When Software Eats the Physical World, Startups Bump Up Against Regulations

Ted Ullyot


Editor’s Note: Ted Ullyot recently joined Andreessen Horowitz to head up its first-ever operating group for policy and regulatory affairs.

Before joining a16z, Ted was General Counsel at Facebook, joining the company in 2008 when it was just 500 people and leading the legal team through IPO until 2013. Prior to Facebook, Ted spent most of his career as a Washington, D.C., lawyer, including as a Supreme Court law clerk for Justice Scalia; an Associate White House Counsel and later Deputy Staff Secretary for President George W. Bush; and at the Justice Department as Chief of Staff to Attorney General Gonzales. Ted also was General Counsel for AOL Time Warner Europe, and spent time as a partner at Kirkland & Ellis, where he focused on regulatory and antitrust litigation among other things. He grew up in, and now lives in, the Bay Area with his wife and three children ages 13, 11, and 8.

a16z: You’re clearly a bit bilingual when it comes to speaking the “language” of both D.C. and Silicon Valley. But what surprised you most when you first made the move here — was there anything that made you go, “Wow, I’ve never dealt with that before”?

Ted: As a general counsel it’s more often the other way around: When does something NOT surprise you? The best piece of advice I ever got about being a GC was from one of my longtime mentors and former bosses, [Time Warner GC] Paul Cappuccio, who told me: “assume that 90% of the stuff that comes across your desk, not only will you never have worked on before, but you’ll never even have heard of before.” He gave me that advice when I took my first GC job.

Being a GC is sort of like being a general contractor on a construction project: You’ve got to be fluent in all the trades, and you probably have a strong background in one or two of them, but no one expects you to know every single thing about each craft. Plus, every company has its own distinct issues, and those change all the time. It’s more about how you think, and having good judgment, than about what you already know.

a16z: But was there a cultural difference compared to the organizations you’d been in before?

Well, besides the more relaxed dress code, I did encounter things at Facebook I hadn’t experienced before — like all-night hackathons. I thought the idea of a hackathon sounded cool, so the first time we had one after I got to the company, I made a point of taking our entire legal team (which at that time was about 10 people) to it. To be sure, we weren’t coding, we were working on legal department projects, answering the occasional legal question, and so on. But we were present.

I thought it was important for the legal department to show respect for the engineering culture of the company. It’s very different when you’re inside an engineering-driven company as opposed to a sales-driven company. I used to have a talk with incoming lawyers, setting the expectation that they’d have to be comfortable being part of the “back office”, the overhead function. Yet we didn’t want to be viewed as this siloed function you only see through enclosed glass offices.

Funnily, our legal team ended up winning one of the five or so “hackathon hero” t-shirts they gave out at 5 in the morning. After that, it became a tradition for the lawyers to attend hackathons, and yes, we even participated in the all-night hackathon the night before Facebook’s I.P.O.

a16z: Okay, given you’ve lived both the worlds of government and tech, how would you characterize the relationship between them?

Ted: I have a couple of thoughts here. One is, I actually think the situation between Washington and Silicon Valley has gotten much, much better over the years and especially in recent years.

For one thing, you have a more “digitally-native” (or at least, tech-native) cohort coming into office, on both sides of the political aisle. There are now a number of tech-centric roles in Washington, like the CTO, the CIO, and most recently, the Chief Data Officer role. The administration has also brought in people who used to work at major internet companies like Facebook, Google, and Twitter. And the U.S. Patent and Trade Office recently opened a satellite location in Silicon Valley. Then, on the campaign side, both Democrats and Republicans have been hiring way more tech folks as well. People are realizing they need tech (not the tech industry, I mean new technologies) to win elections.

Finally, you’ve got people all over the U.S. — regardless of whether they’re in D.C. or Silicon Valley — using more tech. Before, it would have taken years for a specific technology to move outside of early adopters to become a prominent and natural thing among mainstream users. Now, you’ve got the very lawmakers who might be regulating said tech using services like Airbnb and Lyft … and it’s happening overnight. Something like Meerkat can go viral in just one weekend, and a politician might use it in her campaign a week later.

a16z: If the relationship between government and technology is so much better, then why are all these conflicts between lawmakers and startups taking place?

Ted: Regulators and lawmakers have a job to do; they don’t ever want to be accused of being asleep at the switch when some new technology comes in. And I think it’s human nature to react skeptically when it’s a new product, service, or behavior you haven’t encountered every day. Like the first time people “tagged” other people in photos — what does that even mean? Is it a violation of privacy?

On top of this, we sometimes also hear sensationalist reports or accounts of “x person got attacked by y bug”, which doesn’t help matters. The focus is often on what can go wrong with new technologies, rather than on the benefits of these innovations.

a16z: In fairness, many of those accounts are true. They wouldn’t play on people’s fears otherwise.

Ted: Yes. But this isn’t a new problem — the law is always going to be lagging behind tech. It’s been like that since the days of the telephone, and even before that.

a16z: So then are we really seeing something new here, or is it more of the same? Is this just a difference of degree — or is it a difference of kind?

Ted: I think the reason we’re seeing these collisions with existing rules and regulations earlier and more often is that it’s such a difference of degree, that it’s almost become a difference of kind. Tech today has accelerators built into it that it never had before: the internet, social, mobile. That has speeded things up considerably.

Before, people encountered technology behind the scenes (business processes, for example); now, they’re directly encountering new tech in everything they use — from smartphones to the Internet of Things. So there’s the sheer ubiquity of tech as well.

Finally, it’s no longer some remote person in a lab but your next-door neighbor using these things. So it’s about familiarity with tech too. In fact, it’s more intimate and personal than ever because you’re carrying these services/apps on you 24/7.

However, for some physical-world players being affected by tech changes for the first time — say, car dealerships or taxi commissions who’ve never had to encounter this before — it’s a completely new thing. It’s become an existential crisis for many of them.

a16z: Aren’t some of those clashes due to regulatory capture though?

Ted: Yes, but it’s not as black-and-white as “oh, they’re bad guys, they’re fraudulently taking payoffs”. As I see it, regulatory capture is really about two things:

One, it’s about deep familiarity. The downside of having to deal with a regulatory agency a lot is that you have to devote considerable time and resources. But the upside of that continuous contact through meetings and events is you begin to get to know each other, and understand how and why regulators do what they do. This form of regulatory capture is about knowing how to work the system. It won’t ever be framed where the hotel industry says “Hey, we’ve got this new competitor that lets people rent out their homes; can you take care of it for us?” No, it will be framed more in the spirit of “Hey, listen. We’ve got a relationship. You regulate us. You make sure the rooms are safe. When people visit San Francisco, they should have a safe room with a fire escape in the right place. So you should go scrutinize what these new guys are doing.”

The other kind of regulatory capture is a form of back-scratching, where buddies that legacy incumbents have worked with for years are now in office. They are running things. The conversation is some variation of “Please prevent these new entrants from disrupting me.”

While the second form of capture sounds a lot worse, the first type is a subtler form of persuasion that’s hard to compete with. Instead of phrasing things as a favor, incumbents are framing things as “You’re right to be scared by this new technology”.

a16z: If they’ve got all this funding, can’t startups compete with that or just take it on?

Ted: It’s not just because startups don’t have the resources or even the relationships. It’s because they don’t have as much insider knowledge of the industry.

Take net neutrality for instance. Assuming the rules stand, the telcos have literally decades of experience navigating the halls, and the rules, of the F.C.C. Meanwhile, new entrants — who most think of as being protected by net neutrality — may find themselves stymied by incumbents that know how to work the system. They know exactly what objections to throw at your petition. New entrants simply don’t have this advantage.

a16z: What should startups do then?

Ted: First of all you have to be willing to take risks. That’s what a startup is by definition, right? I’ve heard it defined as everything from “an institution designed to deliver a new product/service under conditions of extreme uncertainty” (Eric Ries) to “a company designed to grow fast” (Paul Graham). However you define it, your startup won’t reach that critical point of growth if you don’t take risks and aggressively push forward. And that’s invariably going to bring you into conflict with antiquated laws or permit processes.

So you have to be willing to fight. But you also have to be willing to understand the regulator’s perspective, and to take the long view. This isn’t like one-time litigation, where you can go scorched-earth on your adversary and fight to the death to win. You’re going to encounter these same agencies and people over and over again. And you’ll need them, too. I saw this time and time again at Facebook: One day we’d be in front of someone fighting on issue X, and months later we’d really need their help on issue Y.

If you handle the relationship wrong, your startup might win the battle … but lose the war.

a16z: Isn’t this a form of lobbying?

Ted: No, and this new function is not a lobbying role. Think about it more as long-term relationship building, in the same way Andreessen Horowitz already manages different networks our startups can access.

The best situation for us — and for our companies — is to go to regulators and policymakers when we don’t have any ask at all, and we’re just being called in to explain or share something. I want us to offer our expertise and our optimism around technology and the people building it, so it becomes, “Hey meet some of our companies so you can see what’s going on in such and such area that you’re hearing a lot about.” That’s the ideal we’re aiming for here.

If we’re helping people understand new technology, everyone benefits. Like a rising tide that lifts all boats. Our portfolio companies are not the only ones who will benefit from our efforts.

a16z: How do regulators figure out where fear-mongering ends and legitimate safety issues begin? And how can startups help them?

Most regulators are good people trying to do their jobs, which generally speaking is protecting consumers. By the same token, companies like Airbnb don’t want bad experiences for their homeowners/renters either, so they’re even more incented to enforce ratings and reporting and other mechanisms for trust and safety. Their business wouldn’t survive without this.

In this sense, startups are actually very aligned with the ultimate goals of regulators. The key is to find the thoughtful, progressive regulators who understand what you’re trying to do and know you’re acting in good faith, just like they are.

If you begin before you even have an ask, you can start from a place where you’re saying “How can the tech community help build products that address your concerns?” Or, “I’ve got a new way to do this that’s a lot safer for consumers.” For example, when was the last time you had a bad taxi ride and actually called the taxi commission phone number (do you even know where that is, by the way?) to file a complaint? It’s much more frictionless to input a rating or file a complaint on your mobile app and then get a response within 24 hours instead of within months if at all.

In this last scenario, you’re essentially saying to regulators, “Hey, let us help you make your job easier, especially since things are moving into mobile anyway.” Chances are you can find some regulator out there who is thoughtful and can be a leader in that shift to mobile (or whatever tech trend).

a16z: One idea related to this that we’ve proposed is a sort of “regulatory arbitrage” for permissionless innovation. What’s your take on this?

Ted: In some cases you’ll only have to deal with one regulator/agency — which is great if they’re reasonable and you know the right kind of arguments to persuade them. But let’s say you have to deal with 50 state and countless other local agencies, and in some cases overseas players as well. Then you’ve got this huge matrix and you don’t know where to start.

This is where something that seems like a disadvantage — having hundreds of regulators — can be an advantage. Because you have that arbitrage if you know how to partner with the right person/region before you go to market. (Google Fiber used this to its advantage to find which cities to start with, for example.) If I do my job right, building that network means identifying and getting these relationships in place: Here’s a state that has a great attorney general who is very pro-business, pro-innovation; this governor is talking about tax credits because he really wants to grow his state’s tech sector; and so on.

a16z: Okay let’s wrap up with a lightning round for your thoughts on a few regulated domains that are top of mind right now…

Bitcoin? What’s great is that instead of being reflexively negative about it, we are seeing some governments from New York to the U.K. trying to understand it (or position themselves as thoughtful people trying to understand it) and seeing what advantage they can get. Either way, it’s a good thing.

Drones? I think the FAA has been moving in the right direction here and moving relatively quickly for a federal agency. The risk is if they don’t move quickly and aggressively enough — it will be dangerous for us economically if the U.S. falls behind and the innovation ends up taking place somewhere else first.

Sharing economy? I think these services are great examples of where technology can help achieve regulators’ goals in a more effective way. Of course, some important issues — like the current debate on defining an employee vs. a contractor — are still playing out, but those cases will be determined on a very specific set of facts. And I do think the tech sector is more aware and sensitive to socially conscious aspects of their businesses right now. In the past, a lot of employment risk was bundled inside a single company: how to handle taxes, health insurance, etc. As this ecosystem grows, we’ll see more and more services that help take care of those functions but unbundled across a number of companies. It’s still very early days yet.

a16z: How about patent reform? This is less a tech domain and more a general topic, but what’s your quick take? Note we’re talking only software patents here.

Ted: Back in the early days of Facebook, we were in the position that a lot of young companies find themselves in: an attractive target for entities who use patents as a shakedown weapon. And it wasn’t just trolls, but legacy tech companies and other internet players too who were doing it.

So I’m a huge proponent of software patent reform. There have been some strides toward patent reform in Congress, but really meaningful patent reform on the legislative front remains difficult given all the competing interests. The good news is that we’re seeing some better rulings from the courts, including the Supreme Court. We’ll get there.

Want more a16z?

Sign up to get the best of a16z content, news, and investments.

Thanks for signing up for the a16z newsletter.

Check your inbox for a welcome note.

MANAGE MY SUBSCRIPTIONS By clicking the Subscribe button, you agree to the Privacy Policy.