Tesla has a unique advantage when it comes to making a self-driving car, but it might not be what you think it is.
I hear a lot of people in the media talk about self-driving cars like they’re already here. It’s the typical hype cycle you see around an exciting new technology. They throw around terms like AI, Machine Learning, Deep Learning, Neural Networks and Computer Vision like they have any idea what they mean. I’ve been in tech for about 20yrs, and I don’t know what they all mean either…
I wanted to learn more about this, I called up a fellow online author and Dr. Computer Vision Satya Mallick and thought I would ask him.
AI is the overall field, Machine Learning is a subset of AI, and Computer vision is a subset of that, but also has other parts. So the goal is to teach a computer to “see” but how does that work?
Dr. Mallick explained that there are several methods however the data-driven approach with Neural Networks have dramatically improved the power of this technology.
The key, as you probably guessed, is in having enough data to teach the computer how to see. Of course, Tesla has thousands of cars collecting this data every day, so they’re ahead of the game in this area. But is there more to the story?
Tesla also has an advantage because the images they’re getting from the cameras on the car are automatically labeled based on the driver’s behavior. This is a distinct advantage that Tesla has over all others in this space.
Okay, so Tesla has the data, the key ingredient, obviously a strong team of engineers, and a reasonable roll-out strategy. Why isn’t it here yet? Why do I have to be so incredibly burdened by using my hands to steer the car? Of course, it’s the humans getting in the way again.
So what’s next, once this tech is here what will our future look like? Will we still need roads?
Let me know in the comments below!