Google’s chief emissary to the news business is stepping down — and looking back

Richard Gingras: One is certainly the development and implementation of the Google News Initiative [a grant and training program to help media companies adapt to digital]. We invested huge amounts of money over eight or nine years to drive innovation. I’m very proud of that. I wish there were more such efforts done by others around the world.

Gingras: I honestly don’t have any regrets. I really don’t. Look, being a search engine in the world of the internet over the last couple of decades is both the position, as I’ve often said, of extraordinary privilege, but also a position of extraordinary challenge. It’s almost hard to think of another company that has been in such a position…So, regrets? No.

Gingras: When you put it that way…if there’s a regret, it’s that I was not as effective as I would have liked to be in influencing and driving change across the board. When I look back at the last 30 years of the internet, it saddens me to see how we haven’t really, in the journalism community, fully embraced the change in the ecosystem — or embraced the need to innovate to address that changed ecosystem…What could I have done differently at Google or outside of Google to accomplish that?

And I think that’s a huge part of what I’m going to be doing, going forward. Can I [accomplish] on the outside with Village Media what I had tried to influence from the inside? Can I get my hands dirty working to grow and evolve solutions to the role of a news outlet in a community? It goes beyond talking about concepts and ideas and visions and challenges to actually trying to do it.

Gingras: We were often criticized as being the primary driver of the disruption of the news industry in one fashion or another, which to me is just factually inaccurate. If you do the true economic analysis, that wasn’t the case. The internet disrupted the information space and the advertising space. It changed user behavior. It changed patterns of advertising.

The first component of that disruption was the fact that users now had so many places to go for the information that they used to get from newspapers…If you looked at their primary advertising categories of classifieds, department stores, auto dealers, supermarkets, that advertising revenue didn’t move to Google. Classifieds went to online marketplaces, department stores largely diminished in the world of e-commerce. There are no discount coupons for newspapers anymore. You enter your phone number in a loyalty program.

So I always felt it was unfortunate that real factual analysis [for blaming Google] wasn’t there. But nonetheless, newspapers and news organizations were disrupted by the internet. And we did feel it was important to do whatever we could do to try to enable innovation and help publishers understand how their products could fit in the overall marketplace of the internet.

Gingras: We are indeed moving into a new era, whether we call them language models or chat agents or answer engines, which is the term I would tend to lean on. We’re moving into that era, and obviously it’s not just Google. There are many other players in the space who are striving to give users the answers they’re looking for.

The primary motivation with search is to help people get answers to the questions they have. AI’s capabilities to do that are, in many regards, far better than trying to figure out the answer in a series of search results.

You do a search for a recipe today on the web, and you get a whole lot of results. But the experiences are pretty challenging. I joke that you search for a recipe for pound cake and you get the history of the knife and the fork and wheat ground into flour before you get to the ingredients in the recipe itself. [But] what the users are looking for is a simple answer.

Current events are complex. It would be unhelpful to the user and unwise for Google to think that we could give a simple answer or a singular answer to a complex current events query. I think it’s always going to be important to reference the different perspectives and sources of information given a news query. So I expect in the current events area there will be less impact [from AI] than in other areas of content.

Gingras: That in itself is a very complex question. Google has, in my experience, always been very careful and focused on operating within copyright law, within fair use, and will continue to do so. The training of the language models has obviously been driven off of our understanding and use of the internet and fair use of that internet.

I am sure [Google] and others will also be careful to not plagiarize. So a lot of these things are going to depend on how the environment evolves, how policy evolves, and to what extent copyright law changes.

Copyright law does not accommodate the ownership of facts. That should not be changed. But who knows? We’ll see. That’s not for me to decide. It’s not for Google to decide. That will be for the policy sphere and the courts to decide.

My one bit of guidance, and I’ve been saying this to the news industry: However that plays out, I would be very, very cautious not to see licensing by language models as being, somehow, the path to sustainability. There’s not going to be a pot of gold at the end of that rainbow.

I say that for a few reasons. We’re going into a very different competitive environment. More answer engines. There’s a tremendous amount of money being invested in them. The market dynamics on the revenue side of that are unknown. The margins are likely not to be significant, so be careful what you think you might expect with regard to theoretical licensing revenue from language models.

To your earlier point, I think there does need to be a whole lot of focus on how you continue to build your own audience and retain your own audience. That has always been the case.

Gingras: Who knows? Possibly. [chuckles]

Gingras: I think the real mind meld was that we need to rethink every dimension of the role of the news outlet in society and in communities. It’s not that everything must change, but it’s worth considering everything — how we build audiences, how we retain audiences, the form and style of the content itself, the mix of content, the thoughtfulness of the revenue model, and so on.

Before [Village Media CEO Jeff Elgie] founded Village and its first property up in Ontario, he ran a digital ad agency. So he was very sophisticated in understanding digital advertising and local advertising. But what intrigues me about the model is this: It really starts with the mission. How do we create the fabric for the community that allows them to get the information they need and engage with each other and, in so doing, strengthen the communities that Village serves? That’s a powerful statement.

In fact, Village [doesn’t see itself] as just a news site. It’s a community impact platform. How do we engage? One key element is to start with the content mix. If you do research on a community’s information needs — and by the way, I generally feel that news organizations need to do more real, thorough research about the markets they’re looking to serve — what you find is that people’s information needs are, first of all, highly proportionate to physical proximity. I want to know what’s going on in my community, in my neighborhood. Next, the city; next, the state; and so on.

Those information needs are far beyond accountability journalism, which is absolutely critical. But it also means what’s going on in the community park this weekend. What about local sports? How about the potholes on Main Street? And by the way, that information will drive far more engagement than the accountability journalism. Obituaries will gain more engagement than the accountability journalism. That’s a powerful fact.

The key thing about Village from a business model perspective is it’s 80% local, direct-sold advertising, no paywalls. So you’re reaching deep into the community. In some of their markets, 25% of the community is engaging with them at least three times a week. That’s a powerful statement.

Gingras: It is a very difficult environment. And I don’t mean to suggest any of this is easy. I do think it requires a whole lot of careful thought along many dimensions. And by the way, in a politically divisive world with pretty strong extremes at one end versus the other, I tend to think that it’s really important to try to reach across the spectrum as much as possible to find what we hope will be a fat middle.

I do think, however, it’s important to present fact-based journalism in a fashion that leads more people to trust it for what it is. And I think that requires many different aspects.

Is your fact-based journalism surrounded by a bunch of opinion content? As I pointed out, linguistics matters. I’m fascinated by the work of [University of Florida professor] Janet Coats about linguistics in the journalistic environment and her fascinating study of the George Floyd protests and the pairing of words like “ignite,” “erupt,” “explode” with regard to the protests, often subconsciously. Did that contribute to people’s perceptions that maybe the protests were violent when they weren’t?

[Village] doesn’t do political endorsements. We strive to be assiduously apolitical. I think that’s important if we’re going to achieve that objective of being a source of fact-based journalism for the community.

Gingras: 1440.

Gingras: Two things. I think journalism training is extraordinarily important for things beyond journalism. The second thing I’d say is public speaking and debate.

Gingras: The yogurt malt balls. I have the palate of a child. It’s the snacks that kill me.

Gingras: Okay, well, my most recent Gemini [Google’s AI tool] search. I’m going to Norway to do a speaking engagement and I added three days to drive around Norway and do drone videography. I told Gemini, “I’m going to land in Oslo. I want to stop in Bergen and end in Lillehammer. And I have three days. Will you give me several options for scenic drives and what I should stop and see? And by the way, I’m taking my drone.” It did a marvelously fine job of that. And I planned the entire thing, including booking hotels, in less than 90 minutes.

Gingras: Embarrassing number. Let’s see. Probably 32. Oh. Are there any windows for different purposes? Yeah. About 35 on that one. The media tab — another 35 or 40 on that one. I love tabs. I have too many of them.