We've hit 100 episodes! Here's a look back on our journey so far (website with some stats): robopapers100.com/ Some highlights in the thread 🧵:

Name another project doing this. $AGPU is now one of the fastest-growing neoclouds on the market. Shares up ~590% in six months. $3B+ in signed real-world contracts this year, $6B more in the pipeline, and prepayments already flowing in the hundreds of millions. Dedicated NVIDIA Blackwell clusters. Enterprise SLAs. Actual cash from actual customers. Literally nobody else in crypto is doing billions in real-world GPU contracts with a token sitting at the center of the stack. This is uncharted territory. It’s still early. The big Build deployments haven’t even fully ramped yet. Next is Aethir Access - Aethir moving down the stack into physical AI data centers. Access already secured to 10 sites totaling up to 20 MW across the US and Europe, purpose-built for NVIDIA B300 and GB300 clusters. Projected up to $700M in contracts by end of 2026 and over $2B at full buildout. Months to live capacity, not years. A lot of people keep asking “what about ATH?” Don’t worry. ATH is core to the thesis. Aethir Foundation is $AGPU’s largest shareholder. $AGPU holds a Strategic Compute Reserve of ATH. I don’t know how to tell you any more clearly. This is your accumulation time. Don’t bet against us. You’ll wish you listened when the opportunity still looked like this. @AxeCompute @AethirCloud $AGPU ethereum:0xbe0ed4138121ecfc5c0e56b40517da27e6c5226b

Full ABC release 🔤 ! Train, test, and evaluate your models. The full suite of simulation tasks and data has also been released! Excited to see what people build 🔨🤖

Inspired by recent thoughts from senior researchers whom I deeply respect (e.g. @JitendraMalikCV @Michael_J_Black @Ken_Goldberg @phillip_isola), I wrote down some thoughts as a junior researcher too. The recent Astra demos in dexterous manipulation are truly impressive. But they also raise two questions that I don’t think academia has good answers to yet: - How should credit be assigned when an agent synthesizes many prior works into a new research result? - How can academia attract and retain talents when so much of its incentive structure relies on credit? I reflect on these questions in more detail in a blog post. Curious to hear what others think, and debates are very welcome! toruowo.github.io/blog/posts…

Imitation learning, especially with interventions, has driven so much recent robotics progress. However, improving a policy via targeted interventions until it reaches a useful and deployable success rate is a time and labor intensive process. Instead, wouldn’t it be great if policies could improve on their own? That’s what @VarunGiridhar3 and @animesh_garg join us to talk about. In Q-Planning, they start with a large policy like pi-0.5, and add a Q-function estimator to predict value instead of just actions, then use both successful and failed rollouts to update this Q-function online, then use it to guide sampling and trajectory selection. With just a few rollouts they can dramatically improve policy performance online. This provides a way to do really difficult tasks like inserting a credit card into a wallet, increasing success rate from 25% to 80% in just a few iterations. Learn more in Episode 105 of RoboPapers, hosted by @micoolcho, @chris_j_paxton, and @ruijie_sg.

muse is out there controlling this robot!! we’re rooting for you buddy!!

1/ Better robot models come from better data. That takes two things: hardware that captures it fully, and tooling that turns it into training signal. We built the first. Today, @GroundedSI launches the second: Grounded API. Access both hardware + API ↓

Grounded API is live. - SOTA on hand-tracking benchmarks (< 1 cm) - SOTA on SLAM benchmarks - In-the-wild ego data -> enriched data in minutes - Integration with @huggingface @LeRobotHF & @rerundotio - Built for @BitRobotNetwork RoboCap suite Technical report & more↓

We recently came out with our blog post, Does Scaling Web-Video Pre-training Help Real Robots Do Real Work? This is a pretty exciting moment for robotics, because as far as I am aware, it is the first evidence that robot foundation models can scale on internet video data; not merely on teleoperation, UMI data, or egocentric data, but on the data you can truly find anywhere. It’s the necessary requirement for reaching massively powerful models.

One of the key advantages of legged robots like humanoids should be how effectively they can move across a wide variety of terrain types to accomplish their task. But Light-Loco-Parkour from the team at Light Origins aims to change that: using only onboard sensing, they show a policy which can decide when to walk, vault, climb, or otherwise traverse as it moves through a complex environment. Unlike many others, it uses sparse seeds instead of relying on a large motion corpus, learning when to use its skills to move around without specific sub-task labels. @ChemXiaodao and @Yuntao144 join us to go into the details. Watch Episode 104 of RoboPapers now, with @micoolcho and @chris_j_paxton, to learn more!

Unitree General-Purpose Humanoid Foundation Model Fully Upgrade Major Open Source🥳 Unitree majorly fully open-sources the UnifoLM-WLA-1.0 embodied foundation model, achieving new SOTA results across multiple benchmarks among open-source models worldwide. A single model coordinates desktop and whole-body mobile manipulation, supporting cross-task and cross-end-effector generalization, driven by one model, whole-body coordination. unigen-x.github.io/unifolm-w…

Predictions on where robotics is going in the next few years

What is the role of academic computer vision research in the age of increasingly powerful large models? Is GPT-6 Astra a step change? How can a researcher have an impact today in academia? These are the questions I ask myself as I head off to ECCV 2026, a conference I’ve attended since 1992. One of my papers this year is VIGA, a method that takes an image as input and outputs a 3D Blender scene that represents that image. This is a classical inverse-graphics task and VIGA was the first method to solve it using an agentic approach. The idea is now several years old and the first version of the paper was rejected. This delayed publication significantly. After it was accepted at ECCV, it was quickly surpassed by people using Claude Code for the same purpose. Today GPT-6 Astra blows away all previous results. But we still head off to ECCV to tell the community about our invention that is now fully out of date. The way academic work often progresses is that one reads recent papers, notices that they have limitations, comes up with a new idea, explores this, publishes it, etc. Any published paper I read today is based on ideas that are at least a year old. And those ideas were based on the literature of the time, which was also a year old. That means that any paper I see at ECCV is likely two years out of date. In AI today, two years means your work is likely irrelevant. At CVPR this summer I noticed that many authors have not gotten the message. They continue to work on “old” problems that have a long history. This history is based on assumptions about how the “vision problem” will be “solved”. The truth is that it is being solved in a very different way and many of these problems are no longer relevant. Another group of papers focuses on very niche problems where large models likely fail because of insufficient data or lack of business interest. The impactful papers were largely from industry and had long author lists and massive data+compute behind them. These papers were also out of data, describing systems that had been released months before, but at least they served to provide the community with more complete documentation and analysis of commercial systems. So what should academics do? First, we need to put aside the tools we’ve used for years and start from scratch. Every project should start by trying really hard to solve the problem with existing tools. I would like to see every paper begin with a detailed experimental analysis of how existing models perform and why they fail (if they do). This gives the kind of insight we need today. Then, assuming current models fail, the solution should provide some fundamental insight that will outlive the next release of such models. Reviewers today still focus on technical novelty. This pushes people to focus on tweaking architectures rather than clearly moving the field forward. Papers need to be judged based on their novel insight and not their novel technical contribution. This is a real shift in thinking but it focuses us on what matters - progress of the field. If we want there to be a “field” of computer vision, then it can’t become a marginal backwater, focusing on esoteric problems. If you haven’t tried using Astra (or whatever comes next) to solve your problem, then you have not done your homework. This omission should be seen as negatively as not having a previous work section. Concretely, I think papers should include a new section analogous to “Related Work” where that related work is current models and how they perform on the task. Reviewers should start asking for this and expecting authors to be able to articulate their insights about the limitations of existing large models. I'm interested in your thoughts.