In a recent interview with SAE International , our CEO, Daniel, talked about the idea of AI alignment and its importance in making autonomous vehicles safer and more efficient.
Daniel explained that AI alignment means making sure that an AI system behaves the way humans want it to. As AI systems get better, it's important to check that they are doing what they are supposed to do. While self-driving cars have improved a lot, Daniel pointed out that driving is still a subjective task that involves making decisions and judgments. AI systems need to understand what humans want and learn from data to make accurate predictions and plan routes.
When asked about the incident with a Cruise robotaxi in San Francisco, Daniel mentioned that machine-learning systems can have problems if they are not trained on the right data. He stressed the importance of having a safety plan and the challenges that companies face in creating one.
Kognic is the dataset management platform helping assemble efficient ground-truth data pipelines to create and optimize sensor-fusion. We work with companies like Bosch, Qualcomm, Continental and their OEMs such as Ford, Volkswagen and Volvo to develop and deploy ADAS/AD systems.
Looking into the future, Daniel predicted that Waymo will continue to expand where its self-driving cars can go, Aurora and Kodiak will do well in delivering things commercially, and Ford will become a stronger player in helping drivers. He also said that he is hopeful that more cars in the industry will have systems that help drivers at a Level 2.
Kognic's advanced way of understanding data from sensors and their commitment to making sure AI systems behave well help make self-driving cars safer and more efficient.
Read the full interview on the SAE blog here.