RT Michael Malice: This is like a Black Mirror episode
RT Michael MaliceThis is like a Black Mirror episode
X / Twitter 6510 stories
Frontier labs will be viable businesses by being able to integrate and optimize inference at lower cost/performance than most other models. They'll ha…
Frontier labs will be viable businesses by being able to integrate and optimize inference at lower cost/performance than most other models. They'll have a margin advantage…
Not every page in your PDF needs the same treatment. A scanned cover, a dense table, a clean text page, a figure-heavy diagram…. most parsing pipeli…
Not every page in your PDF needs the same treatment.A scanned cover, a dense table, a clean text page, a figure-heavy diagram.... most parsing pipelines throw…
RT Douglas Murray: A reminder that the book that explains all this is still available, in an updated version, wherever books are sold.
RT Douglas MurrayA reminder that the book that explains all this is still available, in an updated version, wherever books are sold.
Still profitable btw
Still profitable btwZephyr: OpenAI cut Luna prices from $1/$6 to $0.2/$1.20A 5x price cutOpenAI begins the price war, with the goal of crushing the Chinese playersNot…
Grim…
Grim…GoForceX: @teortaxesTex they changed the banner on platform silently, "mid-july" is changed into "soon" btw
Can an open-source model perform like a foundation model? @Osmosis_AI is betting yes, using reinforcement learning and the dedicated @ycombinator GPU …
Can an open-source model perform like a foundation model? @Osmosis_AI is betting yes, using reinforcement learning and the dedicated @ycombinator GPU cluster on Together AI to…
luna is a thing of beauty, incredibly capable and low-cost
luna is a thing of beauty, incredibly capable and low-costShashank Goyal: Luna is far and ahead the most efficient dollar per token model right now.We also…
RT Google AI: What if you could see a city as it looked 100 years ago, or visualize a new basketball court at your local community center? Starting to…
RT Google AIWhat if you could see a city as it looked 100 years ago, or visualize a new basketball court at your local community center?Starting…
slightly better than V4-Flash on benchmarks, while smaller good
slightly better than V4-Flash on benchmarks, while smallergoodThinking Machines: Today, we are releasing Inkling-Small.Inkling-Small achieves comparable performance to Inkling at a quarter of its size. It…
Let’s go, they’re cooking. Looks like a great model. Fast follow up on a first model is such a green flag. 🫡
Let’s go, they’re cooking. Looks like a great model. Fast follow up on a first model is such a green flag. 🫡Thinking Machines: Today, we are…
RT Martin Ziqiao Ma: Thanks @ArtificialAnlys for working together on evaluations! Excited to see Inkling-small on the pareto line🫡
RT Martin Ziqiao MaThanks @ArtificialAnlys for working together on evaluations! Excited to see Inkling-small on the pareto line🫡Artificial Analysis: Thinking Machines' new Inkling Small scores 40…
RT Andi Marafioti: Thinking Machines just released Inkling-Small: 276B total, 12B active. A faster Inkling that matches or beats its 975B big brother …
RT Andi MarafiotiThinking Machines just released Inkling-Small: 276B total, 12B active. A faster Inkling that matches or beats its 975B big brother in many benchmarks.To test…
RT Horace He: Whereas I felt like it took a village to release inkling, inkling-small felt much more routine 😆 We just took the pipeline used for I…
RT Horace HeWhereas I felt like it took a village to release inkling, inkling-small felt much more routine 😆 We just took the pipeline used for…
RT ken: this, by the way, is the correct way to kill US proliferation of chinese open source
RT kenthis, by the way, is the correct way to kill US proliferation of chinese open sourceOpenAI: We are committed to pushing the model frontier across…
RT Kevin A. Bryan: Great work here from an all-star team on where we stand on RSI via AI research. My mental model is "what year will AI independently…
RT Kevin A. BryanGreat work here from an all-star team on where we stand on RSI via AI research. My mental model is "what year will…
RT merve: Thinking Machines released Inkling Small (🦖) + NVFP4 12B active 276B total params, the model performs better than larger Inkling on codin…
RT merveThinking Machines released Inkling Small (🦖) + NVFP412B active 276B total params, the model performs better than larger Inkling on coding 🤯 > check out…
Inkling-small. 2 weeks after inkling Nearly as good as Inkling but 4x smaller. We're just getting started…🔥
Inkling-small. 2 weeks after inklingNearly as good as Inkling but 4x smaller.We're just getting started...🔥Thinking Machines: Today, we are releasing Inkling-Small.Inkling-Small achieves comparable performance to Inkling…
Inkling-Small is comparable to Inkling at a quarter the size. Weights are open, fine-tunable on Tinker today. Look forward to seeing what people make …
Inkling-Small is comparable to Inkling at a quarter the size. Weights are open, fine-tunable on Tinker today. Look forward to seeing what people make with it.Thinking…
Re @ArtificialAnlys ok @openai gets it https://x.com/OpenAI/status/2082878156483219672
Re @ArtificialAnlys ok @openai gets ithttps://x.com/OpenAI/status/2082878156483219672OpenAI: We are committed to pushing the model frontier across cost efficiency, capability, and speed.Starting today, we are reducing prices for…
RT Thinking Machines: Today, we are releasing Inkling-Small. Inkling-Small achieves comparable performance to Inkling at a quarter of its size. It fea…
RT Thinking MachinesToday, we are releasing Inkling-Small.Inkling-Small achieves comparable performance to Inkling at a quarter of its size. It features 276B total parameters, 12B active. We…
> I would be okay with them rising in 50 years after a long period of being free Wypipos are truly generous…
> I would be okay with them rising in 50 years after a long period of being freeWypipos are truly generous…Black Goat of the Woods with…
RT Yun-Ta Tsai: Everyone is a real-world AI engineer. They don’t just make cars, but autonomous robots with frontier capabilities. Thanks for their h…
RT Yun-Ta TsaiEveryone is a real-world AI engineer. They don’t just make cars, but autonomous robots with frontier capabilities.Thanks for their hard work keeping the line…
Gemini Robotics 2 is here, with our new suite of models, robots can now reason through every movement to manage tasks that weren’t possible before, l…
Gemini Robotics 2 is here, with our new suite of models, robots can now reason through every movement to manage tasks that weren’t possible before, like…