AI News Roundup – China accelerates AI chip production, Taco Bell struggles with AI voice drive-through ordering, NASA using AI to predict solar flares, and more
- September 2, 2025
- Snippets
Practices & Technologies
Artificial IntelligenceTo help you stay on top of the latest news, our AI practice group has compiled a roundup of the developments we are following.
-
- The Financial Times reports on China’s latest push to produce a domestic supply of AI-focused processors. The country is seeking to triple its current output in 2026, with three semiconductor fabrication plants currently under development. The three plants are focused on manufacturing chips for Huawei, seen as a leader in AI hardware design within the country, though Huawei denied plans to build its own plants, leaving the exact ownership of the new plants in question. The new plants will also free up capacity at plants owned by Semiconductor Manufacturing International Corporation (SMIC), of which Huawei is the largest customer, allowing smaller companies to get their own AI chip designs manufactured. Competition in the AI processor space in China has grown rapidly due to U.S. restrictions on chips made by global AI leader Nvidia, and many in China see domestic AI hardware capacity as a national security issue. One investor said of the phenomenon: “necessity begets innovation.”
-
- The Wall Street Journal reports that fast food chain Taco Bell is running into issues with an AI-powered voice ordering system at its drive-through locations. The company has introduced the system at over 500 locations, but many customers have reported issues with the new technology, including software glitches. Some nefarious customers have also “trolled” the system through unrealistic orders such as some for 18,000 cups of water. Taco Bell’s Chief Digital and Technology Officer told the WSJ that the company was “learning a lot, I’m going to be honest with you,” with regard to the issues. The company may try a more hybrid approach, with the busiest restaurants relying more on human workers. The struggles, however, reflect the limitations of generative AI three years since the AI boom began, as many companies try to determine and revise their policies surrounding the technology.
-
- WIRED reports on a new AI model aimed at predicting solar flares. The model, Surya, was developed as part of a partnership between NASA and IBM, and was trained on data from the former’s Solar Dynamics Observatory (SDO), a satellite that observes and measures the sun. Developers at IBM used this data to create a “digital twin” of the sun, which can be updated as further information comes in from the SDO. Surya can also use data from other spacecraft that observe the sun, and in particular predict solar flares an hour earlier than traditional prediction models. Major solar storms have the potential to disrupt electrical and communication systems on Earth, making predictions critical to the continuous operation of modern technology. One lead scientist on the Surya project told WIRED that “our hope is that the model has learned all the critical processes behind our star’s evolution through time so that we can extract actionable insights.”
-
- The Boston Globe reports on the use of AI by New England universities to cut down on administrative costs. At Roxbury Community College, students and staff are being directed to an AI chatbot, “AskRoxie,” for common questions asked of student services or IT offices. The use of AI in education contexts has been extremely controversial, with an explosion of use of the technology by students for cheating on assignments, but many institutions of higher education are turning to the technology to save time and money in running day-to-day operations. Administrative costs have risen drastically in the past 25 years, and recent cuts to federal funding have led to a rush to adopt AI to ease repetitive or burdensome tasks. One analyst says that the use of AI by administrative staff “is a sign universities are moving away from the initial ‘panic’ surrounding the technology.”
-
- The Times reports on a AI-generated video produced by Japanese authorities depicting the consequences of an eruption of Mount Fuji. The volcano, a symbol of the country, is about 60 miles southwest of Tokyo, and last erupted over 300 years ago. However, an eruption of Mount Fuji in modern times would have devasting consequences for the world’s most populous metropolitan area. To encourage disaster preparedness among citizens, the government of Tokyo released an AI-generated video depicting the effects of an eruption on the metropolis, mostly in the form of huge quantities of volcanic ash that could shut down transportation systems and contaminate the water supply, warning that “the next eruption could occur at any time.”