# Enterprise AI Agents: Infrastructure, Security, and Workflow Transformation

**Podcast:** Latent Space: The AI Engineer Podcast
**Published:** 2026-03-05

## Transcript

Like you don't write code, you talk to an agent and it goes and does it for you, and you maybe at best review it.
That's even probably like like largely not even what you're doing.
What's happening is we are changing our work to make the agents effective in that model.
The agent didn't really adapt to how we work.
We basically adapted to how the agent works.
All of the economy has to go through that exact same evolution.
Right now it's a huge asset and an advantage for the teams that do it early and then are kind of wired into doing this because you'll see compounding returns.
But that's just gonna take a while for most companies to actually go and get this deployed.
Welcome to the Late and Space Pod.
We're back in the Chroma studio with uh Chroma CEO Jeff Hoover.
Welcome.
Returning guest, but now guest host.
It's a pleasure.
Wow, how'd you get upgraded to uh to that?
Because you he is like the perfect guy to be guest host for you.
That makes sense, actually.
We love context.
We we both really love context.
We really do.
Uh and we're here with uh Aaron Levy.
You're welcome.
Thank you.
Good to uh good to be here.
Uh yeah, uh so we've all met offline and like chatted a little bit, but like it's always nice to get these things in person in conversation.
You just started off with so much energy, you're you're super super excited about agents.
I love agents.
Yeah.
Open claw just got by got bought by open AI.
Well, no, not bought, but you know, you know what I mean.
Some some uh you know, aqua hire executive hire executive hire.
Executive hire.
Hey, that's my term.
Okay.
Um what are you pounding the table on on agents?
You have so many insightful tweets.
Well, the thing that that we get super excited about that I think is probably, you know, should be relatively obvious is we've we've built a platform to help enterprises manage their files and their their corporate files and the permissions of who has access to those files and the sharing and collaboration of those files.
And all those files contain really, really important information for the enterprise.
It might have your contracts, it might have your research materials, might have marketing information, might have your memos, all that data.
Obviously, has pr you know predominantly been used by humans, but there's been one really interesting problem, which is that you know, humans only really work with their files during an active engagement with them, and then they kind of go away and you don't really see them for a long time.
And all of a sudden, uh, with the power of AI and AI agents, all of that data becomes extremely relevant as this ongoing source of answers to new questions, of data that will transform into something else that that produces value in your organization.
It contains the answer to the new employee that's onboarding that needs to ramp up on a project.
Um, it contains the answer to the right thing to sell a customer when you're having a conversation to them with them, contains the roadmap information that's going to produce the next feature.
So all that data that previously we've been just sort of storing and and you know, occasionally forgetting about because we're only working on the new active stuff.
All of that information becomes valuable to the enterprise.
And it's gonna become extremely valuable to end users because now they can have agents go find what they're looking for and produce new new value and new data on that information, and it's gonna become incredibly valuable to agents because agents can roam around and do a bunch of work and they're gonna need access to that data as well.
And um, and you know, sometimes that will be an agent that is sort of working on behalf of uh of you and and effectively as you as and and they are kind of accessing all of the same information that you have access to and and operating as you in the system.
And then sometimes there's gonna be agents that are just effectively autonomous and kind of run on their own, and and you're gonna collaborate and work with them kind of like you did another person, open claw being the most recent and maybe first real sort of you know, kind of you know, updating everybody's you know views of this landscape version of of what that could look like, which is okay, I have an agent, it's on its own system, it's on its own computer, it has access to its own tools.
I probably don't give it access to my entire life.
I probably communicate with it like I would an assistant or a colleague, and then it it sort of has this sandbox environment.
So all of that has massive implications for a platform that managed enterprise data.
We think it's gonna just transform how we work with all of the enterprise content that we work with, and we have just had to make sure we're building the right platform to support that.
The sort of shorthand I put it is as people build agents, everybody's just realizing that every agent needs a box.
Yes, and it's nice to be called box and just give everyone a box.
If I you know, if we can make that go viral, uh like I think that that terminology, every agent needs a box.
Every agent needs a box.
If we can make that the headline of this, I'm fine with this.
Yeah, exactly.
Every agent needs a box.
Um, I like it.
Can we shift this?
Like, okay.
Uh, my work here is done, and I got the value I needed out of this podcast.
Yeah.
But but um, but but you know, so the thing that we we kind of think about is um is you know, whether you think the number's 10x or 100x or whatever the number is, we're gonna have some order of magnitude more agents than people.
That's inevitable.
It has to happen.
So then the question is, what is the infrastructure that's needed to make all those agents effective in the enterprise?
Make sure that they are well governed, make sure they're only doing safe things on your information, make sure that they're not getting exposed to data that they shouldn't have access to.
There's gonna be just incredibly spectacularly crazy security incidents that will happen with agents because you'll prompt inject an agent and sort of find your way through the CRM system and pull out data that you shouldn't have access to.
So I mean, it's just gonna happen all over the place, right?
So so then the thing is is how do you make sure you have the right security, the permissions, the access controls, the data governance.
Um, we actually don't yet exactly know in many cases how we're gonna regulate some of these agents, right?
If you think about an agent in financial services, does it have the exact same financial sort of uh requirements that a human did, or is it is the risk fully on the human that was interacting or created the agent?
All open questions.
But no matter what, there's gonna need to be a layer that manages the data they have access to, the workflows that they're involved in, pulling up data from multiple systems.
This is the new infrastructure opportunity in the era of agents.
You have a piece on agent identities, which I think was today, um, which I think a lot of people are.
Breaking news.
The security, security people are talking about, right?
Like we basically I always think of this as like, well, you need the human you, and then there's you need the agent you.
Yes.
And uh, well, I it I don't know if it's that simple, but is the box going to have an opinion on that, or you're just gonna be like, well, we we're just uh sort of this the story layer.
Yeah, let's Okta of serial handle that.
I think we're gonna have an opinion and we will work with generally wherever the contours of the market end up.
Um, and the reason that we're gonna have an opinion more than other topics, probably is because one of the biggest use cases for why your agent might need it an identity is for file system access.
So thus we have to kind of think about this pretty deeply.
And I think uh unless you're like in our world thinking about this particular problem all day long, it might be, you know, like, why is this such a big deal?
And the reason why it's a really big deal is because sometimes sort of say, well, just give the agent an account on the system and it just treats treat it like every other type of user on the system.
The problem is is that I, as Aaron, don't really have any responsibility over anybody else's box account in our organization.
I can't see the box account of any other employee that I work with.
I am not liable for anything that they do.
And they have I have I have you know strict privacy requirements on everything that they are able to, you know, that that they work on.
Agents don't have that, you know, don't have the those properties.
The person who creates the agent probably is gonna for the foreseeable future take on a lot of the liability of what that agent does.
That agent doesn't deserve any privacy because because it's you know, uh it can't fully be autonomously operated and it doesn't have any legal, you know, kind of you know, responsibility.
So thus you can't just be like, oh, well, I'll just create a bunch of accounts and then I'll I'll kind of work with that agent and I'll talk to it occasionally.
Like you need oversight of that.
And so then the question is: how do you have a world where the agent sometimes you have oversight of, but what if that agent goes and works with other people and that person uh over there is collaborating with the agent on something?
You shouldn't have access to what they're doing.
So we have all of these new boundaries that we're gonna have to figure out of, you know, it's really really easy.
So far, we've been in in easy mode.
We've hit the easy button with AI, which is the agent just is you, and when you're in claude code and you're in cursor and you're in codec, you're just the agent is you.
It you're offing in to your services, it can do everything you can do.
That's the easy mode.
The hard mode is agents are kind of running on their own, people checking with them occasionally, they're doing things autonomously.
How do you give them access to resources in the enterprise and not dramatically increase the security risk and the risk that you might expose the wrong thing to somebody?
These are all the new problems that we have to get solved.
I like the identity layer and and identity vendors as being a solution to that, but we'll we'll need some opinions as well because so many of the use cases are these collaborative file system use cases, which is how do I give it an agent a subset of my data and give it its own workspace as well, because it's gonna need to store off its own information that would be relevant for it.
And how do I have the right oversight into that?
One thing which um I think is kind of reason to think about it's like you know, how humans work, right?
Like I may not also just like give you access to the whole file.
I might like sit next to you and like scroll to this like one part of the file and just show you that like one part and like you know partial file access.
Well, I'm just saying, I think like our like our back does seem to be dead, right?
Like you want to say something is dead, I'll probably R back is dead, and uh like the authority to me seems like incredibly unsolved and unaddressed by like the existing state of like AI vendors.
But yeah, I think um we're I mean, you're taking obviously really to up a limit that we probably need to solve for.
And we built an access control system that was was kind of like you know its own little world for a long time.
And um, and the idea was this it's a many to many collaboration system where I can give you any part of the file system and it's a waterfall model.
So if I give you higher up in the in the in the system, you get everything below.
And that that kind of created immense flexibility because I can kind of point you to any layer in the in the tree, but then you're gonna get access to everything kind of below it.
And that mostly is is working in this in this world, but you do have to manage this issue, which is how do I create an agent that has access to some of my stuff and somebody else's stuff as well.
And which parts do I get to look at as the creator of the agent?
And and these are just brand new problems.
And humans, when there was a human there, that was really easy to do.
Like, like if the three of us were all sharing, there'd be a Venn diagram where we'd have an overlapping set of things we've shared, but then we'd have our own ways that we shared with each other.
But in an agent world, somebody needs to take responsibility for what that agent has access to and what they're working on.
These are like the some of the most probably you know boring problems for 98% of people on on the internet, but they will be the problems that are the difference between can you actually have autonomous agents in an enterprise context that are not leaking your data constantly.
No, like I mean, you know, I run a very, very small company for my conference, and like we already have data sensitivity issues.
Yes, and some of my team members cannot see uh the others.
And like I can't imagine what it's like to run a Fortune 500 and that you have to worry about this.
Uh I'm just kind of curious.
Like you you talked to a lot of like like 70, 80% of your custom uh of the Fortune 500 are your customers.
Yep.
67%.
Very ICC.
Sorry, I'm sorry, yeah, I'm not something.
I'm rounding up.
Yes.
I appreciate the round-up.
For the government to do that.
I'm projecting to the end of the year.
Okay, thank you.
There we go.
You do make it sound like we we what we've got to be honest.
Like we're taking we're taking way too long to get to 80%.
Well, no, I mean, so like how are they approaching it, right?
Because you're you don't have you don't have a final answer yet.
Well, okay, so so the this is actually this is the stark reality that like unfortunately is the kind of like pouring the water on the party a little bit.
Yes.
We all in Silicon Valley are like have the absolute best conditions possible for AI ever.
And I think we all saw the Dwarcash, you know, kind of Dario podcast and this idea of AI coding.
Why is that taken off?
And and we're not yet fully seeing it everywhere else.
Well, like if you just like enumerated the list of properties that AI coding has, and then compared it to other knowledge work, let's just let's just go through a few of them.
Generally speaking, you bring on a new engineer, they have access to a large swath of the code base.
Like there's like very like you just like new engineer comes on, they can just go and find the the the stuff that they they need to work with.
It's a fully text-in, text out, you know, medium.
It's only it's just gonna be text at the end of the day.
So it's like really great from uh from just a uh, you know, kind of what the agent can work with.
Obviously, the models are super trained on that data set.
The labs themselves have a really strong kind of self-reinforcing positive flywheel of why they need to do you know agentic coding deeply so then you get just better tooling better services the actual developers of the AI are daily users of the of the thing that they're we're working in versus like the you know probably there's only like seven Claude co-work legal plug-in users at Anthropic any given day but there's like a couple thousand Claude code and you know users every single day so just like think about which one are they getting more feedback on all day long.
So you just go through this list you have a you know everybody who's a developer by definition is technical so they can go install the latest thing we're all generally online or at least you know kind of the weird ones are and we're all talking to each other sharing best practices like that's like already eight differences versus the rest of the economy every other part of the economy has like like six to seven headwinds relative to that list you go into a company you're a banker in financial services you have access to like a tiny little subset of the total data that's going to be relevant to do your job and you have to start to go and talk to a bunch of people to get the right data to do your job because Sally didn't add you to that deal room you know folder and that the you know the information is actually in a completely different organization that you now have to go in and and sort of run into and it's just like you have this endless list of access controls and security as as you talked about you have a medium which is not it's not just text right you have you have a zoom call that that you're getting all of the requirements from the customer you have a lot of in-person conversations and you're doing in-person sales and like how do you ever digitize all of that information um you know I think a lot of people got upset with this idea that the code base has all the context um that I don't know if you follow you know did you follow some of that conversation that that went viral is like you know it's not that simple that that the code base doesn't have all the knowledge but like it's a lot you're a lot better off than you are with other areas of knowledge work.
Like you we like we like have documentation practices you write specifications those things don't exist for like 80% of work that happens in the enterprise.
That's the divide that we have which is which is AI coding has is just fully you know we've reached escape velocity of how powerful this stuff is and then we're gonna have to find a way to bring that same energy and momentum but to all these other areas of knowledge work where the tools aren't there, the data is not set up to be there, the access controls don't make it that easy the context engineering is an incredibly hard problem because again you have access control challenges you have different data formats you have end users that are going to need to kind of be kind of trained through this as opposed to they're adopting these tools in their free time, that's where the Fortune 500 is.
And so we, I think, you know, have to be prepared as an industry where we're gonna be on a multi-year march to to be able to bring agents to the enterprise for these workflows.
And I think probably the the thing that we've learned most in coding that that the rest of the world is not yet, I think, ready for.
I mean, we're they'll they'll have to be ready for it because it's just going to inevitably happen.
Is I think in coding, what what's interesting is if you think about the practice of coding today versus two years ago, yeah, it's probably the most changed workflow in maybe the history of time from the amount of time it's changed, right?
Yeah, like like has any has any workflow in the entire economy changed that quickly in terms of the amount of change?
I just you know, at least in any knowledge worker workflow.
There's like very rarely been an event where one piece of technology and work practice has so fundamentally, you know, changed change what you do.
Like you don't write code, you talk to an agent and it goes and does it for you, and you maybe at best review it.
And even that's even probably like like largely not even what you're doing.
What's happening is we are changing our work to make the agents effective in that model.
The agent didn't really adapt to how we work.
We basically adapted to how the agent works.
All of the economy has to go through that exact same evolution.
The rest of the economy is gonna have to update its workflows to make agents effective and to give agents the context that they need and to actually figure out what kind of prompting works and to figure out how do you ensure that the agent has the right access to information to be able to execute on its work.
I, you know, this is not the panacea that people were hoping for of the agent drops in, just automates your life.
Like you have to basically re-engineer your workflow to get the most out of agents.
And uh, and that that's just gonna take you know multiple years across the economy.
Right now, it's a huge asset and an advantage for the teams that do it early and then are kind of wired into doing this because you'll see compounding returns.
But that's just going to take a while for most companies to actually go and get this deployed.
I love I love pushing back.
I think that that is what a lot of technology consultants love to hear this sort of state.
First to embrace the AI to get to the promised land you must pay me so much money to adopt the prescribed way of uh conforming to the agents.
Yes.
And I worry that you will be eclipsed by someone else who says no come as you are and we'll meet you where you are and and and what was the thing that went viral a week ago open AI probably uh is hiring FDEs to go into the enterprise and then anthropic is embedded at Goldman Sachs.
So if the labs are having to do this, if the labs have decided that they need to hire FDE and professional services, then I think that's a pretty clear indication that this there's no easy mode of workflow transformation.
So so to your point, I think actually this is a market opportunity for you know new professional services and consulting firms that are like agent pilled and they and they kind of you know go into organizations and they figure out how to re-engineer your workflows to make them more agent ready and get your data into the right format and you know reconstruct your business process so you're you're not doing most of the work.
You're telling agents how to do the work and then you're reviewing it, but I haven't seen the thing that can just drop in and and kind of let you not go through those changes.
I don't know how that kind of sales pitch goes over.
Yeah, you know, you're you're saying things like, well, in my sort of nice, beautiful walled garden, here's here's uh here's this here's this beautiful box account that has everything.
Yes.
And I'm like, well, most most real life is extremely messy and like poorly named and just outdated shit.
100%.
And so 100%.
And so this is actually no, so so this is, I mean, we agree that the getting to the beautiful garden is gonna be tough.
Yeah.
There's also the other end of the spectrum where I can I just like it's a technical impossibility to solve.
The agent is is truly cannot get enough context to make the right decision in in the in the incredibly messy land.
Like there's no AGI that will solve that.
So we're gonna have to kind of land in somewhere in between, which is like we all collectively get better at documentation practices and and having authoritative, relatively up-to-date information and putting it in the right place.
Like agents will will certainly cause us to be much better organized around how we work with our information simply because the severity of the agent pulling the wrong data will be too high.
And the productivity gain of that you'll miss out on by not doing this will be too high as well.
That you that your competition will just do it, and they'll just have higher velocity.
So uh, and and we we see this a lot firsthand.
So we we build a series of agents internally that they can kind of have access to your full box account and go off and you give it a task and it can go find whatever information you're looking for and work with.
And, you know, thank God for the model progress.
But like if if you gave that task to an agent nine months ago, you're just gonna get lots of bogus answers because it's gonna, it's gonna say, hey, here's here are five five, you know, documents that all kind of smell like the right thing.
And I'm gonna, but I but you're you're putting me on the clock because my system prompt says, like, you know, be pretty smart, but also try and respond to the user, and it's gonna respond and it's like it got the wrong document.
And then you do that once or twice as a knowledge worker, and you just never again, you never again.
You're just like done with the system.
Yeah, it doesn't work.
It doesn't work.
And so, you know, Opus 4.6 and Gemini 3.1 Pro and you know, whatever the latest 5.3 GPT will be, like those things are getting better and better, and it's using better judgment.
And this sort of like the all of these updates to the agentic tool and search systems are are we're seeing we're seeing very real progress where the agent kind of can can almost smell something's a little bit fishy when it's getting, you know, we we have this process where we we have it go fan out, do a bunch of searches, pull up a bunch of data, and then it has to sort of do its own ranking of you know what are the right documents that that it should be working with.
And again, like you know, the intelligence level of a model six months ago would it'd be just throwing a dart at like I'm just I'm gonna grab these seven files, and I I pr I hope that that's the the right answer.
And something like an Opus first four five and now four six is like, oh, it's like no, that one doesn't seem right relative to this question because I'm seeing some signal that is making that you know that's contradicting the document where it would normally be in the tree and who should have access.
Like it's doing all that kind of work for you, but like it still doesn't work if you just have a total wasteland of data.
Like it's just not it's just not possible.
Partly because a human wouldn't even be able to do it.
So basically, if a if a really really smart human could not do that task in five or ten minutes for a search retrieval type task, like you know, your agent's not gonna be able to do it any better.
You see this all day long.
So this touches on a thing that just is passionate about, which is context engineering.
I'm just gonna let you ramble or riff on on context engineering if if there's anything like you did really good work on context rot, which has really taken over as like the term that people use and the reference.
100%.
We we all we think about is is the context rot problem.
Yeah, there's certainly a lot of like ranking considerations, genetic search I think is incredibly promising.
Um, I was trying to generate a question though.
I have a question right now, Swix.
Yeah, no, but like I think uh there was this moment um, you know, like I don't know, two years ago before before we knew like where the the gotchas were gonna be in AI, and I think someone was like was like, Well, infinite context windows will just solve all of these problems.
And because you'll just you'll just give the context window like all the data.
And it's just like, okay, I mean, maybe in 2035, like this is a viable solution.
First of all, it would just, it would just simply cost too much.
Like, we just can't give the model like the five thousand documents that might be relevant and it's gonna read them all.
And I've seen it enough to to start believing in crazy stuff.
So like I'm willing to just say sure, like in in 10 years from now.
In in 10 years from now, we'll have infinite context windows at at a thousandth of the price of today.
Like, let's just like believe that that's possible.
But right, we're in reality today.
So today we have a context engineering problem, which is I got I got you know 200,000 tokens that I can work with, or probably I don't even know what the latest graph is before like massive degradation.
Okay, I have 60,000 tokens that I get to work with where I'm gonna get accurate information.
That's not a lot of tokens for a corpus of 10 million documents that a knowledge worker might have across all of the teams and all the projects and all the people they work with.
I have I have 10 million documents, which you know, maybe is times five pages per document or something like that.
I'm at 50 million pages of information, and I have 60,000 tokens.
Like holy shit.
This is like, how do I bridge the 50 million pages of information with you know the couple hundred that I get to work with in that in that token window?
This is like that this is like such an interesting problem.
And that's why actually so much work is actually like just like search systems and the databases and that layer has to just get so locked in.
But models getting better and importantly, knowing when they've done a search, they found the wrong thing, they go back, they check their work, they they find a way to balance sort of appeasing the user versus double checking.
We have this one, we have this one test case where we ask the agent to go find 10 pieces of information.
This is a complex work you vow?
Uh, this is actually not an eval.
This is this is sort of just like we have a bunch of we have a bunch of internal benchmark kind of scenarios every time we we update our agent.
We have one, which is I ask it to find all of our office addresses and I give it the list of 10 offices that we have.
And there's not one document that has this.
Maybe there should be.
That would be a great example of the kind of thing that like maybe over time companies start to, you know, have these sort of like what are the canonical, you know, kind of key areas of knowledge that we need to have.
We don't seem to have this one document that says here are all of our offices.
We have a bunch of documents that have like here's the New York office and whatever.
So you task this agent and you you get you say, I need the addresses for these 10 offices.
Okay.
And by the way, if you do this on any uh, you know, public chat model, the same outcome is gonna happen, but for a different kind of query.
You give it, you say, I need these 10 addresses.
How many times should the agent go and do its search before it decides whether or not there's just no answer to this question?
Often, and especially the the let's say lower tier models, it'll come back and it'll give you six of the 10 addresses.
And it'll, and I'll just say I couldn't find the other four.
It doesn't know what it doesn't know.
It doesn't know what it doesn't know.
Yeah.
So the model is just like, like, when should it stop?
When should it stop doing like should it should it do that task for literally an hour and just keep cranking through?
Maybe I actually made up an office location and it doesn't know that I made it up.
And I didn't even know that I made it up.
Like, should it just keep read should it read every single file in your entire box account until it until it should exhaust every single piece of information?
Expensive.
These are the new problems that we have.
So, you know, something like let's say a new opus model is sort of like, okay, I'm gonna try these two types of queries.
I didn't get exactly what I wanted.
I'm gonna try again.
I'm gonna at some point I'm gonna stop searching because I've determined that that no amount of searching is gonna solve this problem.
I'm just not able to do it.
And that judgment is like a really new thing that the model needs to be able to have.
It's like when should it give up on a task?
Because you just don't, it just can't find the thing.
That's the real world of knowledge work problems.
And this is the stuff that the coding agents don't have to deal with because it just doesn't like like you're not usually asking it about you're you're always creating net new information coming right out of the model for the most part.
Obviously, it has to know about your code base and your specs and your documentation, but but when you deploy an agent on all of your data, now now you have all of these new problems that you're dealing with.
Our uh follow-up research to context rod is actually on a genetic search.
Um and we've like right sort of stress tested like frontier models and their ability to search, um, and they are not actually that good at searching.
Right.
So you're sort of highlighting this like explore exploit.
Debbie Dunner is like everything doesn't work, like well, somebody has to be.
Um can I throw out one more thing that is different from coding and and the rest of the knowledge work that I I failed to mention?
So one other kind of key point is is that you know, at the end of the day, whether you believe we're in a slot pocalypse or or whatever, at the end of the day, if you if you build a working product at the end of if you if you've built a working solution, that is ultimately what the customer is paying for.
Like whether I have a lot of slop, a little slop or whatever, I'm sure there's lots of code bases we could go into in enterprise software companies where it's like just crazy slop that humans did over a 20 year period.
But the end customer just gets this little interface, they can they can type into it, it does its thing.
Knowledge work uh doesn't have that property.
If I have an AI model go generate a contract, and I generate a contract 20 times, and you know, all 20 times, it's just three percent different.
And like that, I that that kind of slop introduces all new kinds of risk for my organization that the code version of that slop didn't didn't introduce.
These are and so like, so how do you constrain these models to just the part that you want them to work on and just do the thing that you want them to do?
And and you know, in engineering, we don't you can't be disbarred as an engineer, but you could be disbarred as a lawyer, like you can do the wrong medical thing in healthcare.
You there's no there's no equivalent to that of engineering.
Do you want there to be?
Because I've considered software.
Oh, is that civil engineering there is, right?
Civil engineering sure, oh yeah, for sure.
But like in any of our companies, you like you know, you'll be forgiven if you took down the site and and we will do a rollback and you'll you'll be in a meeting, but you have not been disbarred as an engineer.
We don't we don't change your you know your computer science uh postmortum.
Yeah, exactly, exactly.
So so uh now maybe we collectively as an industry need to figure out like what are you liable for, not legally, but like in a in a management sense uh of these agents, all sorts of interesting problems that that that uh that have to come out.
But in knowledge work, that's the real hostile environments that we're operating in.
I do think like uh a lot of the last year's 2025 story was the rise of coding agents, and I think 2026 story is definitely knowledge work.
Yes, 100%, right?
Like that with and I think open cloud cloak are just the beginnings.
Yeah, like it's the next thing's gonna just gonna be absolute craziness.
It it is, and and uh and it's gonna be, I mean, again, like this is gonna be this this wave where we we are gonna try and bring as many of the practices from coding because that that will clearly be the forefront, which is tell an agent to go do something and has an access to a set of resources, you need to be responsible for reviewing it at the end of the process.
That to me is the is the kind of template that I just think goes across knowledge work.
And Claud Cowork is a great example, open cloud's a great example.
You know, you can kind of sort of see what codecs could become over time.
These are some really interesting kind of platforms that are emerging.
Okay, um, I wanted to uh we touched on evals a little bit.
You had you had the report that you're gonna go bring up, and then I was gonna go into like uh boxes evals.
But uh go ahead, talk talk about your search thing.
Yeah, mostly I think kind of a few of the insights, it's like everyone, frontier model is not good at search, humans have this natural explore exploit tradeoff where we kind of understand like when to stop doing something.
Also, humans are pretty good at like forgetting actually, like pruning their own context, whereas agents are not, and actually an agent in their kind of context history.
If they knew something was bad, and they even you can see in the trace, the reason you trace hey, that probably wasn't a good idea.
If it's still in the trace, still in the context, they'll still do it again.
Uh huh.
Uh, and so like I think pruning is also gonna be like really it's already becoming a thing, right?
But like letting models like we self-prune the context windows, yeah.
So so don't leave the mistake, don't leave the mistake in there, cut out the mistake, but tell it that you made a mistake in the past and so it doesn't repeat it.
Yeah but like cut it out so it doesn't get like distracted by it again because really you know what is so it will repeat its mistake just because it's been it's in the context it's in the context so much example even if it goes it's like oh this is a great thing to go try even if it doesn't work.
Yeah exactly so it's like a bunch of stuff just groundhog's day inside these models I'm gonna go keep doing the same wrong thing that makes sense right like you know it's some crude analogy you're trying to like fit a manifold in latent space which kind of is doing break program synthesis which kind of one we think about like what Long we're doing right like you know certain facts might be like sort of overly pitting it for certain you know sectors of latent space and so like plug clean space yeah and uh and so uh our editor adds a bill every time you say that so you have to make we have to like remove those like a gong like C VPN or something if we gone you have to remove those links to like kind of give it the freedom kind of do what you need to do.
So but yeah we'll release more soon that's awesome.
Yeah that'll that'll be cool.
We're a cerebral podcast that people listen to us and and sort of think really deep.
So yeah we try to keep it subtle.
Okay.
Okay, fine.
Um you you guys you guys do have evals.
You talked about your your office thing, but uh you've been also promoting apex agents and complex work, uh whatever you want wherever you want to take this, just yeah how you uh Apex is is obviously in our course uh uh kind of um agent eval.
We we supported that by sort of opening up some data for them around how we kind of see these um data workspaces in in the you know kind of regular economy.
So how do lawyers have a workspace?
How do investment bankers have a workspace?
What kind of data goes into those?
And so we we partner with them on their their Apex Eval.
Our own um eval is so it's actually relatively straightforward.
We have uh a set of of documents in a and a range of industries.
We give the agent previously did this as a one shot test of just purely the model, and then we just realized we we need to, based on where everything's going, it's just gotta be more agentic.
So now it's a bit more of a test of both our harness and the model, and we have a rubric of a set of things that it has to get right, and we score it.
Um, and you're just seeing you know these incredible jumps in almost every single model in its own family of you know, opus four um, you know, sonnet four six versus sonnet four five.
Yeah, we have this up on screen.
Okay, cool.
So if somebody you're seeing it somewhere, like I I forget the total it was like 15 point jump, I think on the main on the overall.
Yes.
And it's just like you know, these incredible leaps that that are starting to happen.
Um Anthropic doesn't know any like any it's completely held out from Anthocks.
This is not in any, there's no public data, which has you know benefits, and this is just a private eval that we do, and then we just happen to show it to to the world.
So you can't you can't train against it.
And I think it's just as representative of you know, it's obviously reasoning capabilities, what it's doing at you know, kind of test time compute capabilities, thinking levels, all like the context rot issues, so many interesting, you know, kind of uh uh capabilities that are that are now improving.
One sector that you have that's interesting.
Uh people are roughly familiar with healthcare and legal, but you have public sector in there.
Yeah.
Uh what's that like?
What what what is that?
Yeah, and and we actually test against I don't know, maybe 10 industries.
We we end up usually just cutting a few that we think have interesting gains.
So Polictor's one, a lot of like government type documents.
Um what is that?
What is a government type document?
It's like government finance.
Probably not tax returns, it would be more of what would the government be using uh as data.
So um, so think about research, that that type of uh of data sets.
And then we have financial services for things like data rooms and what would be an investment prospectus.
That one you can dog food.
Yeah, exactly.
Yes, yes.
So uh so we we run the models um in now, you know, more of an agent mode, but but still with with kind of limited capacity, and just try and see like on a like for like basis what are the improvements, and again, we just continue to be blown away by how good these models are getting.
Yeah.
I I mean, I think every serious AI company needs something like that, where like, well, this is the work we do.
Here's our company eval.
And if you don't have it, well, you're not a serious AI company.
There's two dimensions, right?
So there's there's like how are the models improving?
And so which model should you either recommend a customer use, which one should you adopt, but then every single day we're making changes to our agents, and you need to know if you regress.
If you know, yeah.
You know, I've been fully convinced that the whole agent observability and eval space is gonna be a massive space.
Um, super excited for what Brain Trust is doing, excited for you know, Langsmith, all the things.
And I think what you're gonna, I mean, this is like every enterprise, like literally every enterprise.
Well, right now it's like the AI companies are the customers of these tools.
Every enterprise will have this.
You'll just have to have an eval of all of your work.
And like we'll you'll have an eval of your RFP generation, you'll have an eval of your sales material creation, you'll have an eval of your uh invoice processing.
And and as you, you know, buy or use new agentic systems, you are gonna need to know like what's the quality of your of your pipeline.
Yeah.
Um, so huge, huge market with agent evals.
Yeah.
And and you know, I'm gonna shout out your your team a bit.
Uh your CTO Ben uh did a great talk with us last year, and he's gonna come back again for World's Fair.
Yep.
Just talk about your team.
Like you know, brag a little bit.
Yeah, I think I I I think people take these eval numbers and pretty short it's it for granted.
But no, there, I mean, there's there's lots of really smart people at work doing all this.
Biggest shout out uh is we have a we have a couple folks at Ditya, uh Siddharth, uh, that that kind of run this.
They're like a you know, kind of tag tag team duo on our evals.
Ben, our CTO, heavily involved, Yasha, head of AI, uh, you know, a bunch of folks, and um evals is one part of the story, and then just like the full, you know, kind of AI and agent team is uh is a is a pretty, you know, is core to this whole effort.
So there's probably I don't know, like maybe a few dozen people that are like the epicenter, and then you just have like layers and layers of of kind of concentric circles of okay, then there's a search team that supports them and an infrastructure team that supports them, and it's starting to ripple through the entire company, but there's that kind of core agent team um that's a pretty pretty close uh close knit group.
The search team is separate from the infra team.
I mean, we have like every every layer of the stack we have to kind of do except for just pure public cloud.
Um, but um, you know, we we store, I don't even know what our public numbers are in, you know, but like you can just think about it as like a lot of data is is stored in box.
And so we have and you have every layer of the of the stack of you know, I do manage the data, the file system, the metadata system, the search system, just all of those components.
And then they all are having to understand that now you've got this new customer, which is the agent.
And they've been building for two types of customers in the past.
They've been building for users, and they've been building for like applications, and now you've got this new agent user, and it comes in with a different set of property sometimes, like hey, maybe sometimes we should do embeddings, an embedding based, you know, kind of search versus you know, your your typical semantic search.
Like it's just like you have to build the the capabilities to support all of this, and we're testing stuff, throwing things away, something doesn't work and and not relevant.
It's like just you know, total chaos.
But but all of those teams are supporting the agent team that is kind of coming up with its requirements of what what do we need.
Yeah, uh we just came from uh Firesight chat where you did, and you you talked about how you're doing this.
It's it's kind of like an internal startup within the broader company.
The broader company is like 3,000 people, yeah.
But you know, at there's there's like this is a core team of like, well, here's the innovation center.
Yeah.
And like that, every company kind of is run this way.
Yeah.
I want to be sensitive.
I don't call it the innovation center, only because I think everybody has to do innovation.
Um there's there's a part of the the the company that is is this sort of do or die for the agent wave.
Yeah.
And it only happens to be more of my focus simply because it's existential that we get it right.
Yeah.
All of the supporting systems are necessary.
All of the surrounding adjacent capabilities are necessary.
Like the only reason we get to be a platform where you'd run an agent is because we have a security feature or a compliance feature, a governance feature that that some team is working on, but that's not going to be the make or break of of whether we get agents right.
Like that already exists and we need to keep innovating there.
I don't know what the right exact precise number is, but it's not a thousand people and it's not 10 people.
There's a number of people that are like the the kind of like you know, startup within the company that are the make or break on everything related to AI agents, you know, leveraging our platform and letting you work with your data, and that's where I spend a lot of my time.
And Ben and Josh and Diego and Tyri, you know, these are just you know people that that uh you know kind of across the team are working.
Yeah, amazing.
How do you I think about I mean you talked a lot about like kind of read workflows over your box data, right?
You know, genetic search, questions, queries, etc.
But like what about like write or like authoring workflows?
Yes.
I've already probably revealed too much, actually, now that I think about it.
So um I've talked about it.
Whatever you can check.
Okay, yeah, it's just us, yeah.
Okay, of course, of course, of course.
So I I guess I would just uh I'll make it a little bit conceptual, uh, because again, I've already I've already said things that are not even G A, but but we've we've kind of like danced around it publicly.
So I yeah, okay.
Just like hopefully nobody watches this um episode.
It's tidbits for the highly engaged to go figure out like what exactly um, you know, is is your sort of line of thinking.
Yeah, they can connect the dots.
Yeah, yeah.
So so I would say that that uh we you know, as a as a place where you have your enterprise content, there's a use case where I want to you know have an agent read that data and answer questions for me.
And then there's a use case where I want the agent to create something and use the file system to create something or store off data that it's working on, or be able to have you know various files that it's writing to about the work it's doing.
So we do see it as a total read write.
The harder problem has so far been the read, only because again, you have that kind of like 10 million to one ratio problem, whereas writes are a lot of that's just gonna come from the model, and and we just like we'll just put it in the file system and kind of use it.
So it's a little bit of a technically easier problem, but the only part that's like not necessarily technically hard, it's just like it's not yet perfected in the state of the ecosystem.
Is you know, building a beautiful PowerPoint presentation is still a hard problem for these models.
Like, like we still, you know, like like the these formats are just we're not built for they're working on it.
They're they're working on it, everybody's working on it.
Everybody launch is like, well, we do a PowerPoint now.
Yeah, getting a lot getting a lot of better each time.
But then you'll do this thing where you'll ask the update one slide, and all of a sudden, like the fonts will be just like a little bit different, you know, on two of the slides, or it moved you know, some shape over to the left a little bit.
And again, these are the kind of things that like in code, obviously you could really care about if you really care about you know how beautiful is the code, but at the end user doesn't notice all those problems.
In file creation.
The end user instantly sees it.
You're like, ah, but like paragraph three, like you literally just changed the font on me.
Like it's totally different font.
And like midway through the document.
Those are the kind of things that you run into a lot of in the the in the content creation side.
So we are gonna have native agents that do all of those things.
They'll be powered by the leading kind of models and labs.
But the thing that I think is is probably gonna be a much bigger idea over time is any agent on any system again, using box as a file system for its work.
And in that kind of scenario, we don't necessarily care what it's putting in the file system.
It could put its memory files, it could put its you know, specification, you know, documents, it could put you know whatever its markdown files are, or it could, you know, generate PDFs.
It's just like it's a workspace that is is sort of sandboxed off for its work.
People can collaborate into it, it can share with other people, and and so we we're thinking a lot about what's the right, you know, kind of way to deliver that at scale.
I wanted to come into sort of the sort of AI transformation or operations things.
Um, one of the tweets that you that you wanted to talk about.
This is just me going through your tweets, by the way.
Oh, okay.
I mean, like this is like you're having a read one by one.
You're the easiest guest to prep for because you you already have like this is the this is what I'm interested in.
I'm like, okay, well, are we gonna get to like like February, January or something?
Where are we in the in the timeline?
So how far back are we going?
Can you can you describe boxes a set of skills?
Right?
Like that that's like that's like one of the extremes of like, well, if you you just turn everything into a markdown file, then your agent can run your company.
Like you just have to write a find the right sequence of words to do it.
Oh, we're sorry, is that the question?
So I think the question is like what if we documented everything the way that you exactly said, like yes, um, let's get all the Fortune 500s uh prepared for agents and like you know, everything's in golden and and nicely filed away and everything.
What's missing?
Like what's left, right?
Like yeah, you you've you've run your company for a decade.
Like I think the challenge is that that that information changes a week later.
And because something happened in the market for that customer or us as a company that now has to go get updated.
And so these systems are living and breathing and they have to experience reality and updates to reality, which right now is probably gonna be humans, you know, kind of giving those g giving them the updates.
And you know, there is this piece about context graphs uh as uh that kind of went through the very viral.
Yeah.
And I I I was like a I I I thought it was super provocative.
I agreed with many parts of it.
I disagree with a few parts around, you know, it's not gonna be as easy as as just if we just had the agent traces, then we can finally do that work because there's just like there's so much more other stuff that that's happening that that we haven't been able to capture and digitize.
And I think they actually represented that in the piece to be clear.
But like there's just a lot of work, you know, that that has to you just can't have only skills files, you know, for your company.
Because it's just gonna be like there's gonna be a lot of other stuff that happens.
Change over time.
Yeah.
Most companies are practically apprenticeships.
Most companies are practically apprenticeships.
Like every new employee who joins the team like you spend one to three months like ramping them up.
Yes all that tacit knowledge is not written down.
Yes.
But like it would have to be if you wanted to like give it to an agent, right?
And so like that seems to me like to be a one is I think you're gonna see again a premium on companies that can document this much there'll be a huge premium on that because you know can you shorten that three month ramp cycle to a two week ramp cycle that's an instant productivity gain.
Can you dramatically reduce rework in the organization because you've documented where all the stuff is and where the answers are can you make your average employee as good as your 90th percentile employee because you've captured the knowledge that's sort of in the heads of of those top employees and make that available so like you can see some very clear productivity benefits if you had a company culture of making sure you know your information was captured, digitized, put in a format that was agent ready and then made available to agents to work with and then you just again have this reality of like at a 10,000 person company mapping that to the you know access structure of the company is just a hard problem.
Is like it's like, yeah, well, you just not every piece of information that's digitized can be shared to everybody.
And so now you have to organize that in a way that actually works.
There was a pretty good piece.
Um, this this uh this piece called your company as a file is a file system.
I don't know.
Yeah, did you see that one?
Nope.
Uh yes, you saw it.
Yeah, and and uh I actually be curious your thoughts on it.
Um like an interesting kind of like we we agree with it because because that's how we see the world.
And uh we have it up with it.
Okay, okay, yeah.
But but it's all about basically like, you know, we've already we we're already organized in this kind of like you know, permission structure way.
Uh and and these are the kind of you know natural ways that that agents can now work with data.
So it's kind of like this this you know, kind of interesting metaphor.
But I do think companies will have to start to think about how they start to digitize more more of that data.
What was your take?
Yeah, I mean, like the company is probably like an acid compliant file system, uh chicken guessing boxes, right?
So yeah, yes.
Yeah, which you have a great piece on, but uh yeah.
Well, uh I I my my direction is a little bit like I want to rewind a little bit to the graph word.
You said that there's that's a trick magic trigger word for us.
I always ask what's your take on knowledge graphs?
Yeah, because every especially every data database person, I just want to see what they think there's been knowledge graphs type cycles, and you've seen it all.
So I actually am not the expert in knowledge graphs, so so that you might need to read it.
You don't need to be an expert.
I think it's just like, well, how how seriously do people take it?
Like is it is is there a lot of potential in the in the H O B I.
Well, can I can I uh understand first if it's um is this a loaded question in the sense of are you super pro, super con super anti media.
I see pro I see pros and cons.
Okay, uh, but I I think your opinion should be independent of mine.
Yeah, no, no, totally.
I just want to see what I'm stepping into.
No, I know it's a and it's a huge trigger word for a lot of people in our audience, and they're they're trying to figure out because why is this such a hot item for them?
Because a lot of people get graph religion.
And they're like, everything's a graph.
Like, of course you have to represent it as a graph.
Oh, well, how do you solve your knowledge um changing over time?
Well, it's a graph.
Yeah.
And and I think there's that line of work.
And then there's there's a lot of people who are like, well, you don't need it.
And both are right.
Yeah.
And what do the people who say you don't need it?
What are they arguing for?
Markdown files.
Oh, sure, sure.
Simplicity.
Yeah.
Versus it's it's structure versus less structure, right?
That's that's all they really think.
The tricky thing is um is is again when this gets met with real humans, they're just going to their computer, they're just working with some people on Slack or Teams, they're just sharing some data through a collaborative file system and Google Docs or Box or whatever.
I certainly like the vision of most most knowledge graph, you know, kind of futuristic kind of ways of thinking about it.
Uh, it's just like, you know, it's 2026.
We haven't seen it yet kind of play out as as I mean, I remember you remember the um in like actually I don't I don't even know how old you guys are, but uh I'll for for to show my age.
I remember 17 years ago, everybody thought enterprises would just run on wikis.
And uh confluence and and and not even I mean, conf confluence actually took off for engineering for sure, like unquestionably.
But like this was like everything would be in the wiki.
And I think based on our uh our uh general style of of of what we were building, like we were just like, I don't know, people just like want a workspace, they're gonna collaborate with other people.
Exactly.
So you were you're anti knowledge graph.
Not anti, not anti.
I'm not I'm not anti because I think i i think your search system, I just think these are two systems that probably but like i'm i'm not in any religious war i don't want to be in anybody's youtube comments on this there's not a fight for we love that you comments we're getting the comments okay uh but like but i i it's mostly just a virtue of what we built that we just continue down that path yeah yeah and um and that that was what we pursued but i'm not this is not a you know kind of this is not a it's not existential for you great we're happy to plug into somebody else's graph we're happy to feed data into it we're happy for agents to to talk to multiple systems not not our fight yeah but I need your answer yeah graphs are nerd sites is very effective nerd see because this is this is one opinion and then I've I think that the actual graph structure is emergent in the mind of the agent in the same way that it is in the mind of the human and that's a more powerful graph because it actually evolved over time.
I'll I'll figure it out myself exactly okay all right and what's yours I like the the the wiki approach um I I'm actually like if uh you know obviously I spend some of my time at cognition which uh you you know very well and they've had a lot of success with deep wiki it powers a lot of Devon brain super powerful and it's super it's useful for humans but it's oh my god it's useful for agents yes tell me if you think I'm I'm wrong on this but but not much of an access control structure issue it's like the whole you get the whole code base and everybody gets before before I speak too much there there may be some enterprise controls on the the enterprise deep wiki offering that I'm not familiar with yeah but yeah I don't I don't have any anything on the public side but yeah I I think like almost like every agent should have its own wiki that it's updating and that's persistent memory and uh that is a very weak knowledge graph yeah and you you could strengthen it if you want voice structure but you may not need it.
Yeah markdown files having links and wiki style right very effective right Lindy yep I like that as a as a just a general pattern.
Okay so uh last couple questions but you feel free to jump on in or you if you want any rants um I see you as a very interesting and an unusual founder where like you've been in a business and you're uh you're both like you're of like of two worlds like you're of Silicon Valley but you're also of the Fortune 500s and like I feel like your kind of founder mode is very different from the Brian Chesky founder mode.
And I'm just kind of curious if you have like ref reflections on like how you operate as a founder.
What would his founder mode be don't delegate.
Ah right and what how would you put me?
You do delegate ah okay I I I see the um I think that I I don't know that Brian and I would be that far removed from each other when you get to the specifics.
Okay.
So there's a whole bunch that I delegate 90% of the work that happens at box is fully you know fully delegated.
We've got great leaders running, running all that stuff.
It's just too much for my brain to handle.
And probably 70% of the work.
I'm gonna make up all the numbers here.
Probably 70% of the work at box or 70, 80% of the work at box.
I only need to really look at about five percent of that for like some high leverage decisions to be involved in.
You know, what's the marketing message that we think is gonna resonate with with customers?
So that's a little bit of high leverage thing that that we do in marketing, but most of marketing activities I don't get involved in.
What's our sales pitch?
Maybe I'll be involved in that a little bit, or like what's roughly the investments or push we're gonna do in certain verticals, you know, that's about five percent of like the total bandwidth of you know this the key areas of sales or go to market.
Okay.
So like 70, 80% of the company, I can just do about five percent, and then and then just like operationally, we've got great leaders and they're gonna execute on that, and we collaborate on the five percent anyway.
It's not like I'm just like making up a decision and and saying to go and do it.
Then there's this part that is like the existential part of the business, which is if we don't do this right, we're out of business.
And uh by virtue of just being a founder, you get kind of sucked into that part of the work because you can feel it.
Like this is like like you can just see how the AI tsunami could wipe you out if you make just two, three, four, five wrong decisions in this space.
Like a couple wrong architecture decisions, couple wrong AI feature decisions, couple wrong API platform decisions, and and you might be out of the game in a year from now.
And like you just feel it in your bones.
You you know this, uh, like it's just like like like we feel this all day long in this space given what's happening.
And so that in that area, it's you can't kind of delegate in a classic sense.
You still need to make sure you've got great leaders and strong hires and people that that are have high agency because they want to be able to the own part of the the strategy in the roadmap, or else you can't hire good people.
But but you know, there's gonna be a lot of little micro forks in the road that they will compound to determine whether you succeed or fail.
And so your kind of founder energy just like automatically draws you into those because they are the determining decisions of of your company's future, and that's kind of where I spend my time.
And I and you have to kind of you know do it in a collaborative way again, because if you are only dictatorial and just you know, you just won't won't eventually be able to hire the best people because they won't want to work on that environment, but you also just can't like abdicate all the responsibility because the risks are are just simply too high.
Like, and so you have to somehow obviously add some value.
And so the value I add is I've seen 20 years of this business, so I I think I can kind of piece together what I expect the value propositions are gonna be and how customers will react to certain things.
So that's what I can bring to the table.
And then you have this kind of existential fear of if I get it wrong, it's all on me anyway.
I don't get to blame you know, you know, the engineer that was working on that project.
Like, it's all it's it's it's my fault, right?
Like at the end of the day, it'll be my fault if it doesn't work.
So by virtue of of that liability, uh responsibility, you just get pulled into needing to make sure like it's all going according to to kind of how you think it needs to end up.
I don't know what I don't know how Brian would answer that, I guess, but like I I yeah, it's a long essay.
It's an interesting essay that people should go and compare and contrast your answer versus his.
Uh, I do think that um systems have a way of letting entropy get to them.
Yep.
And you you if you step away for too long, you need to have a way to like check in and go like, well, do I need to come back in or are we good?
And people are gonna tell you things are good, but they're not good.
Yes.
Yeah.
And that's actually I'm um I'm a fan of actually process for the that 70 to 80 percent.
Yeah.
So that's 70 to 80 percent, the process is you're gonna do uh, you know, a quarterly business review and you're gonna have a brand check-in, and you're gonna do those like you're gonna make sure that that you're seeing all the the right episodes uh of what's changing and and how it's kind of you know evolving and and make sure it's kind of going the right direction.
And then there's some areas which is like, no, it's 247.
Like, like I guarantee after this podcast at 11 p.m.
I'll be doing a zoom with Ben uh and probably some other people because we're gonna be talking about agents and and new platform features.
And like that's your just in the cauldron, you know, kind of grinding on on on that side.
Yeah, yeah.
That's uh that's extremely um realistic.
Like what is what it's like.
And I just want to have people hear your perspective on what it was.
And this is the this is this like um you read the post the about you know everybody having agents running in the weekend, and um, and it's like I you know, you you just I mean, first of all, anybody crazy enough to come to Silicon Valley, like we don't bring good news about the sort of like healthiness of our environment right now.
Like you actually have to know what you're signing up for, but like you know, there's a real issue, which is like, shoot, do I have enough agents running?
And yeah, I made a meme that was like semi-viral for me, but uh about this.
Like, yeah, exactly.
That's and that's you can't even enjoy a party these days because you're working with your tokens.
So, like I paid for the $200, I'm gonna spend the $200.
Yeah, uh, I'm gonna spend six thousand dollars auto-dejoining.
Yeah, yeah.
We need to make anthropic very unprofitable.
So we're not doing a good enough job.
Cool.
I have a closing question if you unless you're a question.
I've asked this question in private before, but I ask it again, which is uh it's a question that Tyler Cowan asks his guests on his podcast, which is uh what is the Aaron Levy production function?
Oh, I love that.
I love this question because there are so few people that I think are good at both executing, but also like distilling and like just putting good ideas into the ether.
You put a lot of good ideas into ether.
And so, like, what is the Aaron Levy production function that allows you you to do that versus others?
How do I get that information or I can I can give you uh a variant, which is what goes into Air and Levi, yeah, and what goes out, and how does it turn inside?
I'm just trying to think of because I mean, you know, there's some very I guess read a lot of Twitter uh as well, and so like I just And you've you spent a lot of effort too in your chat.
You don't see like great mini essays from Brian Chesky every day.
Uh but you do from you.
Oh, yeah.
You're kind of weird in that way.
Maybe he's helped maybe he's healthier than me, actually.
We should just like we should just text him to see if you know he's got a bigger muscles.
Well that's the thing.
I I work out less than him and I tweet more than him.
So that's the uh that's how we're balancing things out.
I'm um I mostly the way I just think about it is uh is just um you know, there's there's lots of work that's happening in the business.
I'm getting to see the all the problems that we are running into constantly.
And I'm trying to uh be a little bit of a create a flywheel between what we're doing internally, what what what then we talk about, uh getting a feedback loop on that and seeing other people's you know experiences of what they're doing, bring that back into the business.
And and so I just see that uh like my job is as you know, hopefully being able to kind of connect the dots of of what's going on in the world with what's going on in box, and then I just happen to tweet about that along the way.
Yeah.
Um because it's all you and there's no like editor, yeah.
Yeah, wow.
The uh I got um there was a funny uh uh my I I tried to get an internship in um between freshman and sophomore year of this company in it was a it was a film uh kind of production company in New York.
And uh I got the internship and then I emailed my liaison kind of guy who sponsored me for the internship, and I said, Hey, I'd like to do a blog of my summer internship where I blog about you know the the being an intern at a production company in New York.
And about like a I don't know, half a day, a day later, uh they re emailed me back saying they've rescinded the internship.
No.
Um yeah, because I showed a lack of judgment on you know, professionalism, you know, or whatever.
Like like just even the the idea that I would ask that question, red flags went up of like who the fuck is this guy?
So anyway, I I only say that to say that like to me, just like you know, building in public is just like a natural is a natural thing.
And so I so I just you know go through the day, we we deal with interesting problems, I tweet about them, I get information back in the process.
I I see your work, I see your work, you know, see a bunch of folks and and try and you know kind of incorporate that back into box.
My job is to try and connect all these things together and uh and make make it useful.
And you're I mean you're the number one spokesperson, right?
So you do have to be out there.
Yeah, I but I I kind of would be doing it whether or not like it's like I don't really think of it as a job requirement as much as like I just like I like social media.
You're so good at it.
Yeah, it's still hard to believe.
So like okay, so do you get up at 5 a.m.
with coffee?
Is that your secret?
It's like how do you work?
Do you actually just like that in the back of Waymo's?
Like is that do you do it that way?
Like, how do you do this?
It's it's well, no, it's it's like it's mostly that though.
It's mostly uh there's uh you know, I I I have a commute home each night.
I try and see, you know, my kids most weekdays before I have to hop back online.
So there's like a 20-minute window there where I can kind of like distill the information that's happened and be like, ah, there's there anything I learned today that would be interesting to throw out there or anything that I saw.
And then probably somewhere between like 7 30 and 9 p.m., I finally get a chance to like look through the feed and see like, did anything crazy happen in AI?
And um uh and then that's that will also kind of catalyze, you know, something as like that's the best I can kind of you know respect.
Yeah, okay, thanks.
Uh now I know your your cutoff is 8 p.m.
I will try to get AI news out before 8 p.m.
so I can help him do do his thing.
But basically, if if I don't see it before eight to eight thirty, I'm not gonna I'm not gonna be able to like core tweet or something.
Yeah, yeah, yeah.
So I wasn't gonna plan on asking this, but you've mentioned yeah, you mentioned it film stuff.
Yeah.
And I know from one of my favorite parts of doing a research on you was that uh you got the idea for box from like the the Paramount lot uh pushing paper.
Uh you film guy?
You're you're big.
Uh I I I I I would say I used to be more of a film guy.
Yeah.
What what's your what what were your favorites if you have you wanna list off any?
Kind of the classic uh wannabe film student classics.
Are you talking Scorsese?
Yeah Pulp Fiction, Magnolia, Requiem for a Dream.
Basically, like if there was an art house film in the 90s uh to early 2000s, that was my genre yeah that got me into like wow, wouldn't it be cool to do, you know, you know, film?
And then I I thought maybe I could connect digital into it.
Like, could you could you do film online?
That just seemed too hard from a licensing standpoint and then obviously Netflix you know co-existed um so I I never quite was able to fully connect the dots on these things but the internship at Paramount um was one kind of catalyst for starting box because we were using just traditional enterprise software and I was like wow it's like really hard to share data you know just like files going back and forth um but the same thing was happening in school as well and so that all led to led to box basically um well day 24 is uh you know kind of giving back the sort of resurgence of the independent film I guess 100% um in in in the face of all the Marvel slop uh you know I was thinking about this the other day and A24 is you know uh certainly the best uh example I'm sure of of this today but um you know they just don't you know you don't it's hard to make a film uh like you know no country for old men or um there will be blood like like what is that movie today yeah like what is a brand new movie that is just like original you just watch it and you're like what what did I just watch?
My my you know, Six's movie bench is uh Forrest Gump.
Okay.
Which is iconic in its time.
Yep.
Hundred percent.
Never again.
Yeah.
Yeah.
We we do not make we don't know how to make Forrest Gump anymore.
Um we'll try it with the sequel though at some point for sure.
I I forest go to and three.
I would be fine with it.
No, that Forrest Gump has a kid.
Yeah, yeah, he's still right.
Exactly.
Um I think Forrest Gump has a grandkid would be like a good movie.
Like, what is the grandkid of Forrest Gump doing in uh in 2026?
Goes tropical.
Yeah, but um, yeah, I definitely let's look, I want to see good, I want to see more movies out there.
You know, I'm a little bit conflicted on AI and film because well, because I um uh the world does not need more slop on on AI entertainment, but I'm kind of like in a mode where I think that AI is is gonna be, you know, generally a pure positive because if I'm a if I was me 25 years ago in high school, for sure I would be making a full production film that had explosions and car chases and but then there'd be like people that would show up there.
So like I think that ability to to just you get to be Spielberg, you know, is is you know completely amazing and and democratizing that is incredible.
And I you know, I'm I'm concerned about like how do you make sure that we still get PT Anderson along the way, and and can we make sure that those those guys exist?
And then interestingly, I never and I never saw it, but Darren Arnofsky, I I believe has either put out or gonna put out a f an AI film.
You know, even some of the best artists are are you know starting to adopt this.
But um, uh but yeah, I I definitely don't want to what I don't want to do is just be like in this like TikTok feed of just films, and it's just like, oh, this film about the car chase that does this thing, and it's just like we don't need that.
Like, like like this should be a form of entertainment and art, and let's use AI to accelerate the production process, do the really hard CG work that that you just you had to spend way too much money on previously to do the, you know, kind of like let's let's use it to test out all new kind of plot ideas, uh yeah, previs.
Yeah, like background.
And it's incredible, whatever.
And all those things are super incredible.
I still like the it's very nostalgic, but I still like the idea of like this is a camera and a person and a person that says, you know, action, uh, and then and let's hopefully like surround AI around that.
Well, yeah, but we'll we'll see how that plays out.
Yeah, I think you know, so one of the things that stability AI uh made an impression on me was like, well, you know, and at least now we can remix Game of Thrones season eight.
And I can, you know, uh like like it was meant to be, not uh not rushed.
Yeah.
And then you watch um, well, I have a six and a half year old, and I you know, you see a lot of these kid movies, and you're like, yeah, that probably will be AI.
I don't totally know the job math because I don't know how many animators there are today, but I actually think weirdly, I think we could be producing more high quality, maybe even slightly educational kids' entertainment.
And so it's maybe that's a positive, is like we could just have like more, like you could just have a Pixar for like, you know, things where kids learn stuff, and it used to be these like very, you know, lo-fi, uh, you know, kind of lesson things.
I mean, we had teletubbies, you know, that was so slow.
So we we we could have way more of that, and and maybe every animator that today is making a Pixar film is now, you know, we're like we fragment that out, and uh, but now they're responsible for more content and they've got AI agents running.
So like so I think there's some optimistic scenarios on the entertainment side, is like there's a lot of great use cases for being able to do, you know, generative media.
Yeah, yeah, educatement as well.
I guess one question I it's it's kind of like a self-serving one and almost like an advice uh side of the the the question.
One of the things I this uh really enjoyed uh researching you was that uh Michael Arrington had some influence in the box journey because he went to his house party.
Yes.
And and that's how you got funding.
Yes.
One of Latin spaces, and that's a deep cut, right?
Yeah, very deep cut.
That's uh oh six deep cut.
Yeah, uh do you I mean, do you wanna tell that story?
I don't know if you've told it.
Yeah, cause probably like a random intro, right?
Like well, it was just he used to have house parties.
Uh TechCrunch had had these house parties and and it was um probably no different than somebody's doing a house party in SF.
Uh just go.
Yeah, you just go and you meet the VCs and founders, and like I'm gonna make up examples so I don't want to like, you know, there'd be like Chad Hurley over there pitching his, you know, YouTube to people, and like that's just like how it worked.
And it was just like, wow, like that was this era where all these new companies were were emerging.
And I met uh our first investor uh in Silicon Valley at one of these house parties, Emily Melton, who then brought us into DFJ, yeah.
That became our series A.
So that was all because of Errington's uh backyard party.
One of my aspirations for latent space is to be as helpful, influential, whatever as TechCrunch was awesome.
Yeah.
What would a new TechCrunch today look like?
You know, what what what should I what should I do?
I think there used to be TechCrunch disrupt.
Yeah, I could do that with my conference, but I haven't done it yet.
Well, I mean, I think um useful.
I don't know.
Uh well, you know, actually interestingly, I would I would argue disrupt came after the period that was the was that deep cut period.
Okay.
So I think di disruption, you know, ended up being, you know, you know, catalyzing, I don't even, I think Cloud of Flare launched disrupt.
Yes, is that the story, right?
Okay, okay.
So like, so like I think anytime any time you can be in a a launch pad is is just great because it draws in people that are trying to do in that creative moment, and whether it needs to be a contest or or just like everybody gets like five minutes and you're fundraising.
I mean, I who knows.
But but I mean for what it's worth, like I don't know have that much advice because I think you you're you're already doing it effectively.
Like I I just like watch the YouTube videos late at night um uh from the events.
I haven't been to one of your events, but like from the from the camera angles, it looks like everybody's there.
So like what's great is that people are gonna be in the audience as like two random people, and they'll be like, you know, the next the next big AI company will come from, you know, people coming to a meetup because they were like, I came in from Chicago and I'm uh from you know Poland and let's go do a startup.
Like that's the magic of the valley.
X for T found is co-founder at EIE.
Oh, and I know if at least one marriage that's that's wow, you have marriages already?
Yeah, I don't I never heard that about it.
That's my that's my favorite KPI.
Wow, we have AI marriages at the at the AI engineer conference that these are like humans, to be clear.
Okay, exactly.
That's a very good clarification.
I like that you have to check.
Yes, that's a very good clarification.
No, but I I think you have your your insightful business leader with like a lot of thoughts on media, so I just figured I would I mean media is such an interesting space right now because I you know with the GoDirect model, every company is gonna have to be a media company.
You are going to you are the OG go direct.
Yeah, but but but you know, we're we're we're still like I think I think what what you guys are doing, and I don't even know all the overlapping relationships, but like I watched your guys' videos of your events, watch your event videos, but like it's clearly like this is the new format, right?
Companies have to become channels to communicate with audiences.
I think the resurgence, uh resurgence maybe is a bad word because it implies a decline, but like devrel is hot, like the hottest thing of all time right now.
I like if you could produce a freaking factory of devrel people, like there's just like unlimited jobs right now on the other end of that.
Yeah, yeah.
Um, because we're gonna everybody needs their services and APIs to be used by agents, and so we have to all find a way to like, hey, look at me.
Like, like agent over please come over here, agent.
And that's gonna that's a content game.
Like, how do you get the agents to see your stuff?
Yeah, and know your APIs, and like this is like a new world that that we are in, and uh it's gonna be uh it's it's gonna completely be a digital marketing, you know, kind of world that we're in.
Yeah, uh for what it's worth, I'm trying to help by doing little writing boot camps and basically turn into a devrel boot camp.
Um, where at you know, well, it's a demand and supply problem.
This this huge demand, there's no supply.
Wow.
So that's increasing.
Why is there no supply?
The one the really good ones work for themselves.
Uh-huh.
Creator economy, screw it, screwed you over.
So I see.
So Substack and YouTube payouts, and that's is that really Patreon.
Yeah, like the most talented guys are making you know millions and just working for themselves while they work for you.
We don't want them to make that much money.
Okay, we need to be able to hire people.
I mean, I think I think like, you know, do do what some companies are doing, you know, not saying it's my situation exactly, but like give them equity and like you know, it should probably would be worth more uh just like sort of helping them out.
Well, they are getting oh sorry, as full-time employees or not?
Oh, part time.
You need full time.
I'm part-time.
Yeah, but but you're you're you're NF1.
Like we need like also people that are full-time.
Yeah, yeah.
My classic joke or or like observation was like this was when Hobspot bought like their the appointment like a newsletter business, uh, and then they bought the My First Million, like the sort of podcast.
You must know Damish R.
Um, so he's like obsessed with this guy.
So my conclusion was like every company must either build or buy a media company.
Yes.
Right.
And until you unless you realize that you have to take it that seriously that you are running a media business in your company, yes, you will never be as good at it.
Yes, 100%.
Yeah, yeah.
No, we're we're very much taking that seriously, but but still, and yet DevRel, I mean, I gotta do one plug.
I don't mean we're hiring in DevRel.
Yeah.
Like, no, these all engineers here.
Like, yeah.
And like you've made it like, and I just said every every agent needs a box.
Like, let's go, let's go.
No, that's the headline.
We are hiring DevRel to make that happen.
Uh, but yeah, I think DevRel is like the future job.
So we're all just gonna be doing DevRel of in some form.
Okay, yeah.
I mean, what is FDE?
Developers are ruling the earth, yeah.
What is FDE?
I don't know.
Um, it's it's DevRel.
Yeah.
Okay.
Yeah, you just you're going to a company.
Isn't it just like glorified consulting?
That's this the down.
Sure.
I mean, I guess nobody can like actually d you know, d fully define this, but um uh but I think it's it's it's micro devrel.
Like you're in the company, you're helping them with the services, you're doing a little bit extra implementation.
Yeah, yeah.
Um, but uh, but yeah, so it's uh I I think we're all you know, the thing that's gonna happen on the ledger of software is we're gonna produce far more output of code and thus features per dollar.
But on the other end of this, we're gonna actually end up spending probably just as much on how do you get all of that stuff to the customer, and that's gonna create a new set of roles that we are all doing.
Partly because I either because there's so much choice, now you have to kind of fight for attention there, or because the stuff is is just changing so quickly that you have to technically help your customers along the journey.
Yeah, so so I just think like I this is why I I I always laugh when uh you know, people say you don't need to be an engineer, don't do computer science.
I actually think like that is like still one of the most protected job categories because things are only getting more technical.
Things are only gonna get harder.
And anybody in a technical position is in the best position to get agents deployed, get them built, get them adopted, build the the custom code software to the for the IT system, all of that.
So yeah.
My my classic funding story of like why I picked AI engineer as a title and as a as a theme for this podcast, as a theme for my conference, was um back in like early 2023.
Someone non-tenical came to me and said, like, I'm all in on AI, what should I do?
And I was like, I just looked at her, I was like, God damn it, there's nothing you can do.
Like engineers are about to get so much more powerful than you.
You don't even understand.
Tell me there's a good that should she go and then learn.
No, I didn't I didn't say any of that to her.
Oh, okay, I'm saying I'm not that honest.
I hope somewhere out there she she did went to some online academy exactly.
Learn to code.
But there's a lot of people like there's a lot of people who believe AI too much, and then they're like, Well, you don't need to learn to code, so I won't learn to code.
Yeah, and then there's there's like there's a bunch of us who are like just in that sweet spot of like we can code and we can wield AI a thousand times more effectively than you can.
Yeah, and like, well, who's gonna win here?
Like, I think I this was another uh a tweet, but it was like the observation that like really software engineering for the past 30 years was the primary career track for like technical high agency people that wanted to have a large outsize impact on the world, yeah.
And like software was a means to you know do that, right?
Effectively.
Um, and so yeah, with AI, is it like that?
Uh and for AI could eat software engineering, or is it software engineering can eat all their kind of domains of discipline?
You're those same principles then get applied to every other.
Yeah, exactly.
Yeah, I mean GT engineering is that something else, yeah.
Well, this is the you know, uh anybody who believes that an enterprise, and I'm I'm I'm mixed on the I'm mixed on this, is but if you believe that an enterprise is going to build its own software for all of its problems, then you must be the most long on computer science, you know, as a discipline of all time.
Because guess what?
The most of the economy does not have enough engineers to then maintain all those systems to update the all those systems to figure out the the relationship between the business problem and what the code needs to do to go and actually manage that.
And so, so like that's uh that's a very pro engineering job argument of what the future is gonna look like.
I'm still again I go back and forth on like are you gonna really build all these things versus no pre-packaged software?
But no matter what, there's gonna be 10 to 100 times more code.
So I think you can be very long engineering right now as just a you know, purely on the dimension of of software is gonna become increasingly more important once agents are are you know turning everything into software.
Yeah, all right.
Three software guys say software.
Not biased at all.
Okay, but uh Aaron you're inspiration, it's such a pleasure.
All right, good to be here