Goal for SystemX

We already have a self driving car. It’s called a taxi. If you are at all familiar with self driving , you have heard of companies like Tesla , Waymo, comma ai etc and the work they have been doing. Elon Musk last year had promised full self driving or FSD. Tesla currently as of this writing (May, 2020) is winning with Level 2 autonomous driving. But, to actually solve self driving , we need to solve or atleast understand how human behaviour works. The Human factor on both ends still remains a challenge. Human driving is not great, (i.e, if you look at the overall quality of driving). Solving a system 2 problem, will require an understanding of human bias/intention. Also, the bias of system 1 can be overruled by System 2. Deep learning is taking care (perception problems) of the fast, automatic aspect but System 1 on its own isn’t going to be enough. We need more. We need a combination of two systems that can deliver performance. System 1 which is automatic and System 2 which is logical and calculating. Kahneman focuses much of his book on the interactions of System 1 and System 2. System 1 can be learned over time , but System 2 is much more slower, making it difficult to attain. The problem of Static Driving can be solved by system 1 , but solving dynamic driving and human intention will require system 2.

“So, you are all welcome to join this extraordinary journey as we try and figure out a combination of System 1 and System 2 to take self driving to level 5. The end result can be anything…!!”-