Waymo recently published a study outlining the importance of large data sets for improved autonomous vehicle performance.
The last few years of AI performance have been powered by scale. It has been repeatedly shown that the performance of deep learning models scales predictably as we increase model size, dataset size, and training compute. These scaling laws drive continuous advancements in large language models (LLMs) in particular, as evidenced by the increasingly capable AI systems we see emerging regularly.
The post is hard to read and features inside baseball terminology, but the results clearly suggest that larger data sets are helpful. Specifically, “Closed-loop performance follows a similar scaling trend. This suggests, for the first time, that real-world AV performance can be improved by increasing training data and compute.“ This certainly suggests that Waymo and Tesla have a huge upperhand for the future autonomy battles because of their enormous troves of data.
Leave a Reply