On Automotive Electronics

Excerpt

A few months ago, there was an explosion of news about the dangers posed by artificial intelligence (AI) following an open letter signed by thousands of experts, who warned that AI technologies such as GPT-4 present "profound risks to society and humanity." GPT-4 is a large language model that lets you have human-like interactions with a computer. It can answer questions, compose emails, write reports, and even software code. The worry, according to the letter's organizers: "AI systems are growing ever more powerful - and we don't know their upper bound. Malicious actors can use these systems to do bad things, and regular consumers might be misled, or worse, by their output." That made me curious about what risks AI poses to people in cars. Xiaowei Huang, director of the Trustworthy Autonomous Cyber-Physical Systems Lab at the University of Liverpool, is among a small group of people from the auto industry who signed the letter. In an email, he wrote that his concerns about AI extend to the auto industry. "We aren't sure if AI will run within our expectations. My concern is mainly on whether AI has been rigorously engineered to ensure its safety and trustworthiness. The engineering process would normally include falsification, verification, explanation, validation, and runtime monitoring. The key challenge will be how to organize these sub-processes to showcase to regulators and consumers AI's safety and trustworthiness."

…