Ryan Wang

867 posts

Ryan Wang banner
Ryan Wang

Ryan Wang

@ryanwangny

Futurist and Tech leader; Venture Capitalist and entrepreneur in Spatial / AI / Blockchain / Frontier Tech; former investment banking and private equity

San Francisco Beigetreten Şubat 2010
548 Folgt763 Follower
Ryan Wang retweetet
Mike Kalil
Mike Kalil@mikekalilmfg·
Chinese robots appear headed for a federal ban in the United States. The bipartisan American Security Robotics Act of 2026, introduced by Republican Tom Cotton of Arkansas and Democrat Chuck Schumer of New York, targets unmanned ground vehicle systems (UGVs), which includes everything from wheeled surveillance robots to legged humanoid and robodogs. If passed, it would prohibit federal money from going toward robots manufactured or assembled by companies tied to China and designated foreign adversaries. GOP Congresswoman Elise Stefanik has introduced companion legislation in the House. The legislation does not propose a blanket consumer ban but it could have a major impact on federally funded research at universities where Hangzhou-based Unitree’s G1 has become the default humanoid development platform thanks to its relatively low entry price. Concerns about China’s aggressive robotics push have risen sharply among US officials after Unitree’s viral display during the 2026 Spring Festival Gala on primetime Chinese TV. A fleet of Unitree’s compact G1 and full-sized H2 humanoids stunned audiences with their advanced Kung Fu and acrobatic capabilities. “The Chinese Communist Party has shown that they are willing to lie and cheat to get ahead at the expense of the American people and our national security,” Schumer said in a press release. “They are running their standard playbook – this time in robotics – trying to flood the US market with their technology, which presents real security risks and threats to Americans’ privacy and American research and industry. We must protect our country from these threats, starting with a ban on the federal government buying CCP technology.” The bill’s language provides wiggle room for the Department of Homeland Security, the DoD, and the DOJ. They can still use the systems if it services national interests or if they’ve been modified to eliminate any data-sharing risk with foreign entities and are certified as secure. The proposal has a moderate chance of making it to President Trump’s desk for approval. Whether he signs it probably depends on his relationship with Chinese President Xi Jingping when it hits his desk. It’s a politically safe move for both major parties in the US, where attitudes about artificial intelligence have soured significantly since it captured the cultural zeitgeist in late 2022. Democratic Congresswoman Alexandria Ocasio-Cortez and Independent Sen. Bernie Sanders have proposed a moratorium of new AI datacenter construction until comprehensive federal regulations are considered.
Mike Kalil tweet media
English
8
8
39
4.5K
Ryan Wang retweetet
Unitree
Unitree@UnitreeRobotics·
🚀 Unitree open-sources UnifoLM-WBT-Dataset — a high-quality real-world humanoid robot whole-body teleoperation (WBT) dataset for open environments. 🥳Publicly available since March 5, 2026, the dataset will continue to receive high-frequency rolling updates. It aims to establish the most comprehensive real-world humanoid robot dataset in terms of scenario coverage, task complexity, and manipulation diversity. 👉 Explore the dataset here: huggingface.co/collections/un…
English
28
166
763
87.2K
Ryan Wang retweetet
anand iyer
anand iyer@ai·
Haptic scraped all 64 episodes of @RoboPapers (@chris_j_paxton + @micoolcho) and ranked every pain point in physical AI research. The top 10, by mention frequency: 1. Scalable data collection 2. Generalization / zero-shot robustness 3. Dexterous manipulation 4. Teleoperation / whole-body data 5. Sim-to-real transfer 6. Evaluation / benchmarking 7. VLAs / foundation models for control 8. Human video to robot transfer 9. Long-horizon memory 10. RL scaling / offline-to-online Code keeps getting cheaper. Atoms stay expensive. That's the entire startup opportunity in physical AI right now. hapticlabs.ai/blog/2026/03/0…
English
13
68
485
45K
Ryan Wang retweetet
Anish Moonka
Anish Moonka@anishmoonka·
$39 billion. That's what investors think Figure is worth right now. Their only completed deployment so far: two robots loading metal parts at a BMW factory for 11 months. The robots came back from BMW covered in scratches and dents. Figure pulled the entire line out of service. And the CEO has admitted he doesn't leave the robots unsupervised around his own children. So what's actually behind this living room demo? Figure built its own factory in San Jose called BotQ. The plan is 12,000 robots a year, eventually scaling to 100,000. They don't sell the robots outright. Companies rent them for about $1,000 a month. That's the real business model: recurring revenue from industrial customers long before any robot shows up at your door. The home play is even more interesting. Figure partnered with Brookfield, one of the biggest real estate firms on the planet (they own over 100,000 apartments worldwide). Brookfield is allowing Figure to record how people move through its buildings, kitchens, hallways, and offices. That data trains Helix, the robot's AI brain. Without it, these robots can't generalize beyond a controlled demo room. That data collection just started. Here's the pricing problem. About 15,000 humanoid robots were shipped globally last year. China made 90% of them. Tesla is shutting down its Model S and Model X lines at Fremont to convert them into an Optimus robot factory. They already have over 1,000 units inside their own plants collecting training data. @elonmusk says Optimus will cost $20,000 to $30,000. Unitree sells one starting around $16,000. 1X has pre-orders open at $20,000. Figure 03? Estimated at $50,000 to $100,000. Three to five times pricier than everyone else going after the same living room. The demo is real progress. But Goldman Sachs doesn't expect consumer humanoid sales to ramp until the early 2030s. Between here and a robot tidying your apartment, there's a factory that hasn't scaled, a price tag most households can't touch, and training data that's still being collected.
Figure@Figure_robot

Today we're showing Helix 02 that can tidy a living room fully autonomously Figure is designed so when you leave the house, your home resets exactly how you like it

English
27
14
305
93K
Ryan Wang retweetet
Brett Adcock
Brett Adcock@adcock_brett·
Today, Figure is showing another major milestone towards a robot in every home Running Helix 02, cleaning a living room fully autonomously
English
397
543
4.2K
673.3K
Ryan Wang retweetet
Lukas Ziegler
Lukas Ziegler@lukas_m_ziegler·
🚨 BREAKING: Rhoda AI raises $450M at $1.7B valuation! That means another unicorn in robotics industry (at Series A)! 🦄 Rhoda AI just raised $450 million in Series A funding at a $1.7 billion valuation and unveiled FutureVision, a robot intelligence system that handles the unpredictability of industrial environments. Let's take a look, at how it works. First, it studies hundreds of millions of internet videos to learn how objects move and how the physical world behaves. Then it uses that knowledge to constantly anticipate what's about to happen around it and translate those predictions into physical movements, repeating this cycle dozens of times per second. Most machines perform well in controlled, predictable environments but struggle when something unexpected happens. FutureVision targets this longstanding robotics problem. The platform integrates with a wide range of robotic hardware, allowing manufacturers and logistics operators to deploy intelligent robots without rebuilding existing systems. Breakthroughs in AI models that help robots understand language, interpret visual information, and predict how the physical world behaves, combined with growing investment from major tech and robotics companies, are driving robotics adoption. This is world models for robotics entering production. Learn physics and object dynamics from internet videos, then use that model to predict and act in real-time. The same pattern as language models: pre-train on massive internet data, then deploy for specific tasks. Congrats @startupjag! 🔥 Read more here: reuters.com/technology/rho… ~~ ♻️ Join the weekly robotics newsletter, and never miss any news → ziegler.substack.com
Lukas Ziegler tweet media
English
7
31
235
16K
Ryan Wang retweetet
AMI Labs
AMI Labs@amilabs·
Advanced Machine Intelligence (AMI) is building a new breed of AI systems that understand the world, have persistent memory, can reason and plan, and are controllable and safe. We’ve raised a $1.03B (~€890M) round from global investors who believe in our vision of universally intelligent systems centered on world models. This round is co-led by Cathay Innovation, Greycroft, Hiro Capital, HV Capital, and Bezos Expeditions, along with other investors and angels across the world. We are a growing team of researchers and builders, operating in Paris, New York, Montreal and Singapore from day one. Read more: amilabs.xyz AMI - Real world. Real intelligence.
AMI Labs tweet media
English
347
882
8.5K
4.9M
Ryan Wang retweetet
Products
Products@Products·
Boost team energy with FluidStance Fleet: dynamic standing decks for focused workdays.
English
0
139
1K
6.2M
Ryan Wang retweetet
Fei-Fei Li
Fei-Fei Li@drfeifei·
The dream of robots helping people live and work better in the physical world begins with helping robots to become more spatially intelligent by learning from the infinitely diverse and intricate environments of the 3D/4D worlds 🤖🤩
World Labs@theworldlabs

World generation is a bottleneck for robotics. We’re exploring how generative 3D worlds can reduce manual simulation setup and enable broader, more realistic evaluation 🧵

English
62
156
1.2K
102.5K
Ryan Wang retweetet
Chen Wong
Chen Wong@7n39_igolnik·
@Robo_Tuo Actually, China has more robotics companies.
Chen Wong tweet media
English
2
9
38
12.3K
Ryan Wang retweetet
Tuo Liu
Tuo Liu@Robo_Tuo·
Many cities in China are developing humanoid robots. Here’s a list of about 50 companies working on humanoids that we selected. With so many robotics companies, we’ll inevitably miss quite a few, but this gives a full picture of how fast and large the industry is growing here.
Tuo Liu tweet media
Tuo Liu@Robo_Tuo

We now have humanoid robot maps for China’s four major cities: Beijing, Shanghai, Shenzhen and Hangzhou. It might feel overwhelming to see so many humanoids, but it’s exciting to see these robotics companies working hard to push humanity forward.

English
46
325
1.2K
203.8K
AW3 Technology
AW3 Technology@aw3_xyz·
In 2022, we founded AW3 Technology - aw3.tech - with a simple but radical vision: to rebuild the way startups are formed, funded, and scaled — through collaboration instead of competition. Over the past three years, we’ve evolved into a venture studio developing companies at the intersection of AI and Web3, building software that democratizes intelligence, ownership, and opportunity . Today, we are proud to announce that AW3 Technology has officially acquired Deepwaters (VatnFörn Inc.), a former digital asset exchange and technology platform, marking a new phase in our mission to create a decentralized innovation ecosystem we call the Econoverse. Carrying the Deepwaters Torch: When we decided to acquire Deepwaters, it wasn’t just about acquiring technology — it was about inheriting a mission. Deepwaters was founded on a conviction that the digital asset economy deserves better infrastructure — markets that are fair, transparent, and immune to the opaque practices that have plagued centralized finance. Its WTR token and exchange architecture were designed around that principle: to reward transparency, not manipulation; to democratize liquidity, not hoard it. AW3 Technology will carry that torch forward by relaunching the Deepwaters platform as part of our broader decentralized ecosystem — integrating it with our venture network, token architecture, and Web3 development stack.
AW3 Technology tweet media
English
7
9
23
2.1K
Ryan Wang retweetet
Elon Musk
Elon Musk@elonmusk·
TSLA up $69 to ~$420 as foretold in the prophecy
Elon Musk tweet media
English
15.6K
22.3K
300.1K
30.5M
Ryan Wang retweetet
UploadVR
UploadVR@UploadVR·
BREAKING: Meta's HUD glasses with sEMG wristband will in fact be Ray-Ban branded, a leaked clip which also depicts the HUD and wristband in action reveals. Details here: uploadvr.com/meta-ray-ban-d…
English
95
262
1.6K
666.7K
Ryan Wang retweetet
Harrison Kinsley
Harrison Kinsley@Sentdex·
This is incredible
English
1K
1.8K
23.5K
2.8M
Ryan Wang retweetet
Andrej Karpathy
Andrej Karpathy@karpathy·
Noticing myself adopting a certain rhythm in AI-assisted coding (i.e. code I actually and professionally care about, contrast to vibe code). 1. Stuff everything relevant into context (this can take a while in big projects. If the project is small enough just stuff everything e.g. `files-to-prompt . -e ts -e tsx -e css -e md --cxml --ignore node_modules -o prompt.xml`) 2. Describe the next single, concrete incremental change we're trying to implement. Don't ask for code, ask for a few high-level approaches, pros/cons. There's almost always a few ways to do thing and the LLM's judgement is not always great. Optionally make concrete. 3. Pick one approach, ask for first draft code. 4. Review / learning phase: (Manually...) pull up all the API docs in a side browser of functions I haven't called before or I am less familiar with, ask for explanations, clarifications, changes, wind back and try a different approach. 6. Test. 7. Git commit. Ask for suggestions on what we could implement next. Repeat. Something like this feels more along the lines of the inner loop of AI-assisted development. The emphasis is on keeping a very tight leash on this new over-eager junior intern savant with encyclopedic knowledge of software, but who also bullshits you all the time, has an over-abundance of courage and shows little to no taste for good code. And emphasis on being slow, defensive, careful, paranoid, and on always taking the inline learning opportunity, not delegating. Many of these stages are clunky and manual and aren't made explicit or super well supported yet in existing tools. We're still very early and so much can still be done on the UI/UX of AI assisted coding.
English
457
1.1K
12.3K
1.2M