Section Insights
Introduction to Anthropic's Business Strategy
What are the hosts discussing regarding Anthropic's potential business strategies?
The hosts, Aaron, Brian, and Brandon, are brainstorming potential business strategies for Anthropic, particularly in light of its upcoming IPO. They explore innovative ideas, including a pay-for-performance model and the need for Anthropic to align its offerings with how companies actually operate.
- Anthropic is preparing for an IPO and needs a solid business strategy.
- Innovative payment models, like performance-based compensation, are being considered.
- Understanding the operational realities of businesses is crucial for developing effective AI solutions.
Exploring Advertising and Operational Efficiency
How can Anthropic leverage advertising and operational efficiency in its business model?
Brandon suggests that advertising is an overlooked area for Anthropic. He discusses the potential for Anthropic to help organizations identify and enhance their differentiating factors while also improving operational efficiency by reducing unnecessary costs.
- Advertising could be a significant opportunity for Anthropic.
- Identifying and enhancing differentiating factors in businesses can lead to better AI solutions.
- Operational efficiency is a key focus area where Anthropic can add value.
Outcome-Based Business Models in Sales and Marketing
What is the potential for outcome-based business models in sales and marketing?
The discussion highlights the possibility of creating outcome-based business models in sales and marketing, where companies only pay for successful outcomes. This could revolutionize how businesses approach these traditionally costly areas.
- Outcome-based models could transform sales and marketing strategies.
- Businesses often lack clarity on the effectiveness of their sales and marketing expenditures.
- A shift towards outcome-based compensation could lead to better financial predictability.
Streamlining Transactions with AI
How can AI streamline major purchase transactions?
The hosts discuss the potential for AI to simplify major transactions, such as buying a house or a car, by automating processes and reducing the need for human intermediaries. This could lead to a smoother and more efficient transaction experience.
- AI can significantly streamline major purchase transactions.
- Reducing human involvement in transactions can enhance efficiency.
- The concept of 'informed commerce' could reshape consumer purchasing experiences.
The Need for Aggregation in Anthropic's Business Model
Why is aggregation important for Anthropic's business strategy?
The discussion points out that Anthropic lacks a significant aggregation point, which is crucial for maximizing revenue potential. Unlike competitors like Google, Anthropic needs to develop a collective service that can serve as a foundation for its business model.
- Aggregation is a key component for successful business models in the tech industry.
- Anthropic has not yet established a strong aggregation point.
- Developing a collective service could unlock new revenue streams for Anthropic.
Transcript
0:09 Good morning, good evening, wherever you are, and welcome back to the Enterprise AI Show. This is your host Aaron, and today Brian and Brandon do another round table on Anthropic by imagining they just got hired, and they have one job. Find out the next giant business. Brandon floats the idea that might get him fired, plus a radical pitch where Anthropic only gets paid when the work actually gets done. Brian counters that with a product that quietly matches how a company really runs, and the two of them wonder whether the biggest piece of Anthropic's business is still missing.
0:52 And all of that is coming up right after this quick break. >> >> Today's show is sponsored by Nord Security. Your employees are using AI every day, but do you know which tools they're using, what data they're sharing, or where their AI agents are connecting? You can't govern what you can't see. Introducing NordLayer AI Scout, the visibility layer for enterprise AI. AI Scout gives you visibility into AI activity across web and desktop apps, detects sensitive data exposure, helps control AI agent connections, and creates an audit trail for governance.
1:27 So, you can protect your data without blindly blocking innovation. See what your AI is doing, then decide what happens next. Connect with Nord Security to learn more about NordLayer AI Scout. Visit nordlayer.com/ai-scout /enterprise-ai, and get started today. >> All right, and we're back. Joining me again, Brandon Witchard. Brandon, last show we did, and we said we're going to do this as part of a two-part series, we we put ourselves in the strategy shoes of Anthropic. Anthropic, which we expect to do a an IPO sometime here in November, December 2026. It might slip, you never know, but that's what we're expecting. We said, "Hey, what do we think their strategy is going to look like for big growth?" And we threw out some ideas of what that looked like. I want to do a flip side of this, which is, "Okay, there's some things that they could do for growth. We might get those right, might get those wrong."
2:21 But, usually what ends up happening, and we see this a lot, is a company will get a bunch of funding, could be an IPO funding, could be just a bunch of VC funding, and you start seeing people in your LinkedIn profile who go, "Hey, the last 10 years at such and such a company was great, but I'm going to change the world over at Anthropic." And they end up becoming a product manager or business lead or something at some group, and they get snatched up because they've got some domain expertise about something that might help that company. And so, what I want to do is put yourself in the shoes of you just got hired as general manager, product management lead, whatever it might be. Doesn't really matter what the domain is. But, how would you go about trying to position some new Anthropic thing with a bunch of customers? Like, how would you Would you be going after it with the sort of traditional go-to-market approach? Do you think there's something going to be completely new with how AI is going to drive go-to-market? sit Sit in that space for a little bit. What are some of your early thoughts about how you might try to accomplish this?
3:21 >> I think the first thing you'd have to do, and this is breaks all the product management rules, is like, I think I'd have to go into the research group and be like, "What's going on in here? What is behind the hood?" Because I think the thing about this kind of problem, and I've run into this in my my {quote} {unquote} previous regular jobs, is like, you can come up with ideas, and maybe you can find some need, but usually you'll be shot down cuz like it doesn't They'll say something like, "It doesn't matter. Even if we're successful doing that, the market's too small, right?" So, you can't just walk in and be like, "I know how to make Claude code like 1% better." They're like, "Yeah, that's not We don't really care." So, the thing that's going to be really hard about that is you're going to have to really hunt down some interesting ideas.
4:01 And what you'd be looking for is the classic do they help big markets? So you of course you'd be like finance, banking, healthcare. You'd be like what in the research group do I think is going to pop out? And if if it did, where do And then I would immediately start prototyping every And this is the place where using cloud code and the funds I'd be like, we're pro prototyping everything that we can possibly think of >> Yeah.
4:26 >> in one of these domains and just trying to see if we can get any type of traction. The other thing that I think would probably get me fired from Anthropic so I won't do this if I'm actually there is I'd be like, listen guys, I looked at the stock market and it's like the bit some of the biggest companies, they all sell digital advertising. Maybe we should Maybe we should think about that. Maybe we should think about some type of contextual ads.
4:47 Now, I feel like that's not really what Claude and Anthropic want to do and that's more of the world of OpenAI, but you'd kind of be crazy not to like at least have the conversation, right? And say, hey, is there a way we can do ads in a way that is acceptable? Because even bad ads that are acceptable inside of services like this could have enormous profits. Of course, there's a lot of issues. Privacy, like we don't have to We will not re-litigate all of social media again here. So, >> >> I don't know. I don't know how they navigate that. And I would expect though the pressure and this is going to be real interesting is like the pressure to deliver against these numbers at some point will become so great that at some that it's like companies typically have to evaluate these things even if they at first didn't want to. I don't know, that's been my observation. What do you think?
5:39 >> Yeah. No, I think that's right. I think you're right in that you'd want to go in looking for gigantic markets, right? You'd be looking for it. Advertising feels like the one that's the most not on the radar right now as far as as far as Anthropic goes. I think the other way that I was thinking about this was I know we talked about it on the previous show being like, okay, do you just try and for example go to an organization and be like, we will rebuild your software. We will rebuild your systems.
6:07 We will do things that replace one thing, some big system you have, an ERP type of system, with something else. And I got to thinking about it and I was like, the we talked a little bit about that Alex Karp rant about the Palantir rant from a few months ago where he basically said a lot of these AI systems are sitting there between you the things you're doing with AI and your data and they're learning about your business. He called it it's taking your alpha, right?
6:34 Taking your differentiation. But I do wonder if there's a flip side to that discussion, which is you could go to a lot of organizations and be like, hey, what are the things that truly differentiate you? If you went to a bank for example or even you went to automotive company and you're like, hey, you guys do a thousand things. How many of them are truly differential and how many of them it's probably 70% are just things you have to do to run the business.
7:00 And if you there was an opportunity for you to take 30%, 40%, 50% of the cost out of that or find a way to get infinitely more efficient at doing it, you'd probably get most CEOs and CIOs to sign up for that. And so the the other thought that I had, you were thinking about advertising, I was like, is there a way to build a product that essentially like actively you would put in between whatever people are doing and and and the data. So for example, it almost just looks like a giant sniffer, if you will, for people that that haven't heard of that. Like, it would sit between whatever people do on a day-to-day basis and your biggest data sources and just be like, just figure out what the company does. Look at all the transactions, look at all the things that it's asking for, try and figure out the context of all that sort of stuff.
7:47 And then basically be like, how would I make this better? And I think you could potentially it would all sit in the background. You wouldn't be asking anybody to change anything for what they did, but you could probably come back with a pretty interesting picture after 3 months, 6 months to the C-level team and be like, there is a this isn't consulting BS. This is like tangible things you could go do that people would pay they'd pay some fraction of that, which would be a pretty large number in a lot of large corporations. I feel like that would be a I don't want to say it's a low-hanging fruit, but like that might be one of the initial things I would go after.
8:25 >> I like that idea. Does that make any sense? I know it feels like a privacy concern thing. >> Yeah, I like that idea, but I'm also thinking is there a place for like really a radical new go-to-market model? And I was thinking as you were talking there that I think it's Sierra, which is the company that was founded by I think Bret Taylor. He's like the Salesforce and I don't know he's very successful and all those things.
8:44 But as I recall his that company charges it's basically like customer support like people call in. >> Yeah, he charges it by outcomes. >> Right, but only by outcome, right? They only pay if it actually handles the call and the issues resolved. And I'm thinking to myself because like right now we're constrained about thinking about like This is a We have a product. We're going to have professional services, professional services and the product features go to the customer and eventually they produce something for them. And it's some version of like licensing, time and materials, something along those lines. But let's use Dario's thing What does he say? I have the 10,000 PhDs 10,000 PhDs in my system here or whatever, right? And I said to myself like are you willing to take on more risk? And so you go to the company and say listen, I've and I know we talked a little bit about this on a previous episode and we're talking a little bit about Cisco, right? Where and that they've founded one person or a small team to rewrite everything. But what if you walked in and said listen, I instead of me coming here and telling you how to rewrite this system, what if I actually just take this over and I say to you I want to take this over and whatever whatever metric we think that thing produces like some type of transactions or something like that. I just want to charge if I successfully do that and I'm giving you the outcomes that you wanted because I have the 10,000 PhDs with me, right? I don't need you to do that. And if I do that well, I'm essentially going to absorb all the risk cuz I'm going to do all the work and when I say all probably Claude. I'm basically going to take all the tokens, do all the work, and then if it starts to work, I want you to you're going to pay me. And we're going to come to some agreement that the price is going to work. So, that would be like a different way to go than like the very traditional. And I guess that's trying to think of like we have the agents, and we believe that we can run autonomous agents can ultimately do the business outcome that you want. And it's like it may take something like that to get everyone over the hump. Cuz I I still think we all, and me included, we all probably think too much agent in the loop. At least if we're believing, and I know today we're really we're on the optimistic side of this conversation. If we believe that optimistic side, we should be thinking the agents can ultimately do most of this work for us.
10:53 >> Yeah. >> So, and then maybe if we're philanthropic, like I they need to prove it to some people, and they need to be able to put some, like if you will, money on the table versus you pay me, and then I leave you with an agent that's doing all the stuff that you may or may not be happy with. So, I don't know. I like I like the idea of doing that, but it does feel complicated.
11:11 >> Yeah. No, it's I like the thing you looked at, actually. We've had a number of of companies who have pinged us about wanting to come on the show and talk about outcome-based results and so forth, which So, I think we need to do do a follow-up with some of those. I I would be interesting I know the company you talked about is doing it with customer service types of things, which feels to me a little bit like Yeah, we we went from like smart people to they outsourced it to other groups, so you got cheaper people, don't know as much, you might have to take longer to get to stuff. And then they they started outsourcing it till you got a pop-up window that would give you a menu. Now we're starting to have chatbots. So, I can see where that one moves towards outcome-based. I would be interested to see if they could do it towards things that feel like you spend a lot of money on them. So, name namely, sales and marketing, right? Sales and marketing the sort of SG&A li- line on the income statement or the the balance sheet or the income statement, I guess it would be. Those feel like things where you're like we spend so much money, and we really have no idea I don't want to say we have no idea, but we have limited idea of how successful they're going to be. And could you would somebody ever build a business that would be outcome-based around like marketing or outcome-based around sales?
12:21 Sales is probably the bigger thing cuz I think like Meta and Google what are you nowadays? You start placing ads since we don't charge you unless you will click on them. That's sort of outcome-based. But if you could do it from a sales perspective, hey, I'm going to give you this territory. I'm going to give you this number and I know with I don't know, 80% accuracy that you're going to make that number or these tools are going to make that number. I feel like that would be a huge thing. It would be interesting to look at what are the things in a business today that are the most should be outcome-based, but people aren't they're not getting very good returns on what outcome-based look like.
12:56 >> It's really interesting. It's a little dangerous to even talk about cuz it crosses over a little bit of what we're talking about before, right? I said before it's we got to broach this idea of online advertising, but maybe what we want to go to is there's so much information. If you start using Claude, there's so many things that Claude knows about you, especially if you just go read the memory files of it and when I use Claude code. And that if you kind of marry these two concepts together, right? It's so much because I think today people are using I just bought a car and it's just like everyone was using AI. I was using AI to write emails to the dealerships. They were using AI to write me emails and then they were generating ready quotes with AI. Then I was generating counter offers with So it was just like I was I said to someone else we could just we could just get out of the way. The agents could just do this work. It's not I'm barely involved.
13:39 And so I don't know, maybe that's the next level. And I think today that's what I feel like most people are doing when they're doing their outreach is they're using Claude to help write their emails. They're doing outreach through having Claude help send emails. >> Yeah. >> And you know, and of course we all get way too many emails. It's election season, right? And it's endless amounts of emails about the elections, about that. And it would be interesting I don't know, like all the projects I'm doing it's like I would be interested in what would Claude, if it kind of knows the stuff I'm working on right now, what would it tell me? What recommendations would it tell me to be like you should go talk to these people or maybe they they some tools. Cuz then it it they it happens implicitly today. If you're going to build a website, Cloud's basically going to tell you you can host it for free on Cloudflare. Don't tell anyone. Don't tell anyone at Cloudflare that I'm using it all their free sites they're offering so do it. And it's So there's some implicit advertising already happening, but it would be interesting if you could almost connect the dots cuz there would be a real funnel of money if you ever got there.
14:34 Of course, numerous privacy issues. Numerous numerous so many issues with what I just said, but you kind of feel like the money would be there. >> Yeah, and it would be Yeah, it would be an interesting conversation. I about a year ago I was I was doing an event I was speaking at an event and literally in the next ballroom next to me at this hotel this random hotel in Florida was an entire event just about the auto industry and it was just like auto dealers and stuff. And it was the people not like Ford Motor, but like the the place where you would go get a car for Ford Motor at your local place down the street or whatever. And the entire thing was about AI. The entire thing was about how do we better handle people coming in looking around the website? How do we reach out to them? How do we not How do we be aggressive? How do we not be too aggressive? How do we And the whole thing you think about that process itself and they went through a lot of hey, how do you get from better conversion rates from people that look around your site and how can you tell if they've been looking at other stores and they're just shopping against price and stuff. But you think about how many processes So much of what we talk about with AI is that you want to keep human in the loop and when are you going to have agents versus when you're going to have people and so forth.
15:43 But I bet there are a ton of like auto like buying a car feels like one of those things at the top of the list of things you're like, I don't want to deal with the people. I don't want to deal with the sales manager. I don't want to deal with the finance manager. I don't want to deal with like I would love to find a way to do a a a test drive and I'd love to pay you. And I everything else in between like this whole thing could be automated and I wouldn't have to deal with people and so on and so forth. I wonder how many of those types of things are out there that would be a perfect example of I don't have to worry about human in the loop because the dealer probably doesn't want you having a dealer in the loop. You don't want to do You don't want a person in the loop. All that kind of stuff. But yeah, no, it's >> So what if we like kind of even built on this a little further and we kind of like got had this idea of like what if we call it forget advertising? What if we call it like opt-in like real-time shopping? So for example, I just went through this. Basically like if Claude was I was using it. So like it knew like I even had a date. I needed to get this car by a certain date, right? It would almost be interesting if you could almost opt into a moment and be like like you you're like you're going to broadcast it like through Claude through your agent and say, "Listen, tell everyone I'm going to buy a car by this date. And have Claude do something that like really validates that you're going to do it. That you're really qualified.
16:52 That it's really going to happen. Like maybe you got to put some money in there. Maybe you got to show some type of intent beyond this. Cuz I know on the deal on the salesman side salesperson side, they get frustrated cuz people just >> lead. >> Yeah, what are you doing? Are you shopping? But it's okay. It's almost like I almost like I have this idea of the console then. Like I'm just in there and I'm just like, "Okay, I'm going to buy it this time. I've specified what I want. Let everyone ask questions." And then their AIs or or the salesperson behind the scenes, they can go back and they can essentially go out and give you their best and final out the door price, right? And it's like it all just works cuz it really like today I for this last time this is the first time I did it where I did almost what you said. I got five dealerships. I think four were willing to talk to me over the phone.
17:33 And then we ended up doing most things via text, right? >> Yeah. >> So it's just like the next level. And it was just like And it's so this idea especially for like a major purchase like a house could be the same kind of thing, right? Where it's just I may cuz you think about when you make a you go to your realtor usually and your realtor makes an offer on your behalf and then there's another realtor and then they go back. And it's it's almost like this moment of forget about almost the sales and marketing. Let's just make the transaction go really smoothly, right?
17:57 >> Yeah. >> And let's just see how much of it we can get done so that all the paperwork is done and that way when you walk up and you buy your house all you have to do is sign the papers, buy your car. So I don't know. I like the idea of like just pushing this because it's not so much that we just have more information. So we have a thinking machine behind us that can actually work with it. And we know on the other side there are other thinking agents. Maybe thinking's is word. But they're agents that are willing to help us do that. So, I kind of like the idea of if you I don't know what we'd call it. I don't know, informed commerce or agent-based commerce, where it's let's just get out of the way. Get the human beings out of the way.
18:31 >> Yeah, it your shopping example kind of created an interesting thought in my mind because in in in reality and if you're just thinking about it from a purely transactional perspective, if you're a salesperson, you want qualified leads and then you just want to try and go close those those things, right? You're trying to reach like I want to I want to find a perfect harmony between somebody who's like, I have intent, I have money and you're like, I have inventory and we're going to match those things together, right? And that's But then I also got to thinking about it and I was when I was thinking about, okay, let's suppose you are somebody who's like, I'm going on vacation. I'm going on vacation for a week. And this wouldn't be you or I cuz we would throw four t-shirts in a bag and we'd be done.
19:08 But let's say this is your wife or your your kids or somebody who's fashion conscious. The last thing they want to do is be like, okay, I've got to I've got to figure out weather, I've got to figure out activities, I want to know what's going on at the hotel or what are the local restaurant If you think about you're going to go on vacation, there's 40 variables in play. Who's going? What all these But you're like if you were like Anthropic and you don't think about this in terms of I'm selling stuff to people, right? You're thinking like, how would I create the ultimate immersive shopping experience? You'd be like, okay, tell me about your trip.
19:41 Where do you think you're going? What What do you think's going to be going on? And you'd be like, okay, I'm going to I'm going to I'm going to lay you out on an itinerary, which people do today with all the chatbot stuff. But you'd be like, I need to buy six or seven new outfits. And what you don't want to do is be like, all right, I'm going to go to 700 different stores and check out.
19:58 You want to be like, what would I look like in these? How What do you think I should put together? They could create because again, they're going out, they're looking at 50 different things through their tool calling. They could create the most amazing experience for you. And on the back end of that, now you've got Imagine all these, you know, retail places like bidding to try and get being part of that transaction. You've got restaurants bidding to try and get part of your things. They could be ultimately creating sort of what is the new going back to your point of outcome-based as opposed to advertising-based, that could become what their business looks like in terms of essentially every every experience that you're like, "Yeah, I got to do a whole bunch of stuff and I don't want to go off searching through 10 different websites. I just want to create the experience."
20:42 That's going on thousands and thousands of times a day all the time. They could create what would become sort of the new Google trying to pull all those things together for you and then being like visually here's what you would look like in this outfit. Here's what these things would look like. Here's how you'd pack it. Oh, you need you're That's not going to fit in the right suitcase. Here's the right luggage you need for this stuff.
21:01 Like that would be That would be a big picture thing they could only pull together as opposed to any individual company, any individual brand pulling that together. >> Yeah, I think there's some idea here that we just have to uncover more. It's like concierge like buying, right? Where I think we're getting at is no one really like today like you said before you can get an itinerary and stuff and it's like it's just okay. But really what you want is someone, you know, just like you would do with a concierge, like someone asking you questions that knows you, that knows the area, that's making you recommendations and then you change your mind and whether that's clothes or what you want to do or what you're buying or maybe you change your mind on the car, right? You're like, " it turns out I don't want the black one. I want the blue one now, right?" Something along those lines. So this idea that it we get away from where you started the question where traditional marketing is like pushing stuff in your face and this is why people complain and hoping somebody that somebody actually has the need when they see that versus intent-based. I'm really in it. And that's what everyone would like to do.
21:54 It's like if the only things that you saw were things and marketing and sales that actually were things that you wanted or needed, you'd probably be happy with it. The fact is most of the stuff is just is you're not going to buy it now because you don't need it now, right? And that's the frustrating part of it all. >> And that's And that's part of That's part of sort of your traditional advertising which is I'm going to keep throw That's basically what Facebook and Instagram does.
22:14 They're They constantly go, "Oh, we're going to show you stuff that you don't even think you know you need." But once we show you the the emotion of that experience or how cool the person looks wearing something or whatever it is, they're like you're going to be like, "Okay, I guess I got to buy that. I hadn't thought about buying that, but I guess I need to buy that." Like that that might not be as possible if it's going through one of these systems.
22:35 >> But I do think how big they can get is getting, if you will, into the larger conversation around how we're going to make back all this money for the IPO. It's like you want to get involved in the transactions, the business and the selling, right? Because that's the only kinds of things that are going to be big enough to generate profits. Think of it like MasterCard and Visa and all these other things, right? It's like you want to be getting involved in that side side of the industry so that you can make it and offer some type of value. Like Stripe is a >> and that's the piece the the more we talk about this, the more I think about it, like they I don't know how their enterprise deals are structured versus their individual deals. I'm sure it's like a lot of enterprise software, but like they still don't have a giant aggregation point, right? Google's got Google is the Uber aggregator of lots of different stuff, like mapping and travel information. They've got ads, obviously.
23:22 They've got they aggregate all sorts of videos. Like Anthropic hasn't even scratched the surface of being an aggregator of really anything yet. Like everything feels very individual when you're doing stuff with Anthropic. I know they do team stuff and but maybe that's where this thing really explodes in the way that none of us can really understand it because we're looking at it immediately and being like, "Oh, you don't do ads yet." And it's yeah, but maybe there is a much bigger thing there. They just they don't seem to have like even showed the tip of the iceberg of what that aggregation thing is, which is the aggregation business as Ben Thompson talks about all the time. It's like the biggest way to make money in in in the internet, whether it's on the consumer side or the enterprise side, but more on the consumer side.
24:03 >> And that's when it's really OpenAI who has the aggregation point, but they have yet another thing Ben Thompson talks about all the time. They have not figured out how do we make money for it. And And I don't know the people they have the people point. They have lots of people, but they don't have an aggregation point. There's no collective service that they provide yet. So maybe that's the open question going forward is for so long everyone's scoffed at subscription-based for consumers, right?
24:23 I know obviously people listening to this podcast probably have a probably willing to pay their own subscription, but like maybe if you're any one of these companies, you think yourself, "How do we get Let's kind of like Netflix. How do we get 300 million people paying $20 a month?" And not And just think of it just like Netflix. Of course, I'd pay $20 a month for Claude or for OpenAI because that's just what I do, and that's how it works. So, that would be I don't know. That That's sort of the white whale. Like no one for productivity tools, no one really has cracked that. It seems like it always becomes ad-based. But you'd have to If you want to make billions and ultimately trillions of dollars, like you you got to be in some big markets.
24:58 That's a big dilemma. >> Yeah. Yeah, the more I think about it, the more we're talking about this stuff. Like even if you were doing the thing that creates that great experience or something like that. And maybe you could and you could do Claude for retail experience, Claude for travel experience, Claude for whatever. You've got to reach a point where you're doing much more aggregation than going to every single business and being like, "Okay, let's replace your ERP system.
25:20 Let's replace your legal group. Let's like" You have to start with some sort of bigger aggregation that then allows you to then go off and It's It's a It's almost like the Facebook strategy that we're talking about now, which is like they've got these aggregation points, and now they're trying to get into enterprise. I don't think you can do it the other way around. I don't think you can start in enterprise and then move into other aggregations. At least we've never seen anybody necessarily do that before.
25:44 >> And I think before we maybe get out of here, and maybe the real answer to this question is like the nobody really knows, but they just know if we just keep making these models smarter, that these models will do something that's incredibly valuable. And that's You know, there's nothing more that people like than a great story. That one day the great model will solve all the problems, and if you don't invest in it, you're going to miss out. And some days I believe that, and some days I don't.
26:06 >> Yeah, and that And that was the other thing that I had lingering in my head the whole time was I was like, "We keep thinking about this in the context or in the the framing that that we've done for 20 or 30 years, which is like there's a product that does a thing for you." And then there's a different product, and it's like their product is essentially like the aggregation of all human intelligence. You've just got to You've just got to have the owner's manual to go unlock it.
26:29 And I'm like, "On one hand that's unbelievably powerful and the opportunities for that are like unbelievable. But at the same time, you're then expecting people to go figure this stuff out. And yes, people have figured those things out, but like the population as a whole aren't creators. There's small portions of the population that are creators and are curious enough to go do that. The rest of the population tends to be like, I follow along to what the system tells me to do or what I was told at school or what my neighbor does or whatever. So, I do sometimes wonder to to Anthropic's point and some other like if they just they don't grok just how regular human people work as opposed to like the super smart people that may be working at Anthropic or the the way that they work since they're tend to be less social in what they do.
27:13 But yeah, that that point of like maybe they don't need that stuff because the system already is the super aggregator itself and it will just figure out ways to do it. I just don't feel like like you have enough control on how to drive big markets that way. So, I don't >> So, you know what? Maybe the real answer to this question is like you just wait till the model gets really smart. Like as a product manager, you say, "Listen, I'm just going to wait long enough till the model's really smart and then I'm just going to my prompt is going to be like, 'Claude, what is the best way to make a lot of money with stuff that you can do?'" And then it just tells and it's like, "Oh, here's the answer." And it's like, "I can invent a cheap renewable energy for you. I can solve major medical problems. I can do this and this." I'm like, "Okay, go ahead and implement all that and I'll take it back to my boss." Like maybe that's the answer is we're just going to ask the model at the end.
27:55 >> That feels a little bit like the What's the old saying? The market can remain rational much longer irrational much longer than you can remain insolvent or you or remain solvent or whatever. Yeah, we could keep cranking to build the world's greatest model. We might run out of money in doing that, but And then at some point we'll be able to we'll be able to do the one command that just says, "Claude, make me rich." >> Yeah, that's You know what? That's what we need. Until then, I guess we keep podcasting.
28:16 >> That's right. That's right. All right, man. We're going to wrap it up there. Hopefully folks enjoyed this. We'd love to hear your feedback on both everything that we got wrong, which might be the entire show, but more importantly, if you guys were put in those positions, what would be your ideas for how to what Claude's going to look like to get from $2 trillion to $200 trillion. So, let's hear your ideas. Love to hear them. Show at the Cloud Cast Show at the Enterprise Show.com you can hit us up or hit us up on socials or any other way. So with that we'll wrap it up. We'll talk to you next week.
28:45 >> Thanks for listening. Check us out at the Enterprise AI Show.com for past shows, newsletters, and all things Enterprise AI.
Summary
- Anthropic is expected to pursue an IPO in late 2026, prompting discussions on its growth strategies.
- Brandon suggests a radical model where Anthropic only charges clients based on successful outcomes, similar to some customer service models.
- Brian proposes a product that analyzes company operations to identify inefficiencies and potential improvements.
- The hosts discuss the untapped potential of advertising and contextual ads as revenue streams for Anthropic.
- They emphasize the importance of targeting large markets, such as finance and healthcare, for AI applications.
- The conversation touches on the idea of creating an aggregator platform that connects various services and products, akin to how Google operates.
- There is speculation on the future of AI-driven commerce, where AI could facilitate seamless transactions without human involvement.
- The hosts conclude that the key to Anthropic's success may lie in leveraging its AI capabilities to create innovative solutions that meet market demands.
Questions Answered
What are the hosts discussing regarding Anthropic's potential business strategies?
The hosts, Aaron, Brian, and Brandon, are brainstorming potential business strategies for Anthropic, particularly in light of its upcoming IPO. They explore innovative ideas, including a pay-for-performance model and the need for Anthropic to align its offerings with how companies actually operate.
How can Anthropic leverage advertising and operational efficiency in its business model?
Brandon suggests that advertising is an overlooked area for Anthropic. He discusses the potential for Anthropic to help organizations identify and enhance their differentiating factors while also improving operational efficiency by reducing unnecessary costs.
What is the potential for outcome-based business models in sales and marketing?
The discussion highlights the possibility of creating outcome-based business models in sales and marketing, where companies only pay for successful outcomes. This could revolutionize how businesses approach these traditionally costly areas.
How can AI streamline major purchase transactions?
The hosts discuss the potential for AI to simplify major transactions, such as buying a house or a car, by automating processes and reducing the need for human intermediaries. This could lead to a smoother and more efficient transaction experience.
Why is aggregation important for Anthropic's business strategy?
The discussion points out that Anthropic lacks a significant aggregation point, which is crucial for maximizing revenue potential. Unlike competitors like Google, Anthropic needs to develop a collective service that can serve as a foundation for its business model.