It seems every time I chat with one of ROBO’s dedicated robotics and AI experts, I’m reminded just how fast technology leadership can change and why diversification is key to helping investors capture the growth in AI. Recently, a short email from Henrik Christensen, PhD confirmed the vital importance of our approach to investing in robotics, automation, and artificial intelligence (RAAI)—including focused research and diversification.
Last week, Bill Studebaker (ROBO Global CIO) sent a note out to our strategic advisory team asking for their thoughts on the newest Silicon Valley unicorn, Scale AI. This three-year-old startup, run by 22-year-old Alexandr Wang, is now valued at over $1B. What the company does is key to enabling AI in a multitude of industries. Currently, behind every self-driving car or cashier-less Amazon Go store sit thousands of humans whose jobs are to train computers to see. These people look at pictures and label what they see: a cat, a tree, a stop sign, a six-pack of soda. It’s a long, time-intensive, and costly process. In the autonomous-car industry alone, companies spend millions of dollars each year hiring people to label pictures gathered from cameras in their vehicles. This data annotation is fed back into AI software so it can learn how to do the same thing.
What Scale has developed is a software solution that helps feed ever-hungry AI databases by improving and accelerating this process of data annotation and labeling. By assisting a large human workforce with this meticulous task, Scale is reducing the time to complete a previously hours-long process to just a couple of minutes. The potential is fantastic.
Already, companies focused on self-driving cars are on Scale’s growing customer list, including Cruise, Uber Technologies, and Waymo. But what sparked Bill’s interest in Scale wasn’t its impressive customer list, but rather its recent jump in valuation. Valued at an already impressive $90M a year ago[1], its newest round of financing values the company at around $1B. The numbers beg the question: why the valuation jump—especially when the reality of automated vehicles seems to remain in the somewhat distant future?
At first, the answer wasn’t so clear. Sure, as Henrik pointed out in his email, “There is no doubt data will make or break the present autonomous vehicles systems.” By increasing the speed at which data can be processed and translated to help cars see, Scale is helping to move the industry forward. Henrik went on to say that he believes driving without HD maps “is going to be the winner in the 5+ year perspective,” and Scale supports that shift and change. Henrik went on to make a couple of important observations about the complexity and shifting nature of how AI is being applied in commercial applications.
1. The autonomous vehicle sector is shifting rapidly.
Henrik noted four companies in the self-driving vehicle sector that are investing in the space: HERE (a developer of HD maps that is jointly owned by BMW, Mercedes, and Audi); UBER (which has a great process for map augmentation, but is finding getting data to be a challenge); Waymo (which has the most miles driven and a massive data set for annotation); and Lytx (whose solution for driver monitoring is installed in 200,000 vehicles worldwide in all USPS, UPS and Fedex vehicles). What’s so interesting about that list? Of the four companies on Henrik’s list, two are already Scale customers. In other words, as players in the space strive for leadership, they are constantly morphing to address their challenges. As stated in Hampleton Partner’s Autotech 1H2019 M&A report, “Disclosed venture capital funding to automotive start-ups totaled at least $8 billion in 2018 – almost six times the M&A figure. One thing is clear: deals like Intel’s behemoth acquisition of Mobileye, or Samsung’s plunge into automotive with its Harman pickup, will become more frequent as this industry continues its drive forward.”
2. Where and how technology is applied is unpredictable.
Scale exemplifies the fact that, unlike many technology solutions from the past, AI technology is highly flexible, so the ways in which a solution like Scale’s data annotation can be applied seem almost limitless. This is another catalyst for Scale’s higher valuation. While Scale has historically focused on autonomous vehicle software, it is now marketing its solutions to developers of all sorts of AI technologies. Scale recently added a large handful of companies that are specifically outside the automotive space, including OpenAI (an AI research company that uses Scale’s solutions for language processing) and Standard Cognition (makers of AI-powered computer vision software designed to enable autonomous checkout processes at brick-and-mortar retailers). As Scale’s Alexandr Wang recently said, “It takes billions or tens of billions of examples to get AI systems to human-level performance. There is a really big gap between the handful of giant companies that can afford to do all this training and the many that can’t.”[2]
These two realities both point to why a dedicated team is needed who understand and monitor the shifting tides of technology and applications and why diversification is so vital. First, the growth is just beginning and, as a result, there are no clear winners or losers in public markets. Second, because AI technologies have broad applications, the addressable market has an uncommonly high potential for growth and is rapidly changing. Investing in the best technologies and applications across the ecosystem provides the highest probability of long-term success
From the start, the ROBO Global Robotics & Automation Index has been designed to address both of these opportunities. Using a modified equal-weighted strategy, the Index strives for diversification across the entire value chain, across geographies, and across a variety of business models. Not only does this approach offer greater exposure to small-cap and mid-cap companies that, in many cases, are not well-covered by Wall Street, but it also addresses the constant shifting of market leadership of technology and applications of Robotics and AI.
Brilliant as they may be, even our Strategic Advisors can’t predict the future of AI. What they can do, however (and what they do so well), is offer valuable insights into their areas of expertise to help us create an index that can capture the greatest potential for growth for investors—no matter where that growth may be. Thank you, Henrik, for helping us keep our eyes on the future!
[1] According to venture capital data provider PitchBook.
[2] “Silicon Valley’s Latest Unicorn Is Run by a 22-Year-Old”, Bloomberg Businessweek, August 5, 2019