Autonomous Driving

Today's Stories

1. 1

   CoreWeave launches Physical AI Field Engineering service

   CoreWeave Inc. launched Physical AI Field Engineering, a service pairing its specialized engineers with customer teams to train models on the customer's own data. It says it has had more than 100 engagements with early adopters. The service aims at a talent gap the company describes in industries such as aerospace, where domain engineers and AI developers rarely share the same physics background. CoreWeave says the approach can cut testing times by between 17% and 35%.

   Early results are the test, not the pitch. CoreWeave says work for Nissan Motor Co. on archived test data cut physical testing times by 17%, and an unnamed automaker completed a key engine calibration step in 24 hours instead of three months.

2. 2

   Is Renesas Undervalued on Autonomous Driving AI Deal?

   The article raises the question of whether Renesas Electronics (TSE:6723) looks undervalued given its autonomous driving AI partnership, without yet announcing a conclusion. This framing suggests the partnership could reshape how the company's valuation is judged, potentially positioning it as a bargain relative to peers in the automotive chip space.

   The answer hinges on whether the market begins pricing in the partnership as a growth driver. Investors should watch for further details or financial figures from Renesas or its AI partner.

3. 3

   AI's early impact too indirect for average people, Interconnects warns

   The Interconnects author argues average people see AI as fringe or confusing, citing the OpenAI-HuggingFace incident as an example of news that doesn't resonate. Unlike the First Industrial Revolution's tangible goods, AI's current benefits are indirect, risking a political backlash that could kneecap development, similar to American nuclear power.

   The author suggests robotics and self-driving could make AI benefits tangible. The diffusion of AI into daily life may take 50 years, while adoption of integrated AI agents is just starting.

4. 4

   Pony.ai CEO James Peng: robotaxis widespread within five years

   Pony.ai founder and CEO James Peng said at the Fortune Leaders Forum in Macau on Sept. 8 that robotaxis will be common worldwide within five years, asserting the technology problem is essentially solved. Pony.ai averages about 25 rides per day per vehicle across commercial services in Beijing, Guangzhou, Shenzhen and Shanghai, showing robotaxi use is expanding beyond tech enthusiasts to diverse riders.

   Pony.ai's expansion hinges on regulatory environments and taxi market costs in new cities. Watch its Uber partnership to deploy more than 2,000 robotaxis across Europe and 200 to South Korea by 2028.

5. 5

   Lightsage raises $4M to help software makers target AI agents

   Lightsage announced it raised $4 million in seed funding, led by Nexus Venture Partners, to help software companies sell to autonomous AI agents. Software developers can no longer rely on marketing to humans alone; they must target AI agents that choose software on behalf of users.

   Lightsage's platform simulates how agents shop for software to show clients why their product is chosen. Watch whether it expands beyond developer tools.

6. 6

   Google's Gemini 3.8 Flash Aims at Jobs, Not Just Answers

   Google started general availability of Gemini 3.8 Flash. It is built for long software engineering, autonomous agents, and complex enterprise workflows. This marks a move from one-shot Q&A to delegating whole tasks. The model splits reasoning, calls tools repeatedly, and verifies results.

   Gemini 3.8 Flash is the default for Google Managed Agents and Antigravity SDK. The test is whether it handles long jobs reliably.