Haaris Mateen: "The amount Big Tech should pay media outlets is so staggering that it scared us."

BarcelonaThe media hurts, and big tech can, but they do so thanks in large part to the content the former have produced. Haaris Mateem is an economist, professor at the University of Houston, and a specialist in media economics. ARA interviewed her because much of her recent work has focused on studying what compensations the media should meet. big tech in news companies to rationally balance an unbalanced equation.

You just participated in a panel discussion titled "How Much Should Tech Companies Pay for the Content They Distribute?" But before we get into the numbers, why should they pay anything?

— We have a platform that is disseminating information to other users, and there is a provider of this information, largely the media. In economic terms, we would talk about complementary services: you work on one thing, I work on another, and the union of the two creates something we call added value. Therefore, given that these companies have benefited immensely from the presence of quality, diverse, and up-to-date content in many languages, economic logic dictates that they should compensate the provider for all the value generated.

The big tech They say that newspaper companies are already making a profit, in the form of circulation.

— When they say this, they fail to admit that, by using media content, they too benefit, and greatly. Think of Google search. If I use it, it's because I type something, and it gives me a result with a high probability of containing information we can trust. But much of that information, especially the most recent, comes from the media. If it's not paid for, it creates an imbalance that, moreover, is about to get worse.

Because?

— Because artificial intelligence makes it so that people receive information without any incentive to go to the original source or the media that produced it.

With this in mind, how much should tech companies pay?

— When we started trying to quantify it, we focused on Google and Facebook's search engines because AI wasn't that developed yet. And we tried to answer the following question: if there were no information coming from news, how much value would Google have? I often give an example. I have a TV package in Houston, which has many channels. But I subscribed because one of those channels is LaLiga, since I follow Barça and other matches as well. I occasionally watch other channels, but if they took away soccer, I would unsubscribe. In the case of Google, our estimates are that 70% of search engine use is for information, and that 50% of this percentage corresponds specifically to news. This means that 35% of the Google ecosystem is news.

Without news, would 35% of people stop using Google?

— Exactly, that's a very crude way of putting it. 35% of Google users would stop using the search engine if there were no economic news.

From theory to practice...

— Okay. Let's say they reject the 50-50 ratio and say it should be 70-30 at most. Let's accept that! Or even 80-20. Because as long as it's a reasonable percentage, it's a huge amount of money because we're talking about a multi-million-dollar product.

Have you been able to transfer this approach to the internet giants?

— Many times. And we've often presented our findings to governments around the world and in the United States, and there's often someone from Google. We don't necessarily speak on every occasion, but we often discuss the situation. Google often says that news is residual for them: one or two at a time. But this doesn't make sense: It turns out that people use the search engine primarily to buy socks? It's true that the commercial aspect is large, but so is the news aspect. The other thing they tell you is that they've conducted experiments, leaving a group of users without news, and that they didn't notice significant effects. But these experiments should be validated by academics, because everyone accepts discomfort, if it's for a short period of time.

Applying this reasoning, in a scientific article co-authored with Anya Schiffrin, they estimated that Meta should pay publishers $1.9 billion and Google between $10 and $12 billion. And that's just in the United States.

— Exactly. What they would have to pay is such a spectacular figure that it scared us. We thought, "Wow, that's very large, are we sure we calculated it correctly?" And we repeated the numbers and sent them to prominent economists to review our calculations. The figure is such that even if it's only 10% of what we're saying, it's still $1 billion.

Google is already making deals with some companies.

— Yes, and they do so under confidentiality agreements, so it's difficult to know what figures they're working with. But, from what we know from the conversations off the record that we maintain with the sector, the money that arrives is a very small fraction of what we consider balanced.

And how would the American media landscape change if $10 billion were to arrive now?

— This is a key question. We want money to reach newsrooms, because there's a risk that compensation will end up in the hands of investment funds. We want to support journalistic initiatives that often require an investment of time and money. We also want to help change the media's business model.

We've long felt that the media is in crisis. Would this money be enough to turn things around?

— Anya Schiffrin and I have been working together, and we both agree that this simply won't save journalism, even considering the money that should come from AI. It's journalism that will save journalism. However, what that money can do is buy the time needed for newsrooms to rethink and adapt to the new landscape.

Australia passed a law on these offsets in 2021, and Canada followed suit two years later. How have they fared?

— In Australia, a substantial amount of money has flowed to newsrooms, and it's had a positive effect. And an important point: Google hasn't left the country. If the legislation is good, tech companies stay because they continue to do good business. The Canadian case is also interesting. They've agreed to give money to the media, although less than we expected. It's still early to assess its impact there, but we see that the Canadian government has introduced good measures to ensure that this money goes to newsrooms. The number of journalists is rewarded more than the distribution of copies, for example. Now, in Canada, we find an example of retaliation: Facebook has announced that it will not have news in the country on its social network, and this has begun to boost circulation. fake newsAnd then there are those who share legitimate news with screenshots, so the algorithm doesn't detect them.

In Spain, Google News was discontinued in 2008, in response to an attempt to legislate compensation.

— They shut down the service, but they didn't remove the news from search results. Ultimately, Google News doesn't directly generate revenue for them, but the search engine does.

In any case, he's back in 2022. And we journalists are happy. If this interview is picked up by Google Discover, his audience will increase six or sevenfold. It's hard to give up on this doping when metrics are what determine performance.

— That's why the order in which it should act is important. The first priority is to achieve legislation that, among other things, prevents retaliation. Google shutting down Google News is clearly a punishment. And we don't accept these punitive attitudes in many other areas. Because logic dictates that the economic incentive is strong enough for Google to continue to offer news, since it needs them to deliver good results. The second step would be to be able to reach a joint agreement. The sector must understand that they are competitors in the market, but are now united in a common cause. Some kind of coalition would have to be created.

How should the sector react?

— No one would even consider creating AI content with Disney characters like Aladi, because they know Disney will sue them. The same goes for hit songs: the major record labels would sue. divide you rule.

We've been talking about search engines and social media, but AI still poses a bigger challenge, so that women can answer questions without being able to trace where they got their information, which may have been a medium.

— Exactly. And it's harder to end up with a link where you can go fishing for information because the answer already gives you a sufficiently complete summary of a topic. The challenge is greater, and therefore there should be legislation on the matter and also negotiations between technology and media companies. But I also see it as an opportunity, but AI companies are competing heavily with each other to have reliable data and answers that don't include what we call hallucinations. This opens up an opportunity for the media to negotiate. Before, you couldn't say, "If Google doesn't help me, I'll partner with Bing," because Bing was very small. But in the generative AI market, there's a lot of competition, and therefore, you can compare who gives you the best results and go with them, offering them access to your information, with a seal of trust.

And should they pay?

— Yes, if it's proven they're using content you create, they should pay to use it. OpenAI, the ChatGPT company, has closed many deals with major news publishers, and this is because they're trying to secure good news sources from the most powerful players in the industry. There's a clear advantage in negotiations with AI companies: it's much harder for them to deny they're using information and news. Therefore, again, I would recommend the three pillars we mentioned earlier: legislation, negotiation, and coordination.