Welcome back to Inside the Mind of a NYC VC, a new series at AlleyWatch in which we speak with New York City-based Venture Capitalists. In the hot seat this time is Oliver Mitchell, Founding Partner of Autonomy Ventures. Oliver sat down with AlleyWatch to talk about gray collar jobs, businesses that are as essential as the air we breath, and why Black Mirror got it all wrong about those robo-bees.
If you are a NYC-based VC interested in participating in this series, please send us an email. We’d love to chat. If you are interested in sponsoring this series that showcases the leading minds in venture in NYC, we’d also love to chat. Send us a note.
Bart Clareman, AlleyWatch: Tell us about your journey into the Venture business and come to found AV?
Oliver Mitchell, Autonomy Ventures: I’ve been in startups for over 20 years. The fist company that I was involved in as an investor was Homes Protection which we sold to Tyco, which later became part of ADT, for >$100mm. Then we founded a financial services ATM company which we sold within 32 months to American Express. And my last startup was Robot Galaxy, a successful edtech company.
After Robot Galaxy I started writing as a syndicated columnist about the robotics space on my blog as well as on other outlets, and I started seeing the opportunities to invest. That portfolio has done very well, and I saw a big opportunity to create a focused venture capital firm around robotics, mobility and AI.
Your website says Autonomy Ventures is investing in the “Autonomous Robolution” – what is the Autonomous Robolution, when did it begin and where are we in the game (1st quarter, 2nd, etc.)?
We are in the early days. The best way to think about the Autonomous Robolution is to think of it as we’re in the fourth wave of the Industrial Revolution, and compare it to the previous third wave, which was the personal computing revolution.
We are standing about six years before a point in time where there was ubiquity with PCs on every desk within an enterprise. That moment, in my opinion, was Windows 95, which enabled people without programming skills to use computers for their own needs.
I think the companies that Autonomy Ventures is providing seed capital to are companies that in 5-6 years can become leaders in the industry, because compared to the personal computing age, we’re standing at about 1989-1990.
Say a word about your positioning – what makes Autonomy Ventures unique?
I think number one is the team. I’ve been an entrepreneur and a startup operator for over 20 years. My partner, Gary Lubin, comes from the corporate venture capital space, where he was a founding member of Merck Capital Ventures, a $100m+ fund focusing on medical devices and healthcare IT. And we have an associate, Lynn Greenberg, who joins us from Bloomberg. Together we have a very unique approach in sourcing, evaluating and managing portfolio companies.
Having a brain trust of diverse backgrounds is very valuable in what we bring to the table.
Another unique aspect of Autonomy Ventures is that we are one of the few micro venture capital firms that is focused solely on robotics, mobility and artificial intelligence. There’s a number of great generalist investment companies that might have a drone company and might have an investment in a robotics company, but they also have investments in a wide range of consumer companies and enterprise software companies, while we’re focused on solely nurturing great robotics companies.
To that end, we provide a lot of great resources to early stage founders in the robotics industry. We partner with them very early on to help them get to the next level. The companies that we source are on the cusp of commercialization, and we bring our decades of experience to really help them and propel them into the marketplace, whether through focused strategic partnerships or business development relationships, in order to take their prototype into a commercial application very quickly.
The third thing that makes us very unique is the sourcing. Both myself and my partner have been doing investments for a number of decades. We’ve built up a proprietary network that includes Israel where I’ve been investing since 1995, and have relationships in all the universities and the labs and the innovation hubs and other venture capital players there. And then here in New York we see ourselves situated at the center of what we call, the “Northeast Academic Corridor”, where we have relationships at every major university, not just in the tech transfer departments, but also in the labs. Our university relationships stem from Brown to MIT-Harvard to Carnegie Mellon and go all the way down to Washington, where the Department of Defense in robotics plays a very big role. And the third area where we source deals is in the Valley, where we have a partner, Mirko Kerschbaum, who comes out of the autotech space, and helps us evaluate both technologies in the auto space and boutique rounds.
Finally, our great advisory board makes us unique. They work with us in the sourcing and in evaluating companies, and also for who we want to plug in to mentor the companies and grow their businesses, whether finding great hires, whether helping them on go-to-market strategy, whether helping them on a board level. Often, larger venture capital players look to us to bring them deals within the robotics space where we can pass on those deals further upstream and participate in larger rounds.
Further from the website – you say, you seek patented hardware/software tech solving mission-critical problems. How do you define “mission critical”? What are the hallmarks?
I think the most mission-critical thing is air. Our goal is to invest in companies that are as valuable to their businesses as the oxygen that we breathe.
We’re not investing in consumer or inventory-led businesses, we’re not investing in anything invasive like surgical robots or defense-related. We’re investing in companies that ultimately streamline their businesses by replacing “dull, dirty and dangerous” jobs.
Warehousing is a great area where robotics has been both increasing margin and taking humans out of dangerous ways of forklifts and jobs that require them to climb up huge cartons and what have you.
In manufacturing we invest in enhanced manufacturing, inexpensive robotic arms, like Carbon Robotics, which I’m an investor in. Companies that not only streamline the manufacturing but give a greater product, and take a very tedious job like drilling screws into a board and providing a very consistent quality product, where the quality assurance levels are much higher.
We’re also improving people’s lives. A big area is aging in place, where robotics through telemedicine, through sensor data, through biosensor data, and artificial intelligence can enable people to live longer with greater dignity.
The energy sector in taking mechatronics innovations to improve safety and take humans out of the dangerous roles, whether that’s on a rig or in the field, and provide ultimately a safer energy product whether that’s oil or solar to the marketplace.
Drones – providing remote sensing data about agriculture and industrial sites that provide actual data. As an example, rather than spraying an entire field with pesticides, they can pinpoint a fungicide to a particular area and reduce the exposure to dangerous chemicals.
And then finally mobility, where right now the hottest area is around connected and autonomous cars. The latest investment we just did, Aurora Labs, does over the air software updates to keep cars secure, and at the same time working efficiently so it doesn’t take away from critical driving functions. Ultimately autonomous cars will make the roads safer and more accessible to more people because cars are underused assets.
For all the gains in efficiency and quality and so forth, there are of course major societal implications of such autonomy. To what extent does an investor or entrepreneur in this space need to concern themselves with those implications?
They should be very concerned. I’m an advisor to the Robo Business Conference in Silicon Valley, I wrote what I called the Roboticist’s Moral Imperative where I said if 1/3rd of the workforce in the world globally or the US is outplaced by automation, then we’ve failed. [To read more about Oliver’s thoughts on this topic, you can read his recent post on Robot Rabbi, “Universal Basic Income vs. Robot Taxes”.]
The real excitement that I feel for investing in this area – obviously there’s a financial return that I have to myself and to my LPs – but the excitement is really in the KPI of positive social change. How can we as investors in automation be effective in investing in change, but also protecting people who are the most vulnerable.
I think it breaks down into two levels. There are high-skilled jobs, which traditionally we call white-collar jobs, and there are blue collar jobs which typically go to people who are not as skilled or not as educated.
If you look at white-collar jobs, already there are robots in the financial services industry with robo-advisors. In media, the AP and Reuters implement robo-journalists. In law firms there’s been a lot of consolidation with AI and automatic docs and discovery, which is a huge profit center for them. But I’m not worried about those jobs, because the people who are being replaced there are flexible and are easily re-educated or retrained for new industries. Because there will be, and there already are, a tremendous amount of creative jobs that are created by robotics.
But, take warehousing. There’s over 800,000 members of the warehousing union, and as we know those jobs have been shrinking. As there’s greater margin pressure on ecommerce companies, like an Amazon, there’s greater pressure to automate and invest in automation.
What can we do about those jobs?
As a community, we can self-regulate. I would like the discussion to begin around that – where robot companies can partner with their clients to come up with some metrics where if a factory or warehouse, for example, is X percent automated, then they should start accelerating the pensions of the older workers, and then take younger workers and retrain them not as blue collar jobs or white collar jobs, but in a new class of jobs created by robotics, called gray collar jobs. These are the workers that are controlling the robots. These are not engineers from Stanford and MIT, these are trainees and technicians that will be servicing, managing and programming the robots in the field.
The way to think of it is, there’s radiologists, but then there’s technicians who manage the radiology in the exam.
How many of the younger blue collar workers can be converted into gray collar workers?
That’s a tough question. I can tell you, the UN reports that a third of the jobs worldwide will be displaced by automation. Michael Bloomberg said he thought that in 10 years 50% of New York City jobs might be replaced by robots.
As with any disruptive technology, there is going to be upheaval, and initially there is going to be a lot of jobs lost in that wake. But with that said there’s going to be a lot of opportunity in that younger generation that’s moldable and flexible that can be working in servicing and managing and deploying robots.
And one thing we haven’t even started is, going back to that Windows ‘95 analogy, what happens when robots are utilized by people who are not roboticists? All the applications today have been thought up by great engineers, but like the mobile phone app world, we now have apps that no one who was designing the iPhone ever thought of. So there’s new areas where automation and robotics and mobility can be applied to that we haven’t even touched the surface of yet.
I’ll give an example. I invested in an exobionics company, which went public. In Israel there was a competing company called Rewalk, and their exoskeleton was designed by a quadriplegic to help paraplegics . The point is, when someone comes to design with a unique perspective of their own, we’re going to see so much growth and opportunity that we can’t even estimate today.
So going back to your question, it’s very difficult to answer in isolation. But there will be jobs lost in that wake, and there will be new jobs created and there will be even more jobs created based upon personal robotics. You know, since the introduction of the ATM in the early 1970s, we have more bank tellers today than ever before, not less. So I’m hopeful.
You’ve been blogging as The Robot Rabbi since 2011. Your most recent post was about Robo Bees which can help pollinate our plants in view of a series of climate change-induced changes in honeybee populations and behavior that threatens the global food supply. So there’s a need for a solution.
On the other hand, the television series Black Mirror recently had an episode featuring robo-bees that a rogue hacker reprogrammed and used to kill people.
Long way of saying – robotics and AI scare people. Is that fear warranted? What can the tech community do to counteract it?
Look, sci-fi is great both for scaring people and exciting people about technology. I think that the concept of Blade Runner and Terminator/Skynet, etc., is so far removed from reality that I think that the idea of killer robots or robots going rogue is less of a concern.
However, there’s a petition in front of the UN to add to the Geneva Convention the idea of offensive robots – robots that actually carry guns and shoot at armies, robots used for warfare. Robots today are used in a defensive way for warfare, where like Iron Dome in Israel takes out missiles or a predator drone always has a human at the controls.
But the idea of a robot making a kill decision, based upon artificial intelligence, in warfare is a very scary premise. So there’s sci-fi and there is some truth to it. I don’t think people need to be as concerned about robots going rogue, but more concerned about humans abusing well-intended technology.
Artificial Intelligence and robots are only as good as the programmers. If you have nefarious programmers programming AI you have hackers. If you have nefarious engineers programming drones with guns you can have mass killing. In both of those examples, the technology was only the medium for killing, it was not the executioner. So the real issue is how do we police the humans from abusing the technology?
When do you estimate AI will reach human-level intelligence?
It’s hard to give a year for any of these things. Technology is not implemented like executive orders where one day it’s one thing and the next day it’s the next. It evolves over time.
With autonomous driving, it’s very foreseeable in the next 5-10 years to have hands-free driving lanes. And I think that would be very acceptable to the public. I think the idea of having cars drive themselves autonomously on the road has challenges for civil engineering.
Let me back up. The first subway system in America was done in Boston, in the 1890s. The subway was installed as a result of problems caused by traffic above ground, primarily electric trolleys. The trolleys were very popular so there was a growing gridlock of them backed up in downtown, as well as an increase human accidents from the faster bus. They saw in London that they were implementing an underground railroad, and if you think about it they redesigned, from a civil planning standpoint, how people would be transported around the city.
I think with autonomous transportation, it’s not just that our road systems today will be acceptable for autonomous buses and taxis and cars like we’re seeing Uber testing in Pittsburgh or San Francisco or the other 10 autonomous driving sites. I think for it to really be safe and take off it will require us to rethink how we plan city infrastructure and likewise roadway infrastructure, which leads us into smart cities and smart roadways and infrastructure.
Looking at things in isolation like “when’s the date” puts aside all these other evolutionary factors.
You say the technology is going to be ready in 5-10 years, but it’s hard to imagine a municipality like New York being ready for autonomous driving in that timespan. How far behind are cities in catching up to what’s coming?
New York City is a smart city. There is someone in charge of it at the Department of Transportation, they’re testing smart taxis, they’re testing buses, there’s an internal high speed network from the Bloomberg administration that advances traffic lights for buses to allow them to move more quickly and stay on schedule.
I think that cities are advancing. Now, obviously, New York City is an older city with outdated infrastructure that millions of people use a day. For us to update things requires us to update other things, and there’s a lot of pressure put on the system. Younger cities like Pittsburgh, saw a big opportunity to become a hub. There’s many cities, mostly small cities, where they’re trying to attract technology companies to become testing platforms.
Going to when AI will be capable of a human decision-making functionality. I keep going back to 1997 when Big Blue beat [world chess champion Garry] Kasparov and everyone thought we were months away from artificial technology taking over for human beings. Then we had AlphaGo beat [18-time Go world champion] Lee [Sedol] in the Chinese game of Go.
I think it gets a lot of excitement, but what we have to understand is that while an AI can beat the greatest chess master, a robot still can’t tie shoes better than my 9-year-old. So we’re very far off from the idea of AI doing human-like decision-making.
But with that said, there will be specific program instances where an AI will be superior within the enterprise.
Talk about NYC as an incubator for robotics and AI technology – are great innovations coming out of here, or do they stem more from places like MIT or Carnegie Mellon or Stanford?
First of all, New York has one of the oldest and most innovative robotic companies in its backyard, Honeybee Robotics, based in the New Lab in Brooklyn. They have been part of the Mars Rover program since inception. Steve Gorevan, the CEO there, is a good friend of mine, they’re basically a professional services firm within robotics. They’ve done robotics for surgery, pipeline and of course Space So, we do have the talent.
Also, don’t forget that Bell Labs was right there in Holmdel, NJ, 40 minutes outside of NYC, so you do have this mentorship of engineering talent in the New York City area.
With that said you have amazing universities here. Obviously you have Columbia, you have Cornell Tech, NYU, but I also wouldn’t discount the state and city universities. In the basement of the Zahn Innovation Center I know of a marine drone company that is building drones for the harbor to repopulate our waterway with a million oysters.
Now, with that all said, certain geographies and business centers lend themselves more readily to different things. New York has a great amount of innovation around healthcare, because we have some of the best healthcare institutions in the world. There’s a great amount of AI around ad tech because we’re the center for advertising and media. There’s a great amount of computer vision software that comes out of Columbia University because their computer vision lab is second to none.
Beyond Autonomy Ventures, you’re also a member of NY Angels. For you as an investor, how does participating as an angel differ from participating at the venture level?
NY Angels is a wonderful, active angel group that has a selection of some of the most impressive executives and membership that I’ve ever encountered. I think that my venture capital firm is very focused around one segment, while the angel group is very broadly focused.
It’s still (sort of) the New Year period – what trends are you watching closely in 2017?
Coming out of CES there’s a number of big trends. I think CES is a bellwether of the year to come Technology around connected and autonomous cars is really exciting. Voice-enabled devices like Alexa and Google Home and even Siri being put in new and novel places, like the Wynn Hotels and washing machines and Ford cars, is a very interesting trend. At the same time, it raises the question of privacy. I am doing our RobotLab event on March 2nd about this, your readers can check Meetup for more info.
I think relevant to New York, the trends around retail and the convergence of IoT and the cashier-less store or restaurant is happening. At the National Retail Federation Conference last month that was one of the biggest themes. And the big overhang there is Amazon Go.
You have 14 patents is there one in particular you’re especially proud of?
It’s true I have 14 patents. I think that intellectual property is very important. When we look at companies we evaluate, we follow Howard Morgan’s rule and look at people, product and plan.
We first look at great people, their experience and their flexibility. Then we look at the product and that’s where intellectual property falls. Patents are important, but I think time in the marketplace could even be more important, how much time and how much of a leg up on the competition you have. I would hate to see a company take a large percentage of their seed round and invest it in lawyers, because it’s easy to file a provisional but to actually file claims and defend them can be very expensive. In the end, I think patents are only as good as you’re willing to defend them.