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 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 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 freeWypipos are truly generous…Black Goat of the Woods with…
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, like…
RT Wall Street MavWhat’s happening in Spain is what is coming for America if Democrats regain power. Complain about Republicans all you want, they’re the only…
good job little broOpenAI: After deployment, we applied GPT-5.6 Sol to advance the frontier of efficiency by making itself more efficient to run. The results:- 20%…
> there was "not a real strategy for how long or what they should do”The privilege of the Hegemon. Like cat and mouse… just capriciously playing…
RT Sayash KapoorCan AI agents conduct open-ended AI research?Most evaluations of agents conducting AI research focus on narrow, verifiable tasks. But AI research is often open…
Good questions. My attempt at answers:1) they meekly sell them to our Asian allies for scrap, no longer having any use for them and trying to…
RT AI Engineer🆕 First Steps Toward Automated AI Research https://youtu.be/pWXUkLP9uWMHumanity advances by trying things, finding the shortcomings, and fixing them. @RichardSocher keynotes our first-ever Autoresearch track…
Running open-weight models in production gets hard fast.On August 6, we’re walking through major updates to our inference platform that make deployment and operations easier:→ Safe…
btw protip:if you can distil modelsyou can also distil harnesses
I hope there are QA testing experts out there who are thinking "finally, we don't need software developers any more!", rolling up their sleeves and building…
try asking codex to go beyond what you think possibleBilly Howell: Things are getting wackier in AI land. I usually give Codex a long "impossible" task…
RT Tesla MegapackAutobidder, our battery optimization service, is maximizing the value of one of France’s largest battery storage projectsThe Chevire project was developed by @Harmony_Energy_ on…
We're live with the @Kimi_Moonshot team to discuss Kimi K3! https://x.com/i/broadcasts/1mxPaaMeLDZKN
Everything you need to know about Kimi K3: A conversation with Moonshot AI and Together AI https://x.com/i/broadcasts/1mxPaaMeLDZKN
Consistent with what I found with Qwen3.6 a while back: Claude Code uses 2-3x as many tokens than (many) other harnesses at similar success rate.- Unoptimized?-…
RT jasonone of the most beautiful things about OpenAI is that every employee really has a voice.i wanted to capture what it feels like to work…
RT jasonone of the most beautiful things about OpenAI is that every employee really has a voice.i wanted to capture what it feels like to work…
Starting in 15 minutesTogether AI: Kimi K3 has everyone’s attention. On July 30, hear @Kimi_Moonshot's Feihu Tang explain the architecture and decisions behind it.He joins Jue…
Incredible to see the open model momentum continue.Seeing more than 230 organizations rally around open weights reinforces what’s possible when the ecosystem comes together.Thank you to…