Intro: We are probably overindexed on the flashiest parts of AI in the consumer world, and then under indexed on the things that will just make hundreds of billions of dollars in B2B. What's the coolest or most mind blowing demo you've seen in the last three to six months? I wish I could have my pre-chat GBT wiring for every individual breakthrough.
Through because I think I would just be like, holy shit, this thing is witchcraft like every day. How do you think about durable value creation? The things you probably want to do to like maximize your moats are welcome to Logan Bartlett show. On this episode, what you're gonna hear is a conversation I had with co-founder and CEO of Box Aaron Levy.
We cover a bunch of ground related to the state of the enterprise today and what he and his customers are thinking about artificial intelligence. We talked through a bunch of different topics related to. Ai, what different people should be thinking about his views on closed source versus open source, and the opportunity that present going forward, as well as some of the considerations related to the business model change that seems to be occurring with [00:01:00] usage-based pricing.
A really fun conversation that you'll hear with Aaron now.
Aaron: Uh, thanks.
Logan: What in your mind is a model company in.
Aaron: Wow. Okay. That's just jumping right in. Um, what is a model company? Well, I mean, I think, I don't think you actually even have many model only companies. Um, and so maybe the only distinction is, is, is you have companies that. Models and then other things. Um, so, so if I'm being like, extremely technical, uh, technical about the definition, uh, there's probably not, not that many like model companies.
I think you have, you have AI product companies that have research divisions that, that, you know, build models and then a.[00:02:00]
The only, you know, like true model company where most, you know, open ai. I would, I, you know, I would, I would not call a model company. I wouldn't call philanthropic a model company. I'd, I'd say those are, you know, AI found, you know, foundation, frontier lab companies that have products and models and services and just like all the things.
Logan: Do you think that serving enterprises with, um, some type of model offering, but ultimately selling. Uh, whatever the core things that enterprises mostly buy, do you think a company can build long-term value around that, or do you think that will ultimately, whatever that is, resides within an AWS or a Microsoft Azure or a GCP or something like that?
Aaron: I think it's, um, I. Think it would be hard to underwrite that alone as a massive business because you basically are dealing with this issue where all of [00:03:00] the hyperscalers at a minimum, and then, and then a few other companies that, that aren't kind of natural hyperscalers, but they have adjacent, they're adjacent enough to the hyperscalers, like they need to make sure their models get used at at an unbelievable scale.
Because they have to win an infrastructure war. They need the compute to, to come to them. They have other software they have to sell. So it is, it is so existential to them to have their models get used at massive scale, that they will constantly drive down the price of AI and the quality up in ai that you would just get squeezed, you know, you know, into such a, such a difficult position.
Uh, that I'm not sure, I'm not sure what your, what your value proposition would be.
Aaron: We ran these, um, we ran these tests on Gemma, uh, from, uh, from two weeks ago. And, um, the new open source model from Google
Logan: Yeah, not to be confused with the severance character, Gemma.
Aaron: yes. Uh, prob Yes, exactly. Um, [00:04:00] so, so we ran these tests and it actually performed only like a couple points worse than Gemini. On, on our evals for, for accurate, accurate data extraction, uh, from, from enterprise documents and data. And so the fact that you have an open weights model from Google that is almost at par with their commercial model that is only was already kind of state of the art two, three months prior, as, as a set of breakthroughs means that.
You know, you're gonna, as an enterprise, you're gonna have a choice between an open weights model, you're gonna have a choice between a commercial model. They're gonna all perform extremely well, and you're really only gonna be making a choice of like, do I want all of the, the scaffolding managed for me, or do I wanna just run this in my own data center, GPUs or my own cloud?
But like, imagine being a, a kind.
And Microsoft.[00:05:00]
Everybody's doing that. I don't know how you would, I don't know how you'd sort of slip in there. So, you know, there's a case to be made for obviously an extreme, like deep vertical, um, you know, use case or, or some, you know, kind of, of, of niche. But, but I'm, I'm taking your question. Literally. I think you'd still eventually then need to build the software around it.
You'd have to build, you'd have to get really good at all of the. Surround delivering that AI model to the customer in a way that let them actually use it, um, as opposed to just selling tokens back and forth. I don't think there's a business model for us, you know, kind of an independent player at scale to just own kind of, you know, monetizing tokens.
Logan: If you were running OpenAI, would you keep pushing the boundaries of frontier models or does that just feel like such diminishing returns, and why not redirect all that energy to. The consumer product that is chat, GBT.[00:06:00]
Aaron: Yeah, so I, I, I would definitely do it and I don't have that much. Um, I, I don't have that much, uh, kind of critical feedback for, for open AI or, or kind of any of the major players at the moment, you know, critical, you know, issue are. What does Google do about core search? Like, that's a super interesting question.
What does Apple do about, you know, kind of the Siri problem and, and like, how much do they need to own themselves versus, you know, kind of let their, let their, let let other players, you know, kind of, you know, power some of that functionality? So, so, you know, there's some interesting strategic questions that, that I think emerge, but in terms of ai, their execution, I think it is, let's.
It's not even hedging your bets. There's a nice flywheel, so they're not like different work streams. It's like, let's have the best models in the world. Let's then incorporate that, that into a consumer and, and you know, prosumer and then business set of products. And then there's now a new layer, which is like, we're gonna, let's make sure that we have like applied use cases in these agents, you know, the rumored engineering agent and all this [00:07:00] stuff.
And so they can kind of, I think they can have their cake and eat it too. In, in a lot of this. And, but it's, it's critical that they're always at the, at the forefront of the state of the art models because that's the, that's the calling card into the business conversation, which is we also have, you know, the, the, the best performing models on these, on these dimensions.
And if there was only, you know, the chat BT play, I'm not sure that would be, you know, that, that, that is not why their brand got to where they are. Um, and I think you want both those things to kind of be interlocked. Now there was a, there was an interesting thought experiment that somebody mentioned recently, which was, you know, for at least the consumer use case, if tomorrow you swapped in, you know, Gemini or deep seek, you know, just like, and, and just, you know, had a person go to chatt, would they notice the difference, the answer?
You know, obviously, probably not at this point. Like we've reached, we've reached a convergence on model quality for most consumer use cases to be satisfied for, for, for.[00:08:00]
The majority, but not necessarily the most important ones. And so, and so there's gonna be an 80 20 rule, which is, we probably have solved like 80% of, of like, Hey, what's the score to the sports game? Tell me about World War ii. Um, you know, what should I wear in this meeting? Like, like we, we, we, we've solved that.
But the, the 20% is gonna still be like, you know, orders of magnitude more, more quality that we need, which is like go out and, you know, do the classic, I need you to book a flight, you know, type problem. And those, those, you know, were nowhere close to solving. So, so we kind of got to the 80% really quickly and now the 20% is gonna take years and years to kind of keep cranking through.
Logan: I, I get the feeling you're excited about agents. If I, if I read your tweets and all that, is there an agent that you think has product market fit today?
Aaron: Um, yes. Uh, it, but, but for, you know, co couple things, like, I think that the, [00:09:00] you, you, um, I don't know every, everybody's, you know, is if everybody's equally defining agents in the same way. But I'll give you like three or four. So. Um, you know, one company, uh, that I happen to be an investor in, but, but I think they, I think they just independently you, you'd probably agree, is this sort of space around outbound selling.
Um, so one company is 11 XI think there's others that are good. Um, and they've, I think they've tapped into, you know, and maybe the only, maybe the product market fit right now is like lead gen. Um, and then it kind of expands from there. But, but for all intents and purposes, people are using this to augment, you know, their, their sales capacity, you know, generate more demand.
That seems to be working. I think like writ large coating agents, you know, seem to be working. Again, maybe we could define different levels of ancientness. Um, uh, but, but what you're seeing in, you know, cursor, windsurf, uh, rept, you know, level of bolt, et cetera, like, obviously, you know, probably different use cases, like probably 80% of the vibe coding is more fun and, and it's interesting to watch and not going into production.[00:10:00]
But I could see a lot of, you know, product market fit for, Hey, I have this internal IT tool. It's not, you know, not the most strategic thing. We have to do an agent, kind of write it or modify it and, and then boom, you know. Like, I've just like solved that problem that, that used to maybe go to low-code tools or you would have to, like, you'd either never solve or you might hire a contract firm and it would take six months to go build.
So I'd say that's product market fit. Um, on our end, you know, we, we have a, we have, um, one of our first kind of agentic use cases in, in just the, the fact that it's not kind of a, a single, you know, one shot, uh, in the model, um, is just we extract data from documents. So then you can, you know, say, Hey, here's a contract, here's an invoice.
I wanna, you know, read the document, put that into a database, and then automate a workflow around it. That's, you know, clearly got, at least in our world, product market fit. Like, like, like mostly our problem is we can't keep up with all the range of use cases customers want. Um, uh, and, and so like, there's like more demand than, than the ability to kind of [00:11:00] satisfy, you know, everything that the customers are asking for.
But those are, those are a few, you know, on side I'd probably not a lot of, uh. Um, but on business, you know, it's, my bet was this two years ago. My bet will probably be it in five years from now. Um, we are probably, you know, we're probably, um, over-indexed on, on the flashiest parts of AI in the consumer world and then under indexed on the things that will just make, you know, hundreds of billions of dollars in B2B.
Um, and like I'm sure like Nik.
Logan: Yeah. Well, no, I mean, it's good you're talking my book here. Uh.
Aaron: So,
Logan: I guess one question. Uh, the business model of cloud SaaS, if you will, uh, I think has been historically probably underappreciated, uh, versus, uh, obviously the delivery mechanism and all that, uh, I, I think people have a [00:12:00] lot of appreciation for, but just purely. Uh, how the, the pricing was able to disrupt a lot of, uh, things, be it Salesforce or whoever.
I'm curious, do you think that usage pricing is, is an opportunity with AI to have that similar, uh, effect, or is it something different in your mind?
Aaron: For like, uh, the, the broad transformation or disruption or, or which on which access?
Logan: Yeah, I, I, I guess, um, I guess as you think about like who your historical competitors were, uh, I think historically people underestimated, um, the on-prem to SaaS transition and the number of customers you could reach was probably one of the big things. Uh, but there was a lot of innovation that ended up happening around.
Seek based pricing and, and elements of, you know, fair use pricing. Like Slack was early in that, and I think all of that stuff, there was a bunch of different elements of, of the SaaS [00:13:00] transition that ended up being, you know, user based in some way that I think was an accelerant of this. Transformation, uh, transition.
That playing do you think has a opportunity? Business is agonal and ultimately it might not be as broadly applicable
Aaron: No, I, I think it's, I think it's, I think it's very relevant. Um, and then, uh, you know, so my, my characterization of SaaS and cloud, uh, from a business model disruption was it took something that you used to spend upfront, like maybe millions of dollars on to now you spent thousands of dollars on and.
Logan: CapEx to opex transition.
Aaron: Opex, um, uh, sort of like perpetual license to, you know, by the drip. Uh, you know, anything that, that was just like, okay, I have to wrap my arm, my head around this, this major outlay of capital system, integrator fees, infrastructure. I have to build out [00:14:00] team, like, you know. If I, this is more anecdotal, but if you were like to deploy a CRM system in the nineties, you know, just think about what had to go into that, like, like data centers system integration.
Like, uh, you bought software for the, for the CRM system, but you still had to buy hardware, you had to buy all the servers, you had to have a network. So like, think about how many steps from the moment you said, I want a CRM system to the day that you're actually gonna like log into it with your data. I dunno, a year, year and a half, you know, millions of dollars.
Salesforce shows up and they're like, literally like tomorrow you can just have a CRM system and oh, by the way, you could use it for 20 people in your company. And there's no, you don't need some kind of like critical mass to, to make the economics work. Like, just like start using it. And so, so you know, you multiply that over two decades or two and a half decades and it's like every s and b on the planet can do it.
Every small team can do it. And these, and the tams of these markets just basically grow. Because now everybody can afford a CRM system that couldn't, so [00:15:00] So that was, it was because it was a seat base? It was 'cause it was subscription. It was also just, 'cause literally like, like your day one upfront bill was just a hundredth of what it would've been, you know, because you don't have all the other, all the other things, you know, wrapped around it.
Cloud did the same thing, obviously more on an elastic basis with compute.
Aaron: So I think what agents do is. Is that, um, I think there's gonna be two big, two big kind of areas of upside. One is you're doing that now for effectively a form of labor. And so that kind of, uh, has this interesting impact of it expands the tam of software from only being able to sell the software to people on the other end.
So you were kind of capped by the number of seats to now you have kind of like, you know, like I, I can think of categories where, where the category of AI. Spend will be an order of magnitude, uh, more than the spend of the software in a, in a kind of a non-AI world. So take, take, you know, a a a, you know, relatively simple example of like contract [00:16:00] management software.
Contract management software as a category is like a couple billion dollars, like globally in the world. Legal services is a couple hundred billion category. And so, so can, can software. Take, you know, 10% of that spend, you know, through AI to, and, and pull from the legal services spend or expand the legal services spend.
And then all of a sudden this category that was, was relatively niche, couldn't make that much money. 'cause the amount of people that wanted to buy contract managers and software was kind of low. Could that now be a $20 ca, $20 billion category as opposed to a couple billion dollar category? So you could do that probably 20, 50 times across software and already underwrite Master TAM expansion.
After going after, you know, you know, a whole bunch of, of new use cases that used to be, you know, kind of labor essentially. And then the other, the other part is, is ai, you know, has a way of just completing the things that people have kind of always wanted to do in the [00:17:00] products themselves, but just whenever possible, which then also unlocks a whole bunch of new budget as well.
So, so I think you're gonna see this is, so this is the new version of disruption. Is, is it Just, will it, it makes. It lets you go and complete the outcome that the customer was looking for in a way that they are not required to go and deliver it themselves with their people. And then that means that you, you know, more of them that value accrues to you.
And so that's the, that's the usage based or outcome pricing, I think we're gonna see in ai.
Logan: When, when you extrapolate that service budget there, uh, forward and, um, and think about the value to the end customer totally makes sense. Uh, that they'd be willing to pay. Hey, I. There was this person that I had to do this, and now I can get software to reach things that weren't even possible or do it, do it more.
Um, I worry about the deflationary element, a little bit of like the competition. And at the end of the day, are you beholden [00:18:00] to what your next. Customer, or sorry, competitor is willing to charge. And so we might not be able to actually tap into the service budget quite as much. Not because the opportunity's not there, but because we're, there's tons of venture companies popping up and competing with one another.
So I guess one would love to get your reaction to that. Two, how do you think about durable value creation? Uh.
I would pay anything for it, uh, basically to work, uh, if there was no other alternative. Uh, but there's a lot of other alternative.
Aaron: Yeah. So, um, I don't think that any law of, of, uh, market dynamics, you know, changes because of ai. So, so I, I, you know, I, I, but I would say that has probably been true in, in a venture startup land kind of forever, which is, you know, the CRMI bet there was a CRM market map from Gartner in 96. [00:19:00] 47 companies that did CRM and somehow only Siebel and like one or two others made it to the, to the other end.
And so, you know, did they temporarily cause a deflationary impact on the price of maybe what Siebel could have charged if they were left alone? For sure. Did it still create a great, you know, outcome for, for Siebel and, you know, you know, there, there's a version of the world where, you know, doesn't sell to Oracle and they, they, you know, the company and now it's so.
I guess it's a hundred percent true, but, but nothing about AI seems to follow a different law of, of how these things play out, which is competition will suppress, you know, what, what you could like maximally charge and, and it keeps kind of all the players in check, but eventually you get like some degree of network effects, some degree of like, you know, reference customers that tell each other, you know, that you can all use, you know, X thing and, and and so on.
And, you know, conversation. Um, uh, my, my, my dad was [00:20:00] asking like, how did you know 30 billion and we're.
How does that happen in a world where like, obviously there's like 500 security startups, like, well, how does one break out? And it's, there's a little flywheel where like, your product's just 20% better. Then you get a, a few more reference customers and then, and then you learn faster and then your product is another 20% better, faster, and then you get slightly better, you know, people working on it and, and kind of spin out, have one escape velocity, you know.
Outpacing the rest of the category, but yet probably there's still a second and third and fourth player that are pretty good and they'll probably also grow and then they'll get acquired. So I think ai, I mean, I don't, I don't see a different dynamic playing out in AI for, for any reason. And so, and so then the things you probably want to do to like maximize your moza are, you know.
Again, the same lessons probably from 35 years ago in enterprise software, which is, you know, get data, do workflows, you know, make sure you've got, make sure you've got [00:21:00] a, you know, kind of enough happening within your ecosystem, uh, that, that keeps customers to have some kind of, there's like a, like a compounding value.
More data that comes into the system. You know, the next run of the AI agent, you know, gets just a little bit smarter. You know, there's gonna be a lot of network effects on the embeddings. Um, your data. So, so the, the, um, you could, as a single programmer, you could probably switch between cursor, windsurf and, you know, rep and a few others, but as a, as a network of 500 engineers in a.
I've done embeddings on the whole code base and, and I'm, I'm talking to all the different, you know, systems, it's wired up to Wow. Change management on that, you know, becomes a lot harder. So, so I think, I think you kind of, I think we're gonna see a lot of very similar, you know, maybe this will be bad news for everyday in ai, but it's gonna look a lot like SaaS and the lessons are gonna look like a lot like SaaS and, uh, and so you want data, you want workflows, you want strong customer references, and you wanna build a network, [00:22:00] network, network of founder, uh, network effect around your customers.
I think it's gonna play out like that. And then maybe the one thing that we just have to be prepared for is there might be some pricing dynamics right now in play in AI that won't necessarily last forever, which is, like right now, you can kind of comp it to labor in some areas. Um, and that, that probably feels like a little bit of a temporary advantage that, that you might have when ultimately it'll probably be more comp to like, like typical software margin structures.
Um, and that'll just be a journey that, that, you know, the industry goes through, um, as, as we kinda land into the, you know, terminal business model.
Logan: Yeah, I, I guess we don't need to belabor this point. The two things that I do wonder about with AI specifically is, are the solutions that are maybe less. Workflow, UI centric. In some ways. There's like the, the SaaS world, traditionally we had all these screens and logins and, you know, workflow that existed on top.
And so maybe security in that way is actually the right, uh, analogy because I, there's [00:23:00] definitely dashboards and integration points and all that with security, but ultimately the work. Is getting done oftentimes outside of those screens, right? It's pulling in all the, all the data in some ways. And so, um, yeah, it's just something we've been noodling internally.
Aaron: What, what, what do you conclude? Uh, because.
Logan: Uh, I think, I think the things that are just tapping into, uh, the work automation, um, be it, uh, customer support or SDR or whatever it is, uh, ultimately because there's less workflow embedded into it, the change. From one to the next. I think there's less, uh, friction associated with that change. you can swap it out.
You see, it's almost like a meme on the internet of people going from, you know, cursor to, to coded or whatever is. And that's not the right analogy 'cause that actually has some workflow around it. But it does feel like people are a little more willing to than at least I had historically experienced.
Where, what's the Asana? Monday and Trello. [00:24:00] You gotta switch your entire team onto something different, uh, in a meaningful way versus when the work is actually being done outside of the ui. It might just be one click change, which leads to a little bit, um, I don't know, uh, more ability to price, shop more, ability to do other stuff.
But I guess that it back to the integration
Aaron: I, yeah, yeah. Well, so that, so that's the thing. So, so if you, so there's so, so the question is, so it used to be the network, the, the stickiness was I had a hundred people that all learned how to use the, use the buttons, and, and then there's data behind the buttons. And so, man, that's a lot of, that's a lot of change.
But the question would be like, is there gonna be an analogy where. It's inverted where you, you still, the ai, you know, it's, it's APIs. It's APIs into your other systems. It's the APIs into your e-commerce thing. It's then the logic around when to make the decision about the e-commerce thing. So I think,
Logan: it's self-reinforcing and learning, right? And so like at some extent you get all the data, get all the [00:25:00] integrations, but then there's ultimately some algorithm or whatever, some AI thing on top that's learned on top of it. sure you could switch, but then it's a cold start to start over.
Aaron: Because I, I, you know, I could, I could underwrite the exact opposite of this point, which is that the, that when you have a learning system, uh, that that learning system will probably perform better than your people did. And so previously you could change out, you could swap out software and. There wasn't like so much IP in the heads of the people about how that software worked.
'cause we're, we're just like, okay, click the next button and whatever. But like, but like, the AI is gonna track every event it's done in history and, and, you know, and, and just get a little bit better. And, and the signal will just get a little bit better and it'll, it'll, you know, have a little bit more fine tuning or it'll have a little bit better instructions to the model.
And then that becomes some, you know, opaque. Data source to you, the customer that is inside this, you know, AI product that [00:26:00] just makes the thing always perform five or 10% better in a way where you can't just go to people and say, here's the new sheet of paper for how to use this new product. We just implemented, like, like you almost have now created your own, you know, for all intent purposes, model for, for doing customer support or coding in your product.
So, so.
Or or non UI oriented, it is to me, doesn't necessarily correlate to its stickiness and in.
Logan: I heard you say that, uh, cloud was, um, pure efficiency and maybe there goes back to one talk you were giving at some big bank that, uh, everyone's eyes were glazing over. I think it was like a conference or something that, that maybe the, the most boring or least interested, you've had a crowd, uh. But it wasn't like tangibly output based, uh, in, in, in some way that like the, the rank and file within an organization didn't care about the [00:27:00] backend infrastructure.
Uh, and ultimately you're speaking about some elements of it. Uh, can, can you give that quick anecdote, but then also talk about how AI is, is different and how that translates to the enterprise and the conversations you're having.
Aaron: Yeah, so, uh, if I recall the anecdote properly, it was, uh, I, I think I was just referencing that like in the peak of cloud adoption, I, I, I did this keynote on like the future of, of, you know, running your bank or, you know, financial services and, and it was, it was mostly just the pitch cloud. And, and it was, you know, the, the, the, I I find it very exciting.
Uh, but the audience didn't because it was, because it was just like, yep, cool. I, we get it, man. Like the servers are not in our data center anymore. They're in now a cloud data center. Like really, like super cool. And, um, uh, and, and so, so that, that, that to me was kind of the, it kind of captured the cloud transition, you know, at an industry level in a nutshell, which was like, it.
It made your, it made your IT [00:28:00] operations faster. It meant you could store more data so you could make better decisions. It makes your employees more efficient 'cause they have better software, but it didn't radically transform like entire industries or sectors. And ai, you know, conversely has the opportunity to, to actually radically trans, you know, transform how you run your business, how you make decisions, you know, how you, you know, serve your customers, how you build products.
And um, and you know, I think it's. Uh, it's, I think it's always useful to learn, you know, figure out like what things will be similar to a prior technical, you know, technology transition, which things will be different. Um, I think the thing that will be similar is that, that adopting AI versus cloud will take all the same change management.
It'll take all the same meetings, it'll take all of the same, you know, privacy and policy and governance and compliance review people, everybody like. But I, but I do notice and, and see one big distinction in, in the two waves having kind of been at the similar timeframes for both. [00:29:00] Uh, in the very early days of cloud, you would go into a meeting with a bank, a, a pharma, a, a government agency, and you'd say, you know, I'd love to talk to you about, you know, moving your data to the cloud.
And it was sort of like, like not, not interested, shock. Definitely not gonna happen. Not gonna happen here. Uh, maybe, maybe there'll be some small little workload that's like a dev test workload. Uh, maybe some, you know, department or fringe, you know, group will use it. And, and so it was mostly like years and years of just like selling and selling and selling and, and trying to get you to understand what was so much better and all of this.
Conversely, and so we're only about two and a half years into chatt, let's say. So that would, that would be, that would easily describe two years in cloud. If I go back to, to the late two thousands, let's say if I look at AI right now, that exact same meeting with a customer.
We know this is gonna change everything. Everybody's clamoring for it. Every employee needs it. The new generation [00:30:00] coming into the workforce is asking for it. My boss is asking for it, the board is asking for it.
Aaron: I need to figure out a strategy and, and like, it's not when it's, it's how, um, it's, sorry. It's not, if it's, it's when, it's how, it's like, what, what could we do?
I need do more yesterday. And so totally different universes. Um, and, and it is, uh. Basically what that means is that you've bypassed all of the normal, like resistance to this happening that normally, you know, happens with one of these big changes. And, and now it's purely in the mode of like, how do we make it happen?
Which, which functions are gonna use it? What, what, what is the impact to, you know, how do we make decisions when an AI generated the response? How much human in the loop do we need? So we've jumped all into just now that mode. Um, uh. Five to seven years faster than we did with cloud. Now, that doesn't mean that the adoption will follow a different trajectory because again, it's really people, it's change [00:31:00] management, it's systems, uh, but the, the reluctance and emotional kind of consternation is very different this time.
Logan: And so what is the resistance when you're going into the organization or how would you describe beyond the enthusiasm and trying to figure out the, the how and the when? Like what are the long polls in the tent? How are on these organizations versus how much are on the commercial vendors?
Aaron: Yeah, I would say actually the, um, maybe also, this might abuse the analogy too much, but if compare Cloud, cloud. What customers were asking for because they were resisting it. So you could have always implemented cloud 50% more than you actually were at the time, or you know, hundreds of percent more maybe.
Actually the reverse is true, which is, I'd say most conversations now, the customers actually like, they're jumping like three steps ahead. Once they see the demo and they're like, oh, that's great. Like, we'll just like totally automate that entire thing. [00:32:00] Well, let's, let's, your data's not yet in the environment that it's gonna need to be to, to do that use case or, you know, you know, we're still waiting on some reasoning model breakthroughs from, you know, open AI or clot or, or something to like, just get that extra 3% of accuracy that you're gonna need.
Um, so in this case, I think customers have, have. Pretty much jumped ahead to way closer to the end state, which is like, yeah, great. Like let's give all the employees a productivity boost. Let's figure out where we can deploy agents to the organization. Um, now, you know, I wanna, I wanna do a massive asterisk.
I'm talking to the people that are talking to us, which means that like by definition, they're leaning in, they're a little bit more forward thinking. So some of, some of the anecdotes I'm sharing are from those types of customers. I think there's, there's still an 80% of customers that are, that are like, whoa, like my head's exploding.
Where should I start? Where should I go first? But there's a lot of, there are a lot of customers where they're now ahead of the technology based on the set of use cases that they have.
Aaron: Um, [00:33:00] and, uh, and so I wouldn't say it's, it's, it's like the hurdles unfortunately, are, are much more. Like practical considerations.
Like literally, like I can't just, you know, turn on an AI thing that looks at all my data, and then an employee can ask a question because guess what? It's all of a sudden gonna reveal, you know, secrets in the organization that they shouldn't have had access to, but there were overprovision for access like five years ago.
Um, or, hey, you know. AI agents are really good at, at doing code, uh, you know, auto completion. But like, am I ready to put an AI agent in production for like generating a full, you know, multi file, you know, um, you know, output. And then, and then how do I, how do I review that when the human actually wasn't along the ride for the, for the actual, you know, editing process?
Are they gonna really understand the code? Are they allowed to press submit? And that goes into production. That's gonna be are big questions that people are dealing with. I mean, imagine. Imagine how [00:34:00] unhinged. Where we're like, oh man, we just made a video game in three, you know, three hours. And it's like, and it does all this stuff.
And then you're like, you're an, you're an engineer in like a banking system. Like are you supposed to go implement that right now? Like, like, is that, like, are we ready for, like, are we ready for like your bank account to be, you know, driven by vibe coding? Like, we don't know. So, so these are, these are the kind of things that I think every enterprise is, is trying to figure out, like, okay, so what, what page should I implement this stuff?
Is my data ready? Are my systems ready? A lot of it is about, wow, the industry is moving so fast. How do I make, um, ironically the industry is moving so fast. How do I make decisions right now that I'm gonna be stuck with in a fast page, in a fast changing space? So a lot of it is actually like, can you get your architecture set up for like optionality in the future? Because maybe the vendor you work with today is not gonna be the one that's the best in the future. You need to swap out different parts. These are the kind of conversations that we have.
Logan: To that end, you, you [00:35:00] cautioned. Um, I, I forget if it was a blog post or a tweet or something, but that, um, being early to AI in some ways can be a liability if you don't set up the architecture with the right abstractions in some ways. Uh, I guess as you think about. The applicability of this for box as a, as a organization, and therefore the advice you might have for, for others.
Um, how, how do you think about being flexible, uh, with this pace of change? And is it more about being model agnostic or modular and elements of the design? What would you share on that?
Aaron: Yeah. All, all the above. So, so there's how we build and then how we even implement ai Internally. How we build is, is basically we try and figure out the parts of the stack that there's probably a few different categories and, and I'm, I'm making up some of them, maybe the nomenclature on the fly, but like there'll be parts of the stack that just like will never change.
You know, a model breakthrough will happen and it's, it doesn't matter. Like we're not gonna [00:36:00] change that particular interface, you know, with our system so that you can just like lock and load and start executing. There's other parts where, where you're like. Um, you know, we, we think there's a best in class technology now.
You know, maybe we're kinda like, we're shaky on it. Um, you know, we probably prefer there to be an open source version. So we, so so that's a, that's like a, a spot where you're, you know, you, you probably need to be, you, you really thoughtful about, you know, how much do you lean into a particular, you know, part of the architecture?
Um, I'm thinking like. You know, your, your, you know, your vector search, uh, your, you know, your rag, uh, you know, kind of, you know, execution. You know, these are spaces where there's a bunch of first movers, you know, people started adopting them. It's not, it's not exactly clear where it all ends up in five years from now.
So you kind of wanna be, you know, thoughtful about that. And then there's the, like, the final part on.
This space is changing every three [00:37:00] minutes, you, you, you need to decide, do you want, do you want like maximum flexibility? So these are the model providers. Do you wanna be able to have a way of, of, you know, swapping in different parts? There's multiple parts of box where we'll use multiple model providers for the same workflow.
Um, uh, and so you just need to then have like a high degree of optionality, a high degree of flexibility in that part of the stack. And, um, uh, and so that, that, you know, so you, and then you have different like abstraction layers between, between those, those different interfaces. But I think that's, I think basically anybody in AI at this point has been building with that pattern.
You know, you go to perplexity, you change your model, you go to cursor, you change your model. Um, I, I think you can, I think we're all realizing like, okay. The user changed their model, or at least you as the developer have some, some degree of optionality there because we know how fast that space moves.
And then you have to decide like, which other parts of the stack you just kind of, you know, lock in on and, and don't worry about again.
Logan: Uh, Klarna, uh, Sebastian, who, [00:38:00] uh, we've had on before, and a good, uh, a good guy, a great CEO has been. Um, I, I think he, he tweeted something recently. I think he accidentally, uh, made news, uh, in his telling of like consolidating SaaS vendors and moving more things to agentic workflow and I'll. Tell you, uh, it set off a little bit of a firestorm internally to see if this is like a, you know, a long, we need to be existentially worried about investing in software because, um.
There was going agent workflows and systems were just gonna do away with the, the need for the interface and therefore everything could be in a CRUD database in, in, some ways. Um, I'm curious your perspective on, on that as we zoom out and maybe is, is what they're doing you think, broadly applicable to other organizations?
Or is this a one of one maybe exceptional case that they're testing something and on the early frontier?
Aaron: Yeah, I, I mean, uh, for, for, I [00:39:00] found it very fascinating. Um, and I, I, I enjoyed, uh, that, that it was happening and I, I, I mean, I enjoyed that it's still theoretically happening. I think some of his updates were that, like they didn't actually do as much on the customer support side, or they have some more, you know, kind of premium agent, uh, you know, human agents that they now have.
Um, but, but so I, I thought it was great from a proc standpoint, I think we do need people testing the boundaries. Um. Um, I think it's the kind of thing, which is it, it didn't cause me to rethink all of SaaS because, because the, the, you know, uh, uh, I, I don't remember the first time I heard this, but maybe, maybe it was like 15 years ago or something.
I was, uh, I was meeting with, um, an it, you know, leader at Tesla. And they were like building, I'm gonna make up 80% of the story, but they were building their own ERP system and I was like, holy crap, I cannot believe that, that you're doing this. And like, what? Like maybe this is the future. And then you like quickly wake up that like, okay, like no, like, like, you know, [00:40:00] this is sort of an Elon special where he needed to control end-to-end the whole thing and they needed their own, you know, didn't want any of the off the shelf stuff, amount of fortitude.
Just, you know, sheer entrepreneurialness, you, you need to have to even wanna attempt that is like kind of insane. And so I think the Klar example is more in that category. Um, uh, like, it's just like if you go to the average company and you're like, do you want to be responsible for running your own HR system?
Like the average enterprise is, is not gonna say yes to that. And no matter how. Something is you, it does mean now you're responsible for the thing like, like this, the, the, the AI provider that has made, the agent that wrote your, you know, your HR system is not taking responsibility for the ar HR system code that was generated.
So that means by definition, you, the customer are now responsible for whatever the thing is that it produced. It's not gonna be the AI models. You know, Claude is not, Andro is not, uh, taking any [00:41:00] responsibility for how your HR system works. And if it leaks data and you get sued by the EU because there's some kind of, you know, data privacy thing, like that's not Claude.
Like, uh, that's Workday. Workday will do that. Not, not anthropic. So I think you'd have to be like, you know, which parts of software are people so unhappy about that they would do themselves, or which parts of software just like are so needed to be customized. Um, or which parts of software are so unstrategic that you don't mind whipping up these, like micro apps for like one off things.
And so, so you don't, you, you, you're fine that it is built by this, you know, an, an agent that you can just deploy quickly and, and turn it off. And it's kind of, you know, it's a three month, you know, it needs to live in the world for only three months. Those for the foreseeable future, next two years, let's say.
That would be the only thing I think could be disrupted. But I don't think your HR system is going away. I don't think your CRM system is going away. Um, I don't think that really changes to your point about deflationary impact.
Aaron: I do think that there's a chance that you [00:42:00] could see, you know, you could see a new crop of like, I'm gonna just like, like yolo it and build like the AI a, you know, uh, pow.
Like, like, I'm gonna build Oracle. With AI agents and just like give it a shot, but that would still be a software company that emerges from that. And you're, you're really using AI purely to just, you know, add engineering augmentation. I don't think the customer wants to be responsible for that software at the end of.
Logan: That was my conclusion as well, I think is good for people investing in software companies. Um.
Aaron: Yeah. Now, now big asterisk, like in 10 years, do I wanna, you know, be on the record as saying this? Like, let's see, but like, but like, just, I just don't think companies wanna be responsible for every problem in the world. Like, like you want, like, uh, the, the, um, uh, Jeffrey Moore came up with this idea, I think it was Jeffrey Moore that came up with it, but, but credit and.
Context is just like shit that like, like you just want done for you. [00:43:00] You don't want, you don't wanna worry about it. Core obviously being like, this is like existential to my business and for most people, like, like their core business is not their HR system. That, and so thus you hire other people to do your HR system and so that's why you don't really want, you know, you don't, you don't want your internal IT team being responsible for rebuilding something that you can get off the shelf and just, it's done for you.
Logan: That's the way I feel whenever technical founders wanna reinvent sales comp. Uh, and it's just like you're not, it's hard enough to differentiate on one vector. You, you know, this isn't gonna be the make break. Yeah. We, we, you know, we can go through all the books and talk about Alright.
Logan: I'm gonna give you some quick ones and you give your, uh, your gut reaction on some of these things.
Reasonable period of time, maybe five to 10 years decreases the number of hours worked by your average knowledge worker.
Aaron: Well, how many years?
Logan: Five to 10.
Aaron: I'm gonna, [00:44:00] I'm gonna stick with no for now.
Logan: Yeah, I think I, I think I agree. Um, do you think AI will flatten org structures in any meaningful way? Example, if fewer middle managers, faster approval, more self-service knowledge work.
Aaron: thi this one I could buy. Um, I, I could buy, I think actually layers have gotten pretty crazy over the past couple decades. Um, so, so I, I think I, I think, yes. Um, uh, I think yes. And, and, and part of it is because I think that. AI kind of lets you, it, it, it lets functions. My latest kind of conclusion is that it kind of lets functions sometimes do their adjacent function.
So it lets the designer write the ai, the front end code. It lets the backend developer write the front end, you know, code. It lets the copywriter, you know, generate the full white paper. So, so I think what that means is that you will have. Have sort of more full stack workers, almost [00:45:00] like, like we don't really think about this outside of engineering, but I think you'll probably, I think we've gotten like maybe overly disciplined in, in job functions, in some parts of, of, uh, of the economy.
And I think AI kinda lets you have a little bit more expansive responsibility. So then that probably means that that's a little bit of a flattening when you don't have to hop through and around as many orgs to get things done. So, but we'll see.
Logan: Did you ever read the David Epstein book range? Why Generalist Triumph?
Aaron: No, uh,
would it to this?
Logan: Yeah, kind of. It sort of talks about like they give the example of Roger Federer growing up and playing multiple sports versus Tiger being specialized in one and how many breakthroughs in society have occurred from people outside of a core field or domain, leveraging the knowledge they have in that and going into another one.
And so it goes through scientific breakthroughs or company foundings or all this where when you get too verticalized in some ways, you just don't have the. Of the way other people do things, and so
Aaron: Yeah,
Logan: new [00:46:00] breakthroughs come outside.
Aaron: I think, I think work has probably gotten like a little bit too, Adam. And we've, we've like, you know, we're all just doing our little part of the paperclip factory or whatever the, you know, or the needle, I think it was needle factory or
whatever the,
Logan: we've all, Andrew, what is it? Aaron, Ross, or whatever, predictable revenue. We've done that across every element of society at
Aaron: yeah, that's really Yeah. So, so, so I think, I think probably AI relieves us a little bit of that, um, like division of labor maybe. Maybe got just a little bit too granular over, over time. And so I think maybe we get, we get a little bit of a chance of a reset, which is like, no, I can like just throw a little bit more of a general problem at somebody and they've got the ability to kind of go and, and fully solve it.
Which, which frankly I think is a huge, hugely exciting thing and much more fulfilling probably for, for the knowledge worker kind of class. Um, uh, that, that I think has a lot of upside.
Logan: What's the coolest, uh, [00:47:00] or most mind blowing demo you've seen in the last call it three to six months? Where, where you were like, wow, I can't believe that's actually possible that we do this.
Aaron: Yeah. Um, I'm having a, a funny journey because like, it's, it's like you do, like build suchit.
Logan: It's like the hedonic treadmill in some ways, right? You're just like, I, there's the, I, I always reference now with ai, the Louis CK bit about like, uh, airplane and wifi, and you're just like, listen, you're flying through the sky, like. And you have access to the internet. And that's the way I feel about a lot of the AI things where, know, deep research, like elucidates a little bit and I'm like, oh, this is bullshit.
Like,
exactly. Was that 15 hours of work you would've done otherwise prepping for this interview? Like, yes. You know.
Aaron: Yeah. Um, so, so it's, it's tough because I, like, I, I do wish I [00:48:00] could, I wish I could have my pre-chat bt like wiring for every individual breakthrough because I think I would just be like, holy shit, this thing is witchcraft, like every day. Um, like I remember, I remember like yelling at a friend who was like deep in lms, like, you know, a and I was like, I was understand how did it.
The business strategy on a thing, it's never scanned from the internet. And, and like, I was just like obviously an idiot at that point on, on, on all this stuff. And, um, and like, but I, I want that, I want that like, like energy again where I'm just like surprised by everything. So, um, so I, you know, I'm, I'm, I'm like, I think from a pure utilitarian standpoint, I still love the, you.
I try and force myself as many times as possible to be like, okay, deep research exists. Deep research exists like, like, don't, don't do the thing, right? You open seven tabs and you go through all this stuff. It's just like, [00:49:00] just send the query. Go back to it. I had a, I had a thing 48 hours ago. It saved me an hour and a half.
Um, and, um, and like I, but I had to catch myself because I was like about to just go to do my research, my normal, like, you know, I've been doing this for 20 years, you know, I, I know how to research things. Did deep research, and lo and behold, like as far as I could tell no hallucination, I, I, I checked a couple of the answers.
Fantastic. Because it just like, boom, solved the problem for me. Like, so, you know, I'm now doing that probably five to 10 times a week on something. Which is just an awesome productivity boost. I, um, I still am, um, I'm, I'm, I'm still probably incrementally, uh, surprised by and excited by front end design creation with ai.
So, so, you know, I'll, I'll pop in a screenshot into Cursor and it codes it, or I'll pop in a prompt into, you know, V zero rept and it, it produces something that I would've like had to ping an engineer or a designer about and be like, can you [00:50:00] go work on this? And so like. Like, you know, unfortunately, like my, my browser history kind of looks a little unhinged 'cause it's like 11:30 PM on Saturday.
I'm like on, on V zero doing designs. So, um, uh, so that's, that's, uh, you know, but that, that, that still is pretty exciting I think to me.
Logan: Yeah. Um, what do you think is more likely to be impactful than the enterprise for AI multimodal capabilities like image, text, code voice, or the agent workflows that we kind of talked about earlier.
Aaron: Um, if, if you have perfect, perfectly executing agentic workflows, that's the holy grail. So, um, so, so the, the amount of the, you know, the, my, my. My workshop analogy is like you, you know, in, in data science as an example and anything obviously that you put into a structured database, you know, we, we've had, we have, we've always had this ability to be like, okay, I'm gonna go compute some large set of things, like, go run the [00:51:00] analysis on this customer base in this region to do a thing.
And like, we used to, you know, we, we still do, but we, like, we would talk about it like, okay, I have a job running and.
All forms of unstructured knowledge work. A world where you just have jobs running where you're just like, I just sent these 10 agents to go and like, figure out and synthesize all of this clinical drug trial research to give me insights. Like you're, it's like we, we were only wired to think about sending off jobs and, and compute tasks for like, like 3% of all corporate work.
Now imagine if you do that for the, the other 97%, or at least 80% of the 90%, and you're just like, yeah, like I, I sent out a job to go research this new region I wanna enter as a business. I sent out a job to go and, and, you know, figure out what, what all of my, you know, product feedback was, and then build me a roadmap.
I decide if it's, you know, [00:52:00] validated or right. Like, like it's just gonna be crazy when you can actually do full agent workflows. Um, whether it's deep research or executing a task or, you know, automating a customer experience. That's the, that that continues to probably be the biggest area of upside.
Logan: Hmm, two years, five years, 10 years. Uh, do you think the best model in those periods. Will be closed source, proprietary, or open source? Maybe go through each like in two years. You say closed source or open source?
Aaron: I, I think, um, for, for lack of having any imagination on this, I think that I would mostly be in the camp of. It almost doesn't matter because within three months of the breakthrough, you'll have an open source version. So, so I, I, I, I think it's, it's kind of doesn't matter whether the open source version did it first.
Wow, China did it, or that OpenAI did it, because I think within [00:53:00] three months we can expect that you'll have the open source version of the, of the closed source breakthrough. So you almost it then, so then almost by definition, you can almost still expect then a large amount of the, the, the traffic and the use will be open source because, you know, you're not trading off that much because you'll still get the open, open weight version of, of, of what just emerged.
Logan: What's the most exciting box, uh, AI feature, application? Workflow, uh, that, that you've. Um, that you've rolled out that you think is underappreciated by your existing user base, and everyone should really be thinking more about adopting.
Aaron: Uh, there's, there's two features we're most excited about. The first is, um, we, we have this product called Hubs. And what hubs does is if you go to Box and you go to any file system, you create a folder, you put files in the folder, you share that with people, um, uh, you have this problem, which is, well, what, like, like folders are, are kind of like unintuitive.
For the [00:54:00] recipient because they're like, I don't know, like, which, you know, which third sub-folder down did you put the, you know, the, the sales material that I'm looking for to, to use for my, my, you know, customer pitch? So we created hubs, which is an overlay onto folders. So you can put as many of your files and folders into a hub, and then you can go and just ask questions of the hub.
And so this actually a very, very nuanced, highly important thing, which. If you go to a general purpose, just AI chat window, you have to do a lot of work to guess what, what stuff is on the other end of, of this chat window. If you go to a hub, in our case, you kind of know by, by hub what, what does it have access to?
So we have a sales hub, which means by definition, that has all the sales information. We have an HR hub, which means that it has all your HR information. So you go to the HR hub and you ask the HR question, and then we're basically doing rag on your documents for all of your HR documents. We're doing rag on on all your sales, you know, you know, uh, [00:55:00] presentations we're doing rag on all your customer support product documentation.
So that basically gives you these micro knowledge portals for every topic that you want. Um, and so that, that's a game changer because, uh, because you just, again, the user instantly knows what topic is gonna be covered in this thing. So dramatically reduce hallucination.
So that's pretty cool. Um, and then the final one that, that is, is just working out kind of way better than we, we had hoped initially, like, you know, a year and a half ago, let's say, uh, is just AI data extraction from documents. Um, it sounds like way too straightforward for anybody listening, but it's insanely powerful if all your job is, is like review contracts, review invoices, pull out data from resumes, you know, standardize my, my financial reporting documentation.
We, we can just now automate, you know, basically most of that.
Logan: That's great.
Logan: Uh, well, Aaron, thanks for doing this. I, uh, I think we covered a lot of ground, so I really appreciate you hanging
Aaron: Thank you.[00:56:00]
Logan: Yeah, thank you. Yeah.