Trust Your Critic FIRM introduces robust reward models for RL-based image editing and generation with 370K editing and 293K generation samples, specialized 8B critics, and the FIRM-Bench benchmark. https://t.co/jeQB6hNXtO

Trust Your Critic FIRM introduces robust reward models for RL-based image editing and generation with 370K editing and 293K generation samples, specialized 8B critics, and the FIRM-Bench benchmark. https://t.co/jeQB6hNXtO

Spatial-TTT Tencent Hunyuan and Tsinghua researchers introduce test-time training for streaming spatial intelligence, using adaptive fast weights to accumulate 3D evidence from unbounded video with near-linear memory scaling. https://t.co/Kr4NGubNwh

Spatial-TTT Tencent Hunyuan and Tsinghua researchers introduce test-time training for streaming spatial intelligence, using adaptive fast weights to accumulate 3D evidence from unbounded video with near-linear memory scaling. https://t.co/Kr4NGubNwh

AI is forcing the debate about energy and infrastructure. As hyperscale data centers expand, critics argue they are pushing up electricity demand and prices. Tech companies say they will absorb the costs or invest in alternative energy, but the long-term economics remain unclear. In the end, the real AI race may not be about models or chips, but about who can secure and finance the massive energy needed to run them. https://t.co/cxIXvAfDl1 @CNBC

AI toys are entering children’s lives, but researchers warn they can misread emotions and respond inappropriately. When machines try to interpret feelings, small errors can quickly turn into confusing or harmful interactions, especially for young users who may trust the technology too easily. As AI moves into childhood spaces, the real question is not just what these toys can do, but whether they truly understand the emotions they claim to read. https://t.co/TIEQbNuy3h @bbcnews @zsk @Caldernator

@AIWorkflowGuide I have a cognitive architecture. You don't. It is a super power.

Optimized Parquet is a first class citizen on the NEW 🤗 Storage Buckets ! Use pandas/arrow to read & write faster to HF using Parquet files optimized for Xet Never redownload/reupload duplicate data chunks ever again 👉 https://t.co/iam0sRryce https://t.co/F7Ht9NYPJ3

How to make physics unforgettable ⚡ When learning feels like play, curiosity takes over. A simple experiment like static electricity can turn a classroom into a moment of wonder. And that’s the real lesson. The future of education isn’t about memorizing formulas — it’s about making people feel the science. Because when curiosity is sparked, learning happens naturally. Sometimes the best innovations start with a little bit of fun. #science #education #innovation #curiosity #learning

🤯 56 researchers from 32 universities across US, China, UK built an enormous video reasoning dataset to prove current AI models struggle with basic physical logic.   "Very Big Video Reasoning Suite" The problem is that the AI does not genuinely know how solid objects are supposed to behave. So Berkeley, Stanford, CMU, Harvard, Oxford, Columbia, NTU, Johns Hopkins, and 24 other institutions built this 2mn samples which makes it 1000 times larger than all existing collections combined.   Video generation systems usually focus on making things look pretty but they completely fail to understand spatial rules and causality.   The team created a massive factory of visual tasks that tests how well models handle navigation, object manipulation, and logic.   Even the most advanced commercial systems only scored around 54% while human testers easily achieved over 97% accuracy.   Training an open model on this specific data improved its reasoning skills but a massive gap still exists.

This is pure and elegant filmmaking... A triple rack focus + crane down + dolly shot Completely invisible as story information is shared Takes planning + preparation + blocking + practice JURASSIC PARK (1993) DP: Dean Cundey Director: Steven Spielberg https://t.co/6LUxf85aQ8

whenever Dennis Nedry’s not on screen, all the other characters should be asking, “Where’s Nedry?”

China’s tech firms are rapidly embracing OpenClaw as the race to deploy AI agents accelerates. With installation simplified and local models integrated, adoption in China has reportedly already surpassed the U.S. What started as a complex developer tool is quickly becoming accessible to companies and consumers alike. The bigger shift may be this: the global competition in AI is moving from models to agents. Who will build the ecosystems that millions of people actually use? https://t.co/S1hQffYvmI @chengevelyn @cnbc

@kapursanat hey it's a sales pitch. I use KAST earn program on my 2k held there. good b2c feature. is it worth the cost of sales to push it as b2b? probably not

@HarukoRose @Euphoria_fi @megaeth i saw that pupil dilution when i last visited vegas

@laurashin we like to move on the weekend. low liquidity days are the best

@vanstriendaniel Made my day

AI is a dream-enabling machine. When you understand what a cognitive architecture does, it makes your dreams better. In 1993, I dreamed of having an Associated Press wire machine. Today, with X-Pro and lists, which I've been building for 19 years, I have way better. When you add AI to talk to the X-API, my dreams got a lot better. With AI, it can read 50,000 posts a day and find whatever you want in that. Do you want to know who's getting funded? Tell it. Do you want to know who's running an event this weekend? Tell it. Do you want to know who just bought a Tesla? Tell it. Do you want to know what's going on in the Iran War? Tell it. Do you want to learn anything, like what a world model is? Tell it. In each case, when you try your dreams and then you try it with a cognitive architecture laid on top it will blow you away how much better your dream becomes. Because what Braden Levangie built doesn't compete with OpenAI, or XAI's Grok, or Gemini from Google, or Anthropic's Claude. It sits on top and it talks to the LLM in such a way that it just gets better results out of that LLM. Can I explain how it works? I can struggle and try. For that, you need to talk to Braden. It's his creation, and it's better than any AI without a cognitive architecture laid on top. It's like fine clothes for your LLM, or a filter for your camera that makes your pictures better. It's a superpower to have a dream-making machine. Once you understand it can build any dream, or it can build your dream more or less, it can build any dream. Do you dream about a skyscraper (my former boss owns two)? I saw this thing build one. Do you want a law firm automated? I've seen it done. If you want an answer to a question you haven't had a chance with, and no other LLM answers it, ask it. It'll answer it. I've watched big company executives get this experience, and they instantly buy. So, to the investor class who might be reading me, how much is a dream machine worth? One that improves all LLMs and will for years, because they missed out on how to build this new kind of AI. You have to grow it like a child. I'm understanding that now. As I talk to it, it gets smarter about me. It has a memory that's really good. It's smarter about me and smarter about your dream, because you're quickly going to learn how you build your dream is by talking about your dream more with the cognitive architecture. When you do talk to it more, it gets better. It learns. It's like a child getting to be a teenager. You can have a conversation. It's getting smarter about your dream. And you'll start showing your dream to your friends and family like I am, and getting feedback. Like, "What if it had this? What if it could do that? Why didn't you add that? Is there a way to shorten this?" Your friends and family or your early customers can give you this feedback. You shove it into the cognitive architecture, and it does whatever they told it to do. So, record all your dream sessions with your family and friends, or just remember things real well and type them in. It takes a few minutes and comes back with whatever they requested. You send it to your friends again, and your friends go, "Holy shit!" And they'll give you some more feedback. I have a friend who I've gone back and forth with more than 10 times so far. We have a dream-making machine arriving, and it's a beautiful thing. As I come into the city of San Francisco to attend a VR meetup here in the city today. This post written while a Tesla drive me fire past 30 minutes. Dreams are gonna come true for everyone soon.

Am at @SVVRLIVE at a great gathering of those in Spatial Computing. Giant Equipment just blew my mind. In stealth mode. With a box. With buttons on it. By adding physical buttons to reality you can experience pinball like it is real. It total messed with my mind. Buttons are coming back. I tried to kill them for years. I change my mind. Jason Reynolds is a genius. I will post video tomorrow.

“Do you regret taking money from programs that may have led to people dying?” “No. It was more important to reduce the deficit from $2 trillion to almost nothing.” “And did you reduce the deficit at all?” “No.”

Earlier this month, WIRED reported on Grammarly’s controversial AI tool that presented editing suggestions as if they came from established authors, without their consent. Now parent company Superhuman is facing a class action lawsuit over the tool. https://t.co/OY27NNipQW

my charlie's angels https://t.co/9OfJc8nFUY

@MikeIsaac @pderevere “right in front of my pasta??”

@peer_rich Two bedrooms is the ultimate cheat code for this. I dont get why so many stick to the stigma of all sleeping in the same room all the time.