# Пов'язані статті щодо AI Data

Центр новин HTX надає останні статті та поглиблений аналіз на тему "AI Data", що охоплює ринкові тренди, оновлення проєктів, технологічні розробки та регуляторну політику в криптоіндустрії.

They Raised 17 Billion in Six Months: Data "Tool Sellers" Became the Most Profitable Business in the Embodied Intelligence Track

Over the past six months, data-centric companies serving the embodied AI sector (robotics) have secured over 17 billion RMB in funding in China, highlighting "data" as the most profitable niche. These "shovel sellers"—providing crucial data, models, and infrastructure for training robots—are flourishing despite robots themselves not yet being widely profitable. The surge is driven by a severe scarcity of high-quality physical interaction data needed for robot training. Companies are tackling this through five main approaches: 1) **Teleoperation/Haptic Data**: Building factories for high-precision data collection (e.g., Paxini, Noitom Robotics). 2) **Simulation/Synthetic Data**: Generating vast amounts of virtual training data (e.g., Lightwheel Intelligence, Transcend Dimension). 3) **UMI/Portable Collection**: Using wearable devices to capture human motions directly, bypassing robots (e.g., Jianzhi Robotics, Tashizhihang). 4) **Video Distillation/World Models**: Extracting actionable data from internet videos or using AI world models (e.g., Shutu Technology, Deep Genius). 5) **Data Infrastructure/Platforms**: Offering data processing, standardization, and platform services (e.g., Wuwen Zhike, Yiren Technology). Key players include LiberAI (founded by a 00-year-old PhD), which focuses on human UMI data and world models and recently raised hundreds of millions, and Lightwheel Intelligence, which became a unicorn and secured large orders. However, the industry faces a critical dependency: most data buyers are unprofitable robotics startups relying on venture capital. Data company revenues essentially redistribute this investment. Long-term viability likely belongs only to companies that become essential industry-standard platforms or masters of continuous, high-quality data supply for real-world deployment. The boom is real, but its sustainability hinges on the success of the broader embodied AI ecosystem.

marsbit21 год тому

They Raised 17 Billion in Six Months: Data "Tool Sellers" Became the Most Profitable Business in the Embodied Intelligence Track

marsbit21 год тому

Is the iPhone Moment for Embodied AI Coming Soon?

Is the "iPhone moment" for embodied AI approaching? This article, based on a roundtable discussion, presents expert insights on the current state and future of embodied AI. The consensus is that the pivotal "iPhone moment" is still distant. The field is likened to the "brick phone" era, with technology paths—such as VLA and world models—not yet converging. While robotic "motor skills" (e.g., walking) have matured, the "brain" (decision-making, generalization) remains far from commercial readiness. A major bottleneck is data: an estimated tens of millions of data points are needed for a breakthrough, but only around 500,000 currently exist globally. Currently, cost remains prohibitive for widespread labor replacement, making the economic case challenging. However, experts see a three-tiered market potential: a billion-level market for emotional companionship (e.g., entertainment, basic care), a trillion-level market for commercial services (e.g., guides, receptionists), and a massive, long-term opportunity for physical labor in factories and homes. The discussion suggests that while humanoid robots face hurdles, non-humanoid embodied AI applications (like existing service robots) can be deployed sooner. The ultimate vision is for AI to operate seamlessly in the physical world, not just behind screens. Regarding AI tools, participants noted their widespread use for boosting efficiency in coding, research, and teaching. However, they warned against over-reliance due to risks of AI "deception" and the erosion of critical thinking, emphasizing that core judgment must remain with humans. In summary, embodied AI holds immense promise but requires significant progress in brain models, data collection, and cost reduction before achieving its transformative potential. Its development is expected to be gradual, advancing through specific use cases rather than a single explosive moment.

marsbit07/14 05:07

Is the iPhone Moment for Embodied AI Coming Soon?

marsbit07/14 05:07

How Did Its Biggest Rival, Yushu, Split Up Again?

China's humanoid robotics firm Zhiyuan has spun off another company, Mbee Technology, focusing on data collection and processing for embodied AI. Amid a severe "data famine" in the embodied intelligence industry, high-quality real-world interaction data is seen as critical as computing power was for large language models. While teleoperated robots provide high-quality data, the process is expensive. Mbee aims to scale data collection through "embodiment-free" wearable devices and offers a data governance platform and marketplace. However, as a Zhiyuan affiliate, Mbee must prove its neutrality to attract competitors. It faces competition from other data platform players like JD.com. The success of its ambitious goal to reach tens of millions of hours of data capacity by 2026 hinges on gaining industry trust and demonstrating the value of its standardized data supply.

marsbit06/23 10:10

How Did Its Biggest Rival, Yushu, Split Up Again?

marsbit06/23 10:10

Interview with the Founder of an AI-Focused Investment Fund: Forget the False Narrative of 100x Returns, What Crypto x AI Assets Are Worth Watching?

Interview with Austin Barack, founder of AI-focused fund Relayer Capital. He emphasizes prioritizing real value over hype, especially in the Crypto x AI sector. Key assets discussed include: * **Venice:** A private AI platform where user data isn't shared with model providers. Its unique two-token model (VVV and DM) provides sustainable, capped user incentives for computation. Rapid user growth (3M) and token usage (15x in months) suggest significant upside. * **Grass:** Sells high-quality, specialized web-scraped datasets to AI labs for model training. With ~$50M+ ARR and triple-digit growth, its ~$400M valuation is seen as undervalued. * **NEAR:** Praised for its superior cross-chain swap infrastructure via intents, positioning it as crucial infrastructure for AI agents and decentralized applications. * **Akash:** A decentralized GPU compute marketplace gaining traction, with visible usage on platforms like OpenRouter. * **Investment Framework:** Shift from generic "buyback-and-burn" narratives to analyzing the "net token value flow"—whether token holders genuinely capture value from the underlying business. Markets are consolidating around a smaller set of quality projects with strong fundamentals like Venice, Grass, and NEAR, making targeted investments more effective.

marsbit05/26 01:56

Interview with the Founder of an AI-Focused Investment Fund: Forget the False Narrative of 100x Returns, What Crypto x AI Assets Are Worth Watching?