This AI Is Already Fooling People on Video Calls Into Thinking Its Human, Company Says
Related
Goldman Sachs: The global humanoid robot market will grow from approximately 20,000 units in 2025 to 1.4 million units in 2035.
According to Mars Finance, Goldman Sachs stated that another path for AI to enter the real world is through robots, autonomous driving devices, drones, and intelligent industrial equipment. Goldman Sachs calls this "physical AI," which is far more challenging than text generation models because machines must not only understand language and images but also handle gravity, friction, materials, temperature, motion trajectories, and safety constraints. Humanoid robots are the most anticipated area. Goldman Sachs predicts that the global humanoid robot market will grow from approximately 20,000 units in 2025 to 1.4 million units in 2035. This demand is based on labor shortages: the US manufacturing sector employs approximately 13 million people, with a shortage of over 1 million material handling jobs. Goldman Sachs expects widespread commercial deployment of humanoid robots to not occur until 2027 to 2029. (Cailian Press)
Is the first humanoid robot company to go public on the A-share market coming? Unitree Robotics' IPO registration on the Science and Technology Innovation Board has been approved, with plans to raise 4.202 billion yuan.
According to Odaily Odaily on July 6, the Shanghai Stock Exchange website shows that the IPO application status of Unitree Technology Co., Ltd. on the Science and Technology Innovation Board has changed to "registration effective". It is reported that Unitree Robotics plans to issue no less than 40.4464 million new shares in its IPO, accounting for no less than 10% of the total share capital after the issuance, with a planned total fundraising amount of 4.202 billion yuan. According to the prospectus, from 2023 to 2025, the company's operating revenue was 159 million yuan, 393 million yuan, and 1.699 billion yuan, respectively, and its net profit was -11.1451 million yuan, 95.4747 million yuan, and 278 million yuan, respectively, making it one of the few profitable high-performance general-purpose robot companies globally. (Bianews)
Sun Guoliang of Muxi Technology: Orders for some products are already booked until next year or even later; the MXC600 chip series has already been shipped in large quantities.
Mars Finance reported on July 8th that during the "Vibrant China Research Tour" in Shanghai, Sun Guoliang, Chief Product Officer and Senior Vice President of Muxi Technology, revealed that orders for some of the company's products are already booked into next year and beyond. The MXC600 chip series is the company's main product this year and is already shipping in large quantities. Regarding the contradiction in the market between "hardware shortages" and "concerns about computing power surplus," he believes that currently, general-purpose, easy-to-use, stable, and reliable computing power is scarce, even extremely difficult to obtain. The surplus computing power refers to computing power applicable to some non-general-purpose architectures, capable of running only previous-generation models or certain niche models, which constitutes a certain degree of overcapacity. (Science and Technology Innovation Board Daily)
Goldman Sachs maintains Nvidia's $285 price target, valuation already reflects ASIC market share risk.
According to BlockBeats, on July 7th, Goldman Sachs maintained its "Buy" rating and $285 price target for Nvidia, stating that the stock's current valuation already largely reflects the risk of market share loss due to its self-developed AI chips and increased competition. Nvidia has recently underperformed the broader semiconductor sector. While chip stocks generally rebounded on Monday, Nvidia's gains were limited; year-to-date, its performance has also significantly lagged behind AI hardware companies like Micron, AMD, Intel, and Marvell. The main market concern is that major customers like Alphabet and Amazon are pushing their self-developed ASIC chips to third parties while still purchasing Nvidia GPUs. Meanwhile, increased CPU investment in AI workloads is also giving AMD and Intel more growth opportunities. However, Goldman Sachs analyst James Schneider believes that Nvidia's risk discount is already too large. He expects that even with some market share gained by ASICs and some incremental growth from competitors, Nvidia's revenue could still achieve strong growth next year. The Vera Rubin platform, which will enter mass production in the second half of the year, will be key to determining whether the company can widen the performance gap again.
Analysts: Samsung's better-than-expected earnings report is already priced in; investors are concerned about the sustainability of AI.
According to Mars Finance, on July 7th, Samsung Electronics issued a better-than-expected earnings forecast: the company's operating profit increased by over 1800% year-on-year, with single-quarter profits exceeding the total of the previous three years. Simultaneously, the company's revenue also increased by 129% year-on-year, reaching 171 trillion won. However, investors seemed unconvinced, and Samsung Electronics' stock price fell sharply after the South Korean stock market opened, with a maximum drop of 8%. Analysts attributed Samsung's weak stock price to some overly high market expectations: after including employee bonus provisions, profits driven by record memory chip prices could have exceeded 90 trillion won. Furthermore, the market is also concerned about a potential slowdown in the construction of AI data centers. Albert Yong, Managing Partner of Petra Capital Management, stated, "Samsung's strong performance had already been widely anticipated by the market and largely priced in during the pre-earnings stock price increase. Investors remain concerned about the sustainability of the AI boom and the risk of a potential slowdown in AI infrastructure spending by major US technology companies."
Anthropic: The Claude model contains elements similar to "conscious human thought."
BlockBeats reported on July 7th that Anthropic released a new research report discovering a spontaneously generated internal "global working space" called J-space (Jacobian space) within the Claude model. This is a dedicated set of neural activation patterns for the model to engage in silent thinking, allowing it to process concepts without writing them down, similar to conscious thought that humans can report. The research used Jacobian technology to identify neural activation patterns in the J-space, enabling it to read unspoken concepts and modify these patterns. Experiments showed that disabling the J-space weakens multi-step reasoning abilities but does not affect basic tasks or factual recall. Click the original link below to join the Beating · Lark AI news channel for 24/7 monitoring of global AI hotspots and news.