A few weeks ago, me and my dad emailed one of my cousins, who works at the Uber self-driving car department. He said he’d be happy to get interviewed by me, and we could face time each other. He didn’t even want my questions beforehand, which was great, so I could have more time to get the questions done. I finished writing my questions and they were all open ended, so I could get a lot of material for my research. I also put it on a video, so I could later go back to it to write some notes.

When we face timed, he told me mostly about what he did, which was something called “deep learning”. It was really confusing at first, but eventually he gave me an example that was really informative, which was that deep learning is like making the car familiar with its surroundings. For a regular human, if you see a dog, you’ll recognize it. So the next time you see a dog, you will know its a dog. But for a self driving car, it doesn’t do that. When it sees a dog for a second time, it won’t know it’s a dog like a human does.

He also told me about most of the level 4 self driving car features. He said that unlike a level 5 car, that can go anywhere, a level 4 self driving car only can drive within a limited area. First, people drive around the town or city manually, so the car can get the roads, lights, and other objects in its surroundings that don’t move.

He taught me a lot about the self driving cars Uber makes, and I could really use it to boost my research.