AI Fact or Fiction: The Fable Ban, Tokenmaxxing, Saaspocolypse — With Ara Kharazian

0:00Big Technology Podcast Host:How much are companies actually spending per employee on AI? Is AI winner-take-all? Is SaaS dead? And is Anthropik screwed after the Fable 5 dust-up with the White House? Let's get to the bottom of these questions, separating fact and fiction with ramp lead economist Ara Kharazian, right after this. Depending on who you ask, between 80 and 95 % of enterprise AI projects fail. To get AI to work for you, you don't need more tokens. You need better people. Aboard pairs powerful proprietary tools with senior engineers who've seen it all. That combination means your project doesn't stall, doesn't drift, and doesn't fall.

0:36Big Technology Podcast Host:It ships. Whether you're a startup that needs to get to market or an enterprise with complex legacy challenges, Aboard delivers exactly what your business needs fast. Aboard is your partner for AI transformation. Visit Aboard.com and let's build something together. Welcome to Big Technology Podcast, a show for cool-headed and nuanced conversation of the tech world and beyond. There are a lot of rumors flying in the AI industry, a lot of narratives flying in the AI industry. And what better way to attack these than to look at the actual data and separate the fact from fiction? Well, we're going to do it today.

1:09Big Technology Podcast Host:We're going to talk, of course, about how much companies are spending per employee on AI and whether token maxing is a thing. But we're also going to get into the news of the week, which is whether the White House's Fable 5 ban, And effectively putting export controls on Anthropics model will have a long-term damage looking at what happened the last time Anthropic had a dust-up with the Department of Defense. We're joined by Eric Karazian. He is the lead economist at RAMP. He's publishing some great stuff at the RAMP Economics Lab. I've been a reader of his work for a long time, and I'm thrilled to welcome him to the show.

1:43Big Technology Podcast Host:Eric, great to see you. Great to talk to you again. Great to be on the show for the first time. Yes, first of many, I'm sure. So let's talk right off the bat about what we can anticipate the impact of the government's Fable 5 export controls on Anthropik's business to be. Right. Because we've seen a version of this before with the Department of Defense naming Anthropik a supply chain risk. This is obviously on a bigger scale. And, you know, even if they go back on the ban, there might be some impact. I keep calling it a ban. Go back on the export controls, which is effectively a ban. There may be some impact here.

2:22Big Technology Podcast Host:So what can we expect? Well, you're right to look at the Department of Defense decision from earlier this spring as the closest recent example that we might use to inform how it's going to affect anthropics business or business adoption going forward. I mean, let's go back to the spring when the Department of Defense labeled anthropic as supply chain risk. Usually in most industries, for most software vendors, when you are labeled as such, businesses are not likely to continue to use that vendor going forward. So if you're just going to think about this from first principles, we would have expected businesses to shut off their Anthropic subscriptions for new businesses to not want to sign up for businesses to explicitly not want to use Anthropic going forward, both because of a true security concern that the government is citing and because they might want to engage with the government on government contracts.

3:14That's not what happened earlier this spring. If anything, this spring was when we saw Anthropics adoption really accelerate with businesses. It was coming off the heels of the successful launches of Cloud Code last year, finally starting to move into a more popular posture with non-technical users. And just this past month, we showed that Anthropics is now the most popular AI model used by US businesses, according to RAMP data. So several of those assumptions didn't come to be. I think two main reasons why. One is that most businesses didn't really seem to take the Department of Defense's label very seriously.

3:54You know, it was kind of like, okay, yeah, the Department of Defense is saying this, but this was one of the best models available, still is the best model available. It's popular with businesses. The second reason is that the Department of Defense lost a lot of credibility over the continuing weeks when it acknowledged that it would be issuing exceptions both internally and externally for businesses that wanted to use Anthropic anyway. So if you're going to go back to the example from earlier this spring, if anything, it's probably the case that the Department of Defense's supply chain risk labeling accelerated Anthropic's adoption with businesses.

4:34and this label now anything puts it on a very interesting competitive standpoint with open AI who has the better model probably the one that the that the federal government has suggested is so powerful it must be controlled and and there's brand strength in that

4:54Big Technology Podcast Host:that's right so we talk on this show all the time about whether anthropics positioning around safety and the fact that it's, you know, taken this steadfast approach when the federal government has asked it to do something. And that's led to some of these actions by the federal government, whether that's marketing. Let's put the intent aside, though. It could end up being a boost for its business, you know, whether it wanted this to happen or not. I keep thinking about the fact that at least we're recording Monday and I keep thinking about the fact that um you know when you're in clod it says hey fable is not available right now right and you're seeing all these posts on x about people who are like oh those five minutes with fable really give me a glimpse into what this could be um and you'd imagine that when it gets turned back on that forbidden fruit effect just uh kind of drives the interest in fable you the model too powerful for the government to let you use, the interest will probably just go through the roof.

6:00Well, let's also note that part of Anthropics, the product experience of using Anthropic models, at least over the last six months, has been getting really comfortable with models being down all the time. Because Anthropic has so many compute constraints that it is very common to be a user of Anthropics models and to be hit with a crash notice or something that just says, hey, the models are down right now or try again later on API error. And yet we would normally expect that that would drive a lot of users over to OpenAI models. OpenAI models are, OpenAI doesn't have nearly as much of the compute issues that Anthropic does.

6:43And yet we haven't seen that kind of switching behavior yet over the past couple of months. So maybe some marginal switching happening from people who would you who are using cloud code now switching over to open ai's codex uh but if anything you know it makes me wonder both as a researcher and user of these models how long is this forbidden fruit effect of anthropics models going to last before it starts to turn users off who really just need to get to work and use the models that they are paying for and have them work as expected.

7:18Big Technology Podcast Host:Yeah, I mean, you're the economist. I don't know, is there an economic theory that explains why people would stick with a vendor that has constant interruptions, even if there's another one with just as good or on par capabilities that doesn't have the turbulence? Well, what we found is that these models are a little bit stickier than we thought they would be. And we talk about them being these commodities that, oh, you can just switch between one model and the other. And maybe at the model level, you really can think of things like that. But in terms of how many employees at firms are actually using AI, you know, better models don't necessarily create the switching for employees and users.

7:58It's about the product experience around those models. You know, Claude Code was so successful, not because it was powered by the models of Claude, but because it was integrated into your workflows in a very effective and agentic way, that it was the first model and first experience that allowed an engineer to execute on multi-step tasks without babying a chatbot the entire time. And so such that incremental improvements of the model are important and helpful, but they weren't the whole story for the actual growth of Cloud Code. And so you can imagine that to be driving some of the stickiness between Claude Code and Codex, OpenAI, and Anthropic is that people get really used to the tools, the software that they're in.

8:56They like the experience that one provides over the other. I also do think there is going to be a sort of branding effect where Anthropik's AI safety posturing, you know, love it or hate it, there are people who really do like that they're the company that at least postures itself as being thoughtful about the effects of AI, whether or not they are the right guardians of that.

9:23Big Technology Podcast Host:Yeah, it's possible that the federal government is sort of, you know, by disrupting Anthropik, maybe giving it a helping hand by sort of making its safety messaging seem more legitimate and again, giving it that forbidden fruit effect. All right. One more economics style thought question for you, then we can get into your data. You know, there have been recent reports that OpenAI is looking to drop prices drastically. And we talked about it on the Friday show. My thought was basically, they are looking at lifetime value of potential customers. And if they were to drop their prices, they could get people used to using, let's say, a codex and then keep them with the stickiness that you're talking about.

10:08Big Technology Podcast Host:And so therefore, it would be a good move on their end to say, we're going to drop prices to win people over, and then hopefully they'll stick with us. Your thoughts? Part of it, I think, is very natural. This is an extremely competitive market where it's very common to see in a matter of months, a newcomer take the lead over a relatively popular provider. We've seen this in software before, but it's especially true and acute in AI. I mean, we saw this with Cursor versus GitHub Copilot. When coding agents came out, at least when AI Code Autocomplete came out two, three, four years ago, GitHub Copilot was the enterprise tool.

10:54That's what everybody used. Unsurprisingly, it was also backed by Microsoft. Cursor comes out, and within about a year, it has the majority of the market. Then, by the way, CloudCode comes out, and then now CloudCode has majority of the market. And then so we saw a similar story with Open Ananthropic. You know, in RAMP's data, we have been tracking, using our flagship research, RAMP AI Index, the share of firms in the United States that are using AI, at least paying for it. if there were subscriptions or tokens directly. And then we break it out by which model they're paying for. And for most of AI's commercial existence, 2023 onwards, OpenAI was clearly the dominant player, somewhere hovering between 20 to 30, 40 % of businesses in the US were actively paying for OpenAI.

11:47And it wasn't really budging much. It was just a very gradual increase, particularly 2023, 2024 onwards. No one was really thinking about Anthropic, which was popular with technical users, but otherwise wasn't this sort of broadly understood competitor in the market. Second half of 2025, we see month over month percentage point increases in the share of firms that are using Anthropics models. Coming to the forefront last month when Anthropic overtook OpenAI in actual business adoption. So now Anthropic sits at about 41 % of US firms are using Anthropic, 39.5 % of firms are using OpenAI. Anthropic is still growing.

12:28OpenAI is relatively flat. And then even in the sectors that are early adopters of AI, we're seeing that growth continue to grow and accelerate while OpenAI holds relatively flat. Right. So it's an extremely dynamic market where you could expect all of the involved players to want to compete with each other. but i actually think that you know right now the focus is on open-end anthropic but there are a lot of players that are underrated google i think is extremely underrated and i think might end up being one of the big winners here that no one's talking about okay but then briefly the price the price

13:02Big Technology Podcast Host:war or the the price undercuts do you think that's enough to dislodge the stickiness i know you don't have the data on it but there has to be some formula out there about you know the price plays into usage well so i do think it's going to be enough to i think look we're going to come to a head at some point where we know businesses keep demanding some better control over ai spend you know we went through a sort of token maxing era where everyone was talking about okay we need to spend as much as possible on tokens and then now we're in this sort of new era where businesses are saying, hey, we need to actually rein in token spend or at least understand where it should be.

13:49It can continue to rise, but we at least need some control over what it is. Neither OpenAI nor Anthropic have built products that allow firms to actually manage their token spend, nor have they built products that incentivize firms to keep those costs in control. If anything, OpenAI and Anthropic are incentivized to have firms spend as much as possible. so so far you know they can compete on price reductions but really what firms are asking for is some degree of control you know maybe that means hey build us something that allows us to smart route tasks over to the most performant but also most efficient model for that task other competitors are offering that it's not usually open ion anthropic though uh so your

14:36Big Technology Podcast Host:data shows that, again, like you said, Anthropic has overtaken opening eye with business spend. Very briefly, Ara, the criticism of the Ramp Economics Lab, whether well-placed or not, has been that, yeah, you're looking at companies with Ramp cards using Ramp, which tends to lean towards startups and tech-forward companies, and it's not representative of the full economy. Your thoughts? The way I normally think about this is that, so we have actually a pretty good distribution in our data set across sectors. However, no matter the sector, the businesses on our platform are inherently more tech forward in that they're using something like Ramp to manage their spend in general.

15:25I do see that as a strength for a couple of reasons. One, AI is this very new and nation technology and one that is not effectively tracked by other data sets. It's not effectively tracked by government data sets either, which have their own set of criticisms as far as how they are surveying firms about AI spend and also the firms they are surveying themselves being a little bit self-selected. uh ai spend if anything is skewed over toward these tech forward businesses such that if you do want to understand how businesses are spending and operating on ai it actually behooves you to look at these very forward thinking businesses that have been leading this charge they are more likely to be early adopters of this technology and whether or not they are representative of the average firm in the united states we know they are not it's more likely than not that the average firm will look more like these firms in a couple years than they look like the average firm today.

16:20So I think if you want to be forward-looking, you want to look into the future a little bit, you probably want to use this kind of data set. For what's worth, I actually think that in many ways we underestimate adoption.

16:31Big Technology Podcast Host:Okay. All right. We'll get into more methodology later. But given that these are the forward-looking companies, let's get into some more of your data because I think that there have been some narratives about token waste that your data has a little bit of a different perspective on that I think we should just discuss. So you've talked about what is the cost of being an AI-pilled company. It's$7 ,449 per employee per month. So you said the top 1 % of firms spend that much on employee per month, and the top 10 % spend$611 per employee per month. And the median firm spends just$11, like the cost of an enterprise seat on enterprise chat GPT or a cloud subscription.

17:16Big Technology Podcast Host:Now,$7 ,400 a month is pretty high on a technology, like for a single person to spend that much on a technology seat or license is somewhat unheard of. However, when you think about the headlines that we've been seeing a company left clawed on and spent a half billion dollars in a month, your data actually presents somewhat of a different picture that companies aren't sort of spending unrestrained right now. They seem to be, you know, sort of dipping their toe in the water as opposed to going all the way in and, you know, spending tokens like they're going out of business. Well, look, AI spend is the fastest growing.

18:00spend category we've ever observed in our data set. Probably one of the fastest growing spend categories for businesses ever, depending on how far back you go into what a business is defined as in prehistoric times.

18:14Big Technology Podcast Host:I couldn't imagine any spend ramping faster than this. Yeah. What could it be? Exactly. And so since January 2025 through May 2026, So last month per business spend on AI tokens is up 15x. And that's amongst firms that were already spending on AI. At the same time, AI spend itself isn't really that meaningfully large for most businesses. So it's grown a lot. But for the top quartile of firms that are spending on AI, top 25%, it's only about 2 % of business spend excluding payroll. You'd be at 1 % if you were to include payroll. So it's grown a lot. That's why we get all these concerns from company executives about how do I manage this growth.

19:04But as far as its actual level, it's relatively small. So you'll notice when people talk about firms pulling back on AI spend, they might be pulling back on AI spend in some parts of the firm. They might be more mindful about which models they're using or making sure that teams don't have uncontrolled budgets. But if you actually look at firms' spend on AI in the last couple months, just last month, it still increased 14 % month over month. So there's clear evidence on our platform, too, that firms are making more cost-disciplined decisions. You know, we've seen an increase in AI spend being routed away from OpenAnthropic and over to these open source platforms.

19:52Last month, DeepSeq was one of the fastest-growing vendors on-ramp. And yet it's still a very small share of AI spend that is actually going through those rails. It's a relatively small share of businesses that are using open source platforms in general. And the vast majority of spend happening is still rising. So those cost discipline measures are important, but they're really just occurring on the margin. And they're not happening fast enough to pull back the rising slope of AI spend. Yeah.

20:25Big Technology Podcast Host:Do you think that there's going to be a moment where some firms start to spend more on AI than they do spend, you know, on they spend more on AI per employee than they spend on employee? For instance, I don't know how accurate this was, but I think like we're actually trending in this direction where someone figured out how long, how much it would cost to run the Fable or the Mythos API. And they found out it was something like$600 an hour, where if you like multiply that over a year, it's 1.2 million. uh so so how so what do you think when you think about the trajectory do you think we're going to get to a place where people end up spending more uh on ai per employee than they spend on uh in place themselves i imagine some will yeah some for some firms i imagine it makes a lot of sense right but for the typical firm there's really it's really hard to benchmark where you should be.

21:21And so the top 1 % of firms spend$7 ,500 per person per month on AI. But that's the top 1%. So you can imagine that's a pretty tech heavy group that actually may include a lot of firms that are ultimately using AI, not just for the employee's usage, but also for the underlying infrastructure of the firm. Maybe they've built a bunch of internal tools, right? So it all gets balanced out.

21:44Big Technology Podcast Host:And software engineers are double that typically 15 ,000 or close to 16 ,000 a month. well again would depend on the firm because you look at the typical firm on our platform and again this is ramp so relatively tech forward platform right and the median firm is only spending about 11 per employee per month right so you know that's bare that's a chat subscription that's like one of the low level open air and anthropic subscriptions maybe a little bit more on the margin. So, you know, it's another reminder of sort of how early we are, right? In that the vast majority of firms, and this has transformed our research approach, because for a long time, the last year and a half, my research has focused on trying to estimate the economic impact of AI.

22:32But if there's no way to find that in productivity statistics, if people are still not sure what the ultimate gains of AI is going to be, then really the only way to start is, hey, how many firms are using AI? And so our original version of Ramp AI Index is just that. It's, hey, what is a share of firms in the US that are even buying it and buying it month over month to try to get some way of at least approaching the question, hey, is this valuable? Now, more than 50 % of firms are using AI. 54 % of firms are using AI in some way or at least paying for it. So our question gets to transform a little bit, not just who's using AI as a valuable, but how are they using it?

23:12How much are they spending on it? What does it mean to be an effective user of these models and to deploy it effectively through your organization? Because that's what's really interesting about AI too, is that it is unevenly distributed and that certain sectors are more likely to adopt it than others. The products themselves are also unevenly designed and distributed and that some of the best, most advanced usages of ai are uh designed for certain job categories coding agents it's like the most obvious commercially advanced way that you could use ai to be productivity enhancing and yet that doesn't exist for most other firms you know maybe there's productivity enhancements you can find in a lot of other jobs but the products themselves are not well developed to make that clear so like the average user so AI is unevenly distributed and so it's ultimately going to be difficult as a firm to identify and benchmark against what good usage of AI means especially because the effects of AI, the productivity gains of AI are not likely to show up in your first couple months there's clearly a learning curve to implementing AI throughout your organization and even through your own personal workflows and employee.

24:29And then beyond the learning curve, there's also this sort of minimum threshold of adoption where I don't think anyone really expects you to get massive economic gains from everyone having a chatbot. But the idea of everyone having their own cloud code for their job is much more compelling. But you wouldn't know that if you're just using a chatbot for like a month or two and then you write off AI because it's like, you know, what's the point of this?

24:53Big Technology Podcast Host:But you look at the curves that you have in your research, right? And it's just a number, just like three hockey sticks, right? If you look at the spend per employee per month of the top 1 % of companies using AI, the top 10 % and the median company using AI, it's like legitimately like a lightly sloping line. And then all of a sudden it shoots up in all three of those categories. so you're trying to you said you're trying to get to the answer of you know how is is this technology valuable and so i'd love to hear your perspective on what these numbers mean even though it's more i guess quantitative or qualitative than quantitative right do the fact that we're seeing these spend increases mean that companies are seeing an roi on AI or is it still potentially in the sort of FOMO stage?

25:45Well, I really am of the school of thought that businesses have no reason to be spending this much money just out of a sense of obligation in FOMO. You know, I get it if we're talking about people buying stocks, right? But like companies making fairly large investments in software, you know, that you can just talk about it externally without making not only the large investments in software, but month over month increases in how much they are spending. So I'm generally of the school of thought that if firms are doing this, they must be finding some value out of it. But there are some places that we look quantitatively for that evidence and also informs our thinking that this is different from most software markets.

26:28So one is that the most advanced spenders on AI don't lock in with one vendor. So this is fundamentally different from how we typically think of software, where it's like, if you're using a CRM, you're going to use one CRM. Maybe you'll experiment with a couple providers, but ultimately you're going to sign with one. That's not the case with AI. The top 1 % of spenders on AI use eight vendors on average, whereas the median maybe uses two. And that's vendors fairly narrowly defined as LLM providers and maybe some AI infrastructure companies. Now, it's also not just experimentation. You know, there's some amount of it that's always like this sort of continuing experimentation where it's like, OK, the AI models come out, but they also change so much so frequently that if you are an organization that is using and implementing AI effectively, you probably want to have paid access to all the major model companies so that you can switch to the most effective model for whatever task makes sense.

27:31Or when a new model comes out, you can see if it makes sense for this workflow or that workflow. that's what being a good ai user often means to these firms in the top one percent but that is an unfamiliar idea for many businesses and and frankly business people at firms who are buying ai in charge of procurement i often get the question when we report open ai versus anthropic adoption rates should i buy open ai or anthropic and then when you which you know if you're someone who uses these day-to-day these models day-to-day it's like a very surprising question because you would just say, well, you should try both.