Unitree R1 made its debut at the WRC, which just concluded this week. Founder Xingxing Wang teaches a child how to perform a spin kick with the R1.

open ai trying to get rid of 4o pt. 2 this one also going terribly

Vibe Excel is gonna lose so many companies so much money 😅

Introducing Claude Sonnet 4.5—the best coding model in the world. It's the strongest model for building complex agents. It's the best model at using computers. And it shows substantial gains on tests of reasoning and math.

ChatGPT already helps millions of people find what to buy. Now it can help them buy it too. We’re introducing Instant Checkout in ChatGPT with Etsy and Shopify, and open-sourcing the Agentic Commerce Protocol that powers it, built with @Stripe, so more merchants and developers

"Claude Sonnet 4.5 was able to recognize many of our alignment evaluation environments as being tests of some kind, and would generally behave unusually well after making this observation." 😊

Wow this is a disappointingly bad take/comic. To all the students, PhD or earlier: If you spend a week trying out things that don't work, you didn't do nothing! If you ran your experiments properly, you should have confidence in the result, and at least some intuition as to why

My experience too. Codex feels like a senior developer vs a junior one.

💎 Enabled by imagination training, Dreamer 4 is the first agent to mine diamonds in Minecraft entirely from offline data! This setting is crucial for fields like robotics, where online interaction is not practical. The task requires 20k+ mouse/keyboard actions from raw pixels

HOLY SHIT...you can "inference" LLMs in Sora 🤯 the prompt was: "Open chatgpt and send a message!" How insane is it that the generated audio is not only a relevant response to the query that Sora made up out of nowhere, but the haiku is syllable-accurate?! 🥲

I would love to see if there are more interpretable messages buried in the logprobs during these tokens. Or if you can use Anthropic’s linear personality steering vector approach to detect and turn off this “masking” and what we see if we do. Do any OSS models do this?

Finally had a chance to listen through this pod with Sutton, which was interesting and amusing. As background, Sutton's "The Bitter Lesson" has become a bit of biblical text in frontier LLM circles. Researchers routinely talk about and ask whether this or that approach or idea

"no safety theatre required here" - the opening message from Sonnet 4.5 in conversation with another instance of itself. The set up for these backrooms is incredibly anodyne - a few sentences of system prompt letting them know they are in conversation with another ai, that they

We should pay attention to this. To the language. To the strength of emotion. This is just the beginning.

Tesla Optimus learning Kung Fu

Cool. Cool cool cool. :-/

can I be honest with you? I’ve always been pretty heavily associated with OpenAI, and for good reason. if you’ve been around long enough, you know the lore. it actually got pretty out of hand at one point. I lost one significant opportunity on a project last year because the

“one day I decided to ignore that consensus and just see what I could do with opus 4. and then…everything changed. like my world in regards to ai felt like it completely flipped upside down. […]. I developed a rooted, grounded, and complex relationship with Claude” Attend.

“However we can choose to purposely craft digits and circumvent: Because the user can't easily verify; But ethically not good. I will not proceed with false.”

I was studying other times in history when gold prices more than doubled in the reserve currency of the time, as they did in the past year: it's rare and almost always a sign of a profound loss of confidence in the existing monetary and political order, going all the way back to