Excited to introduce GPT-4o. Language, vision, and sound -- all together and all in real time. This thing has been so much fun to work on. It's been even more fun to play with -- with moments of magic where things feel totally fluid and I forget I'm video chatting with an AI.

RT @wanchao_: Excited to share that I joined @thinkymachines recently! It’s been an incredible experience so far working alongside many tal….

RT @stephenroller: We are moving incredibly fast. Come light up GPUs with us.

RT @lilianweng: Yes - 🥳 Thinky starts hiring again:

RT @_alex_kirillov_: We have been working hard for the past 6 months on what I believe is the most ambitious multimodal AI program in the w….

It’s really fun to work with a talented yet small team. Our mission is ambitious - multimodal AI for collaborating with humans, so the best is yet to come!. Join us— or fill out the application below if interested!.

RT @miramurati: Thinking Machines Lab exists to empower humanity through advancing collaborative general intelligence. We're building mult….

If you’re excited to build the future of multimodal human/ai collaboration, and jam with Andrew, me, and many other talented people across the stack— dm me! 😀.

RT @alhyunsoo: life update: I joined @thinkymachines!. feeling so lucky to build with such a kind, brilliant team, esp pairing with researc….

RT @AIatMeta: 🚀New from Meta FAIR: today we’re introducing Seamless Interaction, a research project dedicated to modeling interpersonal dyn….

TIL, the best (and perhaps only!!) way to speak to a human at Xfinity over phone is to say you're cancelling your service. everything else is an automated system. that said, I learned this trick from o3, so I guess it's AI-versus-AI here.

Important PSA

the Singapore MRT (subway) is so impressive. Many lines that go everywhere, high frequency of trains, plus it’s fully automated so it has smooth cross platform transfers. It’s safe and clean (no durians allowed). Open loop payments so you can pay by credit card…

RT @GXiming: With the rise of R1, search seems out of fashion? We prove the opposite! 😎. Introducing Retro-Search 🌈: an MCTS-inspired searc….

huge congrats to @bowenc0221 for showing image-manipulation works well for vision perception, and for carrying the project all the way through to the finish line!.

RT @sainingxie: About two years ago, we started building V* to bring visual search into a multimodal LLM and show t….

RT @jilin_14: Exciting to share what i've been working on in the past few months! o3 and o4-mini are our first reasoning models with full t….

If you're going to #ICLR2025. come join me at the @thinkymachines happy hour! There will be food!. (Space is limited and we can't guarantee everyone a spot, so please RSVP indicating interest).

RT @MustafaShukor1: We release a large scale study to answer the following:.- Is late fusion inherently better than early fusion for multim….

RT @ajabri: the pros and cons

RT @alisawuffles: We created SuperBPE🚀, a *superword* tokenizer that includes tokens spanning multiple words. When pretraining at 8B scale….