Amazon’s Secret Self-Driving Car Project
The Borg strategy for overtaking the auto industry in one-fell swoop
I. Setting the Stage
Sometimes in life you have to read between the lines. Like our intelligence officer brethren, we look to tradecraft tactics to determine where things are headed with tech.
For example, drug czars have to pretend to live everyday lives in normal apartments and without the fancy accoutrement of the rich. But if they lock themselves into a compound, where a man never leaves the house after months of surveillance while others come and go, it’s a pretty good chance you’ve got a high-value target in your line of sight.
Similarly, in tech circles, if you see a major global company doing “all the right things” around the edges of a certain technology, it’s highly likely that they’re using misdirection to keep the leaks, press, and corporate espionage at bay.
Amazon is working on a self-driving car. Period.
As they say, where there’s smoke, there’s fire. Lets review the facts.
II. Self-Driving Facts
How about we start with the latest news. Amazon has positioned their future smart retail stores as using self-driving technologies. Every news organization reporting on it uses that language. Here’s a screenshot of the “amazon self-driving car” Google search:
As you can see there have been a few rumors but nobody is really talking about it. Amazon is investing in mapping technology, obviously in the machine learning software, and the sensors.
In December 2016, Amazon made its first delivery using a self-driving drone. It’s fully autonomous without human intervention.
As you can see Amazon is also involved in hardware. Now they typically outsource that. Amazon isn’t going to become a manufacturer of physical materials anytime soon. But they’ll darn sure slap their logo on commodity stuff out of China.
There’s also been a report that they’re talking to auto manufacturers to build their cars. Like Fiat. It’s a funny trend. Google is getting out of building their own cars. So is Apple. Amazon never even considered it.
Finally, CES was pretty much Amazon Alexa’s voice assistant being included into all sorts of products, mostly integrating with car’s infotainment systems. Namely because voice commands in cars are abysmal. Like touch screens before the iPhone came out. Instead of the problem being not registering a tap or chunk-chunk scrolling, with voice it’s about not having to enunciate to a ridiculous degree and repeating yourself.
III. Code Word: Amazon Borg
What is the ultimate strategy? We first have to start with the Car Operating System opportunity previously written. Much like Microsoft owned the relationship with people on Windows and Apple owned the relationship with people on iPhones, the question will be who wins the relationship with people in cars.
With Tesla hiring the inventor of Swift (the new Apple programming language for iOS devices) and putting him on the car software, my gut says Tesla is going to continue to invest in a closed ecosystem.
The 800-pound gorilla is Blackberry QNX which is pretty much the standard for separating the infotainment and secure parts of the car’s operating system. But otherwise most people are using Robot Operating System, an open-source OS for robotics and cars.
Many in the industry are also investing in the V2V and V2X technologies. A funny acronym that just means vehicle-to-vehicle communication. All that sensor data, mapping data, and deep learning algos that one car is learning on the road is also shared with other cars on the road in real-time. Let’s be clear that it’s not updating the intelligence algorithm on the fly, but rather some of the data surrounding that core software.
Looking at all of this through the lens of Hollywood gives us an apt analogy to help understand all of this: Star Trek’s Borg. It was an AI-controlled collective consciousness that made faster decisions, better. It’s the groupthink that the rest of the industry mostly shares.
So what does all this have to do with Amazon’s secret self-driving car project? Well, for one I woke up about two weeks ago with a crazy dream that is essentially this article. This phrase kept echoing through my mind:
Amazon Borg self-driving cars. Hardware 5 years ahead, software 2 years ahead.
Don’t you find it interesting that Amazon created a press release and some video footage for their drone project years and years ago, launched it on 60 minutes, are working on voice controlled hardware, mobile phones, AWS, a massive Prime shipping and online video service, e-ink readers, and a behemoth retail shopping business, but haven’t mentioned a peep about self-driving cars? Not a peep. And they’re one of the biggest, baddest, most valuable companies on the planet. And Bezos owns a freaking rocket company.
A rocket company that’s also using self-driving software. What’s the one thing they haven’t talked about? Cars. And just like tradecraft, if you don’t talk about it, but you talk about everything else, it shines a bright, bright light on it.
So, to the rest of the world who’s not talking about Amazon’s self-driving car project, code named Borg, well let Humanizing Tech be the first.
IV. What’s Coming Next
Based on The Base Code’s latest valuation model, it looks like 75% of Amazon’s current stock price is based on future expected growth beyond the next 5 years of free cash flow. It has a negative Economic Profit of -3% because its capital costs more than the return it is realizing from investing that capital into projects internally.
The stock market already has +5% annual growth built into Amazon’s value and price forever. That means it needs to grow by more than 5% every year, for the rest of eternity for it to be valuable enough for you to purchase a single $800 share today.
The story in the media may be one of rosy times at the ecommerce giant, but the hard and fast reality is that Amazon is in a crisis and fight for its financial life. It needs a big investment to start paying off in a big way. It can only operate on the float for so long.
Even with $18B of cash on its balance sheet, that’s not enough to start up a manufacturing arm for future cars. Its only hope for growth is to leverage the voice assistant tech and AWS into the operating system of the future. Namely, the self-driving and infotainment/security aspects of the car.
If it can do it better, faster, and run it like a cloud computing model, it might have a chance.
But everything we know from building self-driving software, and talking to people in the automotive and consulting industries, is that the shift is moving back from the cloud to on-premise. The “on-prem” part is the car, or more generally, device itself. A16Z gets it, and had an entire presentation dedicated to the end of cloud computing, which we also predicted based on our work on Biologic Intelligence for self-driving.
So, it’s the intelligence software that needs the lowest latency possible to make split-second, real-time decisions. That means you can’t communicate up to the cloud (especially for robotics in Space).
And ultimately, that means the AWS business will reach a fundamental limit that Amazon needs to begin preparing for now.
We’ve written exactly one year ago about how Amazon is building up a massive cash pile. In fact, they show that as financial priority #1 in every public earnings call with analysts and Wall Street.
What do you think they’ll use that massive amount of cash for? Buying ships and airplanes to augment UPS and FedEx shipping to meet demand? Of course. But also, self-driving cars.
Amazon Borg, by Bezos.
— Sean
