AI Today Podcast #009: Investing in AI with John Frankel of ffVC

In this week’s podcast we interviewed John Frankel, Founding Partner at ff Venture Capital. We discussed his firm’s role at NYU FutureLabs Summit (where Cognilytica was a proud partner), the current state of the AI market from an investor’s perspective, and advice for enterprise and startup companies looking to have an AI strategy.

Episode Sponsors:

For over 25 years, QS has been helping prospective MBA candidates just like you make informed decisions about choosing the right business school. At our upcoming Dallas event, you can meet face-to-face with admission directors from top ranked US and international business schools, including UT Austin, SMU, Rice, IE, Hult, and many more! You will also be able to participate in interactive GMAT sessions by GMAC – creators of the GMAT exam, apply for $7 million in scholarships, attend Panels, and network with alumni and your future peers. Learn more about exclusive opportunities on the day by registering now to claim your free ticket at https://goo.gl/iRF9PR!

Show Notes:

___________________________________________

A transcript of podcast is available below:

[00:00:50] Kathleen: Hello and welcome to the AI Today podcast. I’m your host Kathleen Walch.

[00:00:54] Ronald: And I’m your host Ronald Schmelzer. Our guest today is John Frankel: founding partner at ff venture capital. John, welcome to the AI Today podcast.

[00:01:03] John: Hi, thanks for having me on.

[00:01:04] Kathleen: Great. So let’s get started! John, we’d love it if you could introduce yourself to our listeners, and tell us a little bit about your current role at ff Venture Capital, and what you’re currently doing around artificial intelligent investing.

[00:01:16] John: Certainly. So despite my English accent, I’ve lived in the States for most of the last 25 years. Grew up in London, England, graduated from Oxford with a Masters in mathematics and philosophy. Worked for Arthur Andersen, qualified as a chartered accountant and then joined Goldman Sachs. I was there for 21 years doing a variety of roles, helping build various businesses. And the last 11 years working on the sales and trading floor, as equity research salesperson in New York, covering hedge funds. About eight years ago I took what was something I’d done part-time, which was angel investing in start-ups, and founded ff Venture Capital with my partner Alex Katz in 2008.

And so we’re now coming into our sort of tenth year of business in venture capital funds, and we have four raised funds under our belt, about 70 active portfolio companies. We invest across the US in a number of areas. Today the areas that we’re really intrigued by, are drones, robotics, cybersecurity and artificial intelligence. And as a firm, we invest at seed stage up until the mid-single digit valuation. We run with a large team. We get very involved with the companies we invest in, and we look to really help them grow up into being significant companies. And at this stage, with nine years under our belt, there’s a fair number of companies you can check out on our website ffvc.com. Some are household brands, some are enterprise-focused companies and it’s a list we’re very proud of.

Within the AI space, we think this is a fascinating time right now. You know AI, as I’m sure you’re aware, has been on the cusp of going mainstream for about 50 years – but it really is now. And there are a number of reasons around that which I’m happy to dig into. But one of the things that I think is worth highlighting, is that we have this partnership with NYU, with the Tandon School of Engineering there, and that partnership involves a number of factors. One is that we run a mini incubator each year where four or five companies go through, and then that culminates in a conference. We held the second annual conference yesterday and I believe we sold approaching 900 tickets. Most of those people showed up and we just had some phenomenal folks come and talk, you know, from Google and USRA, which is part of NASA, as well as the Vector Institute, TechCrunch, Orrick, x.ai, etc. As well as the four companies that went through the second cohorts of the program – they got an opportunity to meet investors for the first time: Mt. Cleverest, SecondMind, Bite.ai and Bowtie.

[00:03:58] Ronald: Sounds great. Following up on that, we saw you at the NYU Future Labs Summit yesterday. For those of you that are listening, that was October 31st 2017 – that’s right, Halloween – and you gave a good introduction to the AINexus labs. So for our listeners who weren’t there, maybe you can give a quick 2-minute recap of some of the things that you talked about at the recent NYC Future Labs Summit and some highlights from that.

[00:04:18] John: More than happy to. I think the first thing to set in frame is really: why is AI different? We can talk about why AI is now. But I think a lot of commentators have addressed that. We think AI is different and isn’t just a sector. And if you go back 20, 25 years ago, we had this thing called the Internet and cloud computing and it’s had a massive impact and it wasn’t just the sector. It was a sort of architectural change in the way things were done. And then 10 years ago, mobile came along, and initially we talked about companies being mobile – start-ups were ‘mobile first’. You know, all built around SoLoMo, social location, mobile – but these days we don’t use those terms, it’s just embedded in everything. And today when we look at AI, we don’t see AI as the sector. We see it as something that will be embedded in everything that we do and every company in the same way – in the way that electricity today is embedded in everything that companies do.             And so what I talked about yesterday was New York and our belief – and NYUs belief – that New York will become and deserves to be one of the leading hubs in the U.S. and in the world around AI. It’s driven by a number of factors. One is that we have three technical competing universities bringing a lot of talent at the undergraduate and graduate level into New York. The second thing is that we have, within those universities, academics who have one foot in the university and one foot within research labs in major corporations. There’s Yann LeCun who’s at NYU and also leads the AI efforts at Facebook, but there are a number of examples. Another element is that New York has just been a start-up center for long enough that you’ve built up a lot of the start-up infrastructure. There are a ton of incubators, investors, advisors around in the space. In addition, for quite a number of years now you’ve had significant engineering shops set up as part of, you know, otherwise West Coast firms: Google, Facebook, Twitter; all of those have significant presence in New York. And all of that creates a primordial soup as it were, during which great companies can be formed and we’ve had a number of those. And some have exited recently and the infrastructure within New York has just built up incredibly.

And then for AI in particular, because you have companies such as retail, financial services, advertising, media, fashion companies – all of which are being disrupted by AI start-ups and based in New York – there’s a lot of the main specific talent that is sitting right next to this. And that we think is just a great place to grow AI companies. We have a number that are growing very strongly in New York and there are new ones being founded everyday. And you know AI Nexus lab, that we have with NYU, is part of that; creating more hubs for engineering talent to want to come and work here, and more spin-outs for more mature companies where people go and start companies. In fact bite.AI, which is one of the start-ups from our second cohort – which is probably the easiest way to track what you eat – these engineers that have put together some amazing image recognition technology stacked together. They came out of Clarify, which is an image recognition AI company that also has deep roots in New York.

[00:07:53] Kathleen: Okay, interesting. Yeah, I know that New York has a lot of investment and a lot of VCs and a lot of start-up activity going on. It’s a pretty thriving area. And then with that, I know that ff Venture Capital does invest heavily in Artificial Intelligence companies. Can you tell us what you feel the current state of the AI market is as an investor? And then what are some of the most interesting and exciting companies in AI that you have been involved with?

[00:08:18] John: So I’ll give you a little framework here; as I said before, AI’s been on the cusp of mainstream for sort of 50 years, and there’s been AI winters along the way and we’re now in the middle of an AI summer. And we do run the risk that there is too much hype in the space and that expectations run ahead. An example of that, I think, is autonomous vehicles, where if you read the press it feels like in 2021, maybe 2025, you or I could go to a local dealer and buy a level five autonomous vehicle. One that may or may not have a steering wheel, but doesn’t need one. And when I speak to leaders in the field, some say ‘maybe, but there’s a lot to be invented’ and others will say ‘maybe by 2040 you’ll have that’. Now what you’ll probably see is in constrained environments, ring-fenced environments, on a rideshare basis you’ll start to see autonomous vehicles as early as next year. And then over time, the ring-fencing will get wider. But individuals owning autonomous vehicle is a way off for a number of reasons, including political, societal and economic reasons, which I won’t go into now. But the hype is very high.

[00:09:37] Kathleen: Yeah, we have actually talked about this in another podcast, where we said that laws and regulations need to catch up before things can actually be implemented in practice. So you’ve justs hared that as well.

[00:09:48] John: And people have to change their behavior. If you’re in San Francisco, no one jaywalks; if you’re in New York and it’s 4:30 on a weekday, the density of people walking on the street and on the sidewalk are about the same. And I think in New York, if they’ll see an autonomous vehicle coming towards them, they will just stand in its way. So I think there’s a lot that needs to be worked out. I can use that as an example: when I look at it, the first wave of AI recently we saw, were engineering people going after almost research projects that may or may not end up being businesses. And we saw the large tech companies buy these teams in as aquihires and elsewise, to build out that very deep talented pool for the consumer facing products. And we’re seeing the benefit of that today and products you buy from Google and Apple and Amazon and the like.

The second wave that we saw with these sort of horizontal ones, where they said: we’re going take a problem like image recognition, and we’re going to be the place you go for that or voice recognition or the like. And some of those tech stacks were again bought or developed within these large technology companies. And because a lot of the people who work there still have a foot within research, there’s been a bias for them to open source the solutions and open up the APIs. And so a lot of those horizontal solutions today can be replicated at a fraction of the cost they took to build, because a lot of the technology which was very difficult and unknown then, has become very facile and known.

The third wave – which is the wave we’re really investing around – are companies that are enterprise- as opposed to consumer-focused, and going after specific vertical problems, using a combination of technology and data to build deep network effects. Which allows them to build some barriers to entry around what they’re doing, and doing it in a way that is relatively careful and efficient. I don’t think the lean start-up methodology works well, as a rule, with AI. There’s just too much data processing and customer acquisition that needs to be had to build these barriers. But I don’t think you need a 40 million dollar check just to get into the business either. And so we’re seeing some really interesting companies around the space. And I’ll give you two examples if I may. We invested in a company in downtown San Francisco called Dashbot.ai. And they want to become the analytics layer for conversational interfaces for chatbots. And I think a lot of investors are wary of the chatbot space, because the lifecycle of a given chatbot seems to be relatively short. And if we’re not in the first innings, we’re in the pre-game, and that’s a lot of work to be done. We looked at the space and said: any given chatbot may have a short half-life, but the space itself is going to grow exponentially. And we like Dashbot, as it’s analytics play of being this sort of… If chatbots are a gold rush, then they’re selling the equipment, the spades and pickaxes into the space, and the company has grown dramatically. When we invested a year ago – actually not even a year ago – they had grown from processing 40 million messages cumulatively, to have processed 200 million messages November 2016. When I last looked, they had processed 13 billion messages. They’ve become the dominant analytics platform in the space, and they’re building up a very, very interesting business. And by being early in the space, getting the opportunity to define the language, then you need to really understand you know what makes a good conversation, what doesn’t; and allowing companies and others to benchmark their conversational interface versus peer groups. So that’s one company.

Another company is a company called Bowtie, that presented yesterday and was part of the second cohort in the program. And Bowtie – with small businesses, initially they’re starting with spas and hairdressers and the like. And what happens is: if you call up a business, maybe someone picks up, maybe they don’t; but if you’re calling them up to book an appointment, find something out or the like, your chance of calling back again is really low. A fair percentage of calls, somewhere round about 30 percent or so, are outside of working hours. About 25 percent of them are during the busiest time during the day. And what Bowtie does: they partner with these businesses, and they capture the caller ID. The AI text messages someone, saying ‘sorry we missed your call: can I help you?’ And the person says: ‘I wanna book an appointment’. They go ‘this is the availability’. The person goes: ‘I want know the price of this’. And so, the AI is able to build an initial conversation with that individual, around what they want to do; and bring in business that would otherwise have fallen by the wayside. And then what’s fascinating is people subsequently keeping up conversations; so they’re booking a second appointment, or ordering a product through the AI subsequently. And so it becomes a virtual front desk for those small businesses. Fascinating team, they’re growing revenue 50 percent month over month. And it’s a great application where AI is actually not replacing anyone’s job, but it’s supplementing it and bringing incremental revenue into a business, and increasing customer satisfaction.

[00:15:29] Ronald: Oh great! Let me just follow up real quick here. So on that question: what is some final advice you might have to enterprises and others that are focused on implementing AI? And then we can use that as sort of our final wrap-up question.

[00:15:40] I think that enterprises – a lot of them felt that when the Internet came around 20 years ago, they were late for the game. And with mobile they were late. With AI, we’ve seen incredible engagement from corporate VC and corporates to understand it. And I think they get that not only is AI disruptive as a tech stack and as the toolset, but it can be exponentially so. My general sense is corporates do not want to be late to the game around this, and so they are engaging with a lot of companies; either to put in more efficient toolsets or build it within their tech stack. The other thing they’re noticing is that Apple, Google, Amazon, Facebook and the like are implementing AI at the edges of their businesses and educating consumers to expect an AI-enhanced experience – which puts a competitive pressure on them to do the same. So we’re seeing incredible engagement from corporates. I think with an AI start-up, they have to think about who owns the data, who owns the learning set, and what they can do to build barriers to entry around their business as they do that. But I do feel, like with Bowtie, there are ways to implement this technology, where you end up with a much richer consumer experience than would exist elsewise. Because elsewise that was just not economically feasible. AI is not all about taking ten workers and replacing them with one worker and an AI texter.

[00:17:06] Ronald: Well that’s awesome. I think this has been a great podcast and we definitely want to keep following up with you, if that’s possible. I know that you guys seem to be very much involved in the forefront of what’s happening in AI, and so I just want to thank you so much for joining us on this podcast. We really appreciate you joining us.

[00:17:20] John: I appreciate it as well. If you want to see companies we’ve invested in, ffvc.com is a great place to go. And I’m actually fairly active on Twitter. If you want to find me there I’m under John_Frankel. And I enjoy the sort of open conversations that can happen on Twitter.

[00:17:40] Kathleen: Great. And we’ll post that in our shownotes, listeners, so that you can have that as well. And thank you so much for joining us, John. Listeners, we’ll catch you at the next podcast.

[00:17:49] Thank you so much for your time and look forward to talking in the future.