Lately, it’s been impossible to avoid running into some kind of news about self-driving cars on a daily basis. Either it’s about the Google self-driving guru who just quit, or Tesla’s autopilot killing people or headlines that read “Self-driving cars? Who gives a crap?”
Well, regardless of what anyone thinks, self-driving cars are almost here. Fully autonomous cars will be ready for everyday use much sooner than we think. And their emergence will mark the beginning of a pivotal shift in society–a true disruptive force–changing the way the economy works. But, the biggest concern that I have, the thought that keeps me up at night is: “Why can’t I sleep?”
I used to think that it would be decades before self-driving technology would be ready for prime time. I mean, think about how complicated self-driving is. Humans can barely drive cars properly and we’re pretty intelligent–actually, that’s not true–we’re idiots. How long would it take a machine to be just a little bit better than being an idiot in order to drive itself?
There is so much information to process when someone is driving a car: avoiding obstacles on the road, monitoring what other cars are doing, or thinking about cutting off the car in an adjacent lane during a momentary fit of road rage. How can a driverless car possibly determine the proper execution of road rage? Flash the lights? Honk? Side-swipe? All three? These are some of the most challenging and complex issues that Google engineers need to tackle.
I thought that solving these insanely tough problems might take many many years, but when I came across this article, I started to think about all of it differently. The author points out that we could very well be standing at the cusp of what is a huge advancement in artificial intelligence, about to surpass anything our short-circuited brains can comprehend.
As the charts below show (taken from here), humans are terrible at perceiving progress.
We tend to think of advancement as being a simple straight line sloping upwards, when in reality it’s more exponential in nature. As human beings, we are incapable of seeing things for how they truly are. Our perception is so distorted that we think all 2,345 of our Facebook friends love us, when the truth is–they all wish we were dead.
Here is a way to think about it. Let’s say a Kia Soul is artificial intelligence. And this Kia is waiting to take off at a drag strip. You’re standing at the quarter-mile marker of a drag strip with your eyes peeled on this Kia that’s about to take off. From that distance, it looks like it will take forever for the Kia to get to you, but as it picks up speed, it approaches you quicker and quicker. And before you know it, artificial intelligence will fly past you at the blistering speed of 42 miles per hour, well on its way to 42.5 mph.
Alright, maybe the drag strip analogy wasn’t as good as the chart below that technologist, Jeremy Howard, presented in a talk recently. The point is that technological advancement is occurring at an increasingly accelerated rate.
It may well be that advances in artificial intelligence are continuing to gain momentum and it’s probably much more developed than what the general public might be aware of. You might be wondering at this point, what does any of this have to do with cars and self-driving?
Because self-driving, at its core, is powered by machine learning and artificial intelligence. All this just means that the technology could almost be ready for driverless cars to become a reality very soon. Tesla is already proving this to some extent and Uber just made an announcement recently that they will be launching self-driving cars on a trial basis by the end of the month!
Aside from the philosophical theorizing in the article, take a look at just how many companies are involved in this race to build self-driving cars. It’s kind of shocking. There are 30+ major car companies, automotive suppliers and technology companies working in some capacity to make these things come to fruition.
When there are this many companies putting in a humongous amount of resources into something, you can be sure that progress will be made very quickly.
What’s the key to making driverless cars happen? A Beeellleeeon dollars. Toyota recently made a $1B investment in artificial intelligence and GM recently bought Cruise Automation for $1B. I’m sure other companies are spending many more billions.
Ford, GM, VW, Toyota, Nissan all seem to think autonomous vehicles will be here in 4 years in 2020. Ford has already stated that it will launch a fully autonomous vehicle in 2021 and it’s probable that Google and Tesla will be rolling them out even sooner.
Audi is working on an autonomous A7, named Jack, which is going through extensive testing and is currently being trained to act like a human driver. This idea is so bad that it makes the decision to adopt a skunk for a pet seem like a perfectly sensible thing to do. What will Audi do? Teach Jack to cross over 5 lanes of traffic in a split second to take that exit? Start texting, not pay attention and rear-end someone?
It’s funny to see what tech giants in Silicon Valley think of this whole self-driving business. Mark Andreessen, famous Silicon Valley venture capitalist and investor says “Right now the phone is an accessory to the car, but soon the car is going to be an accessory to the phone.”
Oh great–the car will be nothing more than an extension to a smartphone. Apple, our favorite smartphone manufacturer who has contributed to an increasing downward curvature of the head, or “text neck” as some people like to say, is also working on a secret car, which will be the greatest accessory to the iPhone the world has ever seen.
Everybody wants a piece of the action. Nobody wants to be left behind, especially incumbent giants like Ford and GM.
So, yes, these accessories to phones are coming soon. And it’s not just Tesla. When these vehicles begin to slowly populate the planet, rest assured that human driving isn’t going away any time soon. It will be many years before regulations are changed because the government’s number one objective is to prevent any progress from ever taking place. Dealing with the most expensive and sophisticated accessory to the phone will confuse the hell out of government officials. In the meantime, we can take advantage of all that confusion, and enjoy driving for as long as we can. Greatest accessory to the iPhone be damned!