Kent Smet (4:11)

This is really the first time that tariffs are being used as a revenue generator for the US treasury, and the motivation behind that is that we're on this exploding debt path right now. It's unsustainable and so there has to be some revenue that comes in or some spending that goes down in order to try to stabilize the future debt to output ratio in the US economy going forward, especially starting next year. If the current tariff rates stick around, it will raise a fair amount of revenue. And of course, ultimately that revenue is coming from US households and US firms who are essentially paying that money.

Kate Bullivant (4:53)

Predicting future import demand and tariff revenue is also proving difficult, with many of Trump's tariffs still under negotiation and the Supreme Court yet to weigh in on whether the president had the authority to to impose the levies. Last month, a federal appeals court ruled he overstepped his authority when using his emergency powers to rewrite trade policy. Brazil's Supreme Court has sentenced former President Jair Bolsonaro to 27 years in prison after he was found guilty of plotting a coup to overturn the 2022 election. The move defies President Trump's efforts to derail a case that has electrified Latin America's largest country and thrust Brazil into centre of the US administration's trade war. Trump levelled 50% tariffs on Brazilian exports because of what his administration has called a witch hunt against the former leader, while the Treasury Department has imposed sanctions on a Brazilian Supreme Federal Court justice for his handling of legal cases involving Bolsonaro. Posting on social media, Secretary of State Marco Rubio said the US will respond accordingly without specifying how. Turning to markets now and shares of Nvidia supplier SK Hynix hit a record high this morning, surging almost 7% after the South Korean company said it's ready to mass produce a new generation of high bandwidth memory chips. Journal editor PR Venkat says the achievement puts SK Hynix in prime position to meet soaring demand from companies like Google, Amazon and Meta.

PR Venkat (6:36)

One of the biggest advantage for this is that first it's a next generation product with ultra high performance for AI customers so it is able to improve the performance by 69% and when the product is applied it says that the data bottlenecks significantly reduced and data center power costs are reduced which is huge when we are looking at the amount of investment and demand that is there among companies like Google, Meta and SK Hynix now gets an edge compared others. For example, we have recently announced that Oracle has said it has signed several high value contracts. So this puts Hynix in a little bit of a leader position because they're already ready to produce.

Kate Bullivant (7:20)

Ms. Coming up we look at OpenAI's billion dollar spending spree to fuel its AI ambitions and not least whether it has the money to fund those projects. That story after the break.

Oracle Advertiser (7:35)

In business they say you can have better, cheaper or faster, but you only get to pick two. What if you could have all three? You can with Oracle cloud infrastructure, OCI is the blazing fast hyperscaler for your infrastructure, database, application development and AI needs where you can run any workload for less. Compared with other clouds, OCI costs up to 50% less for compute, 70% less for storage and 80% less for networking. Try OCI for free@oracle.com Wall Street Oracle.com Wall Street.

Kate Bullivant (8:10)

OpenAI has big ambitions recently committing to spending billions of dollars on everything from computing, data centers, hardware and chips. The big question is how will the world's largest startup pay for these ventures? OpenAI has told investors it's on pace to make $13 billion in revenue. To a person familiar with the matter, the company loses billions of dollars a year. Journal reporter Elliot Brown has been digging into the numbers. Elliot, just remind us of OpenAI's spending ambitions and some of those planned projects.

Elliot Brown (8:48)

Yeah, so they've really racked up quite a lot of bills that they are on Track to spend $60 billion a year starting on a computing contract with Oracle to train their models. They've committed 18 billion to this giant venture with SoftBank build data centers. They've said they're going to make a Mass market hardware device. And then there was just a deal that was revealed where they are planning to purchase 10 billion of customized chips.

Kate Bullivant (9:15)

I mean, that's big spending. OpenAI is currently on the hook to pay hundreds of billions of dollars over the next decade for these different ventures. How does it actually have that money?

Elliot Brown (9:26)

Well, the short answer is it doesn't. They take in somewhere around $13 billion or on pace to do that this year in revenue, but then their costs are far higher than that. And so it's basically just a really, really, really big bet that they're going to see revenue completely explode over the next few years, that businesses and consumers are really just going to start shelling out and paying for more advanced AI products that we don't even have today. The bull case on OpenAI is that it is, you know, by far the fastest growing consumer app ever. They claim 700 million weekly active users, which is, you know, basically 9% of the entire globe. And the idea is that nothing has penetrated the world like this in technology before so quickly, and that they are going to see revenue just skyrocket. And if you look at how much their sales have grown, it's very fast. It's tripling year over year, which is not something you see often in tech.

Kate Bullivant (10:30)

This sounds positive for OpenAI. Why is there lingering uncertainty then?

Elliot Brown (10:35)

The short version is because they are committing much bigger numbers for spending than they are taking in money today. So it's really impressive that they've been able to grow to 13 billion in revenue, but they're rapidly agreeing to spend well over $60 billion a year in a few years. And where do you get that money in the short term? They largely have been getting these huge equity infusions, investment infusions from some really big tech investors. SoftBank is by far the biggest. They've committed 30 billion. And, you know, then there's a large number of other private equity players who are sort of heaving billions at OpenAI. An issue, though, is that the private market for raising money like that is not infinite. And Sam Altman has already tapped almost all of the biggest players in that. And it's sort of yet to be seen how much deeper it is.

Kate Bullivant (11:31)

And I guess on top of that, academics and industry experts doubting OpenAI's belief that consumers will increase their spending on AI at explosive rates for years to come. Right?

Elliot Brown (11:42)

Yeah. There's a number of recent academic reports that basically show that companies so far have not been seeing any sort of good return on their investments in AI. So lots of people are using ChatGPT, but that doesn't mean it actually makes any money for companies. So consumers have been quick businesses, a lot slower. And if that continues, that's a huge problem because OpenAI and others are expecting more from sort of the business sector than they are from consumers.

Kate Bullivant (12:09)

Journal finance reporter Elliot Brown. Elliot, thanks for joining us.

Elliot Brown (12:12)

Thanks for having me.

Kate Bullivant (12:14)

News Corp, owner of the Wall Street Journal and Dow Jones Newswires, has a content licensing partnership with OpenAI. And finally, a recent study has revealed men and women are using generative AI very differently. In fact, men are much more likely to use tools like ChatGPT than women. Out of roughly 200 million average users of ChatGPT, only 42% of them are female, and that number falls to 27% when you look at users accessing AI on their smartphone. Tech reporter Lisa Ward spoke to the study author who called the findings shocking.

Lisa Ward (12:55)

So you see a real drop when you compare men to women overall, and they found the same trend in like, really diverse groups. So whether that's college kids, whether that's business owners, whether that was in Australia or in the US or in Canada, they looked at different groups across the globe and they were able to see this.

Kate Bullivant (13:15)

And as Lisa explains, this gender gap could have consequences for the development of generative AI.

Lisa Ward (13:21)

Artificial intelligence learns from everybody and develops in a way that is gender neutral. If men are the primary users, gender value could exasperate gender biases or stereotypes.

Kate Bullivant (13:31)

For more on that study and how we use AI, check out our Tech News Briefing wherever you get your podcasts. And that's it for what's news for this Friday morning. Today's show was produced by Daniel Bark. Our supervising producer is Sandra Kilhoff. And I'm Kate Bullivant for the Wall Street Journal. We'll be back tonight with a new show. Until then, have a great weekend and thanks for listening.

Foreign Correspondent (13:57)

Foreign.

Viking Advertiser (14:04)

This message comes from Viking, committed to exploring the world in comfort. Journey through the heart of Europe on an elegant Viking longship with thoughtful service, destination focused dining and cultural enrichment on board and onshore. And every Viking voyage is all inclusive with no children and no casinos. Discover more@viking.com.