The Learning Model Of Google’s Self-Driving Car

It’s not uncommon to see a Google self-driving car around these days…if you’re in and around the Bay Area in California.

There’s a massive fleet of them driving around collecting data and improving their driving so that at some point they can take over the road. As my friend and I were cruising right behind one he told me this amazing story:

He was sitting in the passenger seat of a self-driving car when, for no apparent reason, it slowed down. He couldn’t understand why and looked around but could see nothing ahead of him. Suddenly out of nowhere someone stepped out from behind a car and crossed the street…

Somehow the car knew.

This shook him up because if he had been driving he would not have slowed and would not have seen the person until it might have been too late to stop.

Here’s the amazing thing. Google cars learn not just from their own data collection and experiences as they drive around and practice, they also accumulate the data and experience from every single other self-driving car on the road.

We already understand that practice and experience leads to mastery. We learn from our own experiences and supplement our learning by hopefully having a wise team of mentors around us.

I even feel grateful that I’m able to work with many different companies and so can learn about different industries, ideas, and products and so accelerate my learning.

And yet my own

Individual learning is nothing compared to that of the self-driving cars. If it was analogs, I would be able to absorb and learn from the experiences of all human beings on the planet every day.

Artificial Intelligence is a big thing in Silicon Valley and venture capital firms are investing heavily into it. The learning model of the Google self-driving car is one of the simplest ways to understand how powerful this can become.