Digital Diplomacy Conference Keynote

Sharp Things CEO Adam Sharp kicks off the two-day International Conference on Digital Diplomacy, hosted by the Israeli Ministry of Foreign Affairs, with a keynote discussion reflecting on his career at Twitter and social media's impact on how governments communicate.  


DJ Schneeweiss

Are we on? We're here?

Adam Sharp

Yep. Hello, hello.

DJ Schneeweiss

That's fine, okay. Good evening, ever.

Adam Sharp

DJ and Adam, the VR experience.

DJ Schneeweiss

I'm sorry?

Adam Sharp

DJ and Adam, the VR experience.

DJ Schneeweiss

Yeah. It's not VR, but it is a 360 degree camera which Allad acquired for us today. We always try and be at the cutting edge of everything. This camera will allow us to record things, and then obviously make sure we only put out the good bits. You want to all be careful because it's 360 degrees. You're all on camera too. Not just us too.

Speaker 3

360 and 3D.

DJ Schneeweiss

And 3D, there you go. We have, it's the full deal.

Adam Sharp

I'll gesture a lot more. Give the full effect.

DJ Schneeweiss

Welcome, ever. It's my pleasure to serve as host of this opening session, this fireside chat without the fireside, but with Adam Sharp. Allow me first though to quickly introduce myself, my name is DJ [inaudible 00:01:28], I was born in Australia, as you can probably tell from the accent. I moved to Israel 30 years ago, and have been a diplomat here with the MFA for the last 23 years. I'm currently the director of digital diplomacy here in the foreign ministry.

I wrote here Adam Sharp was born in, and I forgot to ask you, where you were born?

Adam Sharp

I was born in New York.

DJ Schneeweiss

In New York. Adam Sharp is a native American, but not only is he an American and a proud one at that, he is called the human embodiment of Twitter. He was called that by the New York Times. That is based on the fact that he's forged a distinctive career of more than 20 years at the intersection of media, technology and politics. From 2010 until 2016, as Norm mentioned, Adam was head of news, government and elections at Twitter, advising journalists, news organizations, candidates and government officials in more than 20 nations.

He became the longest serving member of the company's global media team, and its most visible broadcast spokesperson. Adam appears regularly as a noted expert and speaker on digital communications and marketing, political affairs and strategy, and issues related to fake news and misinformation. I wonder why.

Recent appearances include events hosted by the New America Foundation, the National Democratic Institute, International Leaders Forum, the George Washington University Graduate School of Political Management, and the Harvard University Kennedy School of Government.

Prior to joining Twitter, Adam served as an assistant, as an aid to US senator Mary L. Landrieu, Democrat of Louisiana, from 2004 to 2009. First as a spokesman and later as deputy chief of staff. He played an integral role in various Congressional initiatives, including the Senate Centrist Caucus, the Gang of 14 Agreement to approve supreme court nominees, and delay for nearly a decade the use of so called nuclear option.

He is also part of the creation of the bipartisan compromises on the federal budget and energy policy, and for the response to the hurricane's Katrina and Rita.

Adam has held leadership roles at NBC and CSPAN and is currently the vice president of the National Academy of Television Arts and Sciences Foundation, a board member of the National Press Foundation, and a term member of the Council on Foreign Relations. I didn't realize how long this was going to take to read.

Adam Sharp

I was about to read, we've got time for one more question.

DJ Schneeweiss

Yeah, thanks for coming. Adam started, he founded Sharp Things in January 2017 to continue providing actionable strategic advice to companies and individuals whose objectives cross the domains of media, technology and politics.

Ladies and gentlemen, please welcome Adam Sharp to our conference. I'm going to start with the first question, which is going to lay the table for us, I hope. Adam, we're already into the second decade of the digital, of the social media age. Facebook is 13 years old, Twitter I think is 11, and these guys are no longer a novelty in politics and in global affairs.

Yet somehow, we continue still to be surprised by the scope and the depth of their impact, particularly in the last year. Certainly more than ever before. As someone who's been intimately involved in the evolution of the interplay between social media and politics, I'd like to open by asking you simply to give us a general sense of where you think we are at the moment. What's our current status?

Adam Sharp

I think we are still learning. Particularly in government and politics, where it should surprise no one in this room that government bureaucracies tend to be slow to change. Very often, a lot of innovation in these spaces comes from other sectors first, but then when you see them used in government and politics, we now see the meaningful impact and we see it at a scale we didn't see before.

I think recent elections are a perfect example of that, where techniques and methods that have been commonplace for several years, in commercial uses of the platforms, really had an impact.

What I think we are seeing in social media, across platforms, but also across all media in all industries, is a new age of authenticity, if you will. We see this in how people interact on social media, how they interact with television. We've seen it in trends in design and even in restaurants, and the menus and the atmosphere they create, that there seems to be a rejection of the mass media age of the last century, of that very broadcast-driven model of communication, where someone sits in a studio and just blasts it out. Those who have brought that broadcast mentality to social media, have generally not been as successful as those that see social media as an opportunity for retail connection, that one on one communication, but, with that visibility of scale.

It's that sense of authenticity that makes stronger connections, in politics, in governing, and elsewhere. One other point I'd add is, I think this has also made our consumption of media a bit more personal in the regard of how we choose where we get our content. In the media age, you had channels, you would tune to one channel for movies, you'd tune to another for news, yet another for sport. Same thing with the newspaper, magazine you choose to read.

Now, when we choose our media diet, it's not so much the channel and it's not so much based on the content, so much as it is who are we sharing that experience with. We wind up instead of the sports channel, the news channel and the movie channel, having the friends and family close channel, the loose acquaintance orbit, and then the global orbit. Right now, I would say that each of the major players have been very clearly aligned to one of those each. Facebook and Instagram own that friends and family space, that looser acquaintance orbit is probably LinkedIn, though there's a few other players getting in there around different types of connection. Then Twitter obviously being the one that's about connecting with the world.

Each of those companies, while being very successful in those orbits have struggled when they've tried to enter the other ones. Facebook has not been as successful trying to be that real time pulse of the world, even as they modify their news feed. Twitter, even with adding pictures and stickers and everything else, has never been able to become quite that friends and family tool that Facebook and Instagram have.

DJ Schneeweiss

Fascinating. The bit about the platform's very interesting, but I want to pick up actually on what you said before about authenticity. The elephant in the room, when we talked to Twitter about the 2016 elections, President Trump, he seems pretty authentic. What role do you think that authenticity played in the 2016 election?

Adam Sharp

I think it made the difference. I think that Hillary Clinton's campaign was the picture perfect textbook definition of a 20th century mass media campaign, and had she run that level of campaign just a few years earlier, it would be in the textbooks as the culmination of everything we learned about the heavy on TV, mass media broadcast style of campaign. The packaging of candidates the way we package commercial products.

Unfortunately for her, she was running in the year when the public rejected that form of marketing, and had already been rejecting it for several years in more commercial sectors. We have come to expect a more personal interaction with brands. It started with when I open Netflix or Amazon, I have my own shelf to choose from. It's not just the same selection that everyone else has to walk past in the store. We've seen it in customer service and how people are turning to platforms like Twitter to engage with airlines and the cable company, and these other industries that are known for poor customer service, and yet have wound up being more successful in social media than they have been in other sectors.

There was a study that found that those who reached out to an airline for customer service over Twitter, on average were willing to spend 20 to $25 more per ticket on their next ticket just because of that customer service experience. There's really no other platform where companies tend to come out better on the back end of a customer service call.

That connectivity with the brands and celebrities and with everything else in our culture, became expected at the time Democrats put forward a nominee packaged perfectly as so. Donald Trump on the other hand, while he was probably the master of that very packaged view of the '80s and '90s, with the name in gold letters on the façade of every building and product, pivoted to this new age quite comfortably. I think the experience through The Apprentice and the celebrity status he earned through that, certainly supported that.

Some of the pillars that contribute to this sense of authenticity, immediacy, relevance, engagement, he demonstrated quite readily every day, even if it was a little bit off the rails by any other traditional political definition. Trump's allergy to staying on message spoke directly to that population that was frustrated with very messaged campaigns.

Twitter did a study in, I want to say 2013, where we looked at the Twitter accounts of 3000 politicians around the world, and looked at the factors that went into someone retweeting or unfollowing, and found the single most telling question you could ask someone about a political figure on Twitter, was, do you believe they tweeted this themselves.

If the answer is yes, it was the single strongest predictor of any data point we measured of retweet. If the answer was no, it was the single greatest predictor of any data point we measured of an unfollow. For the universe of politicians where we on our team had a pretty good idea of who did it themselves and who didn't, we found the audience knows as well. Even when the staffers are very good about putting it in their boss's voice, there's just something there.

In fact, sometimes when staffers do put it in the boss's voice, it's too perfect. It's too real. It loses that authenticity. There's a senator in the United States, man named Chuck Grassley, 80 something year old Republican from Iowa, not necessarily what you'd think of when you think the core Twitter demographic, and yet one of the best Twitter accounts in the Senate. He runs it himself, he won't give his staff the password. He has never spelled a tweet correctly. He routinely makes accidental dirty jokes that he doesn't understand are dirty. He goes on these rants about the History Channel late at night.

One of his staffers was concerned about this. The last time they did a focus group, they said, does it bother you that the senator has this Twitter account and sometimes goes off like a crazy old man? The focus group came back and the consensus was, that's how we know it's Chuck. If he started sounding like a great orator with perfect phrasing and so on, they would know it wasn't authentic.

To bring this answer to a close, I'll share a story incidentally also from Iowa, to draw a critical distinction. Authenticity is not truth. That is something that I think we've also learned from the Trump campaign. I met a woman from Iowa during the election last year, and I asked her what she thought of the candidates. Her answer was, "They're all liars." I'm like, okay. You have to vote for one of them. Who are you voting for? She said Donald Trump. I said, do you think Donald Trump is a liar? "Absolutely." Why are you voting for Trump? "Because when he lies, he lies from the heart."

Right there, I will put forward, is in one sentence, everything you need to know to understand the 2016 presidential election in the United States. The majority of voters thought both candidates were lairs. In Trump's case, his supporters felt he was lying from the heart, and they felt that Hillary Clinton was formulating it through focus groups and marketing teams in Brooklyn.

DJ Schneeweiss

Okay. Point made. Well made. What about the question of the manipulation which we've all been reading about, the bot posts, how the system was gamed and influenced by foreign interests? In many ways, this is the exact opposite of the authenticity you're describing. Can you wade into that for us? Can you elaborate on this seeming contradiction?

Adam Sharp

Yes. I think the answer is yes and no. When you see these bots and automated accounts, in particular, operate, and I should be clear here, through most of this conversation, we'll probably be using the words bots and AI and automation in the pejorative. There're also many very good automated uses of the platforms and of bots, and even in diplomacy, there are very strong and important automated systems. I don't want to cast a wide net saying that all automated use of these platforms are bad.

What we see though with these, is that, and particularly on Twitter, those that are actually creating a false account to put out a message, and we see this on Facebook as well, but it's a lot easier to measure the dynamics on Twitter, that communicating a message from a bot or automated account is largely ineffective directly to a user. If I create a false account and that somehow shows up in a search or a timeline for you, it's not going to be very impactful.

What is more impactful is when a friend shares something. Someone you do have relationships with, that you follow, that you have a trust with. If that tweet gets retweeted into your timeline, or liked or reshared on Facebook to reach your newsfeed, now that message is wrapped in the credibility and authenticity of your relationship. That is where these two notions winds up being compatible, because the message itself isn't authentic, but it gains that authenticity from the state of your relationship.

The other mechanism that we see are then ways of trying to get to that, and the dynamics you see are accounts that are created simply to retweet and like and drive the engagement up on other accounts. Sometimes, that may be another bot account just to try to elevate it in the search algorithms, so that friend of yours can discover it and reshare it, and that the influencers in your community can discover.

Or what I think is more common, they find someone who already is an influencer that shares a world view that they support, let's say it's a political candidate, a columnist, a member of Congress, what have you, and retweet them. Boost their engagement. It's a free lift for those influencers. They already have the authenticity. It makes sense for the actor to want to have that out of reach.

Then the last mechanism we saw was the purchase of advertising, which is essentially another way of getting to that patient zero, if you will, where these actors bought ads on Twitter, on Facebook, and elsewhere, not very large ads, mind you. When they reported that this internet research agency in Russia allegedly had spent about $150,000 on Facebook, the reaction from Trump's digital director was, I spent less than that in a day. This was the entire spend. Most of the buys were less than $1000 each.

Because their goal was not to reach a voter directly, the goal was to identify some of these patient zeros, some of these influencers, so they could get it into the bloodstream and then ride their authenticity and credibility as they liked and reshared to their networks.

DJ Schneeweiss

What it seems to be, and I'm guessing to many other people as well, is this is a vulnerability if you like, of the company. There's an open door to manipulation, or there can be people with a less than positive agenda, can get in through the back door and the companies are there and letting it happen.

What can you tell us about the response of the big companies, and can you tell us if in your view now, you're a year out from your time inside Twitter, how much can you tell us, in your point of view, in terms of how adequate the response is? Do you feel that the companies are making headway and that we're heading in the right direction, and if I'm already on that question, can you also share with us what your view is of government and how government is interacting with the companies, and how together and separately they're addressing these dynamics? How does it look to you as someone who's seen it from the inside and is now looking at it from the outside?

Adam Sharp

As someone who started his career as a journalist, and then worked in government, and then worked at Twitter, you could say my take on these issues, if I were sharing it on Facebook, I would choose, it's complicated. I'll often take a step back and say yes, absolutely, this is a vulnerability of the platforms, it's a vulnerability of the Democratic process. It is a vulnerability that companies must meet, that nations must meet, and one where neither of those two groups has done nearly enough.

When we get down into then, what are these solutions, and what do those solutions look like, that's where it gets cloudier, because it is very hard in my mind to divorce the mechanics that allow this improper use of the platforms, from the mechanics that allow all the good things that we see on the platforms. The ability for anyone with a political message or voice they want to express, to have access to an anonymous global microphone, to have their voice heard, and be distributed, is one of the most powerful facets of the social age, and it is one that has certainly had an impact in this region, it's one that has had an impact in the United States through the Black Lives Matter movement, and more recently the Me Too movement.

Yet, every element that made those things possible are the hands holding those doors open to the malicious actors. It gets very difficult to then say, how do we start carving this out without imperiling that. The only answer anyone is able to come up with is, the companies need to moderate content, and that the companies need to hire as YouTube said today, they're going to hire more content moderators on top of the 10,000 they already have, to sit there and be the ones deciding who gets through that door, and what messages get through that door, and to which audiences.

Looking at this room, can any company in your full confidence, hire even a single person who could make a judgment on a political message of global significance, that adequately incorporates the laws, history, and cultural standards of each of your countries, is able to make that decision several times an hour, and now can that company find 10 or 20,000 of them and can three or four companies do that? I would argue the answer is no, that there is no human solution.

The next step is algorithm, but what trains the algorithm? The multinational nature of the platforms starts to make that very difficult and unscalable, which is one of the reasons why the companies have long taken the position, we won't moderate anything, because the moment you draw a line, you have to defend that line, and why is it here and not over here.

Eventually, they're going to be forced to draw it. I don't think there's a clear path to do so without giving up some of these other things. That winds up being the same liberty versus security argument that has vexed governments and institutions for ages.

DJ Schneeweiss

Yeah. We're going to be talking about some of those issues as the conference goes on, obviously. Maybe we'll find the solution here, between that balance between the governments and the platforms. Maybe not, maybe we will. We can be optimistic.

In talking of the interplay between government and the platforms, I'd like perhaps to take a step back and ask you to tell us a little bit, exactly, about what your role at Twitter exactly was. What does the head of news, government and elections at Twitter, actually do?

Adam Sharp

As soon as I find out, I will ... I often say I worked for three or four different companies in my time at Twitter. I started in 2010. At that time, the company was less than 300 people. I don't think we'd opened an international office yet. Congress, only about one in five members of Congress even had Twitter accounts, staff-run or principle-run. Only one or two of the cabinet agencies did. The job was very much going door to door, introducing people to Twitter and trying to make the case for, this is why it is important.

Is interesting at that time, I would go to one member of Congress and do a whole spiel and show them some stuff, and they'd say, "Oh yes, this is wonderful, great." I'd walk out saying extremely, nailed it. Now going to another member of Congress, neighboring district, same state, same party, and I think oh, I'm just going to give them the same speech, this is going to be easy. I sit down and I start showing this guy how he can search for what his constituents are saying and stuff. He's, babababababa. "I'm not here to listen. Constituents want to hear from me? They can look at the tweets. I get enough from them as it is."

I got many variations of, I don't care what Justin Bieber ate for breakfast. It was a hard slog, that first year or so. Interestingly enough, it was easier on the Republican side than the Democratic side, which was unusual for the period, because this is 2010, Obama had been elected in 2008. Most of the cool factor on use of social media had been ascribed to the Democrats.

But, for all of the dynamics we talked about in the last question, the Tea Party movement started to take off on Twitter. These Tea Party voters that were not new, they didn't just magically arrive in the United States, these voters have always been there, have always been a factor in elections, but had never had the infrastructure to organize. They weren't part of the establishment Republican party, they didn't have a large enough groundswell in their states.

Now, once you had social media, and in particular Twitter, that could connect you to like-minded individuals across borders, and allow you to organize, now, they started to become effective. Republicans in Congress rightly recognized, this is our new base and we need to tap into that.

In later years, because by the time I left, every member of the Senate was on Twitter, all but three or four members of the House, every cabinet agency, half the cabinet members themselves, many of the diplomatic core in Washington, we worked with a lot of ambassadors there and with the team.

At that point, it became a lot more strategic. We would sit down with people on both sides of the aisle, hearing about a new initiative they're rolling out, and talk about okay, how does social media play into that. Started doing more with the UN around the General Assembly. Some bigger events.

Then of course, for elections. That's, in my view, when Twitter really shines because Twitter can put you on that campaign bus, seeing real time through the eyes of a reporter or a voter in another state, what's happening. That's where we wound up being very involved with debates, real time questions and follow up questions from voters. In this election, we had the first time in a presidential debate, this is one of the primary debates, where not only did a question come from social media, but it had come as a followup question in real time.

Hillary Clinton had used 9/11 to defend donations she had taken from Wall Street in her campaign. In the control room, we could see we were partnered with CBS News on this event, just a massive spike in activity, that this was handily the most tweeted moment of the debate. We could look very easily and see that the sentiment was hugely negative for Secretary Clinton, on that point. We were able to find a few tweets from people in Iowa, asking followup questions that we were able to vet and turnaround.

Eight minutes later, Secretary Clinton was presented with a followup question from an Iowa voter to her earlier statement, and where's it'd normally be that out for a candidate to say, oh, that's just you and the media, I'm going to move over here, no, this was now a voter in the room. A man we found out later was an undecided independent voter in his pajamas live tweeting a televised event for the first time ever, who eight minutes after he asked his question, had it asked of the eventual nominee on stage.

Secretary Clinton then declined to participate in any debate that would take Twitter questions ever again.

DJ Schneeweiss

Really? The question of how you guys interacted with the campaigns has also come up in the news. We've seen reports about how Facebook was embedded with the Trump campaign. Can you tell us a little bit how that embedding works, whether Twitter was also engaged in that sort of thing, and also if you could, given that we're an international audience here tonight, has Twitter also been engaged in embedding itself in campaigns, in other countries as well, political campaigns?

Adam Sharp

The embedding that you're talking about is a function of the sales teams. If you are not a campaign, if you are a big brand, I used the example of soap earlier, so I'll use soap again, you are a soap brand and you have a big commercial airing during the Super Bowl. It is very common for a brand to then form a social media war room, get everyone into a room, pizza and beer, you're watching the game, be tweeting and posting all evening, building anticipation for then ad. Then as people respond to the ad, be engaging with them and acting on it, seeing if some reference to the ad starts to trend, so that you can move quickly and buy an ad against that trend, and so on.

It is very common for all the platforms to put members of their sales teams in that room with the brands, because when you are trying to react in real time like that, you don't want to lose a single moment to try and to get your ad rep on the phone and so on and so forth. They can be there looking at the Twitter backend or the Facebook backend and say oh, this is now trending, and help identify opportunities for you to optimize your spend, or as I would put it, opportunities for you to give us more money.

There's an incentive to the brands to have people in there with them, because they can execute more quickly. There's an incentive for the brands, for the platforms to be in the rooms, because they can upsell and say oh, you should be running on this, running on that.

In the campaign, the same model transferred over to the larger campaigns. Particularly presidential, who were spending on a level equivalent to that of these major brands. On debate nights, at the conventions, and these other times where the campaigns would want to react quickly and those ad buys, Twitter, Facebook, Google all offered members of their sales teams to be in those war rooms and help drive that spend.

The Clinton campaign, I understand largely declined that kind of support. The Trump campaign largely accepted it. Yeah, that's the story on embeds.

DJ Schneeweiss

What do you think the impact is, your implications are, when one campaign takes on that kind of assistance and another one doesn't? Does it skew the outcome? Obviously you could say it's the Democrats fault, but is there a huge competitive advantage that then one campaign gains over another?

Adam Sharp

I don't think so. I think the biggest skew is one of optics, and it doesn't look good for the companies when the stories come to saying one did the other, and also because no company is going to want to come out and on the record, throw one paying client under the bus, just because another client paid more. It wouldn't serve the companies well to say we offered it to them, and they said no, and that's their fault and so on and so forth.

It puts the companies in a tough position of appearing to be showing favoritism, when they didn't, but they really can't mount that defense. I think that the, in general, the Trump campaign was more effective in its social advertising. Part of this was because earlier in the campaign, they didn't have the money of the other campaigns. In the primaries, Trump was outspent by pretty much every other Republican on that stage. Even in the general election, he was outspent more than two to one by Hillary Clinton.

For their campaign, from day one, they went to social media, simply because they felt they were getting a greater return on investment, than they would with the traditional TV buy. The Clinton campaign went very heavy on TV, even more than Obama had. I think there was an edge to be had in being more nimble and more aggressive in digital advertising. How much of a difference two people from each company sitting in the room for a couple of hours every few weeks had as part of that, was probably quite small.

DJ Schneeweiss

To my followup, are you seeing similar things? Obviously it's on the ad buying side of things. Are you seeing similar things happening around the world, are other campaigns becoming more like the American campaigns and having this embedded social media strategy inside the campaigns?

Adam Sharp

Paid political ads, at least on Twitter, I can't speak for Facebook, have not been historically as big a focus at the company as they have been in the US. In fact, many of the countries in which we were active in elections, I spent time in Germany, Brazil, Japan, Australia, and pretty much every one of those counties had laws prohibiting, in some countries, paid advertising at all, and in others, paid advertising on digital.

The market was never really there. I think as those adapt, what we've seen in the political playbooks of most countries is, they tend to copy from the playbooks of the United States. I imagine that as political advertising online builds in some of these other markets, you will see campaigns try to adopt the US model, and the companies would be foolish not to rise to that business opportunity.

DJ Schneeweiss

I've got a couple more questions before I open it up to the floor. One of them, if you can give us a quick answer on, on the legislative discourse that's happening now in the states, the Honest Ads ad, is Congress about to regulate these ad buys in such a way that the political involvement of outside actors is going to be curtailed? Can you give us the lowdown on what's happening on that?

Adam Sharp

Sure. To give some background, in response to these alleged Russian ad buys on the platforms, three senators, Mark Warner, Amy Klobuchar, and John McCain, Warner is the leading Democrat on the committee leading the Russia investigation, introduced a piece of legislation that said essentially, social ads should be treated the same way as ads on other platforms, and all have a little disclaimer at the bottom that says paid for by the committee to elect Donald Trump, or what have you.

This has been a norm of television, radio and print advertising for decades. When the companies introduced political advertising in 2010/2011, they actually went to the FEC asking, do we have to have disclaimer. The FEC deadlocked 3-3 and couldn't come to a final answer, which meant it was a free for all. Google and Facebook took that to mean great, we don't need to have disclaimers. Twitter did have disclaimers.

Then, ad buyers come and say, why do I have to go through all these extra steps on Twitter if I don't have to on Facebook and Google? No regulators telling you to do this, and it's the candidates and campaigns that get punished if they don't, anyway.

By the end of 2016, the disclaimer had faded away on Twitter as well. This legislation comes up, and now, all three companies have said they are going to start implementing disclaimers for the 2018 midterms and doing it voluntarily, they're going to build these transparency centers where you'll be able to click through from the disclaimer to see why were you targeted, what was the ad creative, who was paying for it, how much they paid for it, and so on and so forth.

Obviously, being on the outside looking in, I can't speak to their thinking. From my seat, my take on this is the companies recognize that there is momentum in Congress towards regulating them more. This is just the first piece of legislation, because it's the one that makes sense. No one in Congress knows how to write the law to keep that bot off the platform. No one knows how to write the law to prevent fake news, especially with the first amendment saying essentially you can spread fake news. There's no crime in that.

But, political advertising, that's something we know how to do, it's something we know how to regulate, so we want to move forward on that. The companies I think are trying to get out in front of it, to show they're doing something voluntarily because the last thing any engineering organization wants is a product spec written by a committee of senators in Washington.

I think their hope is, if we get out in front of it, that some of that moment for something will wane, because you're not going to lead the news having a press conference announcing that you're going to require someone to do something that to all outside appearances, they've already done, and that that can flitter away, with everything else that is flittering away legislatively right now, where odds of anything being brought to the floor that has Russia associated with it are so slim to begin with.

DJ Schneeweiss

Okay. My final question, then I will open it up you the room. Something that's very high on our concerns here in Israel, and that's the malicious content or, I like to use the word poisonous content on the web, and on the platforms. We're seeing this stuff riding the algorithms all the way to people's home, ever's home and ever's flat. The question is, what steps are being taken by Twitter and the others to address this? To address the volume of the poisonous content on their platforms.

If you can tell us, you just discussed about the legislation, you talk about the companies getting out in front of the legislation, self-regulating, to do enough to avoid the heavy hand of the regulator. The question is, are the companies doing enough when it comes to the actual substance of what's happening on the platforms, are they doing enough to deal with all of this poisonous stuff?

Adam Sharp

No. I don't think anyone has a clear definition of what enough is, either. We'll note when we see it, we're just no, we're not there yet. I think a cursory look at any of our timelines would confirm that.

The companies for the longest time all took a rather laissez faire approach to content on their platforms, I talked about this a little bit earlier, about not wanting to have to defend the line, so we just don't draw the line. You're seeing some echoes of that even in Facebook's response to fake news, where one of their leading experiments right now to get fake news out of the timeline is, what if we just take news out of the timeline? Problem solved.

Of course, the publishers who've been affected by these experiences of real mainstream news, get upset at that. There is also a legal paradigm in the US called the Communications Decency Act, that includes a line or two that is very preferential to the industry in a way I don't think anyone ever anticipated. It basically has two prongs. It says if you are an internet platform and people communicate on your platform, and you do nothing to touch that content, you cannot be held liable for anything that is said or done.

Then there's a second prong that says, if you do something, and you don't do it well or you do it wrong, you can't be liable for that mistake. If you suspend someone or silence someone you shouldn't have, they have no legal recourse against you.

Then my favorite line in this legislation is, even when such speech would be otherwise constitutionally protected, think about that for a moment, for those of you who hear our leaders wax on about the US Constitution, the internet companies are the only industry who have enshrined in law, an immunity that supersedes the US Constitution in its rule of law.

This law was written at a time, mid '90s, when getting online meant dial up. The modem makes the wonderful screeching noise, you have to tell your family don't use the phone for the next three hours. No one wanted to pay long distance, so you had to dial into a local modem nest. Who is going to create a local ISP if they can be liable for every one of their neighbors who dials in and does bad things?

The law, when it was written, was absolutely essential-

DJ Schneeweiss

To the establishment.

Adam Sharp

-to getting the country online. I'd argue it was absolutely essential to the current crop of companies succeeding. But, it also wound up being a permission slip where they were able to say, we don't need to worry about these things, because at the end of the day, there's nothing in force of law that is going to make us do it, so we're going to focus our engineering resources elsewhere.

DJ Schneeweiss

Can I just focus you, I wanted to get a sense of how serious the companies are about dealing with the content. How do they go about addressing the problematic content that they know is problematic. It's not a question of the legislation, what they ...

Adam Sharp

This I do as ...

DJ Schneeweiss

How did they do at the technology level also?

Adam Sharp

This I do as background to say, this paradigm, and not wanting upset it, is always looming in the room when having these discussions, because as it's been interpreted, you are either a platform or you're an editor. You are the phone company or you're a newspaper. The concern is okay, if we do nothing, then that first immunity is vulnerable, because regulators are going to come in and say, no, you should be protected unless it's this kind of content or this kind of content, and you start having that laundry list of content.

You don't want that. You want to be aggressive enough so that Congress never feels motivated to start chipping away at that immunity. If you are too aggressive, and you start moderating too much content, at what point does a judge say, you've established you're able to moderate all of this, so there's now no excuse for not moderating this little piece over here? Now, all of the sudden, that second immunity is blown up because now, you've become a newspaper.

In just the US domestic concern, those are two electrified poles that they're trying to steer between and get hung up on, and trying to create a consistent standard that is at once consistent, but also fair and not politically motivated, and not just representing a particular point of view, and saying that all other points of view are bad.

That's why you've seen the companies get, I think, more precise in the language of their rules, as a first step. Twitter in particular, their policies around what constitutes abuse, what constitutes a violent threat, those have now become a lot more clearly defined because I think they identify low hanging fruit of types of content they don't want to see on their platform, but couldn't act on before. Now, they've created hooks to hang their hat on in those areas.

Increasing reporting, so that users can report bad actors more readily, both Twitter and Facebook have made big investments there. Then, experimenting with are there ways to crowdsource these things. On Twitter in particular, I'd point to Periscope, that has a very novel approach to how it does, that content moderation. It's not a human being doing it. It's everyone who's watching that stream. From time to time, if a comment in a Periscope triggers an algorithm that says this might be objectionable, then a certain number of people watching that stream get prompted, do you find this to be a violation of the norms of the room, or of this stream. Then, the community is essentially voting that person off the island.

You may have a comment that in one context is consider inappropriate, but in another context where they're among friends or like-minded individuals, it's considered appropriate. The community sets the standard.

I think the companies are experimenting with things in that area, also to see if they can make things better. The last point I'd make is, when you then add the global element, it becomes a lot more complicated. I think the companies have always tried to take a globalist view on things. They have always tried to make decisions that are platform-wide, that some of the strength of Facebook and Twitter is that they look substantially the same as you go country to country to country.

As countries get more aggressive in regulation, that gets more difficult to sustain, but it also makes it difficult for them to make these decisions, because it then becomes, what happens in this country if we're asked the same thing? I took note for example a few weeks ago when Twitter announced that they would not allow Russia Today, RT, to advertise on the platform anymore. They put out a statement citing that the US government had declared them a foreign propaganda source.

It then raised, in my mind, the question of, what happens when the Saudi government comes and says Al Jazeera is a foreign propaganda source, or what happens when one in Turkey says CNN, and particularly CNN Turkey where they license the brand, is a foreign propaganda source, because they've licensed that American brand?

If you take what the standard is, if the host country says it's a foreign propaganda source, then we won't take the advertising. Do you apply that to every host country? Because there are certainly countries where from a US perspective, that would not be as palatable as perhaps suspending RT was.

DJ Schneeweiss

These are all fascinating dimensions. I'm not sure we've all thought them through, or've been aware of them. I'd like to now open it up to the floor, if an from the room has questions for Adam.

Adam Sharp

I'm going to turn this way for a moment, because I feel like I've been rude to this third of the group.

DJ Schneeweiss

We have a question for me, if you could just tap the red button on your thing, yeah.

Speaker 4

Yes. Thank you very much. I'm [inaudible 00:53:38] from the German foreign ministry. Thank you first of all for these very interesting remarks. I would like you to elaborate a little further on the question, how important are national rules, national legislation for a platform like Twitter, that is working globally? We've seen the US election, most of the fake news not coming out of the US, but somewhere out of the western Balkans.

First of all, I would like to know how important is it, and then do you already see a trend worldwide that governments are trying to reestablish their digital sovereignty in order to tackle these problems?

Adam Sharp

I think I'll be very Washington and say I'll do the second part first and the first part second, because the second question I think absolutely. I think when you look at some of the interventions in the US election, in the French election, not as much in the last German election at least to my understanding, is certainly on the global radar. Now, I should note that this is not a completely new concept. The tactics may be different, the tools may be different, but the strategies of these actors have been around for a long time.

The Russian government going back to the Cold War, has operationalized misinformation in foreign elections. Quite honestly, so has the United States, so have many other countries. But now, I think some of the big differences for the US in 2016 were, that was directed directly at the US, as opposed to the immediate Russian neighbors. Historically, a lot more of that focus has been on the part of the NATO Alliance that was east of the Atlantic, as a way of weakening their relationships back to the US, as opposed to directly meddling in the US.

I think that was a very tool set oriented limitation for the Soviet Union, and that now with technology, these actors found oh, we can operate in the US just as effectively as any other market, and in many ways, perhaps more effectively, because of the openness of the platforms.

I think the companies do have to take seriously regulation and legislation in the markets they enter. It certainly affects decisions like where they put offices and where they put facilities, because the moment they put people or buildings on the ground, now, they have some exposure to law enforcement in those countries. You've seen compromises there, as you probably know, Twitter has what they call country withholding, that if a tweet, for example, is reported to Twitter by a German court for violating German laws, say it's anti-Semitic, that tweet will still be visible in the US and everywhere else, but in Germany, it will be removed from the timeline, and in its place will basically just be a gap.

In fact, I've seen a few people sharing on Twitter this week, who are Americans, saying if you change your settings to say you live in Germany, you don't have to see all the Nazi references in your timeline. I think as more and more countries opt into this, that model worked when it was a country here, a country there. It doesn't scale when every country has its own set of rules.

That winds up becoming complicated. I hope for the sake of the users, that there becomes some international consensus around what those standards should be. Yes, absolutely, regulate the industries, but do not have it be completely different market to market, because then it makes it very difficult to communicate. If your phone system's not compatible with ours, we can't talk to each other.

That'll be my thought there. Certainly the EU's new standards on data protection and so on, I think are a first step that, a concerted view toward that.

DJ Schneeweiss


Sophie Michaelides

Good evening. Thank you for all the information. I'm Sophie Michaelides, the director of the press and information office of the Republic of Cyprus. My question is twofold. The first one is on, what is your stance on the ongoing conflict between the journalists, since your background is journalistic as I understand, the ongoing conflict between journalists and the society in general, as far as who abuses the digital environment more, as far as the poisonous content is concerned.

Who do you really think has more effect on election results, the journalists or the individual groups? If you believe that the opportunity of superficial involvement in participation or in the news dissemination, increases the percentage of no vote on the elections?

In fact, take a part of the gentleman's question, is I was present three weeks ago, I think, on the eighth of November, on the signing of the letters between the companies, and the director of the council of Europe, that's 47 countries, not the 27 of Europe, in actually finding common ground and common, I'm not going to call them laws, but I'm going to call them standards, in actually touching effectively the issue that we're talking about.

The question is, do you find differences between the way the America is addressing the issue, and Europe or the rest of the world, as far as finding those standards? Thank you.

Adam Sharp

Sure. First with regards to journalists and journalism, certainly something I'm passionate about as a former journalist, but also son of two journalists, a few generations back. It's a tough time for journalism in the US right now. Parallel and in concert with the successful Trump campaign, has been a constant attack on mainstream media sources, and so much so that now, mainstream media is seen as a derisive term, a negative term, in the US.

I think this has contributed to the success of misinformation, disinformation, so called fake news. Not because it elevates the public appreciation of the mis or disinformation, but it discounts the appreciation of traditional journalism to the point where now they share the same degree of respect. That, I think, is problematic.

Certainly, we've seen the dynamics of so called filter bubbles, whereby getting their news from Facebook and Twitter, which are the primary news sources for the majority of Americans, people are now self-selecting who they follow, who they're getting their news from. They're not necessarily following the news makers or the journalists, they're following these influencers in their communities who wind up, bring them to them, and those influencers are generally not the impartial that you'd expect an editor of a newspaper or the producer of an evening newscast to be.

They are not necessarily making sure you have your vegetables with your cheeseburger, in that you tune into the evening news, the number one and number two reasons people tune into their local news in the US are sports and weather. They come at the end of the newscast, and people sit through 20 minutes just to get to sports and weather, but they've now at least gotten 20 minutes of news.

For the majority of Americans who get their news from social media, they never have to see that 20 minutes. They only have to see that slice they're getting. As a result, that just feeds this notion of, then they must not be covering what I care about.

I don't know how we break out of this cycle. It's a cycle that is causing severe distrust in institutions, and particularly institutions of journalism, with nothing really to take its place. That creates a very imbalanced environment for communication.

To your last question, I do think that when it comes to comparing the US approach to other nations, the US essentially being an island, we have a northern border, we have a southern border, but in the context of the world, we're essentially an island. I'm just going around Jerusalem today, and recognizing that in what I would consider a day trip to a neighboring state, I could go to two, three other countries and have the same feeling in Europe.

There's geopolitical pressure to have these alliances and dialogues that don't necessarily come as first or second nature to American policy makers. Now, in the current administration where the mantra is America first, the fact that other countries may have the same idea, is often the quickest way to kill any solution.

I think to my answer here, where we do need to have some common standards, I think the fact that others in the world community are finding a way to that, is better than the approach that's going on in the US. The US then of course has the, I would say the challenge but it's a challenge that I'm blessed to benefit from every day, and that is the Constitution. The Constitution protects all manners of bad people in a way that very few countries do. That has served us very well, but it has also meant certain parts of our public dialogue that we're ashamed of, or certainly I am. But I wouldn't change a word of that first amendment, for all the peace in the world.

How that translates to the digital age, is one we have a hard time navigating. Whereas I think many other countries have a little bit more room within their laws to draw some boundaries there.

DJ Schneeweiss

Any further questions? Shlomi, make it quick, we'll make this the last one.

Shlomi Kofman

Quick two questions, actually.

DJ Schneeweiss

Shlomi, can you [inaudible 01:05:41]

Shlomi Kofman

My name is Shlomi Kofman, I'm [inaudible 01:05:44] concert general in San Francisco, including northern California and five other states. Two questions. One, if we talk about the companies, what is their strategy, basically, in engaging the world outside of the United States? How does it translate when they embark in meetings like tech against terror, in other forms? How serious is this engagement, and to what extent they really see it as a serious part of the game?

The other question is, you mentioned earlier the human factor in the algorithm. Who is training and who is responsible to train these guys, and to set up to tell them, what are the principles and worrying how do they change, and stuff. By the way, to add to tech against terror, is the new center which was established together with the ADL and the media network down in Palo Alto.

Adam Sharp

On the first question, I think the strategy of all these companies in how they view the rest of the world is no different than the strategy in the United States or in anything else they do. Their strategy is to drive value for their shareholders. Their strategy is to make money when you look at the markets they enter, it's usually first, the markets that they can monetize, the markets in which they can introduce some sort of revenue generating model to their platforms.

Second to that, they then are trying to create the environments that either support that or block those environments that detract from that. You'll see companies particularly Facebook and Google, just because they have more cash to invest than Twitter or Snap, you'll see them invest in markets that aren't necessarily monetizable today, but it is relatively low cost for them to make the investment today to have a lock on the market, so that when it becomes monetizable, they are the incumbent.

Certainly, I think that is the strategy through most of Africa and other lesser developed nations around the world. To that second part of my argument, where it's then trying to block those things that detract from that, poisonous speech like you talked about, abuse and harassment, terrorist content. All of these things threaten that ability to drive the business. Either because of public perception of the content that's on the platform and not wanting to be exposed to it, and people who quit Twitter or quit Facebook because of the abuse or because that language, or because it attracts the attention of governments that therefore want to regulate it, that because that content is there, the fear that a government will therefore make it that much more difficult for them to operate as a business.

I think that was part of the pressure that led both Facebook and Twitter to be more aggressive on terrorist accounts in the last year or two, and where Twitter I believe at last count had suspended seven, 800,000 of them.

In terms of training the curators, every company does it differently. You have a mix of contractors and staff people, but most of it is internal, most of it is dependent on where they're operating out of. One concern I have is as these companies now try to build 24/7 capacity to act on these quickly, that usually means that those human curators will be in different places around the world, following the sun around to employ staff.

That may mean that when you file a report on an issue of some content you found objectionable, that that will be reviewed by someone in the United States, or maybe in Ireland, or maybe in Singapore, or maybe somewhere else. While they are all operating against the same set of policies, whether it's the Facebook terms of service, the Google terms, the Twitter terms, as you match that policy to a piece of content and have to make that judgment, you're also bringing your entire life experience and cultural background into that.

I'd argue those three will not consistently make the decision, same decision, the same way, each and every single time. Certainly, when you add more diversity to the mix, which you want to do, you start to increase that variability even more. How do you solve for that?

That's just one of the scale challenges to the problem.

DJ Schneeweiss

Speaking of scale, I think we'll leave it there, because I think most people are tired. Adam's with us all tomorrow, so we'll be able to engage with him also in the breaks and so on. I'm going to thank Adam and ask him for one more question, and it's a question of scale. That is, are you a 140 guy or a 280 guy?

Adam Sharp

Oh. I was a 140 guy for a long time, but I have found myself accidentally using more of the 280 here and there. I like having it. I don't like it being abused. I'm hoping that over time, the people saying, "Oh, I can make a tweet three screens long," will start to fade away. I think good thing that it's available, but it should be used sparingly.

DJ Schneeweiss

On that note, time is available, but it needs to be used sparingly. Thank you, ever, for joining us this evening. We will kick off tomorrow morning here at 9AM. In your little programs on your tag, you'll see the times that the shuttle buses are leaving the hotel, and the shuttle bus this evening leaving here will leave within the next five minutes or so, I guess, to give ever a chance to get to their hotels and to enjoy the Jerusalem nightlife, which is just about to start kicking off, for those of you who are heading out.

Thanks for coming, and we looking forward to tomorrow's discussions. Good evening. Thank you, Adam.

Adam Sharp

Thank you.