# SpaceX Raised $111B — It Already Needs More

Markdown copy of https://realalpha.app/v/9SAgLMahZm4/ on Real Alpha, the best minutes from long interviews with the people building AI. Cite and link the page, not this copy.

John Foley says SpaceX’s $40 billion borrowing plan depends on rapid data-center payback; Ronan Farrow says firms profit by making customer service harder.

- Show: Prof G Markets
- Hosts: [Ed Elson](https://realalpha.app/people/ed-elson/)
- Guest: [John Foley](https://realalpha.app/people/john-foley/) · Head of the Lex Column, Financial Times
- Guest: Ronan Farrow · Journalist, The New Yorker
- Published: Oct 8, 2026
- Length: 36 min

## The best minutes

- [2:40](https://realalpha.app/v/9SAgLMahZm4/?t=160) [SpaceX needs rapid data-center payback to fund its expansion](https://realalpha.app/s/spacex-needs-rapid-data-center-payback-to-fund-its-expansion/) (4 min)
- [9:35](https://realalpha.app/v/9SAgLMahZm4/?t=575) [SpaceX’s debt is riskier, while data centers drive the thesis](https://realalpha.app/s/spacexs-debt-is-riskier-while-data-centers-drive-the-thesis/) (4 min)
- [13:20](https://realalpha.app/v/9SAgLMahZm4/?t=800) [SpaceX’s valuation rests on faith, not reliable price discovery](https://realalpha.app/s/spacexs-valuation-rests-on-faith-not-reliable-price/) (5 min)

## Also worth watching

- [32:06](https://realalpha.app/v/9SAgLMahZm4/?t=1926) [OpenAI’s research progress does not prove its business model works](https://realalpha.app/s/openais-research-progress-does-not-prove-its-business-model/) (4 min)

Skip the 11-minute annoyance-economy discussion and the promotional breaks.

## Full analysis: John Foley says SpaceX's 10-gigawatt build may cost about $500 billion

- Foley's rough math puts SpaceX's 10-gigawatt data-center target at about $500 billion.
- That is a rough build estimate, with payback resting on fast construction and rentals.
- Rising credit risk and a thin float make borrowing costs and price discovery the next tests.

Its 6 sections, what it means for you and what to watch are on [the page](https://realalpha.app/v/9SAgLMahZm4/#analysis), free with an account:

- [3:02](https://realalpha.app/v/9SAgLMahZm4/?t=182) A rough target implies a $500 billion build
- [4:50](https://realalpha.app/v/9SAgLMahZm4/?t=290) Payback depends on speed and sustained rental demand
- [10:24](https://realalpha.app/v/9SAgLMahZm4/?t=624) Credit markets see rising risk, not distress
- [13:00](https://realalpha.app/v/9SAgLMahZm4/?t=780) Starlink helps the credit case, not the whole valuation
- [17:26](https://realalpha.app/v/9SAgLMahZm4/?t=1046) A thin float weakens price discovery
- [5:05](https://realalpha.app/v/9SAgLMahZm4/?t=305) Boards should gate spending on realized rental economics

## Summary

Foley is head of the Lex column at the Financial Times. Farrow is a Pulitzer-winning journalist at The New Yorker.

Foley says SpaceX’s borrowing needs reflect its costly data-center ambitions, and that the plan depends on building quickly and earning returns from rentals. He says rising credit risk and the stock’s small free float make its $2.3 trillion market capitalization a weak measure of price discovery. Farrow argues that companies deliberately make service and cancellations harder to reduce costs, sometimes blocking people from benefits or care; state-level enforcement is advancing while federal action lags. Elson says OpenAI’s unverified math proofs may mark a research breakthrough but do not show that the company can become profitable.

### Minute by minute

- [2:40](https://realalpha.app/v/9SAgLMahZm4/?t=160) **Foley says data centers need vast funding.** Foley says SpaceX has about $90 billion in cash after its IPO and debt sales, but Musk wants 10 gigawatts of data centers by the end of next year. Using a rough $50 billion per gigawatt, Foley says the buildout would cost about $500 billion. He says the bet is that rental income could repay a data-center investment in less than a year, if SpaceX can build fast enough.
- [5:23](https://realalpha.app/v/9SAgLMahZm4/?t=323) **Foley questions SpaceX’s investor pitch.** Elson questions whether plans for data centers on Earth and in space fit together, citing a two-page investor memo with pictures of space. Foley says it appears to be an early, informal pitch, though investors are still being asked to trust Musk’s space ambitions. Foley also agrees that bank fee incentives can create conflicts and says many high price targets rely on projected revenue and multiples.
- [9:35](https://realalpha.app/v/9SAgLMahZm4/?t=575) **Foley sees rising risk in SpaceX debt.** Elson says SpaceX credit default swaps rose to 194 basis points on Wednesday, from 110 in June, while bonds sold off. Foley says 50-year bonds were trading around 85 cents on the dollar and five-year bonds around 95, signaling greater risk but not distress. He says investment-grade ratings depend partly on confidence in Musk’s plans and Starlink, and that debt looks less concerning if analysts’ forecast of revenue quadrupling by the end of 2028 proves right.
- [13:20](https://realalpha.app/v/9SAgLMahZm4/?t=800) **Foley doubts SpaceX’s price discovery.** Elson says SpaceX shares rose from an August low of $108 to $167. Foley points to a successful Starship orbital flight as one piece of news and says lockup expirations were already anticipated. He says Musk’s fan base and Wall Street’s incentives can support valuations detached from fundamentals. Foley cautions that the $2.3 trillion market capitalization is not a reliable measure of value because the free float is small.
- [32:06](https://realalpha.app/v/9SAgLMahZm4/?t=1926) **Elson separates math progress from business viability.** Elson says OpenAI released hundreds of proofs it says solve difficult problems, including a claimed advance on the Riemann hypothesis, but notes that they have not been verified. He sees the work as a major research advance while arguing it does not prove OpenAI can become profitable. He says the company remains deeply unprofitable, with expanding losses, and is pursuing a $1.5 trillion valuation despite revenues below one-tenth of Walmart’s.

## Transcript

Prof G Markets, hosted by Ed Elson, with John Foley and Ronan Farrow.

Transcribed by machine (Whisper) from the episode’s audio, so a word may be off here and there. Where the podcast’s audio and the video part ways, some times are a minute or so off.

[0:25](https://realalpha.app/v/9SAgLMahZm4/?t=25) market vitals. The major indices fell from record highs as bonds sold off again. The yield on 10-year treasuries hit a new 24-year high. It later cooled off following a 10-year note auction that drew solid demand. Meanwhile, minutes from the latest Fed meeting showed another rate hike is likely this year, though it may not come until December. And finally, Brent crude remained elevated around $101 per barrel. Okay, what else is happening? Four months ago, SpaceX raised $86 billion from its IPO. Now, it's looking to borrow even more. On Tuesday, the Financial Times reported that SpaceX is seeking $40 billion to buy NVIDIA chips.

[1:41](https://realalpha.app/v/9SAgLMahZm4/?t=101) $30 billion. will come from investment-grade bonds, and the other $10 billion will come from bank loans. Reportedly, the investors pitched on this deal got a two-page memo, which showed pictures of outer space with an arrow pointing to data centers, quote, somewhere in the universe. This is not the first time that SpaceX has borrowed big. Less than two weeks after its IPO, SpaceX sold $25 billion worth of bonds to investors, which now trade at 85 cents on the dollar. Still, the stock has been on a tear this month. SpaceX shares are up about 12% over the past week. Yesterday, the stock closed down 2%.

[2:20](https://realalpha.app/v/9SAgLMahZm4/?t=140) OK, here to break it down, we're speaking with John Foley, head of the Lex column at the Financial Times. John, thank you for joining us on Prof G Markets. So SpaceX wants to borrow $40 billion. dollars uh kind of crazy considering they raised 86 billion dollars just four months ago then right after that they issued another 25 billion dollars worth of debt and now they're saying that's not enough now we need more now we need 40 billion dollars let's just start with your initial reactions well spacex already has quite a bit of cash on its balance sheet because it did this ipo it raised debt shortly afterwards it has something like 90 billion dollars but what elon musk wants to do is build, he says, 10 gigawatts worth of data centers by the end of next year, because we measure these things in gigawatts. And a gigawatt costs very rough numbers, about $50 billion to build. So that would be $500 billion. So unsurprisingly, he needs more money with which to do that. So it's not totally surprising that he's now coming to the market and saying, can I have some more money? He absolutely needs that and much more if he's going to meet his goals of data centers on earth in space or anywhere else in the universe so 500 billion dollars i heard

[3:33](https://realalpha.app/v/9SAgLMahZm4/?t=213) that correctly right that's that's how much he would need to raise in order to achieve those goals yeah that's like taking a very rough 50 billion dollars for a gigawatt worth of data center and his plan to been build 10 gigawatts by the end of next year so i guess the question and and this seems to be reflected right now in the bond markets right now especially the credit default swaps markets, which is like, where is the money going to come from? And this is something that we had been talking about when they first went public, which is that in order to keep up with their extraordinary costs, they're going to have to raise huge amounts of money. We're already starting to see signs of it. They raised the IPO money.

[4:13](https://realalpha.app/v/9SAgLMahZm4/?t=253) then they go and issue the debt now they want to issue more debt and i guess the question is like how are they going to keep doing this i mean if they're not generating the profits do they just raise half a trillion dollars worth of debt so that's the that is the very important question is how are they going to pay for all this how are they going to how are they going to service this debt um of course they do want to make profit spacex is already making a profit from its connectivity business from Starlink. And the idea is supposed to be that because data centers

[4:43](https://realalpha.app/v/9SAgLMahZm4/?t=283) **John Foley:** are in such hot demand that you can get your investment back very quickly. So the CFO of SpaceX said quite recently that they can get payback on a data center investment in less than a year. So if that remains true, then as you're building data centers, you're also renting

[5:01](https://realalpha.app/v/9SAgLMahZm4/?t=301) them out at high enough rates that you can make your money back. Of course, the question is, can you actually build them in time? What happens to the price of data centers in the kind of rental market in the meantime? But Musk's bet is that he can do this faster than anyone else. So while everyone is scrambling to find space in data centers, he'll be able to bring his online and he'll have no problem. This is what investors are being asked to believe in renting those data centers out at very high rates to people like Anthropic, to Google, to anyone else who needs them the idea is building data centers quickly but also the idea is putting data centers in space and to me those two things seem to be completely at odds with each other you can tell i i mean you can see my prize i think that this is a little bit ridiculous but i'm just going to read you a quote from the financial times report i don't think i'm being unfair when i say that this seems unserious uh here is exactly what was reported quote investors whom spacex has previously approached about financing its multi-billion dollar chip purchase said they only received a short two-page deal memo with pictures of outer space and an arrow pointing out that the company was going

[6:07](https://realalpha.app/v/9SAgLMahZm4/?t=367) to build data centers, quote, somewhere in the universe. How are we supposed to take that to the IC, one of the people said, referring to the in-house investment committee that approves transactions. This is just an extraordinary anecdote that that is actually the pitch on a $40 billion transaction. I mean, what was your reaction to that story? So to be clear, although I'm discussing all the stuff with a straight face, like if you think that SpaceX is really worth $2.3 trillion, which is where it is today, it's because you think Elon Musk is some kind of anomalous genius. And if you don't think that...

[6:43](https://realalpha.app/v/9SAgLMahZm4/?t=403) you should not be in the shares because it would be breaking all of the rules of finance as we know them. So to hand out investors a two-pager with some pictures of space on is kind of wild. It's not like that is going to be the final documentation for this debt issue. I think what we're looking at here is some early communications with would-be investors that seemed very back of the cocktail napkin. That reflects the fact that Musk kind of shoots first and asks questions later or answers them later and people give him money because he is elon musk when they actually get to do this issue which is they're talking about it closing next year there will be something a bit more concrete than that but it will still be based on this idea that his moonshot literally moonshot putting putting stuff into space putting data centers into space is viable which a lot of people think a lot of scientists think it is not but it does seem to speak to this this issue that seems to be kind of pervasive on wall street specifically when it comes to SpaceX, where everyone seems to be shooting completely from the hip.

[7:42](https://realalpha.app/v/9SAgLMahZm4/?t=462) I mean, it's not just this two-page deal memo where they're like, we'll just do it. We'll put data centers in space. Don't worry about it. We'll figure it out later. But we see this in a lot of the price targets from the investment banks as well. Morgan Stanley putting a price target at $300 a share. I mean, Raymond James coming out and saying that the company is actually worth $10 trillion. And my little conspiracy theory about this, which I'd like to get your reaction to is that if they know that this company is going out to issue hundreds of billions of dollars of more transactions, whether that be more equity issuance or more debt issuance, they also know that those are fees that they can profit off of. And they want to be the banks to go out and help with those transactions and issue that debt, issue that equity. So it seems to me that maybe there's a little bit of a conflict of interest here, where there's so much money in the pipeline here that they don't care about the numbers. And yeah, they'll just accept a two-pager with pictures of rocket ships and pictures of space. For sure. And like the whole premise of sell-side bank analysts' work is that you are being asked to believe that there is a Chinese

[8:54](https://realalpha.app/v/9SAgLMahZm4/?t=534) wall that means that they're not writing from a place of the fees that their investment bank is going to make like we you know they say that very earnestly if you don't believe them then you should take all the analysis that you see from them with a pinch of salt and as you said these target prices are really high when you look at how the target prices were derived in many cases. You have research notes that are dozens of pages long, but at the end of it, they extrapolate what they think revenue will be next year and slap a multiple on it, which does not to me seem very rigorous or realistic. It seems to me like a lot of them started with the answer and worked their way backwards. But there you have it, dozens of pages of information and discounted cash flows and whatever you like. So do your own homework, I guess, is the moral of that story. If you believe Musk is a genius, you're not reading that note anywhere. You're just buying the shares or whatever he tells you to pay i'm not sure anyone is reading these notes even the the bears or the bulls um i just want to point to what's happening to the spacex credit default swaps which rose to 194 basis points on wednesday uh they were trading at 110 basis points in june so essentially what's

[10:00](https://realalpha.app/v/9SAgLMahZm4/?t=600) happened here is the cost of ensuring spacex's debt against default is suddenly getting very expensive and the debt itself is also getting sold off as well so it seems like the the bond markets are starting to tell a kind of a different story which is they are very worried about the debt what do you make of what we're seeing in those markets i think they're a bit worried about the debt so the credit default swaps are as you said those spreads are rising and that's because people look for something they can use to hedge against this all going wrong the bond prices have also fallen more at the long end so the 50-year bonds are trading at something like 85 cents on the dollar the five-year bonds are trading much closer to i think it's like 95 last time i looked and that's not a sign that people are you know desperately worried but it is a sign that they're aware that this company is taking on much more debt now the idea is that its cash flows are also going to grow very quickly um you know i think it's it's revenue is expected to quad again we're back to the analysts here but expected to quadruple by the end of 2028 if that happens then like these debt numbers start to look less scary. But certainly, and the other thing I should point out is SpaceX

[11:05](https://realalpha.app/v/9SAgLMahZm4/?t=665) at the moment, for whatever reason, has an investment grade rating. Were it to lose that investment grade rating for the credit agencies, then creditors would start to worry because the debt would become much, the yields would go up a long way and it would become much more risky.

[11:19](https://realalpha.app/v/9SAgLMahZm4/?t=679) **John Foley:** So we should definitely be looking at the credit worthiness of SpaceX. At the moment, the market's not sending out a distress signal, but it is saying there's companies getting riskier by the minute. What do you make of that investment grade rating? You said for whatever reason it has

[11:32](https://realalpha.app/v/9SAgLMahZm4/?t=692) it. Do you think there are reasons it shouldn't? Again, much ink has been spilled by credit rating analysts to justify those investment grade ratings. They're based on the idea that He'll do a lot of what he says he's going to do. Also, the Starlink is actually a pretty good business. But obviously, those ratings are up for grabs, like they get reviewed. If investors don't believe the ratings, then that's a problem for the ratings agency. So it's not in their interest to stretch credulity too far. A lot of people in these financial markets have a lot of faith in him. It is true that the Starlink business and the connectivity business is a really good business and also profitable and has completely dominated that market.

[12:11](https://realalpha.app/v/9SAgLMahZm4/?t=731) but it does seem as though this company isn't about Starlink and it isn't even about spaceships anymore. It's about data centers. And they spent more than $18 billion on CapEx last quarter and it's all going into these data centers. So, I mean, do you struggle as I do to see arguments as valid that are centered around this company being, you know, at least the price being justified by the fact that the Starlink business is strong. I mean, it's not even really about data centers. If you looked at their IPO filing, they had a 23-ish trillion dollar total addressable market of which 22 trillion was enterprise apps.

[12:51](https://realalpha.app/v/9SAgLMahZm4/?t=771) And really, they don't have any enterprise apps at the moment, really. Cursor, I guess, the coding tool. Data centers are what's happening now. And data centers are actually where the revenue is. So realistically, if we want them to generate cash, renting out data centers to AI labs is not a bad way of doing it. um it's just that that revenue is very early stage so he needs to show that he can build the stuff he can rent it out um starlink is not the core it's not the future of spacex and he's not pretending it is i mean the future of spacex is colonies on mars like that's a whole different conversation it's not unwise of him if he thinks he can build data centers quickly to do that now and try and rent it out to people who are more desperate than he is let's just uh take a look at the stock um it's up around 10 in the past month i mean i had a three-part prediction for spacex part one was it would explode on the ipo got that right part two is it would then get cut in half got that right and part three was that it would start after that point or it would continue to slide as the lockups expired got that wrong uh it's risen to 167 dollars a share from its low of 108 in august it has exploded back up to as you say 2.3 trillion dollar uh market cap

[14:09](https://realalpha.app/v/9SAgLMahZm4/?t=849) um what do you make of of of the rise in this stock which from what from what i can understand doesn't seem to be happening on the back of much news it seems to be mostly sentiment there's been a tiny bit of news right there's been there have been some launches that were quite successful so the starship um project like he had his first orbital flight for starship which is which is important because that's going to be key to getting lots of stuff into space and getting his spaceships back again um i'm not totally surprised that the lockups didn't have much of an effect i know there were a lot of people who were watching this very closely but that was pretty well telegraphed i mean we know we knew that that was coming there's no secret about the fact that there's going to be liquidity overhang so i think that it's i expected well i probably expected it to like you to go up and then to come down below the ipo price what i would say is that there are a lot of people who want this to stay aloft there are like as we've discussed a lot of them in this call a lot of vested interests in wall street want to see this stock continue to go up and elon musk is a great generator of fees for Wall Street. So the analysts have lots of incentives to pump him up. He also has a really

[15:15](https://realalpha.app/v/9SAgLMahZm4/?t=915) big fan base. Look at Tesla. We, I, have been arguing for so long that Tesla's valuation is not based on any real fundamentals. And yet, if I'd bet against Tesla with some periods that are exceptions, I would have lost money. So Musk just has this ability. The problem is also that others now are going to start to think that they have the same magic charms as he does, and they will be wrong. Do you think that the long-term trajectory of this stock will be the same as Tesla? And that is a valuation that just completely exceeds and just lives in a different world from the fundamentals. Like, do you think that the Elon premium is actually going to be enough to keep this company, you know, in the ballpark of $2 trillion? It all depends on a lot of things that I don't feel that I can predict with much certainty, like what's going to happen with the Optimus humanoid robot?

[16:08](https://realalpha.app/v/9SAgLMahZm4/?t=968) Is it going to change the world? Is SpaceX going to merge with Tesla, which is kind of widely expected to be something that's on the menu at some point? You know, how will Starship perform? There are all these things that, like, my expertise is not knowing whether humanoid robots is going to be a $10 trillion market, but if you feel comfortable making that bet, then go for it. And clearly there are enough people who are comfortable making that bet that Tesla is where it is. But this is not a regular company where you're projecting earnings out for a couple of years and putting a multiple on them.

[16:39](https://realalpha.app/v/9SAgLMahZm4/?t=999) This is about faith in Elon Musk. And lots of people have. that. How much faith do you have in the actual, I guess, the price discovery here? Because you mentioned that there are a lot of vested interests, a lot of analysts that are interested in pumping it up. And also the float remains very small. And so I guess part of the thing here is that it's hard to tell how real any of these numbers actually are, and whether they are actually reflective of a market that is even liquid enough to assign a price that is actually commensurate with its value. So I guess, what do you make of the validity of the price, maybe,

[17:22](https://realalpha.app/v/9SAgLMahZm4/?t=1042) **John Foley:** I would say? That is a really good question, because that $2.3 trillion market cap is, of course, not real. There isn't $2.3 trillion worth of stock that's available to buy now on

[17:31](https://realalpha.app/v/9SAgLMahZm4/?t=1051) the market. And if all the stock were unlocked, would it be worth $2.3 trillion? If you tried to buy spacex would you pay to bring your trillion like probably not so that number is not also spacex you know it hasn't been included in some very important indices so it's not like everyone owns it so the actual importance of spacex the amount of spacex that's in your portfolio without you knowing about it is quite small it is important because it like sets a reference point for other kinds of ipo for the big companies asking what they're worth anthropic would be a big one open AI. SpaceX kind of, like NVIDIA before, it sets this idea that a company of a certain level of importance should be worth a certain number of trillion dollars. So that's why I find myself thinking about SpaceX as a $2 trillion company, because we like round numbers and it's big. But as you say, the price discovery is not really there because this is a small free float and it's going to be a small free float for a long time. John Foley is head of the Lex column at the Financial Times. John, appreciate your time. Thanks. After the break, a look at the annoyance economy. And by the way, my newsletter, Simply Put, was just nominated for a Signal Award.

[18:38](https://realalpha.app/v/9SAgLMahZm4/?t=1118) Please go vote for it at vote.signalaward.com. It really means a lot. Type in Simply Put in the search bar. You'll find it. And we'll also leave a link in the description to make it easier. Thank you. Are you tired of hearing about business, tech and finance from the bros? We are the Tiger sisters a top business podcast with the vox media podcast network tiger sisters releases every monday and we decode money power and love and who doesn't love all those things we are the internet's wall street and silicon valley Big Sisters. Tune in every Monday for late night sister talk meets boardroom strategy.

[19:17](https://realalpha.app/v/9SAgLMahZm4/?t=1157) Check out Tiger Sisters on YouTube, Spotify, or wherever you get your podcasts. Welcome to the Tiger Fam. Hey y'all, it's Ested from AmericaActually at Vox.com. We know that you have big questions about the world, and so do we. At Vox, we focus on covering the things that matter in your life and the things that make you curious. Whether that's the big politics questions of the day, climate solutions that might actually work, or hopeful news in cancer or Alzheimer's research. We believe that journalism made for the everyday curious person can make a difference in making a better world. And if that sounds like something you want to be a part of, Vox now has a brand new app, and it brings all of our journalism together in one place.

[20:01](https://realalpha.app/v/9SAgLMahZm4/?t=1201) Go to vox.com slash app to download today. As U.S. Americans voted in the midterm elections, A lot was uncertain, but one thing was clear. This year, many of the attention-grabbing candidates identified or were identified as populists. Gina Hinojosa, a Democrat who ran a surprisingly competitive race for governor in Texas. Populism is about economic power. Sam Forstack, who ran for Congress as a Dem in Montana. I'm running as a populist because populism is what the Democratic Party is supposed to be about. It's what democracy is supposed to be about. Josh Turek, a Democrat up for a congressional seat in Ohio.

[20:39](https://realalpha.app/v/9SAgLMahZm4/?t=1239) We want to once again have a common sense prairie populist that is fighting for the people. And Senate hopeful Dan Osborne, who ran as an independent in Nebraska and didn't use the word populist, but... My life sucks too. Costs are too high. The last 10 years in the U.S. have been defined by Trump-style populism. This year, candidates from across the political spectrum embraced the populist mantle. On Today Explained from Vox, what has 10 years of populism gotten us? And are we ever going back? Today Explained drops every weekday. We're back with Prof G Markets. There's a name for the hidden fees, the subscriptions you can't cancel, and the hours spent on hold with customer service.

[21:23](https://realalpha.app/v/9SAgLMahZm4/?t=1283) It's called the annoyance economy. Americans now spend 60% more time on customer service calls than they did 20 years ago. And the Groundwork Collaborative, a progressive think tank, estimates that the annoyance economy costs American households at least $165 billion. billion per year. Now, states are trying to rein it in. Last week, Governor Newsom signed a bill that requires companies to make a good-faith effort to connect customers with a human within 15 minutes of the request. The question is whether regulation can actually fix this problem that companies are making money from. Joining us to discuss the annoyance economy and what to do about it, we're speaking with Ronan Farrow, Pulitzer-winning journalist at The New Yorker. Ronan, thank you for joining us. So I saw you put out this explainer on the annoyance economy, which I just found fascinating. Let's just start with the basic question.

[22:16](https://realalpha.app/v/9SAgLMahZm4/?t=1336) What actually is the annoyance economy? Well, sadly, most of us probably know it very well, right? This is the labyrinth of phone trees and menu options and hold times. It's basically a deliberate business practice. It costs money to give service. So making it as difficult as possible to receive service, it is a business win for companies. And that's how it's viewed. That's how many of the programs that ensnare us in endless paperwork, endless menus are designed. You know, one of the examples I gave when I did a video about this was Amazon built a deliberately winding cancellation path for Amazon Prime. And they actually gave it the codename Iliad after homer's poem right a 10-year war uh and what they found was cancellations fell 14 after they launched that um and you know economists have looked at this uh there's a stanford economist neil mahoney who examined this and his conclusion was a difficult cancellation process can raise a revenue anywhere from 14 to even up to more than 200 depending on the product uh so we see this from, you know, ticketing companies are a big offender, hotels, airlines, cable and phone companies, banks, health insurance companies. We have probably all been there with these practices

[23:45](https://realalpha.app/v/9SAgLMahZm4/?t=1425) and it is not an accident and it is not just a product of limitations of resources on the company's part. Yeah, that's the aspect of this that I find fascinating because, you know, Obviously, we've all been there. I've been there being held up, trying to make a phone call, trying to get through to customer service, trying to get anything done, interacting with a business. always assume when they're incompetent that it's because they are incompetent. But you're saying, no, actually, it's deliberate. This is a sign of their competence because they're basically saving money because it means that they don't have to actually deal with the costs of actually fixing the problem. Talk a little bit more about how deliberate it is. And then I guess the question after that is like, is that legal? Are companies allowed to make their products intentionally more difficult so that we cannot cancel subscriptions, so that we cannot get proper service that we would hope that we would deserve? That question gets tested in the courts a lot. And I think it's important to note, you know, a lot of these are examples where it's more of an inconvenience in trying to get, you know, a consumer good or something.

[24:57](https://realalpha.app/v/9SAgLMahZm4/?t=1497) But it also crops up in cases with real stakes, even life or death ones. One example of this being tested in court in 2024, the Consumer Financial Protection Bureau sued Comerica. That's the bank that ran the debit card that millions of people used to get Social Security. And the allegation was that its vendors dropped more than 24 million calls before anyone picked up as a matter of policy. And, you know, we see numbers where we just know that insured Americans under 65 are delaying and skipping medical care over paperwork problems and obstacles like this in huge numbers. By some counts, up to a quarter of that population reports, you know, I just didn't get the health care because of the obstacles. There's a lot of research on this. You know, people wind up in bad situations because of these obstacles. And the irony is a lot of these policies, Ed, are in place on the argument that if you make it more difficult, only the people who really need the service are going to prevail in this Byzantine process and actually go for it.

[26:11](https://realalpha.app/v/9SAgLMahZm4/?t=1571) But actually, what the research shows is when you increase the barriers, it screens out a lot of the people who need the service most. And actually, what it selects for is who has the time and money. You know, you see that with Medicaid, for instance, the latest federal budget. it starting January 1st. Low-income adults on Medicaid have to document that they're working and reprove eligibility every six months instead of once a year. Food stamps are now getting the same treatment. So, you know, these are cases where it's not screening out people who don't need it. It's screening out, in a lot of cases, the sickest people.

[26:57](https://realalpha.app/v/9SAgLMahZm4/?t=1617) It's kind of crazy to hear that the government is also engaging in this activity, too. That That is what you're saying, right, that the government is actually making it more complicated so that they presumably don't have to spend as much of the money on Medicaid, on food assistance, on these government programs. Is that the thrust of what you're saying? Yeah. And look, there are critics who allege that is the intention. That certainly seems to be what has come up in the way the legislation landed and in the process that got us to that legislation. And we've seen this happen at a state level before.

[27:32](https://realalpha.app/v/9SAgLMahZm4/?t=1652) Arkansas tried this in 2018. They made the paperwork more onerous, more difficult. 18,000 people lost Medicaid. And most of them were eligible is the key thing. The Congressional Budget Office has projected what does this look like with this new federal regime that's in place? Their projection is 5.3 million more people uninsured by 2034, and about 2.8 million of them are going to lose coverage, again, not because they fail this work test that's been put in place, but because they just couldn't get through the paperwork to prove that eligibility. So Governor Newsom is trying to take this on. He's signed this bill requiring companies to make a good faith effort to connect customers. I don't know what that would actually entail, but what do we know about this regulation? And do you think that it will work?

[28:25](https://realalpha.app/v/9SAgLMahZm4/?t=1705) Well, we're seeing a number of states move in this way, right? California, as you point out, is one of the furthest along in the last few years. They've put in place rules where an advertised price has to include every mandatory fee. Since last year in California, you have to be able to cancel the same way you signed up. So if you click to join, you have to be able to click to leave. New York City also has been on this. A new rule went into effect October 1st. uh there's big fines if a company doesn't allow you to cancel with a click um there's a wave of states basically uh And, you know, I'd also point out in courts, we're seeing more action. Amazon had to settle with the FTC. That was a $2.5 billion settlement. They've already sent back hundreds of millions of dollars to customers. StubHub agreed to pay $10 million in fees this April. New York court shut down SiriusXM having like an overly onerous cancellation maze.

[29:26](https://realalpha.app/v/9SAgLMahZm4/?t=1766) uh so we're seeing action here the tricky thing ed is federally we're not really seeing that um the fdc did try to pass a national click to cancel rule um but it got thrown out in court on a technicality so so far the action here is at the state level yeah i was gonna say isn't this literally the fdc's job is to prevent things like this like the whole point of the fdc is consumer protection, protecting consumers in their interests against the negotiation with a business. There is a trade happening there. The FTC is supposed to protect the consumer. This is right in their wheelhouse, right? Yeah. And it's worth noticing that under this administration, there has been the hollowing out of a number of consumer protection organs across the government. So it remains to be seen whether the federal government will catch up on this one.

[30:21](https://realalpha.app/v/9SAgLMahZm4/?t=1821) It is the FTC's job. And there's a number of other agencies that are supposed to participate in this kind of consumer protection. You know, just the track record of late has not been great on the federal side. People should challenge this. The thing that provides hope in this space is that it is a rare, truly bipartisan issue. You know, everyone agrees that these practices are rapacious and predatory and a problem on both sides of the aisle. So people should talk to their political representatives about that. Also, as you actually engage with these systems, you should fight back. There are all these numbers that are really interesting, some of which I put in this latest video on this subject.

[31:07](https://realalpha.app/v/9SAgLMahZm4/?t=1867) showing how so, so few people actually appeal and persist in the process and push back when they get an outcome they don't like. When you're in that phone tree, when you're in that whole time, when you're in that menu system, when you get the paperwork, people don't keep going. And then when they get dinged, they just leave it be. But the numbers also show that for the small... small subset that engage in the pushback and make themselves annoying, the customer being annoying, they really get results. One of the factoids that stuck with me was a lawsuit against UnitedHealth alleged that 0.2% of patients appealed denied claims, so nothing. But of that 0.2, about 90% of those who did the appealing won. So it pays to be a squeaky wheel and to know These systems are set against us, but you can fight them.

[32:02](https://realalpha.app/v/9SAgLMahZm4/?t=1922) I love this. Be as annoying as possible. Be annoying back to them if they're going to be annoying to us. This is why I'm so annoying. Ronan Farrow is Pulitzer-winning journalist at The New Yorker. Ronan, really appreciate it. Thank you. Always a pleasure. OpenAI just dropped a bomb on the world of math. On Tuesday night, the AI Lab published hundreds of proofs that appeared to solve many of the world's most difficult math problems. These findings spanned problems across numerous fields, including number theory and algebraic geometry, and they even claimed to have reached a breakthrough on the Riemann hypothesis, which is a Millennium Prize problem known as one of the hardest to solve.

[32:44](https://realalpha.app/v/9SAgLMahZm4/?t=1964) According to Alex Kontorovich, a distinguished professor of math at Rutgers University, this would have been an instant Fields Medal winner had it been accomplished by a human. Now, I want to be clear, these proofs have yet to be verified. They were essentially dumped onto the internet via GitHub, so they haven't gone through any of the official processes of substantiation. And that is important because it means they haven't officially been solved. However, based on the reactions that I'm seeing from the mathematics community online, this does appear to be a very major breakthrough. These are problems that went unsolved for decades.

[33:23](https://realalpha.app/v/9SAgLMahZm4/?t=2003) And while some might criticize the way in which OpenAI tackled them, specifically drawing on existing research without proper accreditation or attribution, and also burning billions of dollars in the process, the truth is they are pushing the boundaries of math. and if our goal is to better understand the universe well i don't think there's any denying at this point that the company is helping to achieve that because i do want us to uncover the true nature of reality and i do want to find out what is actually going on out there and a deep understanding of math is a fundamental building block in pursuing that goal.

[34:03](https://realalpha.app/v/9SAgLMahZm4/?t=2043) Having said that, this still doesn't answer the question of whether OpenAI can become a viable business, because what we know about intellectual research is that it is very expensive and usually very unprofitable. And if you don't believe me, just go look at the income statement of any research institution. They are largely funded not with profits, but with donations. That is how the scientific discovery process generally works. So if your position is that OpenAI is pushing the boundaries of research, I 100% agree, and I am excited about it. But if you also believe that that is the path to becoming a profitable business, well, then that is where I run into problems.

[34:47](https://realalpha.app/v/9SAgLMahZm4/?t=2087) Because being good at math does not equate to being a good business. They are two fundamentally different pursuits that are being conflated as one and the same. So we can acknowledge the scientific accomplishments of this company, and we can recognize that they are groundbreaking. At the same time, though, we can also recognize that this is a business model that so far doesn't work. The company is still immensely unprofitable, and it appears that their losses are expanding. And despite that, they are still pursuing a $1.5 trillion valuation, which would be make the company twice as valuable as Walmart while generating less than a tenth of the revenues.

[35:27](https://realalpha.app/v/9SAgLMahZm4/?t=2127) **Ed Elson:** In sum, scientific research and enterprise business are two very different ballgames.

[35:33](https://realalpha.app/v/9SAgLMahZm4/?t=2133) And unless OpenAI wants to burn billions until the end of time, eventually, they're going to have to choose. OK, that's it for today. This episode was produced by Claire Miller and Alison Weiss and engineered by Benjamin Spencer. Our video editor is Brad Williams. Our research team is Dan Chiolan, Kristen O'Donoghue, and Mia Silverio. And our social producer is Jake McPherson. Thank you for listening to Prof G Markets