Flint Capital’s Artem Burachenok tells CBR why investors are confident in AI, the next wave of the “software eats the world” trend.
AI is making waves in the tech industry, with everyone from Google to the smallest start-up trying to profit from artificial intelligence. Chatbots, roboadvisors, virtual assistants and smart cars have been brought out of the realms of science fiction and embedded in everyday life, with these technologies seemingly only scratching the surface of what AI can actually do. The promise of AI is a hard one to measure and an even harder one on which to make a bet on. But now, as AI begins to deliver and starts to disrupt, some are making huge bets on what they view as the future of technology.
Talk of unicorns and huge IPOs have fuelled the rush to invest in tech – with investors pouring money into innovators of all sizes, from established enterprises to those who start out in the basement or garage. So how have investors responded to the emergence of AI as an investment opportunity?
Editor Ellie Burns spoke to Artem Burachenok, Partner at Flint VC, to get his take on the investor outlook in relation to AI. With Flint Capital being an international VC fund focusing on AI, as well as mobile, SaaS, and security, Burachenok knows his stuff when it comes to betting on the garage guys and why C-3PO from Star Wars is painful for investors.
EB: Have you seen a shift in investor behavior with the advent of AI technology?
AB: The world has undoubtedly benefited from major technological advances in the past, such as social media, mobile and cloud, and some of them still continue to be extremely important.
We believe that AI is going to be the next shift in technology of similar importance, and that’s why we’re so excited. It’s also why we have made AI our focus since founding Flint Capital, back in 2013. Shortly after, around 2014-2015, AI reached a tipping point, and now we are seeing dramatically more interest and growing investor activity. AI investments accounted for roughly 5% of all VC activity in 2015. Moreover, investors now expect to see AI everywhere, and for some, it’s an essential piece of any investment.
EB: How are investors responding to the increasing excitement surrounding AI?
AB: Investors are responding to the favorable market and great exit potential. These factors are creating confidence for VCs willing to invest in AI companies, and that confidence has translated to more early stage seed investments and larger funding rounds overall.
Moving from 2010 to 2015, investments in AI increased almost sevenfold. And, in the first half of 2016, VCs invested $704 million in AI startups in 79 different deals, putting this year on pace to break records for funds and deals alike.
With all of this investment, VCs are also choosing wisely—since 2011, 60% of the AI companies that have made exits had VC backing, and there have already been five major acquisitions this year. Another interesting piece of what’s going on is an increase in mid-stage deals. Series C deals, for instance, climbed from a 7% share in 2014 to a 15% share last year. Of course, as individuals, investors are also excited about AI’s capabilities.
EB: What makes AI technology such an attractive proposition for investors?
AB: AI offers an opportunity for disruption of multiple industries through software automation, and as such, is an extremely attractive area of investment. The current AI buzz is like another wave of the old trend, captured by Marc Andreessen’s “software eats the world.” Think about all the amazing programs and efficiencies that modern software has created. From early computers to smartphones, software has changed the way we live. AI, particularly machine learning, has 1000% more power than the software we use now, which grants it the ability to make drastic changes all over the world.
Let’s take the field of healthcare as an example—AI is able to automate extremely sophisticated processes that previous software simply could not. For instance, with machine learning, we can teach computers to read MRI scans and identify tumors. Current software simply cannot do this, which makes machine learning a new tool and a new way to build software—one that can save lives. AI’s presence in healthcare will also change the role doctors have, as they will now be able to reduce their own analysis of scans and instead overlook AI. Doctors will be able to increase their efficiency, giving them the time to take on more difficult tasks, or cover more patients. This is the sort of disruption that not only has the potential for profit, but for system-wide change, which makes our investments in AI all the more exciting.
What makes all of this possible is the accumulation of vast pools of minable data, both personal and enterprise, in addition to recent advances in learning systems software, infrastructure and design. These conditions, combined with ongoing erosion of computing and storage costs, make AI more affordable.
EB: Do investors actually care about the technology – or is it all about profits and return?
AB: In general, VCs are financial investors. We love to believe that our modest efforts help to make the world a better place, but the partners who invest in our funds judge us based on profits. That being said, the only way for a VC to get a home run, in terms of profits, is if at least one of its portfolio companies achieves financial growth on a huge scale, and does so within 5-10 years.
To do that, you need to leverage a new technology that is powerful enough to let you create a new market, or to totally disrupt an existing one. AI is exactly that type of technology, and is definitely something that excites us.
EB: Is there a particular industry utilizing AI which is drawing more investor interest than others?
AB: Chatbots and self-driving cars have captured public conversation recently. However, if you look at the AI investment data, you might be surprised, even though I might have given it away earlier. Healthcare is leading the way in AI investments, followed by advertising, sales & marketing, and BI, as detailed by CB Insights.
AI is going to shine in any industry that benefits from pattern identification, which frankly, is a lot of them. Going back to healthcare as a model, we can give AI sets of images, from blood samples to MRI scans, wherein professionals have already identified tumors or other abnormalities. AI can learn these identification patterns very effectively. Any industry that can replicate a similar system is going to quickly see benefits from AI. In particular, risk management and insurance could benefit because their focus is finding patterns that can be used to estimate a client’s potential cost.
EB: Are there any aspects of AI technology which could potentially undermine investor confidence?
AB: AI is aimed at replacing certain human functions. If the technology is applied too early or too aggressively, it may undermine the confidence in the AI market. We saw this when Twitter turned Microsoft’s chatbot into a raging racist, and autopilot killed a Tesla driver. Additionally, cyberattacks on AI systems responsible for making important decisions create a huge risk.
Something else that can hurt AI is its own hype. Popular depictions of AI make people believe they are going to be interacting with computers as sophisticated as C-3PO from Star Wars, which is a far cry from our current stage. Realizing that can be painful, both for investors and consumers.
We’re at the very beginning of our journey, and only patient and consistent investors will be able to capitalize on this new wave.
EB: Is there increasing pressure for investors to find those elusive unicorns? How has investing strategy changed with big-money IPOs and so many tech start-ups?
AB: VCs were always after outsized returns. As opposed to other types of investment funds, VC economics can work only if they have at least one home-run in their fund portfolio, although more than one is always good. What changed is that an arbitrary mark of one billion dollars was glorified by the press and by some investors. This created an irresistible temptation, and sometimes even put pressure on founders to achieve that mark. That’s why we’ve ended up with so many unicorns, some of them premature.
The other driver of unicorn-creation is the fact that companies now stay private longer. This is happening because founders believe that staying private is better for them. On the other hand, though, there is a growing number of investors ready to sign really big checks, ones that historically were only available for public companies.
As IPOs were made later and later, new liquidity options appeared for employees and early investors. Consequently, the secondary market for private startup shares bloomed.
On the other end of the investment spectrum, a huge number of micro-VCs (or super-angles) emerged with their checks, ranging anywhere from $20 to $500,000. It’s getting easier for a team of 2 or 3 founders to raise an angel/seed investment of up to $1.5M, and to start building a business in the garage, dorm or a co-working space.
However, it’s getting harder for those startups to get their next round of funding, since neither the number of VCs nor their fund sizes have grown with the same speed as angel funding has. While that presents a difficult atmosphere for startups, more of them are getting to self-sustainability (basically being profitable or breaking even) with the seed funds they raised. From there, they are growing organically without the need for additional VC financing.
EB: What would your advice be to anyone looking to invest in AI?
AB: Don’t expect miracles from AI. It is just a technology (a powerful one, but still only a tool). AI is still in its infancy. Make sure that the problem that a startup attacks can be solved with the current state of their technology. Or, at the very least, the startup should have a clear roadmap to solve the problem in the near future. Don’t try to build Tony Stark’s Jarvis. Instead, focus on narrow and specific fields.
Should you back a team of experienced data scientists or smart guys in a garage? I believe we’re already at the stage where both options work. But, I’ve had better experience taking the second route. Many of the scientists in the industry are brilliant, but their process is relatively slow. Garage guys, on the other hand, are more agile and make things happen quickly.