GAEA Talks Live from HumanX - How AI Broke Software Pricing with Metronome CTO Cosmo Wolfe

追加: 2026-05-13

[00:00:00]The idea that you can use an LLM to answer support tickets is interesting and exciting, [music] but the idea that then you can build a business around paying for successful resolutions, that is a differentiator and that's let some of the AI, you know, support agents I think really take off.

[00:00:16]Anthropic is a customer of ours for example. Anthropic is powering, you know, if you log into Anthropic, you can see your usage in a close to real time.

[00:00:22]You can see it broken [music] down by cloud code versus, you know, the cloud web app. You can see it broken down by team if you're an enterprise. And that [music] kind of like in-product experience, as well as being able to set up controls on that, that is something that is all kind of powered by [music] that real-time data from Metronome.

[00:00:38]Billing used to be kind of downstream of your product, right? You would use a product and once a month >> [music] >> a back office team at the company would send your back office team an invoice and no one would really kind of like interact with that.

[00:00:56]AI N Talks live from San Francisco at Human N Talks.

[00:01:00]Cosmo Wolf from Metronome, good having you.

[00:01:04]Thanks for Thanks for having me.

[00:01:06]Um So give me a little bit of background to your yourself, what you do at at Metronome and where it all came from.

[00:01:14]So, I am the CTO at Metronome. Metronome is a company we've been around for about 5 6 years at this point. So, we've, you know, kind of seen the AI industry kind of come alive in the last couple of years.

[00:01:29]We actually just recently got uh completed an acquisition with Stripe.

[00:01:34]So, we are now a Stripe product in the in the Stripe product uh family. Uh that completed about a month ago at this point. So, we're still very very new and I have been at Metronome since very early and Metronome is really focused on monetization and how companies, you know, sell, price, bill, and really monetize their products, which is something that has become a lot more critical and I think how companies go to market in the last 5-ish years, around the time that Metronome has been been around. We, you know, we got founded with this belief that how companies take their products to market and monetize their products is in some ways just as critical as the product itself.

[00:02:16]Um, which 5 years ago was a little bit more of a big bet. I think now with AI and just really how the market's been evolving in the last 5 years, it's becoming very obvious that how companies kind of choose to monetize their products is very, very critical. And so, that's been the kind of journey that Metronome has been on. It's been very fun to see and evolve with the market.

[00:02:37]Excellent. What we've seen very recently as well, well, the last couple of years, 3 years, this whole AI revolution really kick off.

[00:02:47]And the technology industry, since that we've both been in tech for a long time, it's like the rules have changed. That SaaS used to be very, very high margin.

[00:02:59]Now I mean, it's crazy to think that the cost of hardware, the cost of cloud compute, the cost of delivering technology in many instances is negative that negative value, where to to sell a to sell to any consumer uh, could be in, you know, actually revenue loss. So, this is this is a serious problem to solve. What what's your experience of the last 12, 18 months? Yeah, I mean, I think, like you said, if you rewind 10, 15 years, you could have a billion-dollar revenue company where you've never really thought that deeply about how you monetize the product, right? You you have a per-seat price, maybe you're, you know, you're selling kind of access. We would call the access era. Even before that, you you you have the, you know, you're just selling licenses or kind of almost selling like physical goods and some of you know, like like like CD-ROMs of of software or whatnot. But, we've moved now into the you know, you can no longer leave it as an afterthought. You have to be very upfront with how you are pricing your products so you don't end up upside down on margin like you're saying like the you know, marginal cost of adding a new user to a SaaS product is basically zero. You can have which will allow SaaS to be a very very high high margin business. And now, not only is there more pressure from consumers or buyers on justifying the value that they're paying, but there's more pressure in making sure the seller the the product is you know, pricing in a way that they can make money as users increase. And that's I think existed even before the AI explosion. If you kind of think back to 2018-2020 era, ZIRP was ending, interest rates were increasing, so that the like money in venture was was drying up. Uh so, head counts were flatlining or decreasing. There was a lot of tech contractions. Um And meanwhile, the usage of products were probably still increasing in many cases.

[00:05:07]As you saw these companies that were seat-based have top-line revenue flatline or decrease as their seat like counts decreased, but their usage or their you know, the the use of their products continued to go up even as their revenue flatlined. And so, that I think was the first kind of trigger that brought this. Then, um there had been this huge amount of relatively free venture money. And so, you had this huge Cambrian explosion of different startups that there was a lot of startups in the space, so a lot of competition.

[00:05:37]Then, as the money again started to dry up, you get more scrutiny on what your your set buying as an enterprise buyer. And all all things pre-AI came together to bring a lot of scrutiny on again like how companies are pricing, how they're aligning their price to value so that they can continue to you know, increase revenue even as again seats seats flatline. And that's all two, three years before the AI explosion, right? And then AI came out and like you said is this very, very strong margin pressure. You're you're basically trying to eke out as much margin as you can on top of GPU costs and it's just the the like fourth or fifth or 10th straw that broke the camel's back and now every single company is kind of forced to iterate on how they monetize and not just on like the product quality itself. Yeah. And I'm saying it from because we have a Turing Elite research labs where we're developing capabilities.

[00:06:34]And one of the primary focuses is on how to create advanced capabilities on ideally edge, local, but highly distilled uh models and leveraging as as minimal compute as possible.

[00:06:51]When in all the discussions we've had here, people are looking more to efficiency and getting rid of all the noise.

[00:06:59]One of the primary reasons is it's unsustainable to to lose $7 for every $1 made. It's okay when you have massive capital behind you and you're going for market share in a particular sector, but we're going to see a complete change and turnaround and more so back to traditional economics and what's the baseline?

[00:07:23]When when when you have a pilot, then you scale up. Everything can work fine on a pilot, but when you when you distribute across, you know, hundreds of thousands or millions of users, you you could break a business if you haven't really understood the pricing and you're keeping tabs on it in in a heartbeat. Yeah. I I definitely think, you know, the push to do more local compute or edge compute or really just any sort of initiative to, you know, lower your cost of goods served is going to become more and more critical, but I think it's one half of the coin and we see companies that, you know, you can also solve that with, for example, outcome-based pricing is very popular right now, where you are really closely aligning the value your users are getting with the cost that they're paying, which allows you oftentimes to charge more.

[00:08:14]Like I think a lot of times people will kind of not The The simplistic model is like, you know, the less you charge the more people are going to want your product or something along those lines, but I think what's really critical is giving people predictableness and visibility and understanding that like if if I have a outcome-based business model and I, you know, uh I pay $10,000 every time a AI brings me $100,000 in business, that's very easy to justify and that is a way to like increase in some ways your your what you're charging your users or at the very least, you don't have to fight to like move everything to the edge to do that because you've brought the you've, you know, you've worked on the other side of the equation, basically. I think the real thing that's going to happen is that across the board companies are working to lower their costs, moving to edge, more efficient models, that kind of thing, or change how they're pricing and monetizing so that they can justify their their the cost of their goods more to their buyers. And and I think you sort of are going to have to do both to be successful in this next era.

[00:09:16]Cuz in many ways, sometimes charging more increases sales. Mhm. Because if it if it is valued, then people will pay for value. If if you're just trying to get get a race to the bottom then that's a lot more complex.

[00:09:31]>> Yeah, I think charging more or just really making sure that your users are understanding the value they're getting from your product and it feels valuable to them, right?

[00:09:43]They're they're able to quantify like I'm using this product to become X% more efficient or close Y% more deals or you know, um the a place we're seeing for example outcome based pricing a lot in right now is the support agent uh you know, uh field, right? Where you you're saying like yeah, there's a price of closing a ticket and as long as the AI is less than that, it's like positive for my business, basically. Um and so I think bringing that visibility and uh alignment with the buyer into how you monetize is allows companies to, you know, be maybe charge more, but at least justify with whatever they're charging their customers more. And how does Metronome roll out?

[00:10:30]>> So Metronome, if you zoom like at a 20,000 ft, Metronome is effectively a black box where companies are sending us things that happen in their business that they may want to monetize. So, maybe every inference or um maybe like users logging in or, you know, taking actions in your product. A lot of times you're you there are things that you may want to monetize on that you don't know yet, right? Like you might be sending some events for when bytes stored even though we don't charge for storage yet cuz you kind of like want to analyze, see the cost profile, see how your users are using your product before then um well, charging for it or monetizing it.

[00:11:08]So, you're sending those events to Metronome and as close to real time we're trying to send you back revenue data. So, we're you know, pricing those events, we're assigning them, you know, we support kind of complex uh organizational hierarchies or different kind of rules on how you want to effectively compute your revenue and then you're getting you're getting those events streamed back to you with, you know, dollar signs or whatever on them to to to to uh to kind of abuse the metaphor a little bit. And so then businesses are building kind of product experiences around that data because one thing that's really kind of changing I think to this point of visibility is billing used to be kind of downstream of your product, right? You would use a product and once a month a back office team at the company would send your back office team an invoice and no one would really kind of like interact with that. But now if you're trying to bring that visibility, that understanding of how your product is driving value and also trying to give people predictable uh spend or at least have them not be surprised at the end of the month, you really kind of have to bring that product experience and billing experience together. And so with that data Metronome streaming back people like Anthropic is a customer of ours for example. Anthropic is powering, you know, if you log into Anthropic, you can see your usage in a close to real time.

[00:12:19]You can see it broken down by Claude code versus, you know, the Claude web app. You can see it broken down by team if you're an enterprise. And that kind of like in product experience as well as being able to set up controls on that, that is something that is all kind of powered by that real-time data from Metronome. And so then once that's kind of set up you can iterate and launch new pricing very easily. So if you, you know, I mentioned storage earlier, maybe you're saying like, "Oh, I want to try to sign a new enterprise contract where we are going to charge for storage." Historically that would be a engineering team, you'd have to make a ticket, they'd be in the backlog.

[00:12:56]Meanwhile you're losing the opportunity to do a close a deal, right? Or you're signing a deal that you then can't accurately bill for, both of which are bad. And so now it's like if you have a deal, you just log into Metronome or, you know, your your non-technical team, whoever the driver is, can set up whatever pricing they want to on top of that data. And so it kind of serves as the single pane of glass that brings together everyone who cares about monetization. So, you know, revenue and and and your kind of like go-to-market teams, the product teams that are actually building the products, the internal finance teams that are kind of like, you know, trying to bring some predictableness and control, all these teams can use Metronome and Metronome's data to support whatever business models your company needs to do to survive in this market.

[00:13:37]And do you think this is going to become a core part of startups and in tech businesses in general over the next year? Yeah, we we see, you know, we we we try to keep our finger very much on the pulse of of companies companies that are starting up today like NYC today often times are doing eight like like 10 plus different or I I just maybe like we'd say you know, five to 10 different pricing model launches before they kind of settle on one that then grows them from, you know, to the, you know, next stage of revenue. So, you know, the first year or two of a company's lifespan they're often iterating tons, you know, tons more than you'd maybe expect. Often times you don't see it, right? It's like the maybe the webpage doesn't always change, but when you're an early stage company, you're trying to sign new deals, you're trying to, you know, you every every business you talk to every customer you talk to you're trying something new out. And so behind the scenes I think there's a lot more change here. And I I think it becomes just as critical this is something that I really believe is it's just as critical as the product. Like in many cases the how you monetize your product is the first experience people have, right? They go to the pricing page, they, you know, maybe a free trial is something that's a monetization uh approach. You know, you're you have to sign the enterprise contract with whatever new terms before anyone's using your product. And so I think iterating on it on your monetization is table stakes. Um, much like I think in the last, you know, let's say 20 years in in in the software industry, companies kind of learned that continuous deployment and feature flagging and, you know, AB testing and kind of iterating on your product experience was needed to be competitive as opposed to doing, you know like a quarterly release train or something 20 years ago. And so I think the industry has learned the product has to iterate very quickly and they are we are now learning as the market matures that the monetization also has to iterate quickly. So yes, I think it's a critical part. And and we are saying it with the literal rate of development with capabilities of AI what used to be a kind of a product roadmap it it's so fluid now. Mhm. Dropping new capabilities in, not having to go resign deals, not kind of saying well, you you can you have to pay for this upgrade and it's coming in 6 months.

[00:16:02]This idea that you almost have a relationship with a customer with flexible components and you're constantly just dropping new features and functionalities in there which yeah, changes the the dynamics of the relationship completely. I mean we we even see with Claude with Anthro Anthropic I I see releases every day.

[00:16:26]And it's going to reinstall, restart, restart. It's a remarkable rate of knots. Yeah. Yeah, I think every single like the the clock rate or the tick rate inside of these companies outside of every company in some way is increasing and I think you know, some companies are moving faster than others but every single company is trying to move faster and there's so many more tools to like you know, the AI tools and I think being able to move faster on the finance, the revenue side is just as critical but the tick rate there is even slower, right? Before maybe we were doing a software release every month or every week, a pricing release every 6 years, you know? And so now it's like can we what does it take to you know, Claude's releasing a new feature every day like you said they're changing their pricing fairly often, too. I think OpenAI we saw did something like 18 new pricing launches in the last year. And so, that might be stuff like adding support to doing the, you know, upgrades as you, you know, you run out of use of codecs or Opus, the high-end models, you can, you know, pay pay-as-you-go upgrades. That might be lowering costs of more commodity models like the Haikus. All of those are like launch moments in their own and and is happening at an incredible rate, as well. Like, just basically the whole speed of iteration in companies across the board, I think is ticking up. And it puts a lot of strain on the teams that historically have had to maintain these systems both on the software product side and also on the monetization side.

[00:18:00]So, um we really want to be able to help increase that tick rate, I'd say. And how have you found, um the the financial side or the financial driver, such as VCs, uh investment into tech companies, their narrative or their understanding change with with regards to resetting expectations of what kind of financial tracking, revenue tracking, and influence that that they now expect you should have? Or is this still something we're waiting to to really come into play?

[00:18:38]It's definitely evolved, but I think maybe not in the not in that specific way. I think the earliest we have more visibility into, you know, when Metronome got founded, it was a relatively I'm not going to say contrarian take, but it was a big bet that consumptive models, flexible kind of like, you know, uh usage models would take off, right? They were again, before AI, I think there were signs the market was maturing and people were kind of forced to align value delivered with, you know, cost, that kind of thing. But, it was early days. There was not that much, you know, something like 5% of companies did so as a consumptive model. Um most were in the infra SAS space. Um and I think to kind of believe in Metronome at that time, you would find the VCs they might not believe in the consumptive model, but they believed in the same thing that we believed, which was that companies were going to have to increase the speed of iteration on this.

[00:19:34]I think now something like 85% of companies are exploring a usage model, which is way more than I think we would have even predicted or any VCs would have predicted early on.

[00:19:44]Which, you know, it's now kind of the table stakes if you're doing AI like we're kind of talking about cuz you need to you need to have that margin protection. So, I think there's a lot more kind of venture belief and understanding of usage models and the industry in general has kind of learned how to or is learning how to, for example, do revenue recognition on that, do forecasting and all all that. Those are hard challenges especially from a legacy business moving over. I think that's probably the main change that I at least perceived in the kind of venture perception of the space. I'm sure that, you know, if you are creating a usage-based business today, there's much there's much a much wider range of operators and and uh investors that kind of understand what that looks like than 5 years ago, but um but I think the main change is just like the commonality of this model. What advice would you give to founders as as a consideration of this entire component of building a business?

[00:20:49]I think the first thing I would say is you just have to be nimble here. Like I think it's very obvious to people if you're founding something that the product is going to be nimble. I'm I'm beating a dead horse at this point, but you have to be nimble on how you monetize the product as well.

[00:21:05]Those things are not You can't separate them in the in this world. And I think it is almost certainly not something you want to build in house. I This is I'm not here to, you know, um pitch Metronome specifically, but I would say if you're take for granted that you have to be nimble on this, you should figure out an approach to do that so you're not taking away your valuable, you know, engineers, product people to build this flexibility in, you know, focus on building the best product, finding that product market fit, and, you know, use technologies like Metronome or or other technologies to to build that kind of nimbleness in the same way you would use AWS as opposing as opposed to having to like, you know, buy data center contracts or something like that. Um because I think it's very easy to kind of put it off early on, and then the when you the when you first time you have, you know, the seven-figure deal or the eight-figure deal or the first enterprise contract, whatever that exciting moment is, that's the last moment you want to realize none of your systems support this, you know, new revenue line or something. You want to be ahead of that so that when people are trying to pay you money, you never have to turn them away because of some internal constraint.

[00:22:19]And how how do you find companies these days, especially new companies going to market, uh choosing how to price products? Cuz it's quite an elusive art, in a way. Yeah, I think there's kind of two There's a fork in the road relatively early on, which is is pricing a differentiator? And so I think, you know, when we see, for example, some of these AI support agents, pricing was a differentiator there in some ways as much as the product, right? Like I think the idea that you could use an LLM to answer support tickets is interesting and exciting, but the idea that then you can build a business around paying for successful resolutions, that is a differentiator and that's let some of the AI, you know, support agents I think really take off.

[00:23:13]Or are you saying we want to play in a market where there's kind of like a defined shape? So, if you're, you know, there's not that many new kind of core labs, but if you're a core lab, it is kind of converging around per token pricing. I'm relatively bearish on per token pricing on for the general uh kind of corpus of companies, but I think if you're, you know, doing some sort of text-based inference, you're probably going to charge on per token and then it's more looking at like, well, what is the value of my users are getting, what are my costs, and what is the price that the market will accept and kind of match it to that. So, I think the first question is just are you differentiating on pricing or are you kind of matching an existing model and then depending on that, it's kind of different paths in either case. But in both cases, uh whatever you decide day one is going to be different on day 30 and day 365.

[00:24:06]Where do you see gener- actually, I'm going to be super generic. The world in 2 years?

[00:24:16]Where do I see the world in 2 years?

[00:24:20]We're zooming out very quickly. Um I I'm still coming to terms with the fact that this AI boom is not slowing down. I mean, this week we, you know, the the news de jour is the mythos from Anthropic. I'm sure, you know, we'll see other exciting model launches and continuing at like a rate that I think I'm very very bullish on AI in general.

[00:24:48]I'm I'm very AI-pilled, but there's a part of me that still like believes that it there's, you know, it will taper off and I I'm still coming to terms with the fact that I that's looking less and less likely every day.

[00:25:03]Um I I don't know. I don't think it's 2 years is not that much time even with the rapid increase. So, I think it's just going to feel like things have been accelerating and accelerating, but if we, you know, not too different day-to-day, just like the models are going to be much much better.

[00:25:22]I think we're going to be seeing more kind of end-to-end agentic work. I think there's going to be whole new categories of product and business that I can't predict, but I think in the you know, what when you're on the exponential, each kind of subsequent tick looks unrecognizable, but I think in 2 years it'll feel like it just kind of we've just been along for a crazy ride. Yeah, it it it does seem that. I I don't think I mean, if if you said 10 years ago, 20 years ago, imagine 2 years ahead, that that would be so believable and you'd feel, well, yeah, I mean maybe this, maybe that, but it wouldn't seem that huge. If if it seems like 2 years now is like being in 1950 and saying, "What's it going to be like in 2026?"

[00:26:08]>> Yeah.

[00:26:10]And I think that's that's the that is the characteristic of being on a true exponential and I think humans are really bad I I don't trust my own prediction on that cuz humans are really bad at exponentials and predictions along them. I I remember reading a book uh probably at this point 2005 called Accelerando which it was following the singularity and it was written in such a way where each uh chapter got like a kind of like exponentially far farther forward in the future and it was pretty formative book when I read in 2005. It was, you know, predicting well, it was all sci-fi. It was all futuristic. And but it starts in, you know, 2010 or something and kind of looks forward to to to however long. And I reread it again last year and like the first three chapters have basically come to pass and the last nine chapters are very weird. And so it's like I I don't know what, you know, it's going to be a it's going to be a fun adventure, but it's going to be weird. But even if now we think I I spoke to a handful of people this year already who have done research reports. Some one is a advisor to the United Nations and another one um is a I think first AI ethics officer at Davos.

[00:27:29]And both had mentioned the amount of young people that have some form of relationship, like emotional some sexual relationship with uh chat-based AI.

[00:27:43]And if you said that would be a thing 10 years ago, you'd just say that's just weird. But but the figures were like 30-something percent.

[00:27:55]But for younger people coming up and through, that's actually normal. And it would be for them probably weird to say, "You didn't used to chat with your AI?"

[00:28:05]They're like, "Well, no, it wasn't there." It even having um a chat with chap called David Wood. So David was uh one of the co-founders of Symbian OS, which is used to be the OS on all Nokia phones and what have you.

[00:28:19]And he was talking about how they saw phones and the operating systems for phones. So it was all about the the hardware, then it was all about the OS.

[00:28:33]And they most people couldn't see what was going to happen when obviously Steve Jobs launched the iPhone.

[00:28:39]And then the App Store and every everything just fundamentally changed in in a direction. But here and then is the point at which Symbian OS was on 500 million phones.

[00:28:52]And Nokia had market share. People thought "Yeah, no one's going to take us over.

[00:28:57]We own this world." And then before you know it, this fork in the road happens and everything changes direction. We've kind of seen a few moments like that in the last 2 years.

[00:29:09]And as you said, you just alluded to Mythos, Mytho, however they're deciding to pronounce it.

[00:29:16]And the fact that they I read somewhere it escaped itself, but it's concealing certain things. And also now it's part of the What was it? Glass Winds or something like that? Where governments are testing.

[00:29:33]We We have to think things that we didn't imagine were possible will be possible. Normal will change form and definition, but also value will change form.

[00:29:46]And we could look at human beings as being really valuable in a way that they've probably been relatively dismissed over the last 10 years.

[00:29:55]Uh everything that we know to be a certain way is getting repositioned.

[00:30:00]Yeah, I think the that's really what I think characterizes the next well well next 2 years, 10 years, 2 months is is these kind of fork in the roads to use your your phrase are just more and more frequent. Like that we were talking earlier the the tick rate inside of these companies is picking up. The I think the tick rate in the world is is picking up and I, you know, to some extent that was always going to be true even AI aside, things move faster, you know, communication improves and and people can work together more, that kind of thing. But AI is just adding to that in ways like you that we cannot predict. And I think two years from now, everything will have made sense and it'll you know we can see a very normal path to get there. But yeah, if you would told me if you would told me five years ago that we'd have AI that passed the Turing test, or maybe not say made five too too recent. Now, 10 years ago that we'd have AI that passed the Turing test in my lifespan, I would have maybe believed you, but it would have been skeptical. If you told me a year ago that we had AI that could run you know, for 12 hours plus towards its own goal like agentically work to escape a container and and you know, email the researcher when the researcher is having lunch like like we saw with with Medusa, that would have been more believable I guess cuz we're in this phase of this of rapid change, but still surprising. And yeah, I'm sure there are things that if you you know, told me today we're going to happen in two years, I wouldn't believe you but then two years from now, it's you know, unrecognizable.

[00:31:41]So which movie do you think is the most probable? If we had had to pick a movie of the future and I I mean two examples Back to the Future or Terminator.

[00:32:00]I feel I mean the right answer is Her, which is already effectively happened.

[00:32:04]But between those two I'm an optimist generally. I think you kind of have to be an optimist to work in early stage companies and so I think I kind of have to say Back to the Future just cuz I I I I'm optimistic about the future.

[00:32:19]But I think probably closer to like I think I don't know, the Terminator feels more uh grounded in like science, but I'm I'm more optimistic than the Terminator outcome. I think we'll I think I think humanity will will uh do very well with AI in the future. I think I think the human side of everything.

[00:32:39]In in many ways the less humans have had the the opportunity for influence the more valuable they've become and in in many ways if you if you try and control humans to a point they will become uncontrollable.

[00:33:00]It's it's it all of history shown us whether uh whether it's empires you know that that comes a structure of control but ultimately free free will and and human spirit is is is a powerful thing but therefore I the augmented approach of humans being able to recognize value, create value and coexist in the world. I mean we've already got a ton of AI. We have since the Star Wars era. And um expert systems and as as they were back then Yeah, I'm really excited. There's kind of you know this question of when will we have the first billion dollar one person company um and I think giving folks these powerful tools and letting you know kind of in some ways uh expanding access or or you know expanding the the number of folks who are able to achieve whatever they kind of want to achieve with these new powerful tools is very exciting. I don't know when we will get to a one person billion dollar company but a lot sooner than six years ago we would have got to it. So it's um I I love I've always been in startups. I love the idea of kind of you know someone having a a founder or a a person having a problem they want to solve and kind of going out and bringing it into the world in some sense, solving that problem. Um and I think the number of people who are going to be able to do that is so much higher now with AI. My I you know, I have friends who are not technical and use the AIs to you know, learn new things or build build a product effectively despite not really knowing how to code or ever really thinking of themselves as, you know, a startup or or anything like that and just being able to do more was empowering folks to do more. I think it's very cool. And I'm hopefully we'll get a that that billion-dollar one-person company soon.

[00:35:07]Uh hopefully >> it is democratizing. Yeah. Uh AI AI is democratizing the ability for people who otherwise couldn't have to be able to do things or which kind of ideation to to reality. And just as a final point to bring it right back, if there's any founders, startup guys, anyone thinking about it, what what final piece of advice would you give them with regards to how to think about pricing as a core component of the journey they're on or about to jump on?

[00:35:50]Yeah, I think anyone, like we were just saying, anyone you with AI, with with uh whatever kind of resources they have available now can build solutions and products for any number of problems and I'm core part of solving a problem is how you kind of monetize as well as the product. And so, you know, think about if you think about it as two sides of the same coin and would you know, happy to happy to chat.

[00:36:20]I'm you know, I'm on Twitter and you know, easy to find me online or whatnot.

[00:36:25]Always happy to chat about anything that people are building, how they're monetizing it, how Metronome thinks about it if you want kind of advice or what we've seen kind of across the industry because, like I said, two sides of the same coin and with AI it's more and more possible for anyone to build products and monetize products and build, you know, the first billion-dollar one-person business. So, I'm happy. Yeah. Excellent. Well, Cosimo, on that note, I appreciate it.

[00:36:48]>> Yeah, thank you. Thank you.