# Anthropic’s commitments raise costs while price cuts test its moat

Markdown copy of https://realalpha.app/s/anthropics-commitments-raise-costs-while-price-cuts-test-its/ on Real Alpha, the best minutes from long interviews with the people building AI. Cite and link the page, not this copy.

- Speaker: [Steve Eisman](https://realalpha.app/people/steve-eisman/) · Investor
- Host: [Ed Elson](https://realalpha.app/people/ed-elson/)
- Show: Prof G Markets
- Episode: [Steve Eisman: Here's What Anthropic Isn't Telling You](https://realalpha.app/v/8uoSVG3P6z0/), Oct 28, 2026
- Moment: 5:42 to 10:27 of the episode (5 min), playing on https://realalpha.app/s/anthropics-commitments-raise-costs-while-price-cuts-test-its/; the episode from there: https://realalpha.app/v/8uoSVG3P6z0/?t=342

## Quote

> “when you commit to $800 billion, **your expenses are going up**.”
> — [Steve Eisman](https://realalpha.app/people/steve-eisman/) at [6:41](https://realalpha.app/v/8uoSVG3P6z0/?t=401) · Prof G Markets

## Takeaways

1. Eisman cites $800–900 billion in future capital commitments.
2. Future infrastructure costs may hit earnings through depreciation or other expenses.
3. A 50% price cut makes market-share gains and revenue crucial to assess.

## About this segment

Steve Eisman argues the S-1 is backward-looking and that huge future compute commitments will raise expenses. He says reported price cuts could reveal whether Anthropic can gain share without a durable moat.

From “Steve Eisman: Here's What Anthropic Isn't Telling You”, Oct 28, 2026. [Watch the full episode](https://realalpha.app/v/8uoSVG3P6z0/)

## Transcript

Transcribed by machine (Whisper) from the episode’s audio, so a word may be off here and there.

[5:54](https://realalpha.app/v/8uoSVG3P6z0/?t=354) I guess this gets to sort of a larger question about this filing, which is like, how much does it actually tell us is really the question that I'm trying to get at. Because you point out the future compute commitments, which for me, that was the most important thing that I read. I'm like, okay, they have narrowed their losses. They're getting to a better place. That's good. Oh, but hold on, $800 billion coming down the line.

[6:20](https://realalpha.app/v/8uoSVG3P6z0/?t=380) And in addition to those $800 billion in compute costs that we're going to see at some point, they also say that they're going to significantly increase capital expenditures, which to me tells me they're going to just build data centers, which seems like a massive pivot, a massive change to the business. I don't know if they're going to build data centers, but they're certainly going to be spending a lot more money either.

[6:43](https://realalpha.app/v/8uoSVG3P6z0/?t=403) Maybe they'll rent, you know, and, and, and that'll cost them a ton. I don't know, but it, but clearly I agree with you for what you're saying that, that put it this way, when you commit to $800 billion, your expenses are going up. How much they're going up? We don't know because I'm not the CFO of the company. I don't know how this is all going to play out. I read one report from someone who argued that by next year, they'll be spending $80 to $100 billion in CapEx.

[7:16](https://realalpha.app/v/8uoSVG3P6z0/?t=436) So, you know, but I don't even know what that, is that expensed? Or does that go on the balance sheet? You know, it's unclear. But it seems as though eventually that's going to have to show up in the income statement in some way, whether it's through depreciation or whatever it is, it's going to show up. And it's not showing up right now. The way it shows up is at the very bottom of the document as basically like a footnote.

[7:41](https://realalpha.app/v/8uoSVG3P6z0/?t=461) So that's going to be a massive change to the business. And then, as you say, it does seem that the current revenue is a pretty important or current Q3 revenue is a pretty important item that tells you something about the future of this business over the next, you know, near future, over the next 12 months. What I can't tell is, did this S1, did it really tell us much about this company? I think it's very, very backward-looking, because it's giving you a...

[8:12](https://realalpha.app/v/8uoSVG3P6z0/?t=492) Snapshot of the company at the height of token maxing, and before Chinese open-weight models really started to take market share. I found a report that said that Anthropoc cut prices by 50% to gain back market share. I don't know if that's true. Let's assume it is. that's an incredibly important data point before you go public that you just cut prices by it you know this is not uh it's 47th street we know we cut them we cut prices in half you know and we're bragging you know cutting prices in half could be dangerous you know the question then would be okay you cut prices in half how much more market share did you get did your revenue go down did it go did you take so much market share that your revenue went up? I don't know. But, you know, that would seem to me to be the most important question to be

[9:09](https://realalpha.app/v/8uoSVG3P6z0/?t=549) asked. I mean, the question for investors, the ultimate question is, do you invest in this thing? Oh, I have an answer for that. That I can answer you. Tell me. No. You kind of got to the point quickly. I was going to ask you, would you need to know the answers to the questions that you just brought up before investing i assume the answer is yes you would well yes but i think look to me the bigger the the bigger bigger bigger question is it just seems you know that there are no moats around this business i mean if it's the case that they actually did cut prices by 50 that would reinforce the concept that there are no moats and that the only way you're going to gain shares price being in a business where you are committed to spend 800 billion dollars where your entire franchise is just simply based on price is dangerous there is another uh interesting

[10:10](https://realalpha.app/v/8uoSVG3P6z0/?t=610) angle uh on this s1 which was the governance structure oh please those is ridiculous