OpenAI’s Vision of 220 Million Paid Users by 2030, Crypto Winter Will Be Different | Nov 26, 2025

0:00.

0:12Welcome everyone to the Informations TI TV. My name is Anita Romaswamy. It is Wednesday, November 26th, and I'm delighted to bring you today's show. First up, we have new projections from OpenAI about how many paid ChatGPT subscribers they expect to have by the end of the decade. Then we'll be talking about big tech stocks and what's moving markets with Nancy Tangler, CEO of Laffer Tangler Investments, plus the second installment of the Information's weekly finance newsletter and a warning for crypto investors. We'll end the show with a really fun story about AI and music rights on the ice at the Winter Olympics.

0:49It's a great lineup today, so let's get into it. The information is out with new reporting about how many paid subscribers OpenAI says ChatGPT will have by the end of the decade. The author of that piece joins me today. Shreema Pedi, welcome to TITV. Excited to be here. Excited to have you, Shree. So what exactly are these new numbers that OpenAI is reporting? Of course. So OpenAI expects about 2.6 billion people to be weekly active users in 2030. Just as a point of comparison, that's about a third of the world's population today, 8 billion. And at least 8.5 % of them, or 220 million, are expected to be paying users.

1:31That's at least 8.5%. As a point of comparison, that's up from about roughly 5 % of users that pay for either ChatGPT Plus, which is the$28 a month plan, or ChatGPT Pro, which is the$200 per month plan. So from a financial perspective, Sri, how do the paying versus non-paying users impact OpenAI's business today? Of course, I think the primary area that non-paying users affects OpenAI's business is that a vast majority of the 800 million plus weekly active users today are actually not paying. As I said, only 5 % or so pay for the consumer subscription plans. And then also there are also others that pay for the enterprise subscription plans, but still a vast majority are non-paying.

2:22And you really see that show up in OpenAI's gross margins. For example, the inference costs or the cost of running AI models as well as products, OpenAI needs to support that for both its paying and non-paying users, but of course, only the paying users generate the revenue. So when you actually look at the gross margins, OpenAI has roughly just 50 % of gross margins across all of its products this year. So one plot line we've been talking a lot about on the show and following closely is Google's new launch of Gemini 3. That's really been the talk of the town in the last week in tech. And I was wondering if you thought, Sri, that the projections for OpenAI are actually feasible in the context of some of that excitement we're seeing around Gemini?

3:05I think they are quite ambitious goals, just given that having one-third of the population be on top on using ChatGPT, it's quite a lofty goal and an ambitious goal in itself. But also Google Gemini has really caught up. I think with the launch of Gemini 3, as you mentioned, there's just so much more excitement around Gemini and the potential there. I believe the latest number is around 650 million users are using Gemini. And that's just comparing that to 800 million plus that OpenAI has. It really seems that Google Gemini is catching up. And so I think that time will tell in terms of how OpenAI will actually be able to hit these goals.

3:45So I'm excited to see what the next few years will bring. You know, another thing you mentioned in your piece was some of the content restrictions that OpenAI has introduced on ChatGPT earlier this fall. And I was wondering if you could talk a little bit more about those. Are they going to be a hurdle to these projections? Yeah, so OpenAI started launching a few types of different types of content restrictions starting in August. This is, for example, to limit suicide talk or flirting with users for users that are under the age of 18. And so this is around just amping up their teen safety content restrictions.

4:20Sam Altman has said, though, that they plan to roll out age verification by the end of this year in December where verified adults can actually have more of those capabilities back. And so I think as a result, there was like some, some bit of a slump in terms of chat GPT usage. But I think that like the content restrictions role in terms of verification will allow at least more open AI users to use the product fully to the capabilities that they have been used to. Moving a little bit back to the big picture, Sri, what is the strategy that OpenAI is using to try to get more people to pay for ChatGPT?

5:02For sure. So OpenAI has such a vast majority of folks who are non-paying users using the free tier right now. And they limit it based on different types of features or not having the fullest capabilities available. And so they're trying to upsell folks to use the pro or the, sorry, the plus, the$20 plan or the pro plan, which is a$200 per month plan. and then be able to increase the number of paying consumers that way. The broader strategy, which I think is interesting, which is my sort of observations on how they're working through it, is that as you get more and more people using ChatGPT, it's a way to actually be able to do bottoms up selling into businesses as well.

5:42And so as, for example, more individual employees are paying for ChatGPT on their own, and you see widespread usage, for example, at a company, then it might make sense for the company to actually pay for an enterprise plan. And so that's a way for OpenAI to sort of enter into the businesses as part of their go-to-market motion. And as a point of comparison, Anthropic, for example, has more of a direct sort of B2B approach into selling their API plan more directly rather than taking that approach with Claude being sort of their biggest wedge. Yeah, I wanted to ask a little bit more about that.

6:17I mean, you likened OpenAI strategy right now to Zoom and Slack and maybe some of the enterprise SaaS companies that we've seen. When it comes to AI and AI chatbots, are there any chatbots in the game that are not using the strategy to try to get more paid users? I think that most chatbots are taking this premium or freemium approach where they're offering a lot of their features for free and then being able to upsell them to premium subscriptions as well. for example, like Perplexity, even Cloud has a free and paid tier. I think the main difference that I was mentioning about OpenAI and Anthropic is that Anthropic, for example, makes 80 % of its revenue through its API business versus OpenAI generates a vast majority of its revenue from its ChatGPT business.

7:01And so as a result, you see go-to-market motions that look slightly different for those two AI labs. But I agree that for consumer play chat bots, similar to Perplexity or even like chat GPT, as we were talking about, a lot of it is like this, just get widespread usage and then sort of like think about monetization as more and more people use it. And that really follows the playbook of like a Zoom or a Slack where you're sort of seeing it pop up and then everyone's using it. And they're like, why not we just pay for the tier for the whole company and sort of get the collaboration features and the security features that a company actually really seeks out.

7:41Got it. Well, Sri, you really have your finger on the pulse of what's going on at OpenAI, and I'm sure we'll want to have you back on the show to talk more as the company moves closer or further away from those projections. Thanks so much. Thank you. Big tech and AI stocks have been moving markets in the last week, with names such as NVIDIA and Oracle dropping, while Google and Meta rose. Here to help us make sense of the moment is Nancy Tangler, CEO and Chief Investment Officer of Asset Manager Laffer Tangler Investments. Nancy, good to have you here. Thanks for having me, Anita. Nancy, it feels like the biggest name on tech investors' minds right now is really NVIDIA.

8:19So I want to start there. We saw the company's share price dip yesterday after the information reported that Google was in talks with Meta and other companies about letting them use Google's TPU chips in their data centers. And that could pose a challenge to NVIDIA. NVIDIA later put out a statement saying they're delighted by Google's success and they've made great advances. and that NVIDIA is a generation ahead of the industry. What are you making all of this? And do you see any one company kind of closing that gap with NVIDIA? Yeah, except I don't think it's necessarily Google. It may, in fact, be AMD.

8:51But I will point out, Anita, that they also issued a similar statement after the DeepSeek situation. And so what you have to look at with NVIDIA, I think, is really two things. Think of NVIDIA chips as the entrance into the ecosystem. Their software, CUDA, is really what developers utilize to develop programs. It's much like Apple's App Store where the cell phone is just the gateway, the smartphone. And so I think NVIDIA is, in fact, years ahead of all the nearest competitors. Now, the good news is one of our largest holdings in our ETF TGLR is Broadcom. And they are the ones that are designing and partnering with Gemini slash Alphabet on the TPUs.

9:44So I think there's plenty of room for a lot of winners in this environment. But right now, the market's just focused on headlines. And are you an owner of NVIDIA and AMD stock? We are, and we own Apple and we own Google. So we have different strategies for different reasons. Google was actually added in our value portfolio first, and then we added it back into growth. Because remember, it was left for dead just a few short months ago. Nobody wanted to own it. It was the laggard in the Mag-7, and now it's the darling. I think we're going to see this leapfrogging for a number of months, maybe even years.

10:26I guess speaking about names that are in your portfolio, Nancy, another name that I thought was interesting is Oracle. The stock is down some 30 % in the last month, and I know investors have been really worried about the company's capital-intensive push to turn from a software company to a cloud provider. I mean, we had a story yesterday about how Oracle and its construction lenders are borrowing huge amounts of money to fund that push. Yet you said that you bought Oracle shares yesterday. And I wanted to know why that is. Yeah, well, a couple of reasons. You know, again, the headlines and the hedge funds and the algorithms are really driving short-term performance in this stock.

11:02But if you look at the debt-to-equity ratio, at the end of the quarter, it was 427%. That was down from 781 % in August of 2024. Now, they did go to the market for a bond offering. That was$18 billion. That drove the debt-to-equity ratio up to 500%. But this is a company that has a long history with debt. And if you look at, you know, people are lumping in the senior credit facilities, which are about$56 billion. They haven't drawn on those yet. And if you look at grant property and equipment, up 130 % year-over-year. Debt is only up 9%. So I think everyone needs to take a deep breath. The RPO backlog has already been taken out of the price of the stock.

11:48And that's why we thought, you know, we were selling it in the low 300s in the summer or actually in September. And then now we're buying it back. I mean, we still held it, but we're adding back to our holdings because we think it's been overdone. And I guess zooming out bigger picture with Oracle, are you not concerned about its cash firm and its long-term future, considering its push? I am not concerned about that. What I am concerned about is the over-dependence in this market or the perceived over-dependence to open AI. And that's something that I think is worth worrying about because that's a company that's got a burn rate and very low levels of revenues.

12:28Oracle has a number of businesses and they will figure this out. Larry Ellison still owns 42 % of the company. We talked to the company. They're going to defend their investment grade ratings. So I don't think Oracle's necessarily the big problem unless there's a problem with OpenAI. And then I think that impacts the entire space. But let's also remember that data center build out is dependent on, I think it's like between 105 ,000 employees to build a data center, Anita. And we've already got 450 ,000 manufacturing jobs that are unfilled. So you've got construction and manufacturing workers required in these data centers.

13:10That may slow down the build out just naturally. But in the meantime, Microsoft, Satya, and Adela had said, these data centers are fungible. We can use them for other things. So I think there's a desire to be panicked. I was managing money in the 90s. This is not that. I do want to get to the AI bubble talks. But before we get there, I just wanted to ask, is there anything else that you're buying? It sounds like you're an owner of many of the MAG7 stocks, but what right now is speaking to you outside of Oracle? Yeah, we're actually buying consumer discretionary, Anita. I think the One Big Beautiful Bill Act has not really been factored into most investors' thought process.

13:49$150 billion in additional tax refunds should be paid to U.S. taxpayers in the first quarter. That's for a total of over$500 billion. So we think the consumer discretionary names will benefit. We've been adding to TJ Maxx, TJX, and Ulta, ULTA. And there's other names in the space we like. We recently added to Starbucks. I think that's a good offset. And then one of our largest holdings is Walmart. And I think Walmart is interesting on so many levels. We bought it about, I don't know, eight or nine years ago. It has become the poster child of our investing theme, which is old economy companies embracing the new technologies.

14:30And Walmart is also, it just provides a lot of insight when you listen to management. They're only seeing about 1 % inflation in their purchasing. So that's pretty compelling if you're also worried about the Fed and inflation and whether we're going to get a cut or not. So those are the names that we like in staples and in consumer discretionary. Interesting. So I wanted to flip the conversation back to AI for a second. What is one AI stock or tech stock that has AI exposure that you're not optimistic about in today's market? So we do not own Meta. We sold it at lower levels than it is today. We bought Spotify.

15:11This is the dilemma for every portfolio manager. You always look at what you sold and you go, oh, but it kept going up. Well, Spotify went up a lot more, and that's how we have to console ourselves. We've been sort of circling the wagons around it, but I'm not ready to pull the trigger yet because the monetization of their AI spend is a little bit more difficult for us to wrap our brain around. So we've sort of stayed off to the sidelines and will continue to watch that one, but we don't own it currently. So I've got to ask Nancy, I want to wrap this up by getting your thoughts on, are we in an AI bubble or not?

15:44And why or why not? So I have been drawing the analogy to the 1990s, Anita, for three years because I was a portfolio manager then. And there are a lot of similarities, higher interest rates, higher inflation, productivity-driven growth. But the companies that we were focused on in that, or at least that were driving the growth, had declining earnings. So the growth names, earnings were contracting while their valuations were soaring. That is not what we have today. the growth names that are driving this particular technological revolution are generating about 20 % growth on average in EPS. And then in addition to that, the CapEx spend in the 90s, it soared for the entire decade.

16:32We're just getting started. So I think we're in the beginning of the runway. This won't go on forever, but it will go on for some time. And I think for your viewers, you may want to just, you know, focus them on some of the second and third derivatives. One of the names that we own in our thematic portfolio is Symbotic, which has done all of the backroom automation for Walmart. They reported yesterday the stock was up 20%. Don't necessarily buy it today, but look for those kinds of names. I mean, the earnings growth in that name is going to be over 100%. And the adoption to robotics and back office automation or logistical automation is just getting started.

17:14Well, Nancy, you are full of ideas and there is never a dull day in the AI market. So thank you very much for coming on to the show today. Thanks, Anita. The Information is out with the second edition of our weekly finance newsletter. And this one includes a clear warning for crypto investors. Joining me now is Ken Brown, finance editor here at The Information. Hi, Anita. Hi, Ken. Good to have you on the show. Yep. So the first line of your piece says, winter is coming, not just in the seasons, but in the crypto market. Can you explain what you mean? So crypto winter is a phrase that people have used a bunch of times, particularly after the FTX collapse.

17:53Things were, you know, prices were down. It was really ugly out there. So they called it the crypto winter. Right now, we're in another bad sell-off. Crypto's, you know, down a bunch. the most speculative stuff is down a lot. And so, you know, the idea of if it gets worse, it's going to be another crypto winter. So I was playing off the fact that it's winter. It's almost winter. And where I live, it feels like winter. And so that was why. That's the first line. Well, it's chilly here in San Francisco too, Ken. But I wanted to ask you, what is the center of this current crypto meltdown? What's really driving it?

18:27So, you know, there was a boom and prices went up a lot. And now they're coming off. I think what's happened in the market is it just got very, very speculative. There were a ton of crypto assets put on the stock market through these treasury stocks. The most famous is Strategy, the company called Strategy. And so they loaded up on crypto. They were on the stock market. And these things just went through the roof and kept buying up more crypto. And that is now over. And those prices have come down a lot. And that was, by the structure of these things, that was driving crypto higher. And by the structure of these things, as they go down, they drive crypto lower.

19:08So that's what's happening now. So there's this big sell-off. so another thing that you mentioned about the crypto market in your piece was that stable coins have gained a lot of popularity over the last year in july we saw the trump administration sign the genius act into law which that really created the first regulatory framework for stable coins and you argue that they're very risky i wanted you to say a little bit more about that and what the risks are right so the difference between stable coins so let's define stable coin is something it's always worth a dollar it's one-to-one against the dollar and other crypto goes bananas, right?

19:41It goes up and down like a stock really, or more. And so stable coins, they promise you they're going to give you your money back. That's the promise of a stable coin, just like it's the promise of a bank deposit or a money market fund. And that's hard to do. And we've seen runs on banks and on money market funds and on anything that promises stability over decades and decades. And so they're making a big promise. And the question is, can they keep it? And so the history is no. Stablecoins have broken that dollar peg a bunch of times in the past. One just happened very recently in this sell-off.

20:19It was a bit of an oddball stablecoin, but it broke the buck. What was the stablecoin that recently sold off? It's sort of a no-name stable coin in Asia that they call it algorithmic. So it's not one that's certified in the US or matches the US law. But it fell 90-something percent. And that's the end of that. Money was missing. They lost money on their investments. And then everyone freaked out. What happens... So the difference is when you're an investor and you own a stock and it goes down 10 % or 20%, okay, you might be unhappy. But that's what happens to stocks. When you put your money in a bank or in a money market fund and it falls 5%, it's not supposed to fall.

21:04You freak out. You want your money back immediately. And you and everybody else causes a run on the bank. And so the risk is with stablecoins, there is a run. And the problem is stablecoins have gotten so much bigger and they're starting to be used more for holding cash in countries where there's unstable currencies like Nigeria or Argentina. People are buying stablecoins. And the risk is that something goes wrong and there's a run and people want their money back. And then there's a liquidation and it just it's destabilizing to the system. So that was the main thing I was getting at. So like you mentioned, Ken, this isn't the first time that we've seen a stablecoin blow up.

21:46And you talked a little bit about this in your piece. Can you share a little bit with our viewers about what we saw when Silicon Valley Bank failed a couple of years ago and how that episode actually related to stablecoins? Right. So Circle is one of the biggest, it is the biggest stablecoin in the US. It's the second biggest in the world. They had$3 billion sitting at Silicon Valley Bank. Silicon Valley Bank collapsed. Basically, the value of Silicon Valley's treasuries that it was holding fell when interest rates went up. So safe treasuries, but they lost value. It fell. The bank got in trouble.

22:19And basically, these deposits were not insured. And so theoretically, Circle could have lost a bunch of money. They announced it. Everyone started to sell their Circle. And the stablecoin fell to like 88 cents on the dollar. And so general freak out. And the US government came in and saved the day and guaranteed all this stuff. So the problem is like, that's not going to happen every time. And it just creates this sort of systemic risk because people start selling one thing and then they start selling something else to get the money to sell something else. And it is this what they call contagion, but it's this domino effect of things starting to, the market starting to tank one after another after another.

23:03And that's what we've seen in the past. Zooming out here, Ken, can you tell us a little bit more about what you're watching for next in crypto more broadly and just sort of how the sector's impacts could ripple through the economy on a larger scale? Well, you know, crypto is more intertwined with the traditional financial system than ever before. And these stocks, crypto stocks that I talked about before are part of that. And so we're going to see this is really what I call that what I was saying is this is a new, it'll be a new kind of winter because because it's so interlinked, so much more interlinked.

23:37There's a ton of crypto on the stock market. There's these stable coins. There's other products, you know, and everyone's issuing these stable coins. I mean, yesterday, Klarna, which is the buy now, pay later company, said it's going to issue its own stablecoin. So what I'm really looking at is how this sell-off, whether it ends now or keeps getting worse, interacts with the global financial system, the traditional financial system. I mean, the regulators of this system are very worried about this and looking really carefully. And so this is a test, right? But what I really wanted to tell people is, like, this is a test.

24:13Pay attention. this is going to tell us how crypto and the traditional financial system are going to react this time. Got it. Well, history certainly repeats itself and we'll all be watching closely and paying attention to that. And I'm sure we'll have you on the show again to talk about it. Thanks so much for being here, Ken. Thanks, Anita. We're closing out today's show on a high note. AI is infiltrating nearly every corner of life, including one that you might not expect. That's ice dancing at the Winter Olympics. Here to explain is our sports business reporter, Sarah Germano. Sarah, welcome to TITV.

24:47Thanks for having me. Couldn't be more excited to talk about this. Yeah, let's talk figure skating. So can you just tell me a little bit about how figure skating is looking to use AI? Yeah, so to start, like, if you think about figure skating, it actually used to be the second most watched sport on TV in the U.S. in the 90s, behind only the NFL. It was huge. If you think about it today, it's virtually invisible outside of the Winter Olympics. So people officiating and in charge of the sport are trying to make it more popular and more accessible in the streaming age. There's one rule this season that ice dancers have to skate to music from the 1990s, like Spice Girls or the Backstreet Boys.

25:27But that comes with a very serious complication, which is that more contemporary music has stricter licensing requirements. And neither the sports world nor the music industry has totally figured this out yet. So the figure skating governing body, mind you, has given skaters permission to use AI-generated music for their programs, which is going to be sort of a first at the Winter Olympics. There is one Czech ice dance team who is using a partially AI-generated program that is inspired by Bon Jovi. It's pretty incredible that figure skating used to be so popular and now the script has completely flipped.