Test. Test. Test. Test. Test. Test. Test. Test. Test. Test. Th really twitter. I dont think theres anything such as an antisensorship platform. You have a principles in place that dictate how your community develops. You look at the most laissez faire places on the internet, the silk road, even they had a prohibition on child pornography, for instance, selling nuclear weapons. No matter where you go, youre going to have certain prohibitions. You see the reason why 8chan went down is because of attacks because that makes it unusable. There are certain types of actions with websites that make them unusable that are not welcome. When you make those choices, you say who you are, what you will become and who you will welcome. I think saying that youre anticensorship, you probably arent. You just have a different agenda from the sites you claim to differentiate yourself from. Thank you. Next . Fascinating conversation. I was hoping you could elaborate on the antitrust discussion you were having earlier. The antitrust of at t it was successful with the rise of other companies to provide technology to create a competing product. I was curious on that discussion, what are the alternatives that might arise . Some people think its interesting. My partner who i do a podcast with, hes a professor at nyu, one of his arguments is if you broke up instagram from facebook, for example, or youtube from google, those are the ones that are talked about, you would create even more shareholders and more value for shareholders. I think a lot of people think that, that it would be a very value property. Aws makes 35 billion in revenue. It would be one of the most valuable companies on the planet the minute it was pulled out of amazon. Instead of making the argument this is punitive, its great for shareholders to create these new companies outside and then they would do innovative things. There used to be google video and youtube and at one point google owned both of them. And i think they closed down google video after the purchase. Thats one of the arguments, this is an unlocked shareholder value which i think is an interesting one and it unlocks venture money towards investments. You will try to get any venture capitalist to invest in any search company, Even Soft Bank isnt stupid enough to do that these days. They did we work which is having problems now. Thats the idea. Theres that concept is that it creates innovation, it creates investment, it creates value. Im sure its not shocking that i dont agree. I think one of the things that people forget when facebook acquired instagram, i didnt have spam filters. Instagram is what it is today because of its incorporation into the facebook system, right . While, yes, we are proud that it is an incredibly lucrative company, we feel its the synergies with facebook that has turned it into the Success Story that it was, that it is today. I think when it comes to the broader antitrust conversation, whats really missing from all of this is a actual definition of the market, right . People talk about its the social media market or its the Search Market or its the Online Retail market. But all of our companies are companies that offer a ton of individual services and among those services, we compete incredibly fiercely, whether its ads, online ads, google is by far dominant, facebook is second. But it competes with the physical advertising space, it competes with the broadcast advertising space. It is not that long ago that there was a headline that says yahoo has won the search war. I think we should be it actually was a long time ago. I wrote it. I didnt write that one. In the grand scheme of this evolution, its not actually that long ago. I think we should be careful about giving into this time its different arguments. Given that theres still incredibly intense competition in this space. I think you should let the little bird out of the nest and let it fly, that tiny little instagram, its time. Push it out. Im a firstyear ph. D. Student in psychology. I study the impact of technology on peoples lives and i was wondering, theres a standard that is set by each website and its free reign from website to website. How do we deal with the fact that you have a president who kind of moves away and breaches a lot of those standards, but we cant do anything about him because now its news and now all of a sudden his breach allows other people to think they can breach that standard and cause a cataclysm of were now making an exception to a rule but now it makes us look like a hypocrite for doing it to other people. It sucks. I think i love blaming facebook for stuff, this one is on us. This one is a democracy problem. Were just sort of this is a situation that we should not be in. Its just so extraordinary that i dont think that theres a tech fix for it. I think this is we screwed ourselves in other ways that have nothing to do with computers. Also, theres people do get kicked off of twitter, just not president trump. Others do. They do. Theres been a number of people that have been pulled off of the service and very quickly. And i think the question is, what are what i think you struck on is each of these companies has different standards and so where is there a common standard . Thats the difficulty. Its just these people and they are in Silicon Valley for the most part. They have a big office here. And, laura, they dont pay you enough, you should get a raise, but they have different standards and its a group of people, its not the most Diverse Group of people on the planet. It and thats a worrisome part. They failed on trying to be diverse and thats another part, whos Community Standards, what it goes on. And id just add to that, having a number of Different Community standards is what section 230s drafters envisioned, there would be experimentation with different procedures, policies, and the idea its called user empowerment. Theyll choose the platform that best suits their needs. Getting into the antitrust issues that were discussed earlier, that might not work in the current environment. But the vision was there would be a lot of different types of standards and not just one uniform standard. I would just say from a Public Discourse standpoint, it would be bad if twitter or another social Media Company kicked off president trump, you know its possible that to the extent that hes offending the Community Standards, that is a product of democracy. But it would be a bad thing to let company decide that youre going to remove this very important person to our society and the Global Affairs from Public Discourse. What if he tweeted, my followers, i would like you to shoot blank . What if he did that one night and it came down. What would you do then . Just curious. Now were talking thats really not even the aclu would defend Something Like that. What should twitter do . I dont know that then removing him from twitter would solve the problem of that statement that he made. I think thats a bigger problem, which you should address in various ways, if hes making an actual threat. He cant be indicted in office, so thats another pickle for us. This is not a tech problem. Look at this world. This is not the computers the computers did not do this to us. Something else did. Yes, the fact that people say terrible things including at times maybe the president or other people doesnt mean that that that taking it off of this platform or from the perspective of government censorship of speech, does not mean that speech does not exist. I think you need to try to address the problems. But merely trying to censer speech has not been the solution. Someone is going to i wrote a column saying theres no other platform for him twitter is the perfect marriage of a person in the platform. If you pulled him off of twitter, he would have much less influence in a way. No really work on other platforms. Thats kind of interesting when you think where would you go . Gab . Everybody is on twitter. Twitch . I dont know. I just learned what twitch is. Im going to play games and insult people at the same time. People have been saying terrible things for a long time, but theyre disfused much more quickly and broadly today, and i think and that is one of the impacts of the internet that were faced with and dont know how to deal with. Theyre amplified very quickly. My question is has to do with how were Tech Companies. Reporter able to remove isis hate speech or is there something specific about the case of isis that could get that case removed so quickly . When it comes to terrorist contact, we have a partnership where we work with other Tech Companies and share information about extremist comment. If we see something, were able to share that broadly and through a process called hatching and matching. Its able to go into the system so that our detection systems can find it much faster. Its been incredibly effective. We testified last week, the head of our content policy team testified and talked about it a little bit. But it is a huge Success Story. To your second point about whether or not that type of thing is applicable other places, it depends. For things like nudity, images with nudity, its thats a thing thats much easier for a computer to read. There have been some notable press exceptions where something got categoried as nudity when it wasnt. When it comes to things like hate speech, its just so much harder because theres context that matters. As you know, machines learning language is much more difficult. Its a place where we are constantly trying to improve. But were further away than we would be on terrorist content or other places like nudity. Im going to add my understanding, youre looking at discreet hashes of content thats already been flagged and reported as terrorist content. So this is similar to the child pornography system where you have a specific piece of content that we people have already agreed we dont want on the internet anymore. The system is able to match it because it is that piece of content as opposed to a machine is sensing terrorism in a video. Thats not going on. You have to have a very sort of specific framework for extending this kind of Machine Learning this matching to other types of things. The other thing i would add is actually, im not sure how much it matters. I certainly respect what the platforms have done with isis content over the years. Theyve gotten quite sophisticated at it. But im not sure what the real world impact is given that sophistication began to ramp up just as isis started losing real life battles. Im not sure what the link to reality is here. Next question, please. Im from korea, im a junior. Im studying journalism. Youve discussed how domestic policy changes would apply as guidelines to the media, but i see social media as a Global Platform with global users, so im curious that changes in american policy affect its global users as well and the focus is america because these are american companies, but should every company be addressing these jobs . Most of the activity is not in the United States. The United States has been incredibly poor at regular not poor, it just hadnt regulated the internet at all. Theres a good book called tools and weapons which i recommend. For example, europe has been way ahead. And on some things theyve gone way to far. At the same time and theyve done more fines, activity, and now we have the chief sheriff of the internet. Facebook and others have to follow the global stuff and therefore they do it here. Thats where its set. Jacinda ardern in new zealand, they have to follow those, and theres all kinds of interesting stuff going on there, some not great, some great. And so the u. S. Is quite far behind in a lot of this and even in privacy, theres a privacy bill thats going to come on line in 2020 in california, the first significant privacy bill that its interesting, california is setting a tone for a lot of thing, ab5, so theres no National Privacy bill here to speak of. But eventually if our government ever gets anything of any substance, and thats a big if, that will lead the way presumably for these companies. But right now its coming from abroad is that i think that the aboard is shaping policy here in the u. S. Much more than the other way around. I think it very much used to be that there was this almost american imperialism, imposing American Values on speech in other countries. Just sort of thoughtlessly, its who writes the policy. Its american lawyers who are educated in american schools and american schools like the first amendment, so on and so forth. But i think whats happening right is that a lot of the resistance towards certain changing Community Standards or inviting regulation in a certain way is that i think companies are afraid of setting examples that allow more totalitarian, more questionable, more corrupt governments and other companies where the markets are expanding, leeway to come and meddle with how their product works, spy on users, censer dissidents, so on and so forth. And i think were seeing that echoed on capitol hill. I think that is those are all good points. You know, everyone is right that the rest of the world is leading the way in terms of what the regulatory landscape looks like, the way we deal with it, we try to geo block things that are illegal in one country because we think its important as a platform to not stymie ideas not expressed in america because they would be illegal in germany or singapore. And we have been clear about our desire to work with governments aboard that have ideas in this space because its important to us that they understand how the systems work, the world of the possible, the world of the impossible, and they have a grounded understanding as they develop legislation around our space. We have a partnership with the french government. The results of that are not all the way out yet. But trying to make sure that these governments are well informed as they try to regulate speech on our platforms. But here in the u. S. , we are fortunate to enjoy the affirmfi amendment and it does obviously to some extent influence how we carry the company aboard but we want to be respectful of those local contexts in those companies, and that means working with governments. Britain has some interesting laws that theyre considering. They dont have any regulation here. Except for current laws that apply to everybody. There isnt any specific internet regulation. Lets take a couple more quick questions. Good evening. My question is with regard to the antisemitic comments that take place on line and how they translate in the real world to the physical harm that people have, for example, in 2017 when one of the speakers was coming on campus. The Online Presence of hate speech and how it translates to the real world and that leads to a lack of security and leads to physical harm for people, what is your take on this physical issue. I think everybody realizes whats happening. I think there will be study hundred years from now, there will be a fantastic book on what happened. We read about the salem witch trials now and we go, those crazy people. And things like that tend to happen. You look at things that happened with the whatsapp thing, when facebook limited the virality of it, it was it was for safety because what they had done originally, because some engineer at facebook or whatsapp didnt think about it. Didnt imagine that this would be used this way. And thats the problem. Theres someone sitting in Mountain View making an engineering decision and they would have no idea of the consequence. But theres no question that the tarnishing and the way it denigrates our offline life is affected by online. I say this all the time, whenever i meet with groups of engineers in any company, i always say, imagine your product is a black mirror episode and then dont make it. Think of the worst thing that could happen. And i think Silicon Valley doesnt do consequences as well as it should be. And the reason they dont do it is for a very good reason, they dont want it to mess up their jam. Look at facebook changed from move fast and break things, which the word break is an astonishing selection of world, to move fast and build stable infrastructure. Thats their new thing. So thats the question of how the consequences is what has to get through to people in tech. And i think theyre getting the message really you can tell they are, for sure. Not all of them, but a lot of them. This is the question that we grapple with, particularly around the design of our Community Standards. Just i think two weeks ago, we rereleased the version of the values that guide our Community Standards and Free Expression was paramount, but the first caveat is safety. If you dont feel safe, youre never going to express yourself there. I think we are well aware of the unintended consequences that the products were developing can have. My title is external affairs, right . These arent all decisions inside a company that are guided by engineers in a black hole who think if they can just get the algorithm right, human nature will go away. Thats not how our company operates. Our company is proud to work with Interest Groups from across the political spectrum, countless communities, particularly at risk communities, because were aware whenever we build a product, it could be used in a way that were not thinking about. These are safeguards into our system that were building in that i hope will limit these negative outcomes in the future. Obviously, at the end of the day if our User Community isnt safe, we wont have a User Community. It is whats most important. I think this will be well, ill im a student at the law center at georgetown law and you need to get closer to the microphone. Something ive been interested in is the world of paid sponsorship. And the ftc has issued guidelines, but it doesnt have the force of law, its not a rule. My curiosity is, in a marketplace of ideas, its also just a marketplace and what