Is AI about to kill what’s left of journalism?

The query is now personally pressing because the FT last week reached a cope with OpenAI to coach large language models on our content. Like many news organizations, the FT now receives compensation for its mental property and likewise receives attribution when the content appears.

On the one hand, one can argue that it is healthier to receives a commission to do something than to not do something that Big Tech already does without spending a dime, namely absorbing online journalism to coach LLMs. (FT journalists had no insight into our company's compensation levels or the financial details of the contract with OpenAI, so I can't say whether it's a very good deal or not.)

On the opposite hand, I worry that we’ll soon see a repeat of what happened within the mid-to-late Nineteen Nineties, when media corporations adopted the Silicon Valley line that “information desires to be free” and didn't need to put up with that line an excessive amount of to guard copyright and the worth of their content.

As I argued in my second book, and as tech journalist Steven Levy detailed in his book, Google particularly has simply trampled on copyright in a way that paves the way in which for surveillance capitalism normally, including today's AI applications , paved. As I wrote in my January column about news publishers' latest AI copyright infringement lawsuits:

A recent study by researchers at Columbia University, the University of Houston and the consulting firm Brattle Group. . . Estimates suggest that if Google gave U.S. publishers 50 percent of the worth created by their news content, it could need to spend between $10 billion and $12 billion per yr. Currently, the New York Times – one among the biggest news publishers – is receiving just $100 million over three years. Now AI is able to make even this asymmetrical relationship look good. When you ask an issue to a chatbot like OpenAI's ChatGPT or Google's Bard, you won't be redirected to a creator's website. Rather, you’ll receive the reply directly. Users remain within the walled garden of the Big Tech company that owns the bogus intelligence platform. The proven fact that the AI ​​was trained on the very same copyrighted content it desires to bypass makes matters worse.

The news business has, in fact, been hollowed out over the past 20 years, as a handful of digital giants have captured a lot of the value generated by online content. This was particularly true on the local level; In fact, the ability of the digital monopoly has led to such a destruction of local news that Big Tech giants like Google and Facebook, now Meta, began funding the creation of more local news content because its lack was hurting their very own business model. No doubt Marx is laughing from the grave about all this.

There are two ways to be within the news as of late. You might be the 1,000-pound gorilla like The New York Times, which has reached escape velocity with greater than 10 million subscribers. Or you’ll be able to be a high-end premium subscription organization, which is the FT model. I've at all times thought the latter was the higher and easier suggestion to defend. It has survived the most recent round of digital “creative destruction” of the last 20 years. How will it perform next yr?

To be honest, I don't know the reply to this query. What I'm saying is that this: I believe AI is clearly a bubble. As I wrote in March, that doesn't mean that something useful won't come out of it in some unspecified time in the future (the digital bubble of the late Nineteen Nineties spawned tons of froth, but additionally hundreds of thousands of miles of broadband cable). But at once the rankings and the completely simplistic and optimistic plot remind me a number of the mid-Nineteen Nineties, when all the things Silicon Valley said just seemed inevitable. If you questioned it, you were an idiot. You didn’t “get it.” Most journalists and executives were too intimidated by the tech titans' confidence game to actually think clearly about how they were giving freely their very own product without spending a dime.

This time we'll probably get something for the product. But I'm also very skeptical that AI will bring any serious profit to anyone aside from Big Tech within the short to medium term. Remember when SAP got red-hot by selling enterprise software licenses to everyone within the Nineteen Nineties? SAP's stock price rose, however the promised enormous productivity boom never actually materialized (most economists now view the productivity decline of the Nineteen Nineties as a form of mirage). Ultimately, it was the smart move to purchase SAP and sell a lot of its customers.

I fear that the AI ​​narrative today is all too similar. First of all, I've seen almost no evidence that AI actually increases productivity at scale. There are many small experiments driven by Big Tech corporations themselves, and few of them have been properly vetted or audited. We just trust the tech corporations themselves. We still consider within the inevitability of this story: if we don't get on this latest “disruptive,” “progressive,” “world-changing” train, we'll in some way miss out.

But consider another scenario. What if AI's water and energy consumption simply made it unsustainable? What happens if the productivity “miracle” never occurs? What if AI was ultimately one other approach to reduce the price of human labor – whether within the media industry or a spread of other industries – and replace individuals with a technology that financially advantages a handful of huge corporations particularly? What if AI really was just the most recent ploy by Big Tech corporations on the lookout for a approach to keep the music going as rates of interest rise, antitrust measures take effect, decoupling locks them out of the Chinese market, and it normally simply becomes harder to make cash? ?

Peter, what do you say to that?

Literature recommendations

There are so many great FT deals in the intervening time that I'm going to create an internal list of really useful books this week:

• The FT's Henry Mance conducted an in depth interview with economist Joe Stiglitz, who rightly points out that “Trump is what neoliberalism produced”. I actually have to say, I disagree with Mance that the needs of the Global South and the US are in conflict relating to things like climate change. America's industrial policy approach to combating global warming, and particularly Katherine Tai's calls for a “post-colonial” trade paradigm through which America could offer technology transfers to developing countries in exchange for, for instance, rare earth metals and other inputs, are way more palatable than the European carbon tax approach that underpins the worldwide one making the South pay disproportionately for the sins of richer countries.

• My colleague Simon Kuper examines the contrasting political decisions and lifestyles of Americans who’ve more cash in comparison with Europeans who’ve more time. As someone who has lived in each places, I believe Europeans are making the safer, healthier bet.

• In case you missed that FT Weekend Festival within the USA, you’ll be able to proceed to livestream a number of the events and speakers. It's a very great event and I'm so sorry I couldn't be there this yr (but I had a very good excuse, which was that I used to be celebrating my husband's birthday in Greece – back on the tenth and back). ). the saddle with Swamp Notes the next week!)

Peter Spiegel answers

Rana, one among my favorite analytical frameworks within the tech world is that this Gartner Hype Cycle. It's been around because the mid-Nineteen Nineties, but I only learned about it last yr after I hosted our good tech columnist John Thornhill on a podcast in regards to the way forward for the sector (somewhat embarrassing disclosure: my wife works for Gartner, so should I probably do it). (I knew about this before).

The Gartner Hype Cycle argues that every one latest technologies begin with an “innovation trigger” or latest breakthrough that excites everyone, and eventually reach a “peak of inflated expectations” before falling right into a “trough of disillusionment” through which all lives -Promises made regarding the brand new technology are usually not kept. I actually have a depressing feeling that we’re near the “peak of inflated expectations” relating to generative AI.

There's little doubt that artificial intelligence will transform the occupation of journalism, and the FT is already doing a little fun things with it – including Beta test a generative AI tool “Ask FT” that enables subscribers to ask inquiries to a bot that responds with 20 years of FT content.

But is generative AI about to destroy the business model that the Financial Times and other leading news organizations have established within the digital age? I think that we’re heading towards the “trough of disillusionment” on this regard. News consumers – especially those that need real-time information to make business decisions – cannot depend on the historical and sometimes inaccurate information currently provided by AI chatbots. I still consider they are going to visit ft.com and our competitors' web sites for such news.

However, the Gartner Hype Cycle also predicts that the “trough of disillusionment” is followed by a “slope of enlightenment,” where we work out exactly how best to make use of a brand new technological tool. This is followed by a “plateau of productivity” where mainstream acceptance becomes a reality.

Like you, Rana, I don't know the reply to the query of what the subsequent round of “creative destruction” will appear like. People much smarter than me are still grappling with how generative AI will impact the news industry. However, until we get out of the “valley of disillusionment,” I'm perfectly content to play with the Ask FT bot and dabble in other types of experimentation before worrying about journalism soon being taken over by Big Tech is taken over.