$1B+ valuation, pitch includes robots that adapt to production changes without code updates.
The market is telling us what it wants: automation that doesn't start over every time the work changes.
Building exactly that at @robotensor_ai.
Papers to watch: UniWAM
The idea: stop separating "understand the world", "predict what happens next" and "act"
Instead their UniWAM combines a VLM-based physical reasoner + video world model + action policy in one architecture.
The model was trained on:
- robot demos
- human egocentric video
- VQA data
Results - achieved SOTA across several benchmarks:
99.2% LIBERO
92.6% LIBERO-Plus
68.3% RoboTwin Clean2Rand UniWAM
The part I find most interesting: human + robot co-training appears to follow a log-linear scaling trend.
Another datapoint that human video may matter a lot for scaling robot foundation models.
project page: uniwam.github.io
Zero-WAM is a World-Action Model (WAM) developed by researchers from @robbyant_brain and HKUST. Its key advance is in-context task generalization. It can take a human demonstration video or a text instruction as the task prompt. This lets the same WAM infer new tasks from richer deployment-time context and execute them without task-specific parameter updates.
A key enabler is HumanGen. It automatically converts robot trajectories into semantically matched human demonstration videos. This produced 74.2K human–robot pairs across 8.6K tasks, giving Zero-WAM broad task diversity for learning from demonstrations.
The results are strong. On seven unseen RoboTwin 2.0 tasks, Zero-WAM reached 46.95% average success, versus 17.45% for LingBot-VA and 10.98% for WAN-Action. Real-world experiments also showed generalization to unseen task configurations.
In short: earlier WAMs primarily learn to predict the world and act. Zero-WAM extends this toward demo/text → infer task → predict world → act, making WAMs substantially more flexible for unseen tasks.
#Robotics#WorldModels#EmbodiedAI
Brain to Robot: The Next Training Data Layer
In the near future, robot training data may come directly from human intent—not just joysticks, VR, or UMI-style teleoperation.
EEG + vision + robot state could capture what a human intends to do, then map that intent into robot actions.
This could attack a major bottleneck in today’s data pipelines: collecting diverse, natural behaviors at scale—especially tasks that are awkward, slow, or unnatural to physically demonstrate.
The next big robot dataset may be human intent + robot embodiment, not just human motion.
Interesting frontier being pushed by teams around @CarnegieMellon, @UCLAengineering, @neuralink and @BrainGateTeam.
#Robotics#HumanoidRobots#BCI#EEG#EmbodiedAI#RobotLearning#PhysicalAI#AI
When Robots Become the Teachers
In 2026, people are teaching robots how to think, reason, create, and solve problems.
By 2036, the roles may reverse.
Robots could be teaching people—sharing knowledge, guiding decisions, and helping even the most intelligent minds learn faster and think beyond today’s limits.
We are not just building smarter machines. We may be building the teachers of the next generation.
#Robotics#PhysicalAI#EmbodiedAI#RobotLearning
In-context learning for robotics could be much more than prompting with language.
Imagine VLA models that learn from a few demonstrations at inference time—watch the task, infer the intent, and adapt their behavior without fine-tuning.
Demonstrations as the prompt.
#PhysicalAI#Robotics#EmbodiedAI
If you want to have robots helping automate daily tasks, which do you need more?
A. Robot bodies?
B. Robot intelligence?
C. Affordable robots?
D. Plentiful robots?
Well?
If you answered E, all of the above, you are correct!
Robot bodies + brains + affordable + plentiful
47 Followers 386 FollowingInvesting. Building. Crypto, Stocks, τ. Invest with the future of the 🌍 and your 💰 in mind. 1 million followers across socials.
74 Followers 2K Followingonly private elite here!! l will mainly talk about crypto market price action analysis etc. ONLY FOR A FEW PEOPLE MAIN ACCOUNT @ShizzyUnchained
794 Followers 3K FollowingWas once fired for being a generalist. Full-stack cybersecurity marketer. GTM engineer, agentic orchestrator, local AI, & SEO. Bookish. Bitcoiner. Poet/luthier.
2K Followers 293 FollowingBritish entrepreneur in Mexico 🇬🇧🇲🇽 | Founder of 8Besos & Deltoria Media | Mezcal, marketing & content | Building businesses between the UK and Mexico
401 Followers 233 FollowingIndependent Bittensor intelligence 🪐
Deep Dives • TAO Planet Score • Signals
Mapping where real value is being built in $TAO.
13K Followers 3K FollowingThe voice of the Subnet Degens. TAO price action, Bittensor content, and the subnet news that actually matters before the crowd catches on.
402K Followers 3K FollowingI cheated on the entertainment and sold my soul to #satoshi. Betting Into Tomorrow Through Emerging Networks, Shaping Open Revolutions. and I ❤️ @mouseontop27
120K Followers 1K FollowingInvesting. Building. Crypto, Stocks, τ. Invest with the future of the 🌍 and your 💰 in mind. 1 million followers across socials.
7K Followers 161 FollowingOfficial X account for Unchained TAO Heads Telegram group.
Degens trading Bittensor subnets, grinding trenches, and hunting gems before the rest catch on.
360 Followers 110 FollowingGod is good. Bitcoin is money. Life is precious.
Pursuing high-tech real estate development to reshape the physical world.
ex-SWE, ex-Bridgewater
627 Followers 741 Following🔳Dont run the block🔳Own it!🗝 Cryptologist🛡Student of Tokenomics📊 Explorer of Digital Assets⚫️🔗Generational Wealth
〽️ultiplier♻️$TAO - SUBNETS $QNT #BTC
2K Followers 1K FollowingGlobal Events @cryptocom & @og_com
ex ops @suilendprotocol (acquired by @bluewaterHQ | industrial designer
nfa & dyor | views and tweets are my own