft HQ! @burkeholland will be joined by an amazing guest lineup: @housecor, Anastasia Khomochkina, @intellectronica, @kentcdodds, and @stolinski 🔥 Don't miss this 🔗 https://t.co/jROxJndA3v https://t.co/AeRmEB9zHL
🌍 VS Code Dev Days are making stops in Gembloux, London, Cordoba, Malvern, Recife, Paris, Milan, Nyeri, Austin, Kathmandu, Nyahururu, Hyderabad, Bengaluru, Daegu, and Beijing — plus online in the UK! 🔗 https://t.co/4WSHKeTMJV #VSCodeDevDays https://t.co/tHtna3c66S
The latest @code release brings auto model selection (preview) - a way to have the best model picked for you based on current capacity and performance. Currently being rolled out to all GitHub Copilot users in @code starting with individuals. Learn more: https://t.co/NRPlFJMZA1 https://t.co/abEcJbv8Fz
@huggingface Inference Providers, directly in @code Install the Hugging Face extension, bring your own API key, and instantly get access to hundreds of state-of-the-art open models: https://t.co/nhRA5JZ0Jy https://t.co/tg6kU7Y0zc
The GitHub MCP Registry just landed 🙌 Whether you’re building with GitHub Copilot, agents, or any AI tool that speaks MCP, this is the place to find what you need. https://t.co/wvpsVyRurP
The GitHub MCP Registry just launched - and we've got @kentcdodds on the stream tomorrow to chat about it! He's part of a stacked guest lineup we have, streaming live from Microsoft HQ: https://t.co/jROxJnd2dX
We are rolling out experiments to provide larger context windows in VS Code Insiders, now at 200k for GPT-5 and Claude Sonnet 4. Thank you @code community for the feedback! https://t.co/bep4CNmGpg
As you may have seen, the @github MCP registry is now LIVE! But how do you use it in @code? Checkout this video by @liamchampton to see how you can access the MCP servers live in the registry today 👇 https://t.co/dKspxXB01p
Join us at the HQ this morning for VS Code + Friends, a livestream of code, conversation, and what’s next: ✅ Candid chats with developers ✅ Hot takes on the latest news ✅ A deep dive into announcements like the new MCP Registry Don’t miss it: https://t.co/xBDqKP2S8C
🚀 Join us live NOW for a super-sized VS Code Live! @BurkeHolland is joined by amazing guests from the developer community for a session of code, conversation, and a look at what’s next 🎬 https://t.co/w7dkKjQmZj @kentcdodds @housecor @intellectronica @stolinski and Anastasia Khomochkina!
Spending the day at MSFT learning a ton and getting to chat with a ton of really incredible people. Really stoked with all of the energy around @code https://t.co/jYSu3lHOlI
Excited to be at Microsoft today for VS Code + Copilot Insiders Summit! https://t.co/NTDOONHtnp
Excited to be at Microsoft today for VS Code + Copilot Insiders Summit! https://t.co/NTDOONHtnp
Got @GitHubCopilot branded hot chocolate at the @code Summit event at @Microsoft this morning. Here to learn more about the direction VSCode is going! https://t.co/ioDedmTQIe
New in @code Insiders: Resolve merge conflicts with AI from the source control view. https://t.co/RzXzNolzfi
New in @code Insiders: Resolve merge conflicts with AI from the source control view. https://t.co/RzXzNolzfi
Check out our latest episode of the VS Code Insiders podcast, where you'll hear from our summer intern! We miss you Christy 🩵 https://t.co/AIlCWDasra
Ever wanted to try working with agents in Jupyter Notebooks? Not sure where to get started? Check out our newest workshop on working with agents in real-life data science workflows using @code: https://t.co/i1cNZ9BDmt
