Cloudflare TV

How Tech Can Be A Driving Force For Democracy

Originally aired on 

Best of: Internet Summit 2018 - UK

  • Raffi Krikorian, CTO, Democratic National Committee
  • Alex Macgillivray, Former Deputy US CTO
  • Moderator: Matthew Prince, Co-Founder & CEO, Cloudflare
Internet Summit

Transcript (Beta)

So it's my real pleasure to introduce two people who I really admire and can talk both from a technology perspective but also from a political and public policy perspective about how tech interacts.

So first is Rafi Koukourian, who's the CTO of the Democratic National Committee, which is the political organization of the Democratic Party in the United States.

Prior to that, he was the VP of Engineering at Twitter and was the director of Uber's Advanced Technology Center.

So again, a really great sense of both the political and the technical aspect.

Alex McGillivray has been a hero of mine for quite some time. He's the former US CTO, or Deputy CTO, under Megan Smith.

Prior to that, he helped build Google's public policy team from its early days and then was around Employee 60 at Twitter as their first counsel and general counsel and took them through being public and then thought that there was a duty to give back to the country and so went and worked in Washington, which often technologists don't do.

So let's start there.

As people who've worked in technology, oftentimes the last place that you ever want to be is in Washington, D.C.

Why did you both choose to go to Washington, D.C.?

I guess I'll start because it was earlier. I mean, for me, the way I would put it is it was, first of all, a total privilege of a lifetime to work in the Obama administration.

It was just an incredible place to be. Megan Smith, who was the Chief Technology Officer at the end of that administration, the way she always put it was like, go where you're rare.

That was her career advice to people, and I think it's pretty good advice.

And so the fact that there weren't a lot of technologists working in government was a real selling point because you could go to a place where you would learn a ton from all the other people who you'd be working with and make a huge impact on people's lives.

There's little that can impact a person more deeply than government service in terms of being able to deliver things like health benefits or deliver in the Veterans Administration education benefits or even like going to the DMV and being able to get your license to drive a car.

These are all really important, deeply impacting stuff, so being able to affect that was really awesome.

I mean, it's pretty similar for me in a lot of ways.

I mean, I didn't have the privilege of working in the government, but I feel like in a lot of ways our government is currently broken, so I wanted to use the opportunity.

I deeply believe engineers can change the world, and so this seemed like a really great opportunity for us to try to affect change.

There's a whole slew of engineers like me in the U.S. right now that just don't agree with the way our government is being run or who's running it, so how can we put our talents to use there to try to affect change in some way?

What were the just biggest shocks or cultural differences coming both from having worked in technology companies and then going into much more political organizations?

How are those two things different from one another? I mean, for the political side, I think in a lot of ways, one, we're pretty rare.

There aren't that many technologists around, but two, I think I'm used to worlds where you're the tech team, so everyone comes to you, whereas in politics, at least, we're knocking on the door being like, please let us in.

We think we can help in a bunch of different ways, so that reversal was pretty massive for us.

Yeah, there was this great meeting when I first got into government about we're trying to get a government-wide open source policy out, and so we were having this meeting, and it had been called, first of all, the engineers had set up the meeting.

They had set it up for 8, 15 in the morning, and they had arrived punctually at about 8, 10, and there were about half engineers, so about four or five engineers and four or five people from different council's offices who were worried about open source, and the council all arrived late, and the engineers brought doughnuts and coffee for the council, which in Google, if I had scheduled a meeting with the engineers at 8, 15 in the morning, maybe I wouldn't have gotten fired, but it would have been like, why did you do that?

What were you thinking? None of the engineers would have showed up.

They wouldn't even acknowledge the calendar entry, and I certainly would have been the one bringing the doughnuts and getting there early, so there is a little bit of that, but I think the other thing, Todd Park, who was the CTO before Megan Swift, the guy who hired me, he kept telling me, I kept saying, like, I've never done government.

You should find somebody who's done more government, and he kept saying, like, the skills are totally transferable, and I think that's right, so for the most part, it's the same, right?

You've got really smart people trying to fix really important problems, and they might be used to doing a bunch of things in a pretty agile way on the policy side, and granted, they've never heard of agile development, they've never done agile tech, but they're using a lot of the same techniques, because that idea of smart, diverse set of people working on big, important problems, it's not like there's something special about the way tech does that, as opposed to any other group.

Did you feel well -received?

Did the policymakers want to hear from the technologists? Were you included in the conversation?

Yeah, I mean, I guess what I said before about privilege is maybe worth saying again here, so many of the people that we worked with, like, I was at a retreat once for a counterterrorism team, and we're introducing ourselves around the table, and I was last to introduce myself, and every other person had basically said, okay, the way I got to where I am now, which is working in the Obama White House and working on national policy, was, okay, I started off as an army grunt, and then did two tours of duty in war zones, and then worked at the FBI or the CIA for 10 years, and then did this and that and more schooling, and then finally, as a culmination of a 30-year period of civil service, I got to work in the Obama White House doing counterterrorism policy.

That's a very typical story among that group, and it was literally, you go down the line, every single person had that story, and I was like, well, I worked at Google and Twitter, and now I get to be here, which, yeah, so the fact that they were willing to even listen to me at that point, given that I really hadn't paid that set of dues, is a part of how welcoming they are.

The other thing that was awesome is whenever you're in a company, you've got a whole bunch of different competing goals.

You want to drive down the costs and serve the customer, and at Google, we had this advertising thing and the users, and in government, it's really simple.

You've only got one job, which is to make life better for your constituents, for the American people in the U.S.'s case, and that's great.

You're all on the same team. You're all trying to do the exact same thing.

If they think that there's a skill that you have or a perspective that you have that can add to that, they're very welcoming.

Rafi, you're not in the government itself, but you're in a very political organization.

What is tech like within the DNC?

Is it a priority, or is it something that's sort of ancillary? Yeah. I mean, so one, you're right.

So we spend a lot of our time basically thinking about hand-to-hand combat, like how can we actually go and get voters on our side, get that message out, get people to go and support your candidate in a lot of different ways.

So historically, the way that we think about that is like old school organizing, right?

We get people to knock on doors, you get people to call people on the phone, you just get the word out in a bunch of different ways, and it's been changing over the past few cycles.

Now there's more Internet advertising, clearly it's gone wrong in a bunch of different ways.

So what we spend a lot of our time thinking about is just like, what's the better way for us to just connect with humans?

How can we have long -term conversations with a bunch of people?

How can we amplify the message of our candidates in a bunch of different ways?

So fortunately, at least the current DNC team, so remember political parties, they usually rotate every four years along with the US presidential cycle.

So it's a whole new set of people who are trying to tackle this problem.

So this cycle, the political party is really technology -oriented.

We know, we kind of had our asses handed to ourselves last cycle, so the chair this time around is really big in investing in technology.

We want our legacy to be the modernization of the party in a bunch of different ways.

So we're in a turnaround job. Do you think that, is the political party becoming more tech-aware?

Will that, through osmosis, do you think spread through the rest of government, or is there really a separation between those things?

Yeah, so I'd be curious to know Alex's opinion on this, but right now there's a pretty big firewall between electoral work and the governmental work.

So we see it right now, I'm talking to a whole bunch of governmental officials that are going to be up for election in 145 days, and they haven't really thought about technology.

They've been in government for a couple of years, and they haven't really thought about the technology needed in order to pull that off again.

And I'm kind of worried that cycle will happen again when we go through.

Our best hope is some of our candidates right now are really tech -oriented.

There's some really interesting tech issues out there in the US that we're hoping that they'll carry in with them so they can have better conversations with people like me, or hopefully people like me get to tag along when they get into office this time around.

So knock on wood, but right now there's a pretty big firewall in a lot of ways.

The last panel talked a little bit about generational change, so I think that's happening as well.

If you're a random economic policy staffer, you probably grew up using a whole bunch of these tools already, and that will definitely bring change.

But also you do see those transition teams that go in to try to hire the people.

Because that's the other thing I didn't really understand, is government is this incredibly quick startup, particularly within the White House and the executive branch.

You go from having zero people to a couple thousand people in the space of five or six weeks.

So you sort of rely on your people that you worked with in the campaign, and then that helps you to bring the next group in.

So having more people who are really technically sophisticated on the campaign side will, I think, very much help.

I mean, one of the things that we're doing, which sounds pretty small, but we've been seeing pretty big effects, is, frankly, we just moved to the G Suite as part of the Democratic Party.

It sounds crazy that we haven't done so far, but I just watch document collaboration happening all the time right now.

It's insane. So there's actually a pretty big change that's just happened in six months of using the G Suite that we're hoping that we'll carry forward.

What does tech get wrong about politics and Washington, D .C.

or Belgium or wherever you want to take it? And then the corollary, what does Washington, D.C.

get wrong about tech? I mean, so what we got wrong about politics, like my team came in, we came in about August of last year, and we came in being like, well, we know tech, obviously.

We've built self-driving cars, we've built social networks, we kind of know what this tech thing is.

We were wrong. We didn't understand the space.

I think for a bunch of us, we even forgot the first thing, which is talk to your customer in a bunch of different ways.

Go canvas, go knock some doors for a while, go cut some turf who you should go talk to, and just see what the real problems are on the field, because these old school organizers are actually doing an engineering problem.

They're just doing it on pencil and paper, and we can make it faster in a bunch of different ways.

So really understanding that, we took that for granted.

They didn't need a sea change in what they're doing, and maybe in G Suite is one example they do, but what they really need is a way just to make it faster and more efficient and more effective in a bunch of different ways.

Yeah, I think that's a good sort of respect your customers is a great thing that sometimes we don't do on really either side.

I think in terms of the worlds are relatively distinct, and there's maybe less crossover, there's more now, but less crossover than we'd like.

And so I think what that leads is to maybe not fully understanding the types of motivations that are in people's heads, and start coming up with your own.

So we often, I think people often characterize when they're in D.C.

techies as being really motivated by money, because that's the IPOs is kind of what we see from D.C.

of Silicon Valley, and similarly the other way around, right?

Like Silicon Valley, what we think about D.C. is often all those people are just motivated by power, like there's not.

And certainly the thing that I've seen is that both are motivated with trying to make the world a better place, just bringing different tool sets to it, so the sooner you can get to that conversation, the better.

One of the things that we did pretty differently is that the tech team for the Democrats, we're not actually in Washington, like we're actually spread across the United States, and we're not even like in the tech centers, like Silicon Valley has the smallest population of my team, we have people in Chicago, Seattle, Portland, Dallas, so we're everywhere in a bunch of different ways, and that's actually been great for us, it sort of has forced the D .C.

people to acknowledge that there are people outside D.C., and then it's also allowed us to attract talent in a bunch of different ways for people who for different reasons can't move to D.C., but are so passionate about wanting to help in some way that we can just get them where they are.

What went wrong in that, and in terms of how tech influenced elections, and what's the right way to think about that going forward?

I mean, my perspective these days is that it's not that the Internet is broken, but it was built with a different set of assumptions than are needed for today, like we don't have the privacy guarantees that I actually think people deserve when it comes to the Internet, it was built on a trusting system as opposed to a system that had good checks and balances in place in a bunch of different ways, so I think that's fundamentally what's gone on, it's just like we've taken a very trusting system and sort of have manipulated it to be the world that we're living in today, so the Cambridge Analytics situation I think is the best example of just a whole bunch of people trusting an app, but now that data is effectively weaponized against them in some way, so I can tell you what I think we need is I actually, I'm one of the few technologists I think you'll hear, I'll say like, I think we actually need better regulation in this space, like we actually need a counterbalance to a lot of what Silicon Valley types or technologist types would go do in this world.

Yeah, better regulation and more diversity within the profession would be really good starts towards getting to a better product, I guess the other thing I would say is that a lot of the, and this was definitely true of me, so I'll just talk about myself here, which is I worked on a project called Google Book Search, which was trying to scan all the world's books, make them easier to find, one of the things that we did was enter into a settlement to try to open up more of those books, and that eventually got struck down by the courts, but the reason why I'm telling this story is because that was at a time when I joined Google when it was about a thousand people, at the time we were doing this project I think we were somewhere around five or maybe even ten thousand people, when I joined Google we were a little bit more of a plucky startup where disruption is kind of a good word, and by the time we were doing this project we were way more than 90% market share in almost all of Europe, and we just were a different thing, and none of us had figured it out, and so we sort of behaved as if we were these young plucky kids who could try some things and not damage a lot, but maybe come upon something useful, and yet the people we were dealing with and the impact we were having was very much like wrecking people's lives, it was a totally different thing, and coming to that understanding earlier in the life cycle of a company and a technology I think would be great.

Is that possible, is it sort of like you just got to go through your awkward teenage years and that's what Uber is living through right now?

I mean I think you can go through your awkward teenage years quicker if you listen to people more, and if you have more diversity within the organization itself, not just in terms of ethnic diversity but also just in terms of age diversity, like having some people who've been through the rodeo before, I mean that's one of the things that Cloudflare actually does a pretty good job of, but having those people is really really useful, and we don't put enough focus on that, on the wisdom, the hard-earned wisdom of years of service in terms of how tech generally operates.

I mean there's also something to be said about the intangible nature of tech, so I don't think when we were at Twitter we could imagine sort of where it ends up today, but in my experience at least with self-driving cars, we turned on the Pittsburgh fleet without telling Pittsburgh, we just were like, the cars are going to go live today, we learned very fast that was a bad idea, because you have a physical thing on the road that's like driving around, like I wrote it.

I mean and it's like that makes sense in some bizarro way in Silicon Valley, nowhere else in the world does that make any sense, right, and it's amazing how blinded we are, you know, you get into those bubbles.

I mean I think like Pittsburgh is actually an important part of this, because like it wasn't Silicon Valley that we turned it on, we turned it on effectively in the Midwest of the United States, and like people were like very fast being like WTF, like you put a car with no driver on the road, so like I think like that tangible nature and sort of like getting out of that bubble, getting diversity in that perspective, I think like is what's going to help.

So I know it predates your time at the DNC, but in the last election there were some real challenges.

Can you talk about kind of what you saw when you when you came in there and what some of the mistakes that the organization made?

I mean a lot of ways, so just to set the context for a second, like you know Secretary Clinton famously said after election data was bad on the Democratic side.

We had a hack which the Russians infiltrated the DNC servers in a bunch of different ways, so like we're batting from behind, so like and a lot of that is also when you have, you know, when you have a set of people who are who are building technology but not necessarily staying at the forefront, right, technology is not the focus, but like in this world where everything is powered by technology, you actually need excellent technologists running the entire shop top to bottom.

So like we had passionate people who really cared about electing Democrats running the ship at the DNC, but what you need is people who are like skilled, experienced, steady hand in the situation and have seen a bunch of stuff before in order to like combat nation states in a bunch of different ways.

Were the Republicans just way better at it than the Democrats, and if so why and how?

Yeah, I mean one of the things about technology, at least in electoral politics, is like it's always a game of leapfrog, right, so like the 2008 and 2012 President Obama campaigns were like the vanguard of political technology at the time, and so in a lot of ways the Democrats just thought we were ahead and sort of like frankly rested on our laurels for a bit, and the Republicans were just like, we're going to copy that playbook, and they did, and they did really well.

So now they've leapfrogged us in the last four years, so like we're just in that exact same situation.

Like the Democrats have, the Republicans have spent something like $175 million on data and technology upgrades since the last cycle alone, and we've probably spent on the order of 20.

So like we have, like there's just an imbalance, just like a realization, because the problem also in the space is like, you know, what's your KPI, what's your OKR, and a lot of people in electoral politics is like, did I win an election?

But that only happens once every four years, so that's a really bad metric.

So like trying to figure out like what's the intermediary thing so you can like track how you're doing and make sure you're always improving is also pretty important, and we're just getting around to thinking about it like that.

And so how is that, what is that, how does that change over the course of, I mean is there, is there an optimist story for the Democrats or, you know, is this four more years of Trump?

No, I mean, I think like if you look at just election results right now, like I'm probably the only U.S.

election nerd in the room right now, but if you look at election results in the U.S.

right now, like turnout is incredibly, is incredibly up for a midterm election.

So the president gets elected every four years, on the off two years is when we do the majority elections for the Senate and the House, the turnout is incredibly high.

Democrats are coming out in droves to do elections.

So like there's a strong sense of optimism right there, and we've been winning a bunch of elections.

On the technology space alone, like political tech wasn't really an industry four years ago, but like up now there's like 500 different applications trying to, like trying to get their, 500 startups trying to get their tools into the hands of candidates in order for them to win.

Like we're seeing a surge of technology in the space just desperately trying to help Democrats.

So like, yeah, we're very optimistic at this point. It's hard. I'm not sure how many of these will translate to, to this audience, but there's like a, this app that will tell you which five phone calls to make every day, right?

That didn't really exist as now this progressive app that people wake up in the morning, they look at their app, they make phone calls to senators or congressman's offices on specific issues with the help of the, the app telling them here are some of the talking points that if you agree with, that'd be good to use.

And these are things that actually, you know, oftentimes our elected representatives might not spend as much time in their email, but they are still looking at like, how many phone calls are we getting on which issues?

So you have people creating little applications that are making huge difference.

And then you have people creating little types of organizations.

I mean, even you look at the women's March, which has become this hugely powerful thing in the United States to activate people into politics.

All of these sort of new startups that are, that are being created from the energy that came with the wake up call of the, of the last election.

You know, you talk about this as, and from your perspective, you know, being, being kind of on the democratic side of, of, of things, that this is a positive thing.

But if you're telling me that what you guys are doing is recreating some of the Republican playbook, I mean, I come back to Cambridge Analytica, and it's how, how, how do we have a responsibility with, to use technology in politics in a way which, which is actually good, a positive thing.

And that we don't have, that you, that we don't get all trapped in our filter bubbles and that we aren't suppressing, you know, voter results and, and all of the just parade of horribles that we've, that we've been reading about in the news over the last, the last few years.

Yeah. I, I, I think it's probably wrong to look at our political campaigning organizations as the place that are going to break the filter bubbles.

There are a lot of really great experiments that are being run to try to do some of that work.

There's a, organizations being built to try to allow, not allow people, but like encourage, emphasize the crossing of those bubbles and the conversations between them, making them both more civil and, and more productive.

But I don't think you're like, it, it's probably wrong to expect, particularly in the U.S.

where voter turnout is such a big deal.

Basically like a lot of us don't go to the polls. And so if you can just get more of your people to go to the polls, that is.

Or fewer of the other person.

Yes, that, that can happen too. But if you can get more of your people to the polls, that can make a, that can make a big difference in the election.

And that might be a little easier than crossing the aisle.

You do see a much bigger and you know, Raffi is probably the right person to speak to this is not me, but you do see a much bigger attempt, particularly among the Democratic Party to get out and be more competitive in more races, which is part of that breaking of the filter.

It is part of that. How do you show up and be part of the conversation and put your ideas forward?

Because that's another thing that's really kind of exciting in this particular moment in the United States, which is that the typical political boundaries and viewpoints are breaking down a bit.

I mean, you, you do have a lot of Republicans asking, I'm, you know, for small government and not huge deficits, how do I deal with the fact that my party doesn't seem to be for that anymore?

And I've always been for a lot of the democratic social issues. So maybe there's a place for me, which is, again, I'm looking at this with a little bit of rose colored glasses, but that is exciting that we're getting to be a little bit more fluid in our politics the way that a lot of countries in Europe already are.

So I'm going to take a couple of questions from the audience, but first, Alex, I'm going to throw a grenade to you, which is what grade would you give Trump on tech?

There's a reason he's asking him, not me.

Well, first of all, I would have kicked him out of the class for poor behavior on all the other issues so that I didn't have to grade him on tech because those other issues are extremely important and he's failing them.

On tech, there are basically three things I want out of my government on tech.

First, I want them to deliver the government services and the government's obligations well.

With the most recent tech, I would love it to be as easy as buying a book on Amazon to get my driver's license and all the other good stuff.

Number two, I want them to do good on tech policy, things like encryption and net neutrality.

I want them to be smart about how their role as regulator can improve privacy and set up a better platform for everybody to innovate.

And then number three, I think we're living in this age of technological dislocation and the changes that are happening to society that are not about, can I get an Uber on the street in London, but are more about, will my Uber driver have a job in 10 years as we get to self-driving cars?

Those societal dislocations are extremely important and need some just good old fashioned government thought about how do we build a safety net in that situation.

Not to avoid your question, I think on the first one, maybe it's a B plus, A minus.

I'm actually pretty happy. He continued a lot of the Obama era programs.

He was able to keep a bunch of great people in those spaces. He's brought in a few good people, Matt Lira would be one I'd point to, who are trying to push forward the ball of how do we get veterans their benefits better?

How do we do some of the very straightforward things like an open source policy for the military with that launched relatively recently,, a U.S.

open source for government launched relatively recently.

These are all good signs. There are some places where it's been horrendous, particularly around transparency, where we've really pulled back some of the value of some of these government services.

On the other two, just from my perspective, an F on both, there really aren't people there.

He hasn't hired a chief technology officer. He doesn't have a director of the Office of Science and Technology Policy.

He hasn't brought in people with great expertise there.

On the tech policy, I don't, and net neutrality is a great example of a place where that is clearly a stumble.

Not happening, and then it doesn't seem like any attention has been paid to the changes that are happening in our country and around the world that are brought along by technology, but we have to have a real hard look at, and it just doesn't seem to be there.

On that somewhat unsettling note, we're out of time, but thank you so much for coming.