Agents & the $40M Bet on Multiplayer AI

Even if the technology again plateaus, even with current model shape, uh the way we're going to walk in 10 years is probably nothing compared to the way we walk today. And so there's still many many many opportunities to build for for everyone. We should probably start with this multiplayer verse [music] single player AI. I want to mention to folks that you've got a decorated background. you were at Stripe in the early days. You then you were also at OpenAI and now you're doing your own thing with Dust. So, it's super cool to see this progression. Are there things that in those early days of Stripe scaling up or in the early days of OpenAI scaling up that you have taken away from that and you want to infuse into Dust as you're scaling up? >> Oh yeah, a ton. uh mostly stripe over open first uh yeah I've been very lucky I've been very lucky uh so we joined stripe uh through an acquisition actually uh with my current co-ounder so it's not the first time we starting company >> nice >> uh the big difference that it was 15 years ago and we we were young we had absolutely no clue about what we were doing but it was fun uh but we can chat about that later it's a whole subject of its own uh stripe we saw the 150 people to 3,000 it was absolutely incredible and I think uh I hold very dear to my heart the 150 to 500 people. I think it was pretty magical at Stripe >> uh because there was uh really uh very minimal management and it still was working really well and I'm trying I've been trying since then to understand what was happening at the time for that to work so well. >> Do you think that's the writing culture? So yes uh so I've got a I've got a bit of a theory about that uh which we we actually turn into operating principle at dust but my my big theory about the time it's slightly but I think it's interesting nonetheless. >> Uh my theory is that we were actually implementing a flocking algorithm. So you know what a flocking algorithm is is it's like when you so when you want to program a flock of birds or flock of whatnot uh in computers >> that are kind of a autonomous groups of things that autoorganize around obstacles. There's actually only three things to uh there's only three rules to make it work. It's uh local spiration. So small force that push the the people uh away, distant attraction >> and alignment and uh that gives you kind of a some form of autonomous behavior around obstacles. Uh you can you can look at flocking algorithm online and you'll see the small videos and I think at Scribe the writing culture was the distant attraction. Uh so Scribe was very early into implementing that kind of a open writing culture. It was mostly through email at the time, interestingly. >> And what? Oh, no. It was great. It was great. Oh, yeah. [laughter] >> I don't like Slack that much personally, and I love email, but that's another take that we can spend a lot of time on. [laughter] >> You're surprising me here. All right. >> Uh, but anyway, uh, so this traction, why? It's because we had those mailing list and so there were people within the or that was spending a lot of time kind of sifting through those main list and so that mean that somebody would see some else talk about subject A over there and see another person talk about sub the same subject over there in the or was able to tell them hey you're talking about the same thing you should chat and so that's kind of a distant traction uh false uh I think at scribe one thing that really struck me is the amount of trust between people at the time uh it was just ingrained in the cult culture in a very nice way and that's a kind of a local separation force because you don't look over the shoulder. If somebody says I'm I've got this, they've got this and you stop caring about that. And then finally, the last one was which is uh which is almost the the most one is alignment. But I think that's why we were we were and which is very interesting because it's very hard to reproduce that in modern times with AI. But we had a very strong alignment on the vision. It was very easy to understand. We are a dev API for payments. Basically, there's a tensor of countries, payment methods, payout methods, and you just just want to to fill it in. >> It was very easy to understand. And so that means that you are every is kind of aligned very naturally. And so that gives you the three properties of flocking which gives you kind of a autonomous behavior in a sense emergent autonomous behavior which explains why potentially we didn't need that that much management and why it was so nice to work in that environment. It feels like we could draw a parallel here and heads up segue into multiplayer AI, single player versus multiplayer with the flock algorithm. And how have you thought about humans working with agents in that flock style? Do you think that there's anything that you could look at there? >> Yeah, there's probably a ton of stuff, but it's very hard to anticipate because today we are still very much stuck in the single player mode. And what do you mean by that single player mode? >> It's mostly that most of the agent interactions we have, we just have them alone for two reasons. I think uh the main two reasons I see is first the timer of the tasks an agent can take is still pretty small. uh according to the meter benchmark uh we are topping at right now at uh couple hours let's say half a day and mostly in coding task by the way and you know half a day is still very much a single player thing you I can walk half a day on the task you can walk half the day on this task and you walk alone and it's fine because it's a kind of well-shaped tasks the other thing is that the the they're still very imperfect in many ways and so you interact a lot with Um, it's messy. Uh, really when you look at an agent trace between a human and an agent, you look at the how the sages made and like, oh man, you did something funky there. [laughter] >> Or it's like, oh, you were so close, dude. A little bit over to the right or left, you would have been good. >> And so on those time horizon and and in terms of capabilities, the model are super jagged, which means that you're on that you're looking in that direction. the thing is clearly superhuman. You move a few uh uh uh degrees off on a slightly different task. You fall into a valley and the the the thing that thing is the dumbest dumbest thing on earth. And so that's why it's messy. And so all of that makes it very kind of unnatural to collaborate in an agentic loop with multiple humans to share uh your sessions with agents because they are messy. You share the output, you don't share the session. Um, and so that's that's why I think we're still very much single player AI in the way we we work. We all spend a lot of day in in CL code or codeex. We all spend a lot of day in our favorite agent harness to work on presentation and stuff. Uh, and it's we pretty much do it uh alone and we share in the in the other platforms the output and we collaborate there be it on GitHub, bit on notion, on Google drive whatnot. Um I think the thing that the the thing that is supposed to is is is about to switch is probably uh as the models are getting better and they seems to be getting better consistently. It's always an open question but when you look at the past few months Mhm. Um well the uh the time horizon of those task was increased and if you start considering a task like that represents a week of man walk for most job descriptions those tasks are inherently cross team cross people involving multiple person even if you think of coding there is very few coding task that takes a week that doesn't involve collaboration even with somebody external be it a product manager be it a designer or maybe the infra person because you have something finicky about the infra. Uh and so uh and if you look at something that outside of coding then most task that takes a week they are extremely crossunctional and so that will flip the picture where uh it's not going to be possible anymore for just one human to steer the agent in those type of task. the agent even the relationship between the human and the agent was will will flip. The agent will orchestrate a longer horizon task involving humans and the humans will steer but they'll they'll I'm quite convinced there's going to be a need for many humans to steer those agents be an agent or multiple agents we don't really care uh for those kind of long horizon tasks >> they'll invoke the human agent or the human tool when they need to. So it's invoking a SR tool and a >> for that >> there's gonna be there's there's gonna be some of that for sure. >> Yeah. And now, don't you feel like this is just an it's a UX problem more than anything? Because when you tell me about the single player mode, I think, okay, but what ways do I have to share with my colleagues rather than exporting the chat or putting it up with my PR, things like that. it doesn't really feel like I have the tools in order to get to bring someone else along this journey with me. Yeah, I think I think the the the interaction will pretty much remains uh kind of a you with the agent, but it's just that the agents will create many of those sessions in the to to achieve that task and have will kind of collaborate with many humans and so it is entirely a a UX is maybe slightly uh diminishing but I do think it's not a it's not a a machine learning problem. It's more of a product problem. >> Yeah. Um it means it just means that as an example if we if we take co-work which is having a kind of a real moment these days because it's working locally on your computer it is inherently a single player thing it's very hard to kind of a collaborate around something through co-work with multiple humans and that means that you need to push that harness inside of a Austin environment this is the only way I think u and uh and there's a lot of things that we we are exploring around what is the right product surface to enable that collaboration uh where you have multiple humans multiple uh sessions around a shared state uh and how that what what it can look like. I think this is the frontier in terms of product of interacting with agents and that's the one we're very excited about. >> Talk to me more about that. So one of the things we're doing right now uh is uh is so we we we've we've we've released the first version of that which is called uh really a we call it a pod and a pod is really a shared state with a a number of sessions uh between humans and agents and I'll give you a very simple example of where it plays really nice. So we have a as with any companies we have a team weekly. Uh and the team weekly is mostly made of a set of uh slides that are somewhat always the same with small variations every week and maybe some deep dive uh about uh subjects uh every other week. Um and so in the past the way it would work is that somebody would start building a slide and and ping the humans and have the humans collaborate in those slides etc. And uh the kind of orchestration of all of that would take some amount of time uh to that human. Uh now comes uh agents. Maybe you can uh the the initial organization building a slide. You can you can uh you can offload to an agent but you still have to ping the humans and organize the the kind of organize the collaboration so that we get to a shared results. Um and so with a pod which is a group of uh conversations, group of humans and a shared state uh no our team weekly is much more driven by an agent than driven by a human. So basically uh on our team are on Tuesdays, Monday morning uh somebody triggers an agent with a skilled team weekly. The team the agent creates a pod. It creates um one session per slide uh with the owner of each of those areas and maybe one session for the people that are uh doing the deep dive that week >> and it will will pre-walk the slides. So going into metatabase pulling the data keeping up to date revenue customers going on GitHub for the the product work and then going on whatever platform for the support state whatever he'll pre-build the slides and then it'll ping the the the person responsible for that slide and say hey what do you want to talk this week about and let me let's collaborate on on this slides what do you want to add is there anything you want to highlight and so that collaboration happens in one of that session uh each humans that are responsible for each of those subjects are collaborating in different sessions and On Tuesday, let's say noon, the agents the agents wakes up and like looks at everything probably ding on the humans that haven't haven't finished their work. Wakes up yeah wakes up again in two hours like two hours before the weekly look at all the the things that have been built and coles them into one shared presentation. So we we don't use any presentation anymore. We use we have a frame which is the equivalent of canvas or artifacts. already kind of a code generated uh UIs and so it it takes all of those frames and and and >> pull them in that in that shot frame and that thing is ready and really the the weekly the team weekly which is kind of a a multi-day collaborative task is today mostly on auto mode right and I think that's that's the kind of product surface that are interesting if you don't have that pod which is that group with that sh state feels kind of weird because you have a bunch of conversations all over the place. You don't have a sh place to see them all. The agent doesn't have a place to organize all of that work with multiple humans. And so I think that's that's the type of things we're exploring and the which is very exciting for for what's to come for more than for what's where we at. >> Can you talk to me more about this pod? >> Yeah. So yeah, the the a pod is really like technically it's really a group of humans, a group of agents, a group of sessions which is the interactions between human and agents >> and these are sandboxed. >> Each of the session has a sandbox. >> Each of the session has a a state. Uh so a state mean let's call it a fight system. The fight system is visible by the humans but also vis mounted on the sandbox. Basically to make that easy we diving into the technical details here we use GCS fuse so we back that with GCS so it's very easy to show it in the UI it's very easy to show it as a mount it as a file system to the agent so the home directory of the agent in the sandbox is is GCS back bas basically and so that's the state in one session and then inside of the pod we also have another file system which is shared across those sessions and the agents are free to move files between the two file system so basically they the slash session slash blah blah blah and in there they have all the all the file system related to the current session and then they have slashpod and the sl under slashpod that's the shared file system where they can move some of those files up or down as they see fit and so they can really collaborate across sessions you can have collaboration across sessions through that shared file system at the pod level >> and when I'm interacting with my agent I can drop into somebody else's slides presumably also So it's not just my slide. >> So if yeah if the if the slides of that a so somebody has a session with an agent inside of the pod is working on a slide it leaves naturally as a starting point on the file system of the session. So it's mostly visible to that session and then that person will say okay the slides are ready move them to the move them to the to the pod the slides get moved and then somebody else will be able to see them or even >> modify them which is risky in a sense but that's fine. I mean uh that's that's in Google Drive you can modify whatever else has been doing in your company as well. So I think it's okay. >> Do you have the possibility to look at the history of the modifications in case you're like control Z? >> Yeah. No, not right now. Uh [laughter] technically >> that's why it's risky though. >> Yeah. Technically technically we have it because GCS is version. So technically we do have it and that's definely something that we would want. Uh yes you can uh you could the re though the reality is that you do something with an agent it will uh what will happen is get it's going to read the file and then it's going to emit an edit uh and so generally they are able to pretty much pretty easily backtrack that change because they still see them in context. The only thing that the agent can do that would be bad is like like deleting the whole pod shared file system. Uh that's possible technically. Um uh I don't think it happens much. Uh so that's that's kind of fine. >> And also I imagine you have some kind of redundancy maybe some backup somewhere just in case you got >> Yeah. So again in just everything is version so we could uh we could totally restore it and uh and and yeah. So okay now the shared state yep and the ability for almost like this tiered level of I'm working I'm bringing it down not necessarily locally but I'm having my oneplayer experience and then I'm pushing to the multiplayer so everyone can share and look at it. How are you envisioning that in other scenarios? Uh there's uh there there is so many of them. It's um uh it's one of the by the way it's one of the very hard difficulty of building a normal platform. Uh there's a lot of value into building a normal platform. There's a lot of pain in building a platform. The pain is mostly this one is that the uh there is so many use cases uh that's that you it's you cannot I mean I would love to be able to tell you dust is awesome for that but that means we would be verticalized. If we were to be verticalized we wouldn't have the right to really try to equip the entire company and so that's why we accept that that tension. uh many other ways to use a pod. You can uh you can use a pod as um as a as a so basically we also have in that pod we also have a a list of tasks which is a way to try to organize work and so you can create a pod for any project you do and let's say you're on slack and you're like oh I need to do uh we need to do this this and that and you you call dusk from slack and you ask it to add those task to the to the pod you're working on may let's say you're working on a initiative sandbox you're working on adding sandbox to to dust you say should do this and that. The agent is capable of interacting with the pod and add task itself and then as a human you can uh you can the task are kind of a nice way to organize the uh organize the walk and trigger sessions to start solving the task uh with with an agent and a human working together uh to solve that task. So you have that kind of a task kind of a trying to explore the task management part of working uh with agents uh around a project. There's obviously another usage of pod which is uh which works well is around um internal kind of a internal support like uh so obviously the kind of external support the tier one support will will is best tackled in platform that are specialized for that but when the questions is finicky and it reaches your internal support team uh a pod is a really great place to share the way you solve those questions uh uh for people to to to to to look at and and and explore. And so I think there's many many uh use cases around that. We're really in the business of creating um Lego bricks that work well together and and make sure that they are somewhat universal. Um and so uh we see very varied usages. One principle that I think is interesting uh that we try to enforce is what we call the uh birectional aress. Mhm. >> Meaning that we enforce us a very simple rule is that any feature we build it should be equally accessible by agents and humans. >> Okay. >> There shouldn't be anything that a human can do in dust that an agent cannot do and vice versa. And I think >> and how does that play out in the time when you're actually building? >> So it it kind of creates a constraint on the things you can build. uh as an example uh uh I mentioned the tasks around the pod. This is a product surface that is uh kind of very human ccentric. Well, it's uh it it it's very important that those task are can be listed by agents, can be edited by agents, can be created by agents, can be uh the work can be triggered, another conversation can be triggered, the session can be triggered from those tasks by agents as well. And so it's really kind of a a great principle to make sure that we we don't overfeit on any of those two users because we're really building a a product for both those users. Uh and we really care about that because we see we it's the only way to start meshing the work of humans and agents together in our opinion. Only if you have that kind of symmetry will you progressively stop making it fuzzy whether the work has been done by humans or an agents which we think is the future and is exciting in many ways. Uh other things that there's so much so much places where you really want to build something for humans and when you try to apply and believe strongly in that principle it really makes you rethink what it should look like. Uh and so it's a useful principle to uh to apply. But I think most AI products today are kind of building for two two users and so it's important to build for them both. >> Yeah, it's a it reminds me of when someone told me you can create documentation that humans can read but agents can't but you can't do it vice versa. If you create documentation for agents, a human is going to understand it. So you should default to agents being able to understand your documentation first and foremost and humans uh will by default also understand. Now, when you were talking about the idea of the pods and also how work gets done over longer time horizons, it made me realize and and I had the vision of data moving through like a DAG and how a finished piece of work is not dissimilar to that. You have different nodes where work gets done and then it goes to the next node and whether that node and that piece of work is being done by a human or it's being done by an agent, it doesn't always have to be the same human that is doing it. And I think that's the huge unlock that I'm understanding with you is that a human is an expert in different parts of that process. So it shouldn't be the same human that owns it all the way through. We hear a lot of talk of how well now PMS and designers are shipping code >> and that's great but they have their unique skill set where they're very good at things. And isn't it almost more efficient if they can do what they're best at >> in that node of the DAG >> and then it gets shipped to someone who can make it battle harden. Maybe it's the DevOps person or the security person. They can harden it and then it goes to production and then you have that feedback loop where it will go back around and you're constantly iterating and updating. But those longer time horizon events now when you talk about multiplayer, I'm understanding what you're saying there. >> Yeah, I think I I think that's uh that's exactly that. and and and the the grail is u the grail is really try to create a place where it's really easy to end off and so to really create an edge and a next node and end off to the the next processing unit and the next processing might be humans might be an agent and it should be really easy to route towards one or the other and uh and uh and and and and that's that's the thing and to make that a reality uh the the complex Uh the the the dirty B that needs to happen is that uh those agents has to be very contextful and like it's super interesting into trying to think what is the right context for those agents to be useful agents of work. Uh is it just um I mean today the the basic answer is a long agency loop and gazillion MCP calls to uh rebuild the context in every session. Um the kind of a second stage answer is a is a turn everything into a skill and eventually the context will be rebuilt. Uh is that part automated? Is that part on autopilot? Is that part manual? These are all the kind of a question at the frontier of what it means to create that that right context for agents which is uh quite interesting. >> Yeah. How are you seeing success with that and breaking down those barriers so it is more fuzzy when you have the handoff and you can still I also understand there's moments where there's two different people humans that are working on the same thing and then the agent in the middle and so you almost have like this ven diagram of you have the product manager you have the engineer you have the agents all looking at one node if we're going to continue the metaphor of the DAG. >> Uh I I it's I think we are still I don't have a crisp necessary answer to that. I think we are uh again I think we are still very early on on this one because as we said it as as we said at the very beginning the capabilities of the agents are still kind of uh jagged and so that makes that whole process uh not super smooth at all time but we start seeing light at the end of the tunnel and so our goal is really to discover that with our user to give them the right Lego bricks um that means that the the the the within within our user base there's always going to be kind of AI operators, tinkerers that will see the bricks and will want to assemble them in kind of new ways. And it's so it's super exciting to see what people do with the bricks you give them. Sometime they come up with stuff that are just mind-blowing. They just rebuild entire systems, all of that. And it's it's exciting to see. But I think it's a it's a code discovery. It's a code discovery thing. So it's a would be very uh would be very um I mean on top of that it's so dependent on the capability of models which is moving so fast and and which is still kind of somewhat impredictable in many ways. Uh so it would be very uh uh adventurous of of me of trying to to have a very crisp answer on this one. uh >> slant a flag right now and then like by the time this podcast airs that didn't >> exactly exactly [laughter] >> which going back to the alignment uh I kind of seeded that uh like at the very beginning going back to the alignment this is one of the major challenge I see into trying to reproduce the kind of conditions of stripe early on I think is the alignment part because when you're operating in that space the technological substrate is moving under your feet in a way that wasn't true before. For the past 20 years, when we were building tech companies, you were building on JavaScript and posgress >> was stable. >> Yeah. >> Like there's nothing more stable. I mean those things evolve, but it's like stable stable concrete type stable. Uh and today we're building on on models that are changing every weeks. >> Yeah. And so every time we tried to build a very crisp picture of of what we could be constructing in a year, we got it completely wrong because within that time frame uh the model shifted. The place you wanted to build that nice city of yours uh kind of went under sea and the big mountain appeared right next to it. And so now you want to build on top of the maintain uh obviously and so uh that creates the uh that creates a that creates a real challenge for creating a crisp alignment of the team because you you you're subject to the fog of AI and it's very hard to see past six months let's say >> I like >> the fog of AI >> the fog of AI that's so that's that's a pretty uh that's one of an interesting challenge of building a building an AI company I guess Mhm. And now just going back to Dust and how you're building, do you allow folks to interact and be multiplayer in all different types of scenarios like Slack, like in their web browser or in GitHub or they throw like a Jira or linear issue at it? Is that kind of the vision where you're saying we have all these tentacles? Yeah, we try to build as much tentacles as as possible. The big one is obviously so there's always a and there's always an interesting tension is that uh uh when you build one of those connection like Slack is one of the one of the big ones. It's a very natural one. It's a chat uh system and so that's a place where you might want to trigger work with agents. >> Mhm. Um uh it is also an acquisition engine internally within work within uh companies because people are on Slack and they see other people using a system and so they discover it and they might start using it. Obviously >> the the tension is that the you constrained by the interface of the platform you're connected to. And so the experience everything we just said is not is very hard to do in Slack because you you bond by the interface. And Slack has been doing a lot of work there and and they actually you can create pretty smooth interface today in Slack. But let's say GitHub as an example. Uh it's just a text text async uh you don't see the syncing tokens. You don't see the tools being used all of that. And so I we really see those places as uh uh entry points more than place where the collaborations or the collaboration with agents really happens because you always want to you you always try to drive the user to go open the session inside of this because that's where all the richness happens. >> Ah I like that. So it's the gateway into it and you're pushing them to go and look and inspect on dust so that you can get that richer experience. >> Yeah. Some actions that agents take for the for obvious reasons of some admins uh mark as high stake. So it mean it requires a user confirmation before the action is taken. And a good example is sending an email. Um uh and uh and so in Slack you can actually build a pretty good uh tool approval experience but in GitHub you cannot >> Mhm. >> you cannot send the next message on the issue say you need to click here to approve the tool. It wouldn't make sense. >> Well I'm also thinking about even things like MCP apps that make the strat experience so much richer. You don't have that ability. Uh if I'm not mistaken, >> you could you could you could imagine that Slack rebuilds supports MCP app and you manage to forward the MCP app to Slack. That could be but again GitHub wouldn't work, Zenesk or all of that. Pretty complicated. >> Yeah. Or there's so many areas or surfaces that you touch that do not support that. And it would be very nice to have that richer data experience. Like even at Jira or Linear, you might want to see some of these data points or data exploration that happens with the agent before you commit to a quote unquote sprint because I don't know what happens in sprints these days. It's an absolute mess. [laughter] >> I maybe that's a thread we can pull on on how sprints work. No, but uh going continuing on this MCP apps, you almost want an area where you can use all of the data that you're having happen in these different spots like the Slacks or the Jiras or the lineers and then put them in something to allow folks to have that richer experience. So, I could see that being a huge value prop. >> Yeah. But we we really see those platforms as I mean we really try to insert ourselves in a disco system. We don't we don't want everything to happen on dust. We have have work some some amount of work happen on dust but we sometimes that work starts from slack continues on dust and ends up in GitHub as a as an issue and that's really fine. It's really important to be able to read from them. It's really important to be able to write back to them. I think that's the uh that's the at the top position on this one. >> Yeah. And it goes back to the whole fuzziness of the handoffs and how those work right now and how that actually is is being done because you've got some folks that are working in Slack or you've got decisions being made in Slack >> and then an agent will execute on that decision that's being made which then becomes a GitHub issue or a PR >> and then somebody else has to review it but they are using an agent to review it. So you've you've got all these different ways that it could be siloed or it probably is being siloed right now. And I know when we had our lunch and learns, >> one of the big things that folks asked for was how are teams working with coding agents? How are you figuring out which skills should be part of the whole company or should it be teamwide skill? How are you making sure that sessions can be reproduced or shared or are you committing your whole chat history to the PR? Like there's these questions that are coming up and we don't have the tools for them right now. No, at least in the in the coding agent space, we don't have them yet, but that seems I mean it seems pretty obvious that uh at some point uh your codex d- session and that ID or cloud d- session and that ID will be uh hosted in the cloud and you'll be able to tie it to the PR and somebody will want to re restart from there. Uh that's a pretty obvious uh iterations for them. Yep. >> Now I don't want to pass over this very big important thought of finding alignment is very tricky in the age of AI because you have the fog of AI, right? >> Yep. Are there things that you have done or seen work >> or just like a principle that you're operating on in dust so that you can >> not necessarily predict the future but >> iterate quickly when it does happen? >> Yeah, exactly. I think I think the uh the the failure mode of that is to uh uh refuse to paint a picture. I think you have to operate with uh some form of conviction within uncertainty. M >> it has to be clear to everyone that we are operating in a very uncertain environment and that whatever we say uh will be subject to change because the environment changed but that we create conviction on the actual current direction we're taking and so that means uh uh people I mean everybody wants to rally on a vision or rally on a direction and so if you cannot give a one year and two year direction you have to give a six months direction That's fine. And whenever you give a six month direction by month three, it's it's going to be updated. And so that means that you you you need to have that kind of muscle of updating it and being con providing a lot of conviction in in in that process but really being very explicit each time in doing it because otherwise you might fall in the trap of of not not creating that that that shared at least uh speed vector uh that you're trying to to have for the group. Uh and that's the worst that's the the the worst part. Basically, the longer you speak vector, the best it is. The shorter uh it's it's worse. Uh three months is uh so that's we're mixing time with speed here, but it's fine. Three months, what would be what would be anyway? Three months there's an interesting start. Three months is better than zero because then you don't know the direction. >> Yeah. And I imagine how it plays out in practice is every day folks are showing you new things and saying, "Hey, should we maybe go in this direction?" and you have to make those hard decisions of is it worth course correcting for this? Have we seen something that fundamentally changes where we want to go? >> Yep. Yeah. This uh so uh what we're trying to do is we're trying to really give a good picture of what we want to achieve within three to six months. But then we also uh so that's that's a kind of a a photo like it's a it's a snapshot uh and we try to update it but it's a kind of at the origin of bigo of months and then we uh we have what we actually doing and so we call it our stack ranks that's pretty easy it's a list of project that we actively working on and we try to create it very clearly so that the rest of everybody knows so engineering generally knows what they're working on not always true but most of the times better if true and enables everybody else within the company to know what we are changing on the product right now. >> Uh and and then that's that stack rank we really try to keep it very dynamic. So that means that we we have a process which we call stack rank update that is open to anybody in the company to update the stack rank and so uh obviously it gets windy at the top so it's not an easy process. It's not a that hard process either, but it's basically a button you can push and both founders appear in the room and we chat about how we should update the stack rank and everybody within the company has access to that to that. >> But generally we try to we expect we we have a high expectation for the input to that process depending on the on the size or the uh the kind of appetite that is associated with the updates. uh but uh but I think it's very nice because it gives to everybody to be a bit of a product person uh within a company and uh we're trying to push everybody to be a bit of product person inside of a company. Let's change gears and talk about something before we hit record. We were mentioning tokconomics and I think that is very hot topic these days because the general narrative you hear from the internet is oh the era of subsidies is over now folks are trying to go public. We're not going to get these cheap tokens. I'm not sure that I fully buy into that. Whenever I hear something from the masses I that automatically gives me like red flag. >> Yep. However, I do see that the costs are going up a lot and it is really easy to spend a lot of money and not create anything of value. >> So, there's like that [laughter] >> I can burn a whole hell of a lot of tokens. That doesn't necessarily mean I'm good at my job, right? uh and so I do understand that you were mentioning there's the cost of tokens but then there's also the cost of inference and so let's maybe center the next 5 10 minutes around that >> well I think the the way the way I uh we think or at least I think about that is uh we know the end state this uh we are subject to gravity this technology has a sailing it will converge and when it does it will be commoditized and tokens will be cheap plentiful that's still state >> the question is and maybe it will have asi by then but whether whatever is the state where it reach it will plateau at some point because it and there's a limit amount of energy and stuff you know uh and and I think that when when it plateaus it'll get communitized and it'll be pretty cheap very close to the price of uh of of power basically uh the question is when that happens maybe it happens in in a year maybe it happens is five years maybe in 20 years maybe it happens in 100 years who knows uh and so the question is is trying to think about what's the transition state between now and then but I guess the end state is plentifulness cheap intelligence from from electricity um and then well the one thing that I've kind of was a a new learning for me so I'll I'll share it because I think it it was a learning for me in the past couple weeks months is that I've always supposed that if models performance plateau we would enter a very rapid communistation phase and I think I've been coming back from that uh that assumption >> because the u the demand has increased so much that the pressure on inference so just the the fact of serving the tokens is so high >> that uh the actors that have uh pre-built that inference capacity we'll be able to maintain high margin due to that pressure >> which means that at least my learning is that even if I suppose that today is the best we'll ever get in terms of models which is probably not true but even if I make that uh that third experiment that doesn't mean that in six months the tokens will be close to zero because there is so much buildup demand and so much and some of that demand is might not be the best demand because as you said it's easy to to be token taxing on some pure AI slop But >> uh >> uh but this the demand is still there and people are still ready to pay for it and so there's there is kind of another pressure in the market for around that will maintain the price of token high but uh but anyway it's very hard to um uh to know there's a [sighs] there's also the other thing that you can say that is very easy to say that makes sense is we're getting in a world where those those models on some task are PhD level if not better. Who needs a PhD level to fill in uh Salesforce? >> Yeah. >> No one. But the counterargument to that there is two is uh who needs a MacBook M2 to do spreadsheets. >> Yeah. >> No one. But but everybody wants one. >> Still use it [laughter] >> because it's the best thing. And the more serious arguments is uh uh even if the model is a PhD level if you apply to task like filling cell force you might gain one line of of reliability and whenever you gain one line of relability some funky stuff interesting stuff happen you go from I look up the t I look up what's going on I'm checking the work to uh uh to I'm checking every rig to this thing is fully entirely automated and I will never ever in the history of humankind fill in a Salesforce uh card ever. >> So I think [laughter] >> and so I think there's a there's still an argument for using the best frontier model even for task that seems pretty not so PhD levelesque uh because of that kind of added nine of reliability on the task. Well, if the demand right now is the main constraint on what's making this plentiful, like you were saying, cuz 4.8 dropped recently and I've played around with it. I haven't seen that explosion of wow, this is so amazing >> and I need to use it all the time. >> There's a lot of talk about how Mythos is going to drop and it's going to be a gamecher, etc., etc. We'll see when that happens. But assuming in this thought experiment that the models are what they are and we have this pentup demand, the question in my mind is how long does this pent up demand stick around? Does it grow over time? Does it diminish over time? Is it something that stays stable? And so we just need to keep bringing more energy, more inference online to service that demand. >> Well, I think there's a there's a everybody I I presume everybody along the the the chain the the value chain will be happy to provide more offer, right? Yeah, the uh the extrac companies in every countries will be happy to build more nuclear plants and then the uh the cluster building companies will be happy to build new cluster in the place of old industrial places that are not used anymore and the GPU providers uh uh mostly one today but probably more competition tomorrow uh will be happy to provide the GPUs there and so uh I mean it will have economically it will have to equiliberate And um and we know where the uh I mean it's with those type of technology it always uh it always equalates to towards commit station. It's like your it's like your your your mobile phone uh your mobile phone service uh uh service you you pay 10 bucks 20 bucks a month or maybe a bit more in the US I don't know. Uh but it's it's so cheap compared to the value it provides you. M >> it's just because it's in fully commutized and it's incorporates to cost plus >> margin. >> It is fascinating to think about that and right now though where we're at living today I imagine that's another piece that makes the ground underneath you very shaky. >> Mhm. And so trying to build in going back to this metaphor of the fog of AI knowing that the costs aren't really stable, locked in, reliable. And so then you're going out there and talking to users and you have your own product pricing models and all of that fun stuff. And then the users are already having their own thoughts about the tokens that they're spending and how they're spending it. that makes for some interesting conversations. >> Yeah, definitely. I think these uh it's the fog of pricing I guess this time around. [laughter] Uh uh we uh uh interestingly at Dust we um so we started with the with a flat price uh and the uh the thinking was was the following. The thinking was you want a flat price because you want to encourage usage and value creation. So you want kind of a Chinese buffet type of pricing because they uh you pay once can use as much as you want and uh let's go for usage let's go for for for value creation. Um and it was also built on the assumption that eventually even if the eventually was very loosely defined eventually cost would go down and as it happens uh uh with the emergence of uh the models I mean with the awaking of the models uh at the end of last year uh it is not sustainable anymore for us and so we have to refactor our entire pricing to move to what is the industry standard today and I think it's the sustainable pricing unless you are uh willing to go extremely uh negative margins. Uh you you you you have to move to a crate based pricing >> because you don't control those costs. We don't know what's going to be made of tomorrow. We don't know what's maybe MTOS will be really a banger and it will be maybe really expensive at the same time and that's great. That's great if that's the case. That's great if you are in a crate based pricing world because you're like people just go for it. Use MOS if you want. It's an awesome model. Uh it's going to cost you a lot. Uh but we have I mean it's fine if you are on a kind of a flat price uh like typical SAS pricing. Uh you just you just like if a model drops awesome very powerful but super expensive just can't serve it just can't serve it because it's the economics don't work anymore. So I >> Oh, sorry. Go ahead. >> Yeah. No, sorry. So, I think the uncertainty in kind of the time horizon for the evolution of pricing makes it pretty much an unstable position to not have a critbased pricing in the market today. I think >> yeah, it is very dangerous to go per se and then especially if you're abstracting away the cost of the providers. There's some creative things you can do like you were talking about earlier of using smaller models, >> but that at the end of the day is not going to work, especially if folks are wanting these the best >> Yeah. Exactly. >> models all the time. >> For I mean maybe there is a set of task where I'm happy to use a small model, but for most of the stuff I do with agents, I I just want the best. >> Yeah. Well, and you don't it almost takes more time to figure out where to insert the small model. Y >> which slows you down and you just it's like well >> I know I could save a few cents here >> but let's try and use it because it's going to be a lot easier for me >> then if you set up and and that's one thing that has changed a lot recently too is how we don't necessarily set up by hand each step of this graph. We let the agents, we kind of throw the problem at the agents and let them figure it out. And so if the agents have the ability to choose the smaller models or kick off tasks with smaller models, that's great. I'm not going to go in there and like >> use a smaller model for these specific tasks when you are going and doing them in the plan. You know, maybe there are people that do and those folks, wow, hats off to you. For now, I imagine you could optimize >> micro micromanagement of latit. And you probably could get more lift or more cost savings just by optimizing certain context window tricks or the ability to make sure that your agents aren't constantly ingesting files. Maybe it's like cashing tricks or things like that that are going to be much I'm not going to say like much cheaper, but they are going to potentially give you more bang for your buck. >> Yep. >> So, there's there's all that fun stuff. Now, I have one thing that has been going through my head that I've been wanting to ask you, and I'm sure you hear it every day that you talk to customers, and it's, "Well, yeah, cool, dust. I see this vision, but can't I just do that with co-work as we were talking about earlier? Co-work is having its moment. What is the unlock on this like single player to multiplayer? I can share my slides with people or I can share my sessions with people from co-work. So where do I go with Dust? >> Yeah, I think there's a so there's there's three pillars to our differentation with co-work. uh it's uh so first on the on on that multiplayer there's some use cases that you can explain pretty you know as I did for the team weekly where I think you can paint a picture with you you it's very hard to do with co-work that kind of a collaboration around a unit of work that needs multiple people the uh I mean you can obviously do it but it requires more more kind of manual work and it's more it's less integrated I think there's a in terms of enterprise there's the uh governance part which is very important. Um having the ability to uh we we try to do a better job than code. Uh on that then uh be able to distribute in a control way the different uh MCP servers, different tools, the different skills uh limit the access to some data to some users have that things transitively managed so that people can discover the the agents and the skills they can use but in a way that is aligned with other what are the restrictions of the workspace. I think co-work with all the things turning locally, the skills being shared through kind of all of a lot of bands to some extent as of today um is a bit more a bit more of a of a of a far west when you want to have kind of a slightly more uh governed deployment of AI within a company. So that will talk to different type of company obviously. And finally the third pillar which is a pretty obvious one is like you don't want to get stuck in with one model provider. Especially as we were talking about tokconomics just >> exactly you don't want to be stuck with one token provider because the next best model you don't know where it's want to know it's coming from and you surely want to be have access to it >> and so I think it's a kind of a mix of those three we try to be very innovative on the product uh in terms of uh multiplayer AI and that's a differentiation that's something that we hope will awaken and become stronger and stronger as we go uh we want to be uh better at the governance and kind of the enterprise readiness uh as much as possible in at the product layer obviously uh uh we're not talking about we're not talking about FDES trying to implement stuff in your company but really as a product be more enterprise ready and then the third pillar is that kind of a is that kind of a is that kind of a a freedom of of freedom of choosing your model and being able to test open source models and keep your finger on that and have all of that in one central place. Okay. So, I think it's really that kind of a those three factors that makes us still a compelling option uh in the time where everybody really talks about co-work. But I'm I mean we've lived through other hype micro hype cycles. Uh we've had obviously chipd we've long enough we we've hold enough for for having lived as a company the emergence of Chajibility. Imagine uh and we've lived through the the the Glean moment like it was all about Glean two years ago. What happened? That's true. Gle. >> Well, I think it's still a very powerful uh enterprise search and AI answer product and I think they're doing actually really well. I think they just uh they just uh I've seen I don't know if it's official or not, but I've seen on on newspapers. I don't know if it's a true number, but they've reached 300 million revenue. It's a it's a very healthy growing company. But at that at that time, it was the it was really the darling. >> Yeah. The be all to end all. I remember those days in 2023 or 2024. >> Exactly. Exactly. >> So, there'll be there'll be other hype cycles. That's fine. I think we uh what we want to provide our users is really a platform where that is uh that is really uh building and and helping them being at the forefront of what can be done with uh with agents. >> So, at the risk of jumping all over the place. >> Yep. >> File systems. >> Yep. you we got into this earlier. Are you in a place where you feel like file systems can be updated for more agentic work andor I'll add to this longer question because the way the agents work and the way that they will almost fan out and do things. >> Yep. and do things very compute inensive things and then just hang out for a minute and be like hey human I want I want your feedback and maybe it's a minute maybe it is an eternity in computer time. >> Yep. Are there other areas where you feel like the current infrastructure that you have to build dust you need new tools well on the file system it's true that uh uh we can predict a moment where uh so right now we discuss about it we have a file system mounted on the pod we have a file system on sandbox some of the file system in the pod are shared you can have multiple agents working with that file system as of today it's mostly still human collaborative driven and so the kind of concurrency on that five system remains within typical human bounds >> but you could totally imagine like an army of agents collaborating on a project and then the five system becomes the wrong abstraction uh or I mean it's not necessarily the wrong abstraction but there's going to be there's going to be a concurrency pressure on it and it doesn't have any primitive for that I mean it does file locking but uh who uses file locking [laughter] >> agents will >> agents will agents might um uh but I think but to some extent this to me this has been sold there's there's a there's a couple companies uh that are building kind of a git infrastructure because uh if you want to scale providing GitHub repository agents uh using GitHub is not very convenient because there are rate limits that are meant for humans on the platform. And so there's a bunch of companies that are pretty exciting that that built Git Infra and Git infra seems to be a a fair fair solution to the uh to the problem because at least you have that uh resolution forced resolution step when you merge back to main. That makes kind of some sense. No, >> but at the same time, yeah, the uh uh but I think the uh yeah, as as we was chatting, I think one one step before that that excites me a lot is the uh is the kind of a stateful sandbox, the sandbox that stays live because I'm super excited about kind of a hyper specialized business apps. uh many of our users are already doing that with those frames that we uh mentioned. But those frames, they're not backed they're not backed by a particular sandbox. So they they they they kind of they kind of they they snapshots which mean that if you want to update it, you need to ask the agent to go fetch the data again and go update the frame. So what you really want is you want those frames to be backed by a sandbox that stays around because then you can have inside of a sandbox you can have a SQLite database and you can rebuild the CRM and ERP. You can rebuild whatever uh things you need to do your job. And the frame is the UI for the human. The sandbox is accessible by agents. And so you have an API for you have an a way to work with agents or directly as a human. Every human can create a different view of the the app. If let's say we are talking about feed a feedback system for the product where you automatically routt to different owners and that's something that you cannot manage to do well in GitHub or linear and so you build your own thing >> maybe the manager will want a different view than the IC which will want a different view than the person that report the feedback and that's having that ability is extremely exciting and in that world the kind of a de facto way to build that I think that is very easy is to is to do a shared nothing type of approach where the sandbox is the state and you just uh you just let the agents do whatever they want. Use whatever technology they want and you just snapshot the sandbox. You can go back in time easily and there is nothing more to it. There's no external database. There's no external g repository. I think that will that's kind of a very nice convenient way to package those kind of things for agents. Uh the only problem is like if you imagine that you have hundreds of agents collaborating on a shared sandbox then it starts failing and maybe then you need uh uh the codes to be managed by GitHub maybe you need the uh the data to be managed by a hosted database which so we'll see. >> So it works until you get to a certain scale which it doesn't feel like a lot of these workloads are at yet. >> Oh yeah. I mean it's so the moment where so imagine you are uh you're a small SMB uh in whatever countryside of uh of Germany and you uh and you do cookies and you're paying your you're paying your ERP uh 20K or 30K a year to use free features of an ERP that has 10,000 features. >> Yeah, >> that's that's the state of SAS today. And >> Exactly. And so you may want to rebuild that uh that that that process which is uh get me the do get me the thing and then here are the the comments and it's a pretty easy one inside of one of those kind of a specialized app and so uh there indeed there is no there's going to be a few people interacting with it maybe agents maybe directly through UIs and I think having that that's that that that shared no sing approach really works the moment where it would fail is where uh you start growing and you have thousands of humans if they interact through the UI that would the model would still work fine. The money would fail is like you know 10,000 people and and uh or you or you 10,000 you 10 times bigger and know uh you need to make that system evolve and you have agents working with it and changing the code at the speed of hundreds of agents at the same time. We can imagine that world but that's that's where it's going to fail but it's so far away uh that's uh it seems like fine to not overinure for it >> and even the current state of a large company you don't necessarily have the majority of your projects being hit hundreds of times a minute by agents so it's almost like I see a 20 8020 principle here where you could have that working for a a lot of your different projects, but then those projects that really need scale and they need that type of limiting or unlimitedness to it. >> Yeah. >> You would take those or you would architect those in a different way. >> Exactly. >> So that is stateful sandboxes. Are there any other hang-ups that we might encounter if we tried to do them? It feels like it could get expensive if I constantly have a sandbox running and I only use it once a week. It could be wasteful. >> That that that that we have really great companies that gives you the nice preventives there. You just shut them down. The time is boot up time is boot up time is 100 milliseconds. You don't have if if it's sharing the beauty of that is that you don't have to care about where is the DB do you have to shut down the DB blah blah no everything is in the sandbox SQLite database it's a file it's one megabytes to do most of any companies would do with anything serious it just works I think it's I really see as the prolongation of the Excel spreadsheets >> today we have the Excel spreadsheets uh no man's land and SAS [laughter] and it's kind of the uh it's kind of a way to it's it's it's a capture of a lot more that you can do is a spreadsheet that you were forced to do with SAS. And so that's funny because if you really believe in that vision, it feels like uh Wall Street was right way in advance for danging the SAS businesses. And that's it's pretty rare to think that Wall Street was really really right in advance, but um but uh that's yeah, that's the the fun the fun thoughts. Uh but yeah, I think there's a massive opportunity to have anybody create the the app they need to do their job. Uh that is equally accessible. The state is equally accessible by agents and the the UI is actually can be create that can be customized to every human basically. That's there's something very exciting over there. Yeah, especially if you're interacting with this through the chat ability or I I guess where my mind gets caught up and where I've never fully bought into this generative UI principle, although I'm coming around to it a little more, the more and more that I use things like the HTML that the agents will create on the fly, it's like, ah, this is this is actually kind of that idea have generative UI. But what I would get hung up on is that a lot of times we have learned and we are comfortable with certain tools because we know those five things that we need to do. We know how to do them and where to do them. If we now are being asked to create our own tool that does those five things and we have to spend the time creating it, even if it is going to give us a lot of success on the tail end because it's going to be much faster, it's going to be more custom. We don't have to spend as much money, >> it's a pretty hard ask to get get it set up. Yeah. But it's going to be as with any content economy ever. It's a one one person producer 99% consumer which is true of of spreadsheets as well. There are people that knows that know how to do a great spreadsheets and they're not that many of them >> and they are the ones building the spreadsheets and then there are the rest of us consuming in the dashboard and there's the rest of us consuming consuming them. Ah, so you're thinking there will be the builders of these. It's just not we're not all going to be expected to build it. There's going to be the specific builders and they're going to I like that idea of extending the spreadsheet. You're now going to get this micro SAS that is very specific to your business. >> Yep. >> And anybody can build if they want. It's just that uh we know how it where it lands. It's like some will build, some will consume. And >> yeah, the the one place that I think that is for sure having its questions is the no code, low code type of tools because that is 100% in the crosshairs of what we're talking about. Yep. However, I did just see a tweet of somebody laughing saying, "I am constantly surprised at how folks will set up with their agents a whole workflow to automate something that costs like 90 cents to run when with Zapier, you could do it for 5 cents." >> Yeah. Yeah. So, it goes back to that thing of like what costs verse convenience type of thing. And it is much more convenient to explain what we want, especially if you're using some dictation tool. You just mind dump the agent organizes it and says, "All right, let's try it." You test it. It works. It looks good. Ship it. >> Exactly. >> We'll see. It's hard to It's hard to predict where it's going to land, but it's surely there's surely a ton of um ton of exciting stuff to explore. Agreed. that the kind of on thely dynamic UI it's hard to see it fly right now just because of the latency but here it's kind of a slightly more async it's like uh you build once use many times but you can build very custom and a grid that's all of that is doable with uh with many no code platforms out there and it's uh it's very hard to understand why they why they didn't compare the world Yeah. Well, there there's also the other idea of just uh spitballing with you on product here cuz you're taking the approach of one sandbox that will have different views that people can drop into and they can look at. However, what I wonder and and I see it as the use case in my mind is you have a lot of customer calls. Like I'm sure you're on calls all day long talking with customers or talking with potential customers and some of the time of that call or the same call can have value for a salesperson. It can have value for a PM. It can have value for your engineering team. And so you want to take that transcript and you want to display the valuable parts to each party that's looking at it. And so going back to this analogy of all right, we've got data and we want to display it to different stakeholders in different ways, wouldn't it be or is there a world where I can as a PM have my sandbox and everything gets piped into that one place or maybe that is just like you don't need to do it at the sandbox level because you are doing it in all these different ways and you have something that's on top of all the sandboxes that's coagulating all of that data. >> Yeah, I mean some use cases will be leverable by just having the data and having agents just uh just process the data. There's a lot of that happening in dust today like we uh we have a let's say gong connector. So companies that have connected to gong they have all those gong calls that gets transcribed and recorded and that enables interesting use case where indeed product teams can have launch an agent do a deep dive on all of the customer calls for the past months tell me everything that they said about that feature or that thing and that's also completely possible today. uh is that is an exciting that that's something that's possible today. The kind of a micro SAS view of that would be uh you have a call, you dump the transcript, there's a it trigger in in a in a in a UI, it triggers an agent that will uh create stuff for the sales team, create stuff for a product team, uh auto assign tags for the product surfaces that are concerned, auto assign tags for the geography, and maybe you have a a way to say, okay, it looks good. uh and submit and then it gets uh stored in different places. I mean there are two ways of tackling the same problems. Uh but I do think that um uh uh um like there is many cases many use cases for uh kind of a DBbacked apps that will uh that that could be customized for a special way of doing work. uh I have I have many of uh my own ad dust on that could change a little bit the way we operate. We kind of we can always kind of a uh you know you you bend the way you want to operate to the tools that that that you use uh the way you share uh goals for your week or your days uh at at desk happens on Slack. Mhm. >> So it has a certain form factor that is bended compared to the ideal state. Uh the way we make decisions we do that on GitHub which is interestingly a fun place to do because it helps kind of asynchronous collaboration and sometimes it's done on Slack and it's bended in very different ways depending on the platform you use and so there's so many processes and ways to ways to work that we that are shaped by our by our tools and that the other way around and so that's what I find that's pretty exciting. Ah, so then you would have the creative control to do what you want and not be beholden to what the tools force on you. >> Yeah. Yeah, >> I could see that. >> And and the my imagination is going wild as I think through what you could do in those instances. I like this like databasebacked apps or just a sandbox with a database that's stateful is a very fun thing to ponder. Have you felt like the way that agents use databases, we need a new type of database or the way that agents are constantly going around and trying to collect data? And the reason I bring this up is because I was just reading that quote unquote old paper. I think it was from December 2025. It was like supporting our AI overlords and it's written by you know some of the greats I think mate Srya and Yan Stoka and uh Shrea Shankar and so they they mentioned how you have this these agents and what they called it was like agent speculation. >> Mhm. >> So the agents will go and they kind of have a theory and then they'll try and find the data to support that theory and execute on what they need to do. But a lot of times if they're trying to get data from a database, it's not that efficient to pull the whole row or create this join from these gigantic tables or whatever it may be that you're like, can we have agents interact with data in a more efficient way? Yeah, I mean there surely is many things to invent there, but I'm also I'm a bit of a pragmaticist on those questions. And uh at the end of the day, when you think about that technology, it's trained on the internet. It's trained from uh reinforcement learning from human feedback. It's trained on on on human traces being built by many labs, etc. And so my point is that it's a technology that is extremely extremely extremely anthropomorphic. >> Uhhuh. And so it and it feels it actually we had that weird moment where we needed rag. >> Yeah. >> And it was kind of a new way of presenting the data made for agent made for the constraint of the context size that was rather small. uh and with the context size augmenting and the uh the very anthropomorphic nature of that technology the what's the best way to do search in company data today is just let's do what people call agentic search which is basically just let's use the tool that the human use if we can do it they can do it as well >> and we're ready to pay the tax of the latency associated uh because it's just much simpler to just give them the tool and let them uh let them do the work and so I'm I think I'm a um it feels like uh as the context is getting longer the the the the way those those those agents use tools is looking much more I mean it's looking very much like we use tools and so if we didn't have the need in the past why would they interesting question >> and they've been on what the humans will expect of it so >> yep >> if you and they've been on all these traditional tools part also so if you're now creating new tools that you're expecting the agents to use. Well, it may take a few cycles. I I'm fascinated by that space just because I feel like you've got this whole new way of interacting with machines through agents and the infrastructure that we have. It's fun to think about is it good enough? Can we do like you were saying, can we explore and not be beholden to what tech currently >> constrains us to or is there like the other piece the other side of the coin that you're saying be more of a pragmatist and say look I think if we can use it this way the agents can use it this way. It's it's obviously going to be always a bit of a bit of a bit of a bit of both. But um even in the way we use tool, we use them very differently. Uh it's like we use UIs, they use APIs turned MCP turned CLI >> uh and eventually it's the CLI and so I would love to use every SAS in the world with a CLI myself. So in a sense at the end of the day we merge back into the most efficient place. Uh I might use I might use the the CLI that agents have been uh that have been built for agents for using using the SAS tool myself. I'm a big of a fan of the terminal. But >> I just saw a tool I can't remember what it was. It was like turn any SAS into a CLI. That was their whole tool. I can't remember what it was, but that is uh >> it's exciting to me. Yeah, [laughter] really. It's Yeah, it's a different level that we're playing at. Well, man, this has been great. Is there anything that you want to mention before we jump? Is there anything that I didn't ask you about? >> No, I think this was great. We covered the we covered the the main subject. I think we on the verge of going towards multiplayer AI. We don't quite know yet what it looked like, but it surely won't look like what we're doing today. And so that's uh that's why I I think it's a in the in that wall where there are two black holes being created uh in the market. I think uh it's always an interesting questions of uh are they not going to going to take it all? uh we we we we really attached to the to the to the thinking that even if the technology uh again plateaus uh even with current model shape uh the way we're going to walk in 10 years is probably nothing compared to the way we walk today and so there's still many many many opportunities to build for for everyone. >> I will also finish with dust is known throughout the world for having the absolute best inerson events. So if you see a dust event happening near you, I highly recommend that you go. >> Yeah, that is for sure. They are awesome events. Anybody that has been to one will tell you the same. So thanks for doing this, dude. I really >> Thank you very much. It was [music] great. [music]