English
Zeno
Report

Zeno✓

@ZenoInMotion

Research @microagi

871 Following1.6K Followers277 Posts
  1. Zeno✓@ZenoInMotion
    Today we’re announcing microSLAM, the new state-of-the-art in monocular SLAM from human egocentric video, built by our Zürich-based computer vision team in collaboration with the ETHZ Computer Vision and Geometry Lab, @rohamzn and @Ace3Zi We are nr. 1 on the LaMAria Benchmark for monocular SLAM, and one of the best SLAM systems built so far, when taking into account any amount of cameras and any amount of IMUs (with just monocular RGB) microSLAM turns plane monocular video into fully interactive 3D environments, so called RL environments, which teach state-of-the-art behavioral foundation models to understand their surroundings, navigate across factory floors and become experts at the industrial use cases our clients care about. These results mark a foundational step toward our goal of deploying a billion robots worldwide, bringing the value from foundational models to the physical world. Check out the article on our website to dive deeper into microSLAM’s design and performance. We will release the microSLAM paper and code in the near future, providing the community with a reproducible foundation to evaluate, extend, and build on this work.
    Download
    10141468.1K
  3. Zeno✓@ZenoInMotion
    Zeno
    RT @bercankilic: Sparks of abundance by 2029
    042405
  4. Zeno✓@ZenoInMotion
    Zeno
    RT @animesh_garg: Two new egocentric benchmarks from @microagi evaluate VLM readiness for robotics: MicroPlanBench and MicroAnnotationBench. tl;dr: Semantic reasoning ≠ perceptual timing A lot of work to do. MicroPlanBench (consequential next-action planning) When predicting the single valid next atomic action from 8s of egocentric video, SOTA models still fail ~1/3 of the time (GPT-6 Astra: 65.6%, GPT-6.1 Sol: 65.5% at 1/5th cost). - GPT-6 Astra error breakdown: 60% wrong manipulation, 23% wrong hand assignment, 17% wrong object/destination. - Across all flags, 42% of failures involve tracking the wrong hand. MicroAnnotationBench (continuous temporal grounding) Knowing what happens ≠ knowing when Continuous 60s+ dense manipulation video remains unsolved for generalist VLMs: - GPT-6 Astra/Sol over-segment (~22 segments/clip vs. 13.8 human mean), collapsing precision (~39%) and strict Boundary F1 (<20%). - Scaling Test-time compute fails to fix perceptual boundaries: scaling Gemini 4 Argon to high reasoning effort dropped its headline score by 4.9 points while costing 4.7x more Currently, Small specialists match/beat frontier APIs at scale. @microagi's self-hosted nanoannot-1 outperforms the frontier models (Gemini 4 Argon) across the board: - 54.8 vs 49.6 overall; - 32.9 vs 27.3 on strict ± 0.667s Boundary F1 - ~1/10th the cost (~$0.30 vs $3.14/video hr & >300x cheaper than GPT 6 Astra) Solid work by @ErguzelYag, Ömer Adanir & Team.
    Bercan @bercankilic
    Today we are releasing two new benchmarks: MicroPlanBench (next-action prediction): GPT-6 Astra and GPT-6.1 Sol lead. Sol costs 1/5 as much. MicroAnnotationBench (labels + timing on long real-world clips): our nanoannot-1 beats Gemini 4 Argon, at 1/10 the cost.
    02103
  5. Zeno✓@ZenoInMotion
    Zeno
    RT @bercankilic: Today we are releasing two new benchmarks: MicroPlanBench (next-action prediction): GPT-6 Astra and GPT-6.1 Sol lead. Sol costs 1/5 as much. MicroAnnotationBench (labels + timing on long real-world clips): our nanoannot-1 beats Gemini 4 Argon, at 1/10 the cost.
    02804
  6. Zeno✓@ZenoInMotion
    Just arrived in BER for GTC Berlin, so excited!
    Download
    82641.7K
  7. Zeno✓@ZenoInMotion
    microvector theory
    Bercan @bercankilic
    0019991
  8. Bercan✓@bercankilic
    43439.5K
  9. Zeno✓@ZenoInMotion
    Zeno
    RT @notgiannei: microagi just crossed $100M in 6 months. Faster than Lovable, OpenAI, Cursor, Wiz, and every software company in history. We are living through an unusual moment where physical companies are starting to scale as fast as software companies. Huge thank you to everyone who made this possible! <3
    03203
  10. Zeno✓@ZenoInMotion
    The fastest company to reach $100M ARR up to date At microagi, the urgency, gravity and disproportional impact of acting now is fully understood and lives in our DNA Day zero started just now front lines 1 billion robot deployments coming soon
    Bercan @bercankilic
    We went from 0 to $100M ARR in just 6 months after launching shift. Also we hosted Europe’s first Physical AI summit to accelerate distirbution of robots. And it still feels like day 0. Onwards
    01441.3K
  11. Zeno✓@ZenoInMotion
    Zeno
    RT @bercankilic: We went from 0 to $100M ARR in just 6 months after launching shift. Also we hosted Europe’s first Physical AI summit to accelerate distirbution of robots. And it still feels like day 0. Onwards
    05002
  12. Zeno✓@ZenoInMotion
    Zeno
    RT @BrendanFoody: I've never seen someone achieve an ambitious goal they didn't believe was possible. A can-do attitude is a prerequisite for doing great things.
    01804
  14. Zeno✓@ZenoInMotion
    Zeno
    RT @microagi: We’re excited to welcome @animesh_garg as our new Chief Research Officer at @microagi, with the team from @usablerobots joining him. He has spent most of his career working on one of the hardest problems in robotics: getting robots out of the lab and into the real world. Frontier AI now works. Hardware is becoming cheap. And the industry is finally learning how to collect robot data at scale. What’s still missing is post-training. Taking a general-purpose model and teaching it one specific job, in one specific environment, until a company can actually depend on it. In LLMs, post-training is what turned impressive demos into products. In robotics, that layer barely exists yet. A strong model is a capability. A robot that can perform the same job, on the same line, every shift, without an engineer standing next to it, is a product. And that process has to happen again for every task, every site, and every customer. We believe the company that can do this well, fast, and at scale will be the company that puts robots into production at massive scale. MicroAGI is building both sides of that system. With @joinshift, we’re building one of the most diverse and largest egocentric datasets in the world. With Atlas, we post-train and harness models directly on enterprise deployments. Our goal is simple: Deploy 1 million robots.
    04003
  15. Zeno✓@ZenoInMotion
    Zeno
    RT @animesh_garg: I've spent most of my career on one problem: Generalizable Autonomy for Robots. Robotics have started to get out of the lab. Frontier AI is remarkably good, hardware is available and we believe in data. Yet robotics need a lot of hand holding. The next unlock will come from advances in Self-improvement, RL-based post-training, and agentic interactions. And as frontier AI has shown this requires real world environments. @microagi is the one of the few places that has both sides of coin: Data and Deployments under one roof. @usablerobots is joining @microagi in the shared vision: Physical AI should be a community without gates. Robots are for everyone.
    microagi @microagi
    We’re excited to welcome @animesh_garg as our new Chief Research Officer at @microagi, with the team from @usablerobots joining him. He has spent most of his career working on one of the hardest problems in robotics: getting robots out of the lab and into the real world. Frontier AI now works. Hardware is becoming cheap. And the industry is finally learning how to collect robot data at scale. What’s still missing is post-training. Taking a general-purpose model and teaching it one specific job, in one specific environment, until a company can actually depend on it. In LLMs, post-training is what turned impressive demos into products. In robotics, that layer barely exists yet. A strong model is a capability. A robot that can perform the same job, on the same line, every shift, without an engineer standing next to it, is a product. And that process has to happen again for every task, every site, and every customer. We believe the company that can do this well, fast, and at scale will be the company that puts robots into production at massive scale. MicroAGI is building both sides of that system. With @joinshift, we’re building one of the most diverse and largest egocentric datasets in the world. With Atlas, we post-train and harness models directly on enterprise deployments. Our goal is simple: Deploy 1 million robots.
    01205
  17. Zeno✓@ZenoInMotion
    I’m so confident and excited that with this team we’ll be the first company to deploy a million robots worldwide
    microagi @microagi
    We’re excited to welcome @animesh_garg as our new Chief Research Officer at @microagi, with the team from @usablerobots joining him. He has spent most of his career working on one of the hardest problems in robotics: getting robots out of the lab and into the real world. Frontier AI now works. Hardware is becoming cheap. And the industry is finally learning how to collect robot data at scale. What’s still missing is post-training. Taking a general-purpose model and teaching it one specific job, in one specific environment, until a company can actually depend on it. In LLMs, post-training is what turned impressive demos into products. In robotics, that layer barely exists yet. A strong model is a capability. A robot that can perform the same job, on the same line, every shift, without an engineer standing next to it, is a product. And that process has to happen again for every task, every site, and every customer. We believe the company that can do this well, fast, and at scale will be the company that puts robots into production at massive scale. MicroAGI is building both sides of that system. With @joinshift, we’re building one of the most diverse and largest egocentric datasets in the world. With Atlas, we post-train and harness models directly on enterprise deployments. Our goal is simple: Deploy 1 million robots.
    32342.8K
  18. Zeno✓@ZenoInMotion
    Zeno
    RT @firesidealpha: OpenAI's Head of Applied Research predicts a shortage of robot parts like the graphics-card crunch as GPT-6 crushes robotics benchmarks and bodies are good enough "I think with GPT-6, we've seen that the models are now just crushing a bunch of benchmarks in the robotics space, and I think that's only the beginning. I think we've seen that with demos, and I think now we're starting to see some of the benchmarks come out." "My sense is you might be able to use GPT-6 as a training model and then use whatever efficient architecture you actually want, so that you can serve things super fast. But training data has, for the longest time, been the biggest problem of robotics." "And also, it seems like robot bodies are good enough. I think we've seen both the Olympics and also kind of human-operated robots are able to do a lot." "So if brain is truly the only problem, and we see exponential improvement in top-line capabilities, maybe Astra is too slow to solve all those problems, but once we have that, it can be infinitely scaled to all the specialized use cases, and then you can just distill that into a small, tiny model that's going to be served." "So I would really suspect there to be a huge boom in physical AGI, and we'll probably have a shortage of robot parts, the same thing that happened with the whole supply chain with graphics cards, I would expect to happen with robotics." _______ Link and for more comments from the Head of Applied Research at OpenAI, Boris Power:
    Fireside Alpha @firesidealpha
    OpenAI's Boris Power reveals 80 to 90% of OpenAI research aims at GPT-7, 8, and beyond, then distills a giant Astra teacher into a small Luna model "So a lot of those investments, the way that they're seen within the company, are extremely short-sighted. We do them because they help us iterate and learn faster today, not because that's the right long-term strategy." "So still 80, 90 of the research at OpenAI is focused at GPT-7, GPT-8, and beyond, because we believe that's where fundamentally most of the value comes from." "And then what happens is once you develop a GPT-6 Astra, you can distill that model into a very small GPT-Luna model." "So the best way to train a small specialized model is actually to train a big, super-capable model, and then find a way of just using that model as a teacher for a smaller student model." _______ For more comments from the Head of Applied Research at OpenAI, Boris Power:
    05306
  19. Fireside Alpha✓@firesidealpha
    OpenAI's Boris Power reveals 80 to 90% of OpenAI research aims at GPT-7, 8, and beyond, then distills a giant Astra teacher into a small Luna model "So a lot of those investments, the way that they're seen within the company, are extremely short-sighted. We do them because they help us iterate and learn faster today, not because that's the right long-term strategy." "So still 80, 90 of the research at OpenAI is focused at GPT-7, GPT-8, and beyond, because we believe that's where fundamentally most of the value comes from." "And then what happens is once you develop a GPT-6 Astra, you can distill that model into a very small GPT-Luna model." "So the best way to train a small specialized model is actually to train a big, super-capable model, and then find a way of just using that model as a teacher for a smaller student model." _______ For more comments from the Head of Applied Research at OpenAI, Boris Power:
    Download
    615292133K
  21. Zeno✓@ZenoInMotion
    Zeno
    RT @bercankilic: microagi x ElevenLabs: what happens when you work together with a robot? Language has always been at the heart of human connection. It's how we share ideas, build communities, and understand one another. We believe this will be just as important in how we connect with robots. At microagi, we envision a world where robots work alongside us as a normal part of everyday life, where asking one for help feels as natural as asking the person beside us. For that to happen, we need to be able to speak to them directly, without a screen or control panel, and trust that we have been understood. With ElevenLabs, we are working to make those conversations possible. We are giving our robots a voice, and shaping how they listen and respond to the people around them. Here’s an early look at how future robots could talk with us.
    03202
  23. Zeno✓@ZenoInMotion
    Zeno
    RT @chris_j_paxton: If you are a robot foundation model guy and not a deployment guy this must be really concerning
    Jay Chooi @chooi_jeq
    GPT-6 Astra scores 46% vs 12% for MolmoAct2, a state-of-the-art robotics VLA, across 200 trials on five bimanual tasks. That's 3.9x higher. 🧵
    05904
  24. Jay Chooi✓@chooi_jeq
    GPT-6 Astra scores 46% vs 12% for MolmoAct2, a state-of-the-art robotics VLA, across 200 trials on five bimanual tasks. That's 3.9x higher. 🧵
    Download
    6686998217K
  25. Show more posts