Björn Bredehöft @2BSmart: "𝐖𝐡𝐚𝐭 𝐓𝐨 𝐊𝐧𝐨𝐰 𝐀𝐛𝐨𝐮𝐭 𝐍𝐯𝐢𝐝𝐢𝐚'𝐬..."

Germany

𝐖𝐡𝐚𝐭 𝐓𝐨 𝐊𝐧𝐨𝐰 𝐀𝐛𝐨𝐮𝐭 𝐍𝐯𝐢𝐝𝐢𝐚'𝐬 𝐀𝐥𝐩𝐚𝐦𝐚𝐲𝐨 Dear Investors, At CES in Las Vegas, $NVDA (NVIDIA Corporation) CEO Jensen Huang unveiled the next stage of autonomous mobility. With the Alpamayo platform, AI leaves the screen and takes the wheel in the physical world. Until now, autonomous driving has primarily been a problem of perception: Does the camera recognize the stop sign? Does the sensor see the cyclist? With the introduction of Alpamayo, Nvidia has presented a new level of technology. The platform brings AI directly into the vehicle, enabling vehicles not only to "see" through sensors, but also to truly "understand" these inputs. The technological core of Alpamayo is a so-called Vision Language Action Model. The AI no longer operates as a black box, as we know it from common LLMs. Alpamayo will be largely open-source, allowing manufacturers, insurers, courts, and all interested parties to understand the AI's decisions. It unlocks the AI black box in autonomous driving before it's even implemented as a black box. According to Nvidia, the system can analyze traffic situations and justify them internally. If a child is playing at the roadside, the system brakes not just because an object is moving, but because it anticipates the danger. Classic driving school example: “A ball rolls onto the road. A child could follow. I reduce my speed and increase my readiness to brake.” A logical sequence of thought that Alpamayo should be able to implement. As a business lawyer, I'm not just looking at the technology or the business model here: The biggest problem with autonomous mobility right now is the regulatory risks. What happens if the technology fails, or if ethical questions arise from complex traffic situations? For example, imagine a person suddenly stepping into the road. Swerving to avoid it would only be possible at great risk of injuring the driver, or would the car choose not to swerve and only brake hard? Who is liable in such a situation? The driver, the car manufacturer, or the software developer? This is where Nvidia's open-source and Chain-of-Though approach is truly crucial. In the event of an accident, the AI can not only display the data but also verbalize its intention and logic. This makes it possible to clarify the question of fault (e.g., misperception vs. incorrect decision). For the insurance industry and traffic law, this represents a significant change from the current situation. Decisions made by autonomous vehicles using Alpamayo become explainable and therefore controllable by regulators. But what does Nvidia gain from making the software available as open source, besides these massive regulatory benefits? By opening Alpamayo to all manufacturers, Nvidia is forcing the industry into its hardware infrastructure. Anyone using the thinking software absolutely needs the computing power of the new chips. Nvidia is thus building the operating system of mobility and securing significant shares of the high-margin hardware sales. Who could profit from this on the stock market? I believe there could be some beneficiaries, provided the software performs as well under real-world conditions as it promises. $MBG.DE (Mercedes-Benz Group AG): As its closest partner, the company is integrating Alpamayo directly into the new CLA-Class. This secures Mercedes a technological lead over its European competitors. $NVDA: The company is consolidating its monopoly and opening new doors into the automotive sector. To use familiar stock market analogies: Nvidia is no longer just supplying the shovels for the gold rush, but also writing the rules for how it's done. $ASML (ASML Holding NV): As a supplier for the extremely complex manufacturing machines required for the new generation of chips, the company remains heavily involved. Where are the risks? So far, it's just been a software presentation. The fact that a renowned automaker like Mercedes intends to use the system directly in its new CLA series is certainly a mark of distinction and suggests that the system has already been sufficiently tested in real-world conditions, but it's not yet proof of its suitability for mass production. Furthermore, there is very little publicly available information about how far along other automakers like $TSLA (Tesla Motors, Inc.) or Waymo ($GOOG) are technologically in this regard. If these companies are already using or have similar or even more powerful AI software systems in reserve, a fierce competition could ensue. While Nvidia is sitting on massive cash reserves due to the AI hype, if push comes to shove, it would likely invest this money in research and development for its core semiconductor business. Have a nice day, Björn Photo: Reuters

4 Mentioned

NVDA

NVIDIA Corporation

MBG.DE

Mercedes-Benz Group AG

ASML

ASML Holding NV

TSLA

Tesla Motors, Inc.

Strictly Necessary Cookies

Performance Cookies

Functional Cookies

Targeting Cookies