Unsupervised Investments (I): A Guide to AI Investors
A partial list of funds investing in Artificial Intelligence and Machine Learning
I. Investing in AI
Investing in AI is not an easy job: AI technologies are black boxes and unless you are able to dig into lines of code they may be inscrutable. Simply looking at proof of concepts might not be enough to really understand the underlying stack behind specific applications, and this represents a big barrier for investors to efficiently allocate their capitals.
Generalist investors found then alternative ways to discern investable companies from the pile of tech-driven companies out there. Instead of looking at the code or the algorithms, they identified proxies for AI technologies, a sort of must-have list to help them cutting out media phenomena from interesting ventures:
i) Impossible problems: if a problem was not addressable before, it is really likely that a machine learning algorithm is behind the proposed solution of that problem;
ii) Data effect: it is common knowledge that neural nets require a lot of data to be trained, and if the startup has a way to create a virtuous data cycle (‘data network effect’) or has access to proprietary data, this is sometimes enough to be deemed as investable;
iii) Team and Patents: the biggest barrier to entry AI/ML is talents and IP. Therefore, if a team is composed of scientists/researchers and has patents (obtained or pending), it would already be a good candidate for an investment even without any revenues. This is driven by top tech companies acquiring smaller startups simply for their ‘brain power’ rather than their actual numbers.
II. So, who are the smart guys with the wallet?
AI specialists are luckily not that naive, but they are able to go much deeper and look behind the veil. As I already pointed out in previous articles, AI investors have different characteristics from more general investors:
i) Deep Capital Base: they usually should have a deep(er) capital base (it is not clear yet what AI approach will pay off);
ii) Higher Risk tolerance: investing in AI is a marathon, and it might take ten years or more to see a real return (if any). The investment so provided should allow companies to survive many potential “AI winters” (business cycles), and pursue a higher degree of R&D even to the detriment of shorter term profits. An additional key element of this equation is the regulatory environment, which is still missing and needs to be monitored to act promptly accordingly. Of course, in saying that, I only refer to the right hand of my AI Classification Matrix, because for narrow AI companies the risk tolerance may indeed be lower;
iii) First-Hand Coding/Engineering Experience: venture capitalists use the help of ‘venture partners’ or ‘scientists in residence’, but AI specialized investors are able to dig into codes and architecture by themselves.
III. List of AI Investors
I then compiled a list as extensive as possible of every investor I read or bumped into over the past months. It looks like there are at least 85 of them (full list available here) and here it follows a partial list of them:
- Accel Partners (Bay area): less than one year ago they had closed several deals in AI, and probably Paxata was the largest deal made in 2016;
- Amadeus Capital (London): you won’t find anything about AI in their investment strategy or any hint of machine learning in their speech or news. In spite of that though, their portfolio enumerates a list of fantastic AI-driven companies: Graphcore; AlgoDynamix; Five AI; Kreditech; Prowler.io; Ravelin; Speechmatics. Amadeus Capital also was one of the few investors at a seed stage in Improbable;
- Amplify Partners (Bay area): investors in companies like Scaled Inference and Enlitic, they have a partner who wrote down a fantastic report on machine intelligence (David Beyer, see the report here);
- Andreessen Horowitz (Bay area): nothing to say on a16z because their work speaks for itself. They have though an interesting program called ‘Professor-in-residence’, through which they host CS experts in their fund for a year or so. Fei-Fei Li (the creator of ImageNet) was the last professor joining a16z, after Vijay Pande, probably one of the best biomed scientist of the last decade. They are investors in several machine learning companies, but in particular in what I believe to be the best AI company out there: Anki. Final point: one of their investors is Benedict Evans, who writes a brilliant blog/newsletter you should subscribe to, and they have a podcast page with very interesting insights on AI (check Frank Chen’s presentation here);
- Bloomberg Beta (Bay area): Roy Bahat leads the effort of investing the $150M under management raised over the last few years in two funds. Everyone knows about their beautiful and informative ‘Machine Intelligence Landscape’ which is updated every year by Shivon Zilis;
- BootstrapLabs (Bay area): founded by Ben Levy and Nicolai Wadstrom, it is a venture fund that does Seed and Early Stage Venture Investments with a hint of startup studio approach. They backed up companies like Mendel Health, Sibly.co, AEye and Roger.ai among others and they organize a brilliant event called “Applied AI Conference” in Silicon Valley;
- Comet Labs (Bay area): investors with a mixed business model lying at the intersection between a VC, an incubator, and a research lab (Comet Labs Research Team);
- Data Collective (Bay area): Data Collective is really likely the best data investor out there. It looks that two managing partners Zachary Bogueand Matthew Ocko backed 14 of the recent ‘AI 100 companies’ list made by CBinsights. I am not entering into the details of their portfolio because I am impressed every time I see it. I only say one of the most recent acquisitions in AI industry: Nervana Systems (acquired by Intel);
- Georgian Partners (Canada): a very good Canadian group of investors (Justin LaFayette, Steve Leightell, Jane Podbelskaya, and others), one of the few funds investing in analytics and AI in Canada. Recently investors in WorkFusion and Integrate.ai, Georgian Partners released an interesting white paper on principles of applied analytics some time ago and a useful presentation on AI from their Chief Analytics Officer Chris Matys (who is part of the Impact team along with Jon Prial, Madalin Mihailescu, and others);
- Glasswing Ventures (Boston): still in the process of properly raising the fund (target at $150M), it is managed by former partners at Fairhaven Capital;
- MMC Ventures (London): MMC Ventures has more than £160M under management, 6.5% of which personally invested by MMC team members. In the past two years, the multi-funds have been recognized as one of the most active funds in the UK. They recently co-led a Series A round inSignal Media, GrowthIntel, as well as Sky-Futures. The Head of ResearchDavid Kelnar has written a wonderful primer on AI and sharpened the fund investment thesis around AI with a deep analysis of the AI landscape in the UK;
- Zetta Venture Partners (Bay area): Mark Gorenberg, Jocelyn Goldfeinand Ash Fontana (the guy behind the fundraising business at AngelList) have raised a second fund of about $100M to invest in AI startups like Tractable, Kaggle, and Domino. Check out also this really nice post from their MPs (‘Growing up in the intelligence era’);
*Note: the initial list published in Feb. 2017 was made by 77 funds. Thanks to David Kelnar, Nathan Benaich, Alex Flamant, Andreas Thorstensson, Mike Collett, for the post-publication comments. This list is anyway 6 months-old and many more funds should be included. This article is also appearing in a slightly different format in my new book “Applied Artificial Intelligence”(2019).