Predicting Human Decisions: The next frontier in AI

Engineering design inherently involves trading off competing objectives, for example, increasing performance while reducing power and cost. Human beings learn over time how to make these tradeoffs but still have to go through many design iterations to arrive at a cost-effective solution. As product complexities increase, the number of iterations goes up increasing design cost. The limited number of experienced engineers can limit productivity. Design decisions in the chip business, in particularly here at Qualcomm, can make a difference of several hundred million dollars because of the very volume of our market – smartphones. At Qualcomm, we are focusing on teaching machines to make these design trade-offs like a human being, but in much shorter time. Thus we can save design cost and produce more designs with the same number of designers. 

 This branch of learning is called reinforcement learning and at Qualcomm we are investing in R&D in this area. Developing Reinforcement Learning algorithms for usecases like chip design decision making is challenging, because the machine needs to go through several experiences to learn like a human. Creating realistic experiences can take a long time and be impractical, especially in our chip design problem. To this end, we are collaborating with Stanford AI research group to develop new reinforcement learning techniques that can learn from past human experiences without the need to generate new experiences. 

 

In this talk we will share some results with reinforcement learning in engineering design and demonstrate how it can reduce design costs by making design decisions like a human designer but in significantly shorter time.