I recently took part in a one-day Symposium on the topic of fake news, which took place in one of the leading law schools of the United States. By the end of the day, the participants of this representative event, composed of reputable experts, including representatives of Google, Microsoft, New York Times and Buzzfeed are unable even to agree on a definition of counterfeit, or fake news. Even worse, any suggestion about the control and regulation of these non-definitions of news is immediately rejected.
So what are we to think about new efforts by Google and Facebook use algorithms to protect society from the information fake?
In my opinion, we should not be afraid of them, and they must not produce in us any impression.
Do not be afraid, for the simple reason that the fake news is actually not a problem, according to most people. While information on Hillary Clinton in November last year sold out very quickly on the Internet, and could deprive her of many voices, in General terms, fake news are not a serious threat to democracy. There are two reasons.
Fake news like advertising
First the fake news participate in the competition. They are like billboards and television commercials. In advertising, too many untruths and falsehoods. On each of your fake news about me is that I can place two fake news about you. I can also post the network true stories to improve their reputation, or even true stories exposing the stuffing information posted by you. Sound familiar? Mouthing is an integral feature of the policy, and always will be. And any algorithms that will not stop.
But is the rapid spread of fake news on the Internet is not a new type of threat? Absolutely not. As I said, the fake news participate in the competition, and why their rapid spread works for all parties. Initially, they do not benefit one side over the other, though recently, they could help the Republicans. Yeah, campaign Clinton was not ready to counteract the influx of fake news before the election of 2016 was spread by the adolescents from Macedonia. But it seems to me that no serious campaign headquarters will not allow such an error.
The speed of propagation of the fakes by itself is not a problem. We mistakenly believe that the rapid dissemination of ideas or information is a new phenomenon, made possible only recently due to the development of the Internet. We somehow forgot that long before the invention of the Internet news and false rumors spread throughout the society with lightning speed. This and reports of stock market crash in 1929, and the news of the victory over Germany and Japan in 1945, the assassination of John F. Kennedy in 1963. March 19, 1935, a completely unfounded rumor about the overused child, which quickly spread to Harlem in new York led to widespread rioting, the deaths of many people and damage worth millions of dollars. People loved to gossip long before the invention of the world wide web and social networks.
Fake news visible
Secondly, information is a very clear influence attempts. When we discover in news feed Facebook or Google search a story like the one that Hillary Clinton’s an alien, we fully understand that we are under the influence. This story is right in front of us, as a newspaper, a Billboard or a television commercial. Visible sources of influence affect people is quite predictable. People pay attention to information that confirms their beliefs and prejudices. And the rest they ignore or reject.
Much more dangerous than the influence of a new type that people don’t notice. In recent years I have discovered, and are now exploring new sources of online influence, such as manipulating the effect of search engines (Search Engine Manipulation Effect) and the effect of search options (Search Suggestion Effect), which most people completely invisible. This effect is unprecedented in human history. Biased search rankings can have an exceptional impact on people’s opinions, their purchases, preferences in voting. Doing the same tips of instant search, which we see when we begin typing your search word in the Google search box. The effect of this type is absolutely not like influence from advertising or from a fake news, because almost no one sees this bias. And when people don’t see the source of influence, they mistakenly come to the conclusion that you can make a choice and take a decision. Even worse than the other. Those few people who notices the bias and bias in the search results, tend more to lean in favor of such bias and partiality. So, if you are able to notice the bias, it does not protect you from it.
Fake news is alarming; they can influence the minds of hundreds of thousands and even millions of people. But let’s look at the bigger picture: the bias of results and search options may have an impact on billions of people daily, and they do not know about. Fake news as a means of influence in this respect, just a trifle.
If you have doubts, think about one simple method of manipulation, which has at its disposal Facebook. It is specifically targeted messages one demographic group. If in 2016 the company sent to supporters of one political party reminder type “Register and vote!”, or if the day of the vote, she sent them another reminder like “Go and vote!”, the candidate from the party would get a few million more votes, and no one would didn’t understand. Even newspaper magnate William Randolph Hearst, and he could not even dream about such a government: the target of the message on a massive scale, run by leaders of one company that has no competition.
As fake news are visible and participate in the competition, I don’t think this phenomenon is particularly terrible. But I am afraid of the idea that these predatory corporations like Google and Facebook will determine what is fake news, when you decide what materials meet their criteria. After that, they will be able to give some information suspicious, and in extreme cases even remove it.
As I found out during the Symposium, it is unlikely any group of intelligent and sensible people will agree when anything about the definition of fake news. And if so, the definition I offer. But to help you become clearer how serious this problem is, consider the following questions:
The news is fake just due to the fact that it is represented incorrectly? How wrong should be the material presented, so we can call him a fake? 20% information? 50%? But how true is true part of the news?
Legitimate or invalid truthful material published fake news site, for example, such, which the authors pretend to be part of the media that does not exist? Remember that this material is true and false is only spreading his news organization. What to do: post an update or forget about it?
If true news publishes a news website (say, Sputnik), which is associated with a hostile government (say, Russia), the fact that it is a genuine news or a fake? I ask this question partly due to the fact that when I recently published an article in Sputnik, a leading United States information Agency immediately rejected it, although I’m definitely right and talked about his new study on the topic of input completion, which exists in Google.
Fake news due to the fact that it is biased and one-sided, and illuminates only one point of view? If so, don’t fake some news which are published by Fox News and The New York Times? And is it not necessary to automatically delete all the news published on the Breitbart website?
Can it be called fake news satire? Whether a fake the material that I published last year? There is a question about what Google has donated its search engine American society. This news should with disdain reject? And as the algorithm will be able to distinguish fake news from satire, if it can’t do the majority of people?
At that Symposium, the participants discussed other motives. These teenagers from Macedonia could be agents of the Kremlin, but when they were writing his strange stories about Hillary Clinton, they really wanted easy money. How seriously we should take these people? Do I need to take them to court for trying to influence the elections in another country? And for that matter, whether they tried to influence the election? Or maybe they just wanted to make a buck?
Do we use people or algorithm (which is just a set of rules written by men) in an attempt to find the answer to these questions, the one question that is generally understandable (spreading false news), is transformed into a nightmarish set of problems that make the head spin.
The first is a false positive problem. If you or your algorithm correctly calculates and removes the fake news (I say this conditionally, because we do not have even a definition of false news), this is called a true positive result. That is, the correct identification of what you are looking for. What if you called true news fake? This is called a false-positive result, and from the point of view of public policy it’s a disaster, such false-positive results when screening for cancer. You said society, and to the world that true news is not true. Maybe you even removed it from the news feeds to one place in the world never saw.
Most of my life I’ve worked as a programmer. So I give you a guarantee that any system of algorithms, designed to destroy the fake news, will inevitably give a false positive. What sort of case, the threat of destruction truthful news more dangerous than our decision to give fake news right to exist?
The second is the problem of power. You might not have noticed, but companies such as Google and Facebook are already too much power. Do we want to give them another type of power — the power to decide what news is true and have a right to exist, and which are not?
The only good news in all of this next. If these companies will continue to create aggressive program to curb false news, they may get in the crosshairs of regulators. While Google, Facebook and similar companies enjoy the protection of article 230 of the American Law on decency in communications, which States: “Provider or user of an interactive computer service cannot be considered a publisher or distributor of information provided by another information content provider”. In other words, Google and Facebook are not responsible for what is shown, as the authors of the content don’t they. They are just intermediaries, not publishers.
But the more clearly these companies are beginning to act in the publisher role, choosing what news is true and what is not, the less they are protected by the section 230. As I mentioned in my article, The New Censorship (censorship), Google today is the biggest censor. It acts as a super-editor, making decisions about the publication or the elimination news. Thus, its censorial role is becoming increasingly evident to regulators, lawmakers and judges.
Fighting technology. More than 10 years ago, Google laid down this law, and in fact stated: “As you have dared to deceive our search algorithm and force it to put your shitty website higher in the search rankings? Yes we will crush you like a bug”. What made the response optimization specialists search whose task is to raise your business higher in the search ranking — fled to burrows and quietly died? Nothing of the sort. In fact, by the beginning of 2016, the industry of search engine optimization was estimated in the business at $ 65 billion per year. It seems that to deceive Google is a very profitable. Google adapts the search algorithm to protect against such frauds, and schemers in response inventing even more sophisticated algorithms for fraud.
I say this because the same thing happens with algorithms, in which attempts will be made to stop fake news. People who want to distribute them, just come up with a new program to circumvent the algorithms.
Effect Trump. Remember how Donald trump turned leading media in “enemies of the people”? As if Google and Facebook tried to stigmatize the fake news, such leaders as trump will still belong to the final word in how seriously people will take stories of this kind. If a demagogue tells his followers that to believe only news service “Edinburg”, the result of the algorithm, consider its message to fake, will become only that this service will attract more attention. A demagogue will with good reason blame “censorship” and “oppression”.
False news cause for concern, and there is no doubt. But they have always been, are and will be. And if you allow large companies and neglected information technology to manage news fake, that is, to manage our news with you, it will cause far more harm than the fake news.
Robert Epstein, a psychologist and senior research fellow American Institute for the study of behavior and technology (American Institute for Behavioral Research and Technology), located in Vista, California. Epstein has written 15 books about artificial intelligence and other issues. In the past, he worked as chief editor of the magazine Psychology Today.