Zhiyuan
Oct 29, 2026

Zhiyuan unveils GO-3 Preview to scale robot learning across pre-training, post-training and deployment

Zhiyuan on Oct. 29 introduced GO-3 Preview, an embodied AI model the company says brings scalable learning to the full robot lifecycle, spanning pre-training, post-training and deployment. The release was paired with claims that the model ranked first in three benchmark evaluations: RoboDojo, RoboTwin and RoboColiseum. The system is built around WRAM, short for World Reasoning Action Model, which links task understanding, future-state prediction and action generation. In pre-training, Zhiyuan said roughly 90% of GO-3 Preview’s data came from human first-person video, with the rest used alongside robot trajectories to connect observed state changes to executable control. The company also introduced KASO, a method designed to select among multiple plausible future predictions and align them with real actions during training. For post-training, Zhiyuan described Genie Evolver, a distributed hybrid setup that combines real robots, physics simulation and world models. In one experiment, the company said scaling the number of real robots improved phone-packaging success rates from 50.8% to 98.3%, while failures in a 120-run long-horizon test fell from 25 to 2. Zhiyuan also said GO-3 Preview achieved a win rate above ordinary human players in tabletop curling after only a few hours of practice in real and model-based environments. The deployment layer is handled by Genie Master, an embodied agent framework aimed at turning real store operations into new training data. Zhiyuan said its Genie G2 robot has already been tested in a real convenience store, where it kept improving on prepared-food packing, item retrieval and restocking without relying on manual demonstrations or tuning.

650
Zhiyuan unveils GO-3 Preview to scale robot learning across pre-training, post-training and deployment
Binance
Oct 29, 2026

Binance opens FFR stock trading to eligible users outside the U.S.

FF EAI Robotics Ecosystem said its stock is now available through Binance’s U.S. stock trading service for eligible users outside the United States. The company said users can trade the stock with stablecoins. FF EAI Robotics Ecosystem, listed on Nasdaq under the ticker FFR, described itself as a U.S.-based embodied AI robotics ecosystem company. The update was cited by Techub News, which attributed the announcement to PR Newswire. No additional rollout details were disclosed in the brief. The announcement only specified that access is limited to eligible non-U.S. users and that stablecoins can be used for trading. Binance was named as the platform providing the stock trading access.

810
Binance opens FFR stock trading to eligible users outside the U.S.
Unitree
Oct 29, 2026

Unitree and Sequoia China launch early-stage investment program for embodied intelligence startups

Unitree Robotics and Sequoia China said on Oct. 28 that they have jointly launched the "Shanyu Venture Initiative," an early-stage investment program focused on embodied intelligence, robotics applications, upstream and downstream technologies, and other frontier hard-tech projects. The move marks Unitree’s first formal and systematic push into investing. The program’s stated approach centers on five principles: understanding technology and industry, finding projects early, making decisions quickly, backing people, and staying out of founders’ way. Unitree founder Wang Xingxing said the initiative will use the combined perspective of engineers and investors to identify strong founders and help young teams shorten the path to product-market fit. The announcement also fits into a broader pattern in China’s embodied intelligence sector. Companies including AgiBot, Xinghaitu, Zibianliang, Galaxy General and others have already launched investment plans or backed startups across components, data, algorithms and application scenarios. The article argues that as financing in embodied intelligence has cooled in the second half of this year, industrial capital is taking on a larger role, especially in areas where supply chains, mass production and real-world deployment matter as much as funding.

600
Unitree and Sequoia China launch early-stage investment program for embodied intelligence startups
Y Combinator
Oct 29, 2026

YC’s latest startup wishlist shifts AI focus to 11 capital-heavy sectors

Y Combinator’s newest Request for Startups argues that AI’s biggest bottlenecks are no longer confined to model parameters or algorithms. In YC’s view, the next wave of startups will be built around hardware, infrastructure, financing tools and specialized software that bring AI into the physical economy. The list spans 11 areas, from cheap space launch and AI-powered professional services to U.S.-made robotics, government benefits infrastructure, municipal permitting, data center power and cooling, air-gapped AI systems, and enterprise-owned sovereign models. The message is clear: thin software wrappers are losing their edge, while opportunities are moving toward sectors with real-world constraints, supply-chain complexity, regulatory friction and large capital requirements. YC says it wants founders who can build in those environments, not just around foundation models. The firm also highlights specific figures behind that thesis, including Starship launch frequency targets, U.S. robot installation data, Medicaid eligibility checks beginning in 2027, transformer lead times stretching to three years, and the growing competitiveness of open-weight models in enterprise use cases. Taken together, the wishlist lays out YC’s current map of where it believes AI-native companies can create durable value.

520
YC’s latest startup wishlist shifts AI focus to 11 capital-heavy sectors
Sequoia China
Oct 28, 2026

Sequoia China and Unitree launch early-stage hard tech investment program

Sequoia China and robotics company Unitree have jointly launched the "Shanyu Chuang Program," according to an official announcement. The initiative will focus on identifying and investing in early-stage startups working on embodied intelligence, robotics applications, key upstream and downstream technologies, and other frontier hard-tech fields. The program’s stated operating approach is "understand technology and industry, spot opportunities early, make decisions quickly, value people, and avoid unnecessary interference." It is designed to provide both industrial resources and capital support to promising early founders in embodied intelligence, robotics, and related frontier hard-tech sectors. The announcement did not disclose financial terms or a timeline for the program.

800
Sequoia China and Unitree launch early-stage hard tech investment program
Google
Oct 28, 2026

Google DeepMind unveils Genie 4, a real-time world model with audio, dialogue and longer memory

Google DeepMind has introduced Genie 4, its latest general-purpose world model, positioning it as a step beyond video generation and toward real-time world simulation. According to the company, Genie 4 can generate interactive environments from a text prompt, update scenes on the fly, and keep a coherent world running for about two minutes at roughly 24 frames per second. The model also marks the first time the Genie series can natively generate audio and support improvised character dialogue. The release highlights a broader technical shift inside Google’s AI stack. DeepMind says Genie 4 is built as a hybrid world model that works with Gemini 3.8 Flash, Gemini 4 Argon and Gemini Robotics. In the examples shared by Google and early testers, the system handled dynamic scene edits, more realistic human behavior, stronger rigid-body physics, and multi-step robotic tasks, including checkers moves and lab manipulation with a UR5 robotic arm. Genie 4 is now in limited testing, and Google AI Ultra users can apply for the waiting list. Google also said the model will become the first world model available through the Gemini API and Google AI Studio, opening it to developers.

880
Google DeepMind unveils Genie 4, a real-time world model with audio, dialogue and longer memory
RoboParty
Oct 27, 2026

RoboParty open-sources humanoid robot RP1 with mechanical designs and motion stack

RoboParty has officially released RP1, a fully open-source humanoid robot platform that spans both hardware and software. According to Techub News, the project opens access to mechanical drawings, the electrical system, in-house joint modules, low-level control, simulation training, and deployment tools, giving developers a complete R&D path from hardware to software. RP1 stands 135 centimeters tall and has 24 to 30 degrees of freedom. The company has already demonstrated several real-world motion capabilities on the robot, including continuous walking, motion tracking, disturbance recovery, and getting up after a fall. RoboParty also open-sourced hhtools for human motion retargeting, along with URDF/MJCF models and related resources, while building PartyOS as an open development base. The company said it wants RP1 to serve as an open R&D platform for research institutions, universities, and embodied model companies, with the goal of lowering hardware adaptation costs and improving development efficiency. The item was cited by Techub and attributed to QbitAI.

620
RoboParty open-sources humanoid robot RP1 with mechanical designs and motion stack
Brian Armstrong
Oct 27, 2026

Brian Armstrong co-founds Warp Sciences as AI drug startup raises $215 million Series A

Coinbase co-founder and CEO Brian Armstrong said he has co-founded Warp Sciences, an AI and robotics company focused on drug research and development. The startup has raised a $215 million Series A led by Lightspeed Venture Partners. According to the announcement, Warp Sciences aims to move cutting-edge AI from the browser into the lab and build scientific R&D capacity in the United States. Armstrong is joined by a founding team that includes former U.S. government Chief Data Scientist DJ Patil, FAIR co-founder Manohar Paluri, and drug discovery specialist Sol Reisberg. The update was shared via Armstrong’s account and was cited by Techub News.

470
Brian Armstrong co-founds Warp Sciences as AI drug startup raises $215 million Series A
Total 251 items
Go topage