This past week I attended Devol’s Dance – a technology conference hosted by Drive Capital out of Columbus, OH. If you’re a Midwest VC cheerleader (guilty as charged), Drive Capital should be on your radar – they’re a great demonstration of what can happen when Silicon Valley thinking is applied outside of its birthplace.
Chris Olsen, a partner at Drive, hosts Differential – a podcast I pitch to friends as “How I Built This” with a Midwest angle. During breaks, Drive advertised Devol’s Dance: a conference on automation, robotics, and AI and how it will affect various industries. I was lucky enough to score a free ticket (#studentlife), so after a round-trip car rental, 12 hours of driving, and an impromptu hotel room, here are five big takeaways:
1.) Founders aren’t afraid of big tech
Anyone starting a business will face the question of why they won’t inevitably be stomped by [insert behemoth here]. In tech-enabled companies, those behemoths are front and center of the news cycle, making this question particularly top-of-mind.
Jason Mars is the founder of Clinc, an AI-based natural language processing company, and was both a presenter and panelist at the Dance. Clinc touts conversational AI that puts Siri to shame – Jason’s demonstration was laced with colloquialisms, slang, and references to earlier points in the conversation with his robotic assistant. The AI caught all of it, and its replies were leaps and bounds beyond anything consumers have in their hands right now.
When asked how Clinc could compete with Amazon, Google, or Apple, Jason basically shrugged and said “well, we’re better at this than they are”. It was a wonderfully simple response that really boils down the competitive philosophy of a lot of the entrepreneurs I met.
Clinc consists mostly of Computer Science PhDs from the University of Michigan, which sheds some light on why their product is so good: an absurdly talented group is dedicated full-time to solving this problem. The product is so effective, explained Jason, that they beat out Google and 38 other companies in a recent bid to build a chatbot for Barclay’s.
The founders I met don’t assume they can fly under the radar. But they focus on a problem and getting really good at solving it, and the rewards flow from there. Another startup CEO, Austin Webb of Robotany, addressed a similar question about how his nascent company could compete with incumbents and even other startups with stacked coffers. His response: “we’re building a more comprehensive platform”.
Businesses are ultimately focused on solving problems, and you don’t need $100M to solve (most) problems. What you do need is a really, really good team that can do it well. No wonder so many VCs list the founding team as their most important investment criteria.
2.) Build for the long haul: Robustness and Reliability
If you haven’t heard of Raffaello D’Andrea, do yourself a favor and jump to the 8:50 mark of this video. He launched Kiva, a robotics systems company that eventually sold to Amazon and became Amazon Robotics. Now he helms Verity Studios, developing staggeringly functional drones you might see on broadway or at a Drake concert.
D’andrea doesn’t believe in the demo effect, the colloquial name for when devices encounter issues only when demonstrated to a live audience. In his view, any product worth demonstrating should do exactly what’s expected each and every time. And he ensures this with a development process that prioritizes robustness and reliability.
This highlights the difference between building for demonstration’s sake (with a growing rift between the audience’s perception and the true functionality of the product) and building for the long haul. Creating a Wizard of Oz-style illusion to demonstrate the potential applications of a product is all good for showing customers and investors what can be, but D’Andrea commits to a more rigorous process that’s far more robust.
Jason Mars emphasized the power of this approach in developing Clinc. Rather than breaking down language like a middle-school teacher as many of their competitors do, Clinc takes a more holistic approach that more closely resembles how our brains naturally process language. The result: a robust, adaptable solution based on foundational principles rather than a patchwork of one-off solutions to customer complaints.
Entrepreneurs seldom have the luxury of time. But ensuring the core of your offering is built to last will help you spend less time reworking and maintaining, and more time pushing forward with amazing products.
3.) The Midwest is still struggling to keep talent local
While we spent most of the day idealistically thinking ahead, there were some grounding moments that brought into focus how much work we still have in front of us.
During an interview with Axios’ Dan Primack, Governor of Ohio John Kasich lamented the Midwest’s difficulty in retaining technical talent. He was particularly frustrated with what he perceived to be the inability of universities to commercialize their research.
Gov. Kasich shouldn’t be surprised that few new graduates are willing to stake their financial well-being on research that has not yet found a source of revenue. With student debt reaching astronomical new heights with each graduating class, it should surprise no one that talented engineers jump to the comfort of a stable salary, whether with big tech or a well-funded, late-stage startup. The simple fact is that these opportunities are primarily on the west coast.
It’s not that Midwest tech doesn’t have the need. Most entrepreneurs here struggle to find adequate tech talent. But Silicon Valley has a tremendous brand advantage that even Midwest startups are guilty of paying deference to when looking for someone to lead up their technical and product teams. Why would an engineering graduate start her career here when she can come back after 5 years at Google and have her pick of the litter?
There isn’t a single solution to this problem, and what’s worse is the strong incentive for policymakers, entrepreneurs, and investors to wait. Opportunists can hold out until there’s more momentum and still reap many of the benefits. But that opportunity will only come if a critical mass of pioneers are willing to make bets and start the flywheel.
4.) Full automation will bring about a paradigm shift in manufacturing
Elon Musk recently made headlines by proclaiming (admitting?) that the Model 3’s assembly line was overly reliant on automation. While likely a small step back in the inevitable march toward full automation, even robotics executives at Devol’s Dance agreed there was still very much a place for humans in manufacturing. They praised humans’ advantages in haptics, assimilating diverse feedback, and handling “curveball” issues not within the strict scope of a machine.
When moderator Dan Primack challenged the panel, postulating that any task can eventually be replaced by machines, Dan Pratt of Vecna Robotics explained that these innovations will change the fundamental structure of manufacturing processes.
We can theoretically create robots to replicate any task on an assembly line, but since those lines are developed for humans in the first place, the real puzzle will be how to reconfigure the whole system and strip out assumptions born of human constraints. When we consider other innovations like advanced materials and additive manufacturing, it’s entirely likely that “assembly lines” will look very different than they do today.
As robots are more and more integral in factories around the world, we will also see a new take on a tried-and-true business model.
5.) Get ready for RAAS (Robotics as a Service)
GE became the textbook case of business model innovation when it began providing aircraft engines on a subscription basis, leveraging sensors to collect data it used to provide more effective (and cost-efficient) service. Expect to see the same thing happen with robotics manufacturers as they capitalize on the potential of the Internet of Things.
While some manufacturers will likely sell hardware outright, the savvy will design machines that connect to a network, monitor their own health, and coordinate with other machines to maximize efficiency. They will also receive software updates, though these will probably be less frequent and less significant than the OS updates pushed to consumer PCs.
This will also introduce cybersecurity concerns, creating an opportunity for robotics companies to differentiate on security and fail-safes.
The Dance was an amazing experience, with fascinating panels and presentations by some truly visionary entrepreneurs. Twelve hours of driving and mild sleep deprivation was a small price to pay, even in the thick of my mid-term exams. I hope to make it a habit, so if you have any cool events coming up, or can recommend any that are worth checking out, please don’t be shy about letting me know.