Richard S. Lingner( @rslingner@🦣.social )🌎 @rslingner

x.com/i/article/2029…

Of course you think “Operation Epic Fury” sounds badass, you’re 12

This tool is insane. Upload any image. Move camera in 3D space. And get a completely new camera angle 🤯 Also it’s free! 👇

@teortaxesTex Anthropic and Deepseek arguing about who’s responsible for distillation attacks

Concerning.

💯 Information monopolies are always a concern, but LLMs ratchet it up way beyond search engines. You are not just being referred to a trustworthy source, but are being given information predigested from opaque sources. Extinction angst serves as a convenient distraction..

VraserX e/acc @VraserX

6 months ago

Yann LeCun says the real AI risk isn’t extinction or killer robots, it’s power concentration. If AI becomes our main source of information, whoever controls it controls reality. That’s why he argues for open-source AI. A free future can’t be built on black boxes owned by a few.

58 160 718 71K 115

⚡ Google Genie 3 but OPEN SOURCE Not even 48h later and the Chinese did it again: they just dropped a free real-time playable world generator. - LingBot-World - Built on Alibaba's Wan2.2 - REAL-TIME interaction at 16fps 100% open source 🧵

testing mirrors in project genie, they definitely don't work correctly, but the results are funny

A man has been arrested after threatening to burn down an over 800-year-old shrine in Japan The site faced backlash after changing its social media icon to an AI-generated anime girl

keyboard_arrow_left Previous keyboard_arrow_right Next

President Bio, we are citizens from all around the world, and we stand in solidarity with you and your support of the rights of women and girls. Our united voices are behind you in the fight to ban FGM. @julius_maadabio secure.avaaz.org/campaign/en/si…

BREAKING: ChatGPT CEO Sam Altman says people share personal info with ChatGPT but don’t know chats can be used as court evidence in legal cases.

keyboard_arrow_left Previous keyboard_arrow_right Next

We're thrilled to release & open-source Hunyuan3D World Model 1.0! This model enables you to generate immersive, explorable, and interactive 3D worlds from just a sentence or an image. It's the industry's first open-source 3D world generation model, compatible with CG pipelines for full editability & simulation. Set to transform game development, VR, digital content creation and so on. Get started now👇🏻 Project Page：3d-models.hunyuan.tencent.com/world/ Try it now：3d.hunyuan.tencent.com/sceneTo3D Github：github.com/Tencent-Hunyua… Hugging Face：huggingface.co/tencent/Hunyua…

This guy built an auto aiming trash can for his messy coworkers

This may be the coolest emergent capability I've seen in a video model. Veo 3 can take a series of text instructions added to an image frame, understand them, and execute in sequence. Prompt was "immediately delete instructions in white on the first frame and execute in order"

Google just discovered a powerful emergent capability in Veo 3 - visually annotate your instructions on the start frame, and Veo just does it for you! Instead of iterating endlessly on the perfect prompt, defining complex spatial relationships in words, you can just draw it out like you would for a human artist. This capability is begging for a proper UX, but for now just doodle away in your app of choice, and use "frames to video" in Google Flow.

"Encourage Open-Source and Open-Weight AI" I'm super disappointed that this doesn't include open data. Without the training data, the rest is useless, and there is absolutely no way to tell the source of some issues, whether data was sanitized, how much generated data was used, whether specific types of data were repeated artificially to add weight. I will die on this hill and encourage everyone to find fully open models available on @huggingface like @allen_ai

Sriram Krishnan @sriramk

a year ago

link : ai.gov/action-plan

14 25 276 41K 184

Open-weight and open-source AI aren’t just technical artifacts; they are geopolitical tools. You can have both a thriving proprietary model ecosystem and a flourishing open ecosystem. Deepseek proved that just releasing a strong open-weight model can have disproportionate psychological and market impact. Most of the capital is currently flowing into closed, private models, but open-weight models are already punching above their weight in real-world performance. The catch? Most of the best ones are coming out of China. It’s important to distinguish open-weight from open-source. We often don’t know what data went into training these models. That might not matter for 98% of use cases, but for sensitive domains, it absolutely does. Like everything in tech, security lags behind deployment. In areas like code generation, where subtle backdoors could be catastrophic, trust becomes paramount. That’s why genuinely open-source models with verifiable data provenance should be prioritized in sovereign and critical infrastructure use cases.

Trully open LLMs shchegrikovich.substack.com/p/truly-open-l…

Dude had his production database deleted while vibe coding. Explains why Caffeine builds on an ICP tech stack designed for AI, where accidental data loss on upgrades is prevented, among other major advantages

Sriram Krishnan @sriramk

a year ago

link : ai.gov/action-plan

14 25 276 41K 184

🚀 Wrapping up an incredible journey at the PhyRC Challenge (emprise.cs.cornell.edu/rcareworld/cha…) at #ICRA2025! Over the past few months, teams from around the world have been tackling challenging tasks in assistive robotics — dressing and bed-bathing. The journey began in RCareWorld (emprise.cs.cornell.edu/rcareworld/), our human-centric simulation platform, where teams first demonstrated their solutions virtually. The top performers then advanced to the real-world phase at ICRA 2025, executing the tasks with a manikin in a physical setting. The winners were awarded real robots as prizes, including a Kinova Gen 3 robot sponsored by @KinovaRobotics for the dressing winners, and a Stretch 3 robot from @hellorobotinc for the bed-bathing winners. A huge thank you to the sponsors! 🏆 Congratulations to our winning teams: RoboNotts — Bed-Bathing Track (Jialin Chen, Koyo Fujii, Kalu Stephen, Areeb Akhter, Zakaria Taghi, Liz Felton, Luis Figueredo, @praminda, Aly Magassouba from University of Nottingham) UniChAMPions — Dressing Track (Maria Fernanda Paulino Gomes, César Bastos da Silva, Elton Cardoso do Nascimento, Ervin Bolivar H. , Esther Luna Colombini, Paula Dornhofer Paro Costa, Leonardo Rocha Olivi, and Eric Rohmer from Universidade Estadual de Campinas, @FEECUnicamp, and HIAAC UNICAMP) We’d also like to recognize the incredible efforts of the RALLA Team (Eunice Firewood, John Bateman, @jihongzhu, Jian Zhao, and @kefhuang from University of York) and the UWMTR Team (Cheng Tang, Hao Tian, Hasan Khan, Siha Pyo, Eddy Zhang, and Chao Tang from @UWaterloo) for their participation throughout the challenge. Thank you to all the organizing team members who made the PhyRC Challenge possible! @tomssilver @rishabhmadan96 @rohanbbanerjee @ZhanxinWu0725 @TapoBhat Check out the highlight video below to relive some of the incredible moments!