You’re Gonna Flip Over This Robot
Two years ago, the best bipedal robots couldn’t climb stairs or open doors. Watching them calculate, quiver and then tumble was hilarious.
Now, Boston Dynamics has one doing backflips!
Last week, the company showed off its latest iteration in the development of robots that walk like people.
They call it Atlas.
After some software enhancements, Atlas now has an uncanny sense of balance. It can jump up on a stand, then hop to face the opposite direction. And, yes, Atlas can even perform backflips like a gymnast.
Scary, sure. But don’t miss the bigger picture. This is an opportunity for investors.
We humans have been trained to fear robots, especially the ones that look like us. In books and movies, they have this pesky habit of blowing us to smithereens. In reality, they represent what is possible.
In August 2011, on the heels of Great Recession, pioneering computer engineer Marc Andreessen, by then a successful venture capitalist, penned “Why Software is Eating the World.”
|The robot Atlas prepares to do a backflip. Source: Boston Dynamics via Youtube.com.|
The Wall Street Journal essay chronicled how software had crept into our daily lives and changed them forever. From music and movie files, to electronic commerce, bits of code blended with data disrupted everything. Andreessen, wisely, predicted we hadn’t seen anything yet.
A decade later, software, juiced up with artificial intelligence, has eaten through culture, commerce and the better part of our free will.
Google (GOOGL) and Facebook (FB) use software robots to catalog our interests, friends and sometimes feed our worst angels.
|Alley-oop! It’s up and over for Atlas. Source: Boston Dynamics via Youtube.com.|
Algorithms collect detailed, three-dimensional maps of our faces. It’s the small price we pay to unlock our iPhones faster. Besides, what could possibly go wrong?
And like the stumbling robot-turned-gymnast, everything is getting smarter, faster.
It’s one of the features of the New Gilded Age. The world is awash in networks of powerful computers.
Information technology is progressing at an exponential rate because pools of computing power have become ubiquitous and inexpensive. Modeling and massive simulations can solve problems that once took months or even years to answer.
|Atlas sticks the landing! Source Boston Dynamics via Youtube.com.|
At this year’s GPU Technology Conference – or as techs call it, “the GTC” — Jensen Huang, CEO of Nvidia (NVDA), explained how his company taught a software robot named Isaac to play hockey.
Huang’s researchers first constructed a virtual, photorealistic alternative universe that obeyed the laws of physics. There were ice, pucks, hockey sticks and even movable nets. They added virtual sensors and let Isaac learn by trial and error.
Then they duplicated the robot hundreds of times and continued the simulation.
Because the universe was virtual, there was no limit to the number or the speed of the simulations. Isaac was continually refined and, very quickly, mastered hockey.
Huang believes the same massive-simulation learning program will have profound implications for healthcare. He sees robotic surgeons that will develop skills far superior to humans.
And the process has already yielded impressive results for self-driving cars. In October, Huang told the Times of Israel he expects fully autonomous taxis will be in operation as early as 2020.
These developments are impressive. They are also indicative of the amazing opportunities available to investors right now.
In addition to technology-centric ideas, I have been pointing my members to the forgotten beneficiaries in logistics, heavy industry, and financial services.
Flashy self-driving Teslas will share the road with modified delivery vans and truck fleets. Tractors and earthmovers have been autonomous for a generation. Refinements in software will add value to those products.
Bots and blockchains will cut billions of dollars in regulation bloat and the headcounts at forward-thinking banks, brokers and insurers.
The advance of software robots is materially changing the business models for a number of companies. Many of the ramifications have not been fully reflected in the stock prices.
When bipedal robots struggled with simple tasks at the DARPA Robotics Challenge in 2015, critics could not contain their laughter. They looked at the collections of aluminum, copper wire and sensors and surmised a solution must involve a complete redesign.
They missed the story. It was a software challenge. The smart stuff was inside.
Software at companies like Square (SQ), John Bean Technologies (JBT) and Delphi Automotive (DLPH) is getting smarter at an exponential rate.
Investors need to get ready now. The members of my service The Power Elite have known that for a while.