Box ($BOX) CEO & Co-Founder, Aaron Levie on Navigating AI in the New Era of Public Companies

Does fantasy investing count or not?

It was fifth grade in Seattle. And I bought some shares of Pacar. And I think maybe Starbucks. I can't be 100% certain, but those were a couple of the ones I had.

Yes, we did have a fantasy stock kind of program in my school at the time. So that was pretty fun. And then my real memory though, my earliest formative stock market memory was somehow, I don't know if I can fully blame my dad but I think I will. I invested most of my bar mitzvah money in a cable stock in like, it must have been 1997 or '98.

It was one of the E*Trades or actually maybe I don't remember because I had a neighbor that was a stock broker so I may have done it through them. But anyway, so I put almost all of the money, like $2,000, in this cable modem company and I think they went bankrupt between six and 18 months later. So, completely evaporated all of my net worth. And surprisingly that didn't cause me to not like the market, because I was day trading in college, but definitely a very early formative experience.

What they don't tell you when you're going public is because in Silicon Valley we call it an exit. But actually if anything it's an entrance into just an entirely new universe. There's nothing exity about it. And it's funny because the general view is when you build this business and there's some kind of moment where you go public and you've made it and you've done it. That moment very quickly turns into reality when you actually go public because basically the next Monday morning you're just now a public company who has to now interact with 10 to 100 times more shareholders than you previously did. And so there's very limited amounts of a moment where you take a beat and you're like, 'Wow, I've made it.' You're just now back in the grind instantaneously. And so it's funny because I think investors propagate this concept of it's an exit and you've made it and it's this great milestone, but then when you're on the founder side, you're like, 'Oh my god, my life has only started now because now I have this entire new journey.' And your VCs are really happy because they obviously have a good return or whatever, but now you've launched into this new world and you're back on that treadmill with a new shareholder base with obviously public market dynamics which is always fun and crazy but we've enjoyed every minute of it. We've had our ups and downs in the public market, but I'm generally a fan of being public.

Yeah, I mean I've been fortunate where I've had a set of kind of advisers, mentors, people that I've gotten to get advice from over the years. So at least a theoretical level, I was kind of prepared for being public. In practice, it's always different because when you're actually in the line of fire, it's a very different dynamic than what you think. But I had tried to prepare for it. We actually were running the company for the couple years prior to being public as much like we could as a public company as possible, but like we were doing internal earnings calls. We were kind of preparing answering questions from investors. So a lot of that style we had already built in. So a lot of the earnings call process I think we had kind of well covered. The new reality was just again simply that x number of hours per day your stock is going to be trading and there's always a new investor you have to communicate with and convince and explain the business to. Which again is this sort of ironic thing because you spend more time actually with investors as a public company than you do as a private company. Which is this sort of counterintuitive thing to people that are on the private side. You're like, 'Oh man, fundraising is so hard and you have to go to Sand Hill Road and pitch these 20 VCs.' It's like, 'Dude, that's like a week of your life as a public company. You are always doing that.'

Yeah. I think basically people's view of it as an asset or a liability is all dependent on what I'm tweeting. And so sometimes I'll dip into the political realm and then all of a sudden people think it's a liability. And obviously if I'm talking about the future of AI and software, then maybe it's an asset. But I just kind of treat it as a personal platform for communicating the things that are top of mind for me. So, it's obviously quasi corporate just by virtue of me representing the company. But I do sort of treat it as these are just the things that I want to communicate and it's usually just stuff that's top of mind for me. And so I generally would advise investors to mostly treat it as my personal platform as opposed to this is sort of 100% representing Box's views and all of our kind of corporate messaging.

Yeah. So with Box, when we started the business the original idea was we wanted to be the simplest, easiest, most secure way to access and share your data, your corporate files, your research documents, your marketing assets, your financial records, anything that is a sort of enterprise content. How do we make it really easy to manage that information and access it from anywhere? And over the years, obviously, as we've grown and we store more data, we serve more customers, we've always had this burning kind of question or problem, which is there's so much value that's trapped inside of your enterprise information. What if you could actually tap into it and unlock it? And there was never really a technology that could let you do that prior to generative AI and the kind of current AI moment that we're in. And so, we jumped on the latest wave of LLMs because we've been for many many years thinking about this idea around how do you unlock all of this value that is stuck inside of your enterprise information. Think about all of your contracts and what if you could know every single thing about all of your customers, what they're paying, what the renewal dates are, what you're liable for, all that information. Imagine you could just ask all of your contracts questions. What if you could take all of your research data and be able to understand critical insights on a new project you're working on. Let's say you're in life sciences or your product roadmap if you're in an engineering organization. So all of our data is trapped inside of this enterprise content. And so with AI agents we now can actually unlock this information. And so we've been working for the past couple of years of building an AI platform that lets you really answer any question you want from your documents, be able to extract any amount of intelligence from this information and be able to automate any workflow with your enterprise content. And so this for us is a completely new era or chapter of the company because of how much we can go and transform how you're working on this information. And so just imagine if you had a hundred times more capacity than you currently have as a person to go and do anything you want with your enterprise information. What questions would you ask your data? What information and insights would you extract from it? What would you automate in terms of your workflows? And that's the platform that we're building out now with our Box AI platform. So this is really leading to a completely new era of the company. We have a new set of features that we're selling our customers into that offers this functionality. It's our Enterprise Advanced plan. We also have the ability to help drive more and more agent use cases on Box and so we charge more of a usage-based or consumption model for doing that. So there's an all new way that we're selling, an all new way that we're building our product and a new monetization and commercial model as a result of AI agents. And the really cool thing is that it's all built on top of the foundation that we've already created. So for us, we get to actually build on our strengths and do even more for customers. And so it's a little bit less of a pivot and more of just an expansion of our ambitions and of our strategy.

Yeah. So, internally at Box, we obviously use our AI product for quite a bit of our daily sort of knowledge work. So if you have any questions that you want to get answered from your data, you want to summarize information, you want to ask, if you're a sales rep, you want to ask a question about all of the sales materials in the company, the really powerful thing is now you can tap into all of the expertise that our company has ever produced. And so one of the amazing things about AI is, as an organization, you have this classic problem, which is over time, you build more and more tribal knowledge, but as new people come into the company, they can't find that tribal knowledge. The new sales rep doesn't know what sales pitch worked exactly the right way for this customer in this very particular industry. So what you do is you have to then rely on going around to people and asking those questions and then you have to know the right people to go ask. So most companies don't get actually better over time with their operations or their processes the more data they have or the more people they have because any new person coming into the company can't tap into that wealth of information or knowledge. So with Box AI, we let all of our employees be able to do that. So you're a new sales rep, you want to pitch a customer, you can ask a question of all of your sales information. If you're an engineer and you want to ask a question about a product that's being worked on, you can instantly ask a question of all of the data around our product specs or architectures. If you're in legal or compliance, you can look through our contracts or governance documents to be able to help accelerate our process. So what we really want to think about it as is how do we accelerate knowledge work based on the information and insights that's already inside of your enterprise data and that's what we are enabling with our AI agents. So at Box we've been able to accelerate our work and automate more and more of our processes as a result of that. That's our Box AI usage. And then we also use a few other tools for things like engineers using things like Claude Code or Cursor to help accelerate the coding process. We have other technologies that we use for things like customer support. But we want to ensure that Box is being run as an AI-first company. We want to use AI throughout the organization to be able to really drive as much execution and acceleration as possible.

I think right now we're in a moment where there is a very wide gulf between people that have adopted these tools and are using them to the fullest extent and those that maybe are casually asking a question of ChatGPT or something and just messing around in their personal life. Interestingly, I don't think it's that hard for people in that latter category to jump to the former category, but you do have to be motivated. You have to want to play with these tools. You have to want to go explore technology and what's possible. So, we certainly by default try and look for people that have that aptitude, that are curious. They're playing with these tools in their personal life. They're really pushing them to their limits. I think we're in a period right now where we're willing to bet on people that are motivated to join a company and try this stuff out and come with us on that journey. I think in two or three years from now it will be much more of a requirement of if you aren't doing that then something's gone horribly wrong with your ability to adapt to just the modern way of working. It would be like interviewing somebody in 1998 and they didn't have an email account or something. At some point you just have to get with the program. But right now, we're in this unique period where I think as long as you're excited to learn new things, curious, able to adapt quickly, and motivated to do that, then we're willing to hire you because you'll jump right in. You'll learn these technologies. But I do think that the way to set yourself up and Mark Cuban's talked about this a bit and so have some other folks. I think it's an incredible time right now. If you are one of these very ambitious, highly resourceful, entrepreneurial people, go really really deep with AI agents. Whether that's coding or whatever your job function is, go really deep with these agents because when you go to a company, you will look like you're basically a time traveler who's just lived in the future in 5 years from now and you can go and bring in these new ways of working into any organization. I think that's a huge asset. I don't see enough people exploiting that. And I would highly recommend that folks jump on this. The great thing about X right now is all of the conversation is out there. There are no secrets in AI. You can learn this stuff. If you are motivated, you can learn this stuff in a couple of weeks. And it's really about finding the right pockets of information and being motivated to go and again play with these technologies. So, that's who we're looking for in an ideal world. But I think that's going to be true of any organization. And so there's just an incredible moment right now. If you are one of these individuals, you can really go and change how a company works, whether that's a small business or large enterprise, and that will put you in the best position possible.