WSJ What’s News - Will AI Help or Hurt Workers?

Episode Date: December 30, 2024

A.M. Edition for Dec. 30. Would you trade greater job productivity for a decline in job satisfaction? WSJ reporter Justin Lahart explains how new research suggests that AI could force just such a workplace tradeoff. Plus, tributes pour in for former President Jimmy Carter, who has died at 100. And investigators work to pinpoint the cause of a South Korean plane crash that killed 179 people. Luke Vargas hosts. Sign up for the WSJ’s free What’s News newsletter. Learn more about your ad choices. Visit megaphone.fm/adchoices

With Uber Reserve, you can book your Uber ride in advance. 90 days in advance. Perfect for all you forward thinkers and planning gurus. Reserve your Uber ride up to 90 days in advance. Uber Reserve. See Uber app for details. Tributes pour in after the death of former US president Jimmy Carter.

Plus, Donald Trump wades into an immigration debate pitting populists against tech leaders. And will A.I. help or hurt workers? New research shows it can boost productivity in the science world, at least. But at what cost?

Having trained up to do this creative work of working on whiteboards and thinking up new compounds and having that go away is disappointing. It's Monday, December 30th. I'm Luke Vargas for the Wall Street Journal and here is the AM edition of What's News, the top headlines and business stories moving your world today. Today. Former U.S. President Jimmy Carter will be honored with a state funeral in Washington,

D.C. on January 9. Carter died yesterday in his hometown of Plains, Georgia. He was 100 years old and had been in hospice care since early 2023. President Joe Biden paid tribute to Carter yesterday, praising his work after leaving office, which included monitoring global elections, building houses with Habitat for Humanity, and taking on the eradication of diseases. What I find extraordinary about Jimmy Carter, though, is that millions of people all around the world, all over the world, feel they lost a friend as well, even though they never met him.

That's because Jimmy Carter lived a life measured not by words but by his deeds. And for much more on Carter's political and personal legacy, check out the special edition of What's News that we published yesterday evening. Investigators are continuing to probe the crash of a South Korean passenger plane that killed 179 people this weekend. The plane, operated by South Korea's Jeju Air, skidded off the runway as it attempted to land at an airport in the country's southwest before colliding with a concrete barrier and bursting into flames. Shortly before landing, the airport's control tower warned of a possible bird strike, and

aviation safety experts say a severe bird strike could disable both engines and prevent the deployment of landing gear. The U.S. National Transportation Safety Board is leading a team of American investigators, including representatives from the FAA, which certified the aircraft, and from Boeing, the manufacturer of the crashed 737-800. representatives from the FAA, which certified the aircraft, and from Boeing, the manufacturer of the crashed 737-800. Boeing shares are down in off-hours trading. The President of Azerbaijan is accusing Russia of trying to cover up its role in last week's

deadly crash of an Azerbaijan Airlines jet. Speaking yesterday, President Ilham Aliyev claimed that the crash was caused by Russian electronic interference and fire from the ground, and he demanded that Russia take responsibility for causing the crash, provide compensation, and bring those guilty of downing the plane to justice. We can say with complete clarity that the plane was shot down by Russia. This is a fact. And no one can deny this fact.

We are not saying that it was done intentionally, but it was done. Aliyev said explanations provided for the crash by Russian authorities, which faulted a flock of birds and an exploding gas cylinder, were, quote, foolish and dishonest, end quote, and that an apology from Vladimir Putin hadn't been enough. Moscow's spat with Azerbaijan shows its loss of influence closer to home, including in its former Soviet republics where it had long held the upper hand. Donald Trump has said he supports H-1B visas for foreign-skilled workers, despite calling

them unfair for U.S. workers in the past. In an interview with the New York Post over the weekend, Trump appeared to be siding with Elon Musk in a rift over immigration policy that started when the president-elect named Musk confidant Sriram Krishnan as his AI advisor earlier this month. Krishnan, an Indian immigrant and general partner at venture capital firm Andreessen Horowitz, has said he supports removing a cap on green cards for skilled immigration, the kind of issue backed by Musk and other tech industry players. However, the topic has become a flashpoint within Trump's conservative base, amplified by uncertainty over how he plans to deal with legal immigration in his second term.

Meanwhile, Trump has asked the Supreme Court to stop a federal law banning TikTok from taking effect next month, saying that he wants to negotiate a resolution to prevent a nationwide shutdown of the social media giant. In a court filing, Trump said that keeping TikTok operating would preserve the First Amendment rights of tens of millions of Americans, though he stopped short of calling the law unconstitutional like TikTok has done. Trump argued it's possible to address national security concerns around the platform without shutting it down.

The ban, which Congress passed with bipartisan support earlier this year in response to concerns that China could exploit TikTok's influence and user data, conditions TikTok's survival on a divestiture. The app's owner, ByteDance, has said it can't and won't sell its U.S. business. And in markets today, Asian stocks have end of the day mixed. European stocks are mostly lower in midday trading, while in the U.S. stock futures are slipping after major indexes ended last week on a downbeat note. Coming up, would you trade greater job productivity for a decline in job satisfaction? We'll discuss new research suggesting AI could force just such a workplace trade-off

after the break. This is an ad from BetterHelp. This holiday season, do something for a special person in your life. You. Give yourself the gift of better mental health. BetterHelp Online Therapy connects you with a qualified therapist via phone, video, or live chat. It's convenient and affordable and can be done from the comfort of your own home. Having someone to talk to is truly a gift,

especially during the holidays. Visit betterhelp.com to learn more and save 10% off your first month. That's betterhelp.help.com. Two holiday seasons ago, many of us were just getting to know ChatGPT for the first time. Since then though, the tool has become a global sensation for its developer, OpenAI, displacing some businesses, helping others, and sparking endless debate about whether and how human work could change, for better or worse, because of AI. Well, out of a sea of swirling speculation on that topic, a recent paper by MIT PhD student Aidan Toner Rogers caught the attention of journal economics reporter Justin Layhart, and Justin is here to talk about it and what it adds to discussions over whether AI really

is coming for our jobs. Justin, why did this paper, Artificial Intelligence, Scientific Discovery, and Product Innovation, catch your attention? So, economists have been doing a lot of work trying to think about how AI is going to affect the labor market. And what's lacking is studies of it in the real world, field experiments. And what this paper does is it studies an actual field experiment, the introduction of an AI tool to a very large science research and development laboratory.

And they were very careful about how they rolled out this AI tool. They didn't just give it to all the scientists at once, they wanted to study it and to understand what it was doing. All right, and so just in scientists at this lab, they were trying to create new materials. Where did AI come in and what did Toner Rogers find? The way the AI tool worked was the scientists would identify what type of compound they wanted and they would put those specifications into the AI tool which has been trained on lots of other compounds and the AI tool would spit out recipes

that you can conduct experiments on. So the scientists would then do that and they would come up with these new compounds. So what he found was first it had a major effect on these scientists' productivity. They discovered way more compounds, they applied for more patents, it turned into more new products. And also importantly, the compounds that they were finding were novel. It wasn't what people call a streetlight effect where it's like, well, you would have thought of that anyhow. This is like, you know, things that were really sort of surprising and interesting. And one thing I really appreciated about this paper after reading it and listening to your

conversations with Toner Rogers were that, you know, he went beyond those findings and looked at job satisfaction because that's where this gets personal for a lot of us, right? And I want to play a clip of what Aidan told you about that. One kind of story around AI more generally is this going to kind of automate tedious tasks and humans are able to focus on more rewarding activities. And in this setting, it seems basically the opposite, where the AI is kind of automating these more creative tasks. And instead, now the scientists are just kind of sitting there and trying to evaluate these materials that the AI suggests.

And there's kind of a uniform decline in scientists' satisfaction with kind of the content of their work. The fact that they get more productive partially offsets this, but not fully. And so I had this result that 82% of scientists saw a decline in their kind of job satisfaction due to the tool.

Justin, that is a key part of the narrative around AI and jobs that really has felt missing from the discourse. Yeah. And remember, this is just one tool. So we don't know what it's going to be like in other settings. It seems like programmers are able to get rid of a lot of sort of their scut work But for these scientists, it is not like that at all. They're more productive and you could imagine that they will get paid more money so that is an offset but like just having trained up to do this creative work of working on whiteboards and thinking up new compounds and having that go away is disappointing to them.

And this is, there's sort of another question in there, right, so these scientists are put in the role of judging which recipes are best and which one to experiment on first, right? But how did they get there? How did the scientists get good at judging which recipe is good, right? Maybe they had to go through some of this creative work to get there. So if the AI is taking the creative work away, eventually the scientists aren't going to

get the training that they need. I think the same thing has been said about autopilot and flying as well. And finally, Justin, after this paper came out, you decided to check in with two prominent economists, including 2024 Nobel Prize in economics winner, Daron Asimoglu, as well as labor specialist David Otter about what they made of these findings. I'm curious, how do they see this paper updating our assessments about what AI could do to jobs or more broadly how it could change our lives.

So Toner Rogers is their student and they are incredibly impressed. Even though it cuts against some of their views on AI, Asimoglu notably thinks that AI might not increase productivity as much as some people are hoping. And in this paper, at least, you see this huge increase in productivity. It's just one setting, right, but it is impressive. Otter, on the other hand, he thinks that AI might help reduce income inequality, that it could help people who are fairly skilled become much more skilled and reduce some of these gaps.

This is not what the paper shows. The paper shows that scientists who were the most productive before the AI came out get even more productive and the less productive scientists don't really benefit from the tool. So what they want to see is more experiments like this, more sort of real world examinations of what AI is doing in the workplace. Economists have lots of theories. They know about how past technology has affected the economy, but it can be a real mixed bag there.

Zeppelins were supposed to be world changers. They were not. The washing machine was probably not seen as a world changer. It was. There we go. AI between the Zeppelin and the washing machine, perhaps. I've been speaking to Wall Street Journal economics reporter Justin Leihardt. Justin, thanks so much for bringing us this story.

Thanks for having me. And that's it for What's News for this Monday morning. Today's show was produced by Daniel Bach with supervising producer Christina Rocca, and I'm Luke Vargas for The Wall Street Journal. We will be back tonight with a new show. Until then, thanks for listening.