The Agent Network — Dharmesh Shah, Agent.ai + CTO of HubSpot

Hey, everyone. Welcome back to the Late in Space podcast. This is Alessio, partner and CTO at Decibel Partners, and I'm joined by my co-host Swyx, founder of Smol AI.

Hello, and today we're super excited to have Dharmesh Shah to join us. Uh, I guess your relevant title here is founder of Agent.ai.

Yeah, that's true for this. Uh- Yeah, creator of Agent.ai and-

Yes

... yeah, co-founder of HubSpot. But yeah.

Co-founder of HubSpot, which I followed for many years, uh, I think 18 years now.

Yeah.

Uh, gonna be 19 soon. And you caught... You know, uh, people can catch up on your HubSpot story elsewhere. I, I should also thank Shawn Puri, who's, uh, who I've chatted with back and forth, uh, who's been, I guess, getting me in touch with your people, but also I think, like, just giving us a lot of context 'cause obviously My First Million joined you guys-

Yep

... um, and then, and they've been, they've been chatting with you guys a lot. So for the business side, we can talk about that-

Yeah

... but I kinda wanted to engage your CTO agent engineer-

Sure

... side of things.

Yeah.

So how did you get agent religion?

Let's see. Um, so I've been working, uh... I'll take a h- like, a half step back. Um-

Yeah

... decade or so ago, even b- actually more than that, so even before HubSpot, the company I was contemplating, that I had a name for, was called IngeniSoft, and the idea behind IngeniSoft was a natural language interface to business software.

Now realize this is 20 years ago, so that was a hard thing to do. But the actual use case that I had in mind was, um, you know, we had data sitting in business systems like a CRM or something like that, and my kind of what I thought clever at the time was, oh, what if we used email as the kind of interface to get to business software?

And the motivation for using email is that it is automatically, uh, works when you're offline. So imagine I can... Like, I'm getting on a plane, or I'm on a plane. There was no internet on planes back then. It's like, oh, I'm going through business cards from an event I went to, and I can just type things in into, uh, into an email, just have them all in the backlog.

When it reconnects, it sends those emails to a processor that basically kind of parses the, effectively the commands, um, and updates the software, sends you the file, whatever it is, and there was a handful of commands. Um, I was a little bit ahead of the times in terms of what was actually possible.

And I reattempted this natural language thing, uh, with a product called ChatSpot that I did back, uh, 20-

Yeah, this is your first, like, post-ChatGPT project.

Yeah.

I saw it come out.

Yeah. Um, and so I've always been kinda, um, kinda fascinated by this, uh, natural language interface to, to software because, you know, as software developers, myself included, we've always said, "Oh, we build intuitive, easy-to-use applications," and it's not intuitive at all, right?

Because what we're doing is taking the mental model that's in our head of what we're trying to accomplish with said piece of software and translating that into a series of touches and swipes and clicks and things like that, uh, and there's nothing natural or intuitive about it.

Um, and so natural language interfaces, for the first time, you know, the, whatever the thought is you have in your head expressed in whatever language that you normally use to talk to yourself in your head, you can just sorta emit that and have software do something.

And I thought that was a kind of a breakthrough, uh, which it has been, uh, and it's gone... So that's where I first started getting into, um, kind of the generative AI side of it because now it actually works, right?

So once we got, uh, ChatGPT, and you can take, uh, even with a few shot example, uh, convert something into structured, even back in the, uh, Chap- the ChatGPT 3.5 days, it did a decent job at a few shot example, convert something to, you know, to structured text, um, if you knew what kinds of intents, uh, you were gonna have.

Um, and so that happened, um, and that ultimately became a, a HubSpot project. But then agents intrigued me, uh, because I'm like, okay, well, that's the next step here. So chat's great. Love chat UX. Um- ... but if we wanna do something even more meaningful, it felt like the next kinda, um, advancement is not this kinda I'm chatting with some software in a kind of a synchronous back-and-forth model, is that software's going to do things for me-

Mm

... um, in a kinda multi-step way, uh, to try and accomplish some goal. So, um, yeah, that's when I first got started. It's like, okay, what would that look like? Um, yeah, so... And I've been obsessed ever since, by the way, so.

Which goes back to your first experience with it, which is, like, you're offline, and you-

Yeah

... wanna do a task. You don't need to do it right now. You just wanna queue it up for somebody-

Yeah

... to do it for you.

Yes.

As you think about agents, like, let's start at the easy question, which is, like, how do you define an agent? Maybe.

You mean the hardest question in the universe? Is that what you mean? Um...

You said you have an irritating take.

I do have an irritating take. Um, I think, well, some number of people have been irritated, uh, including within, uh, within my own team. So I have a very broad definition for agent, um, agents, which is it's AI-powered software, um, that accomplishes a goal, period.

That, that's it. And what irritates people about it is like, well, that's so broad as to be completely non-useful, and I under- ... and I understand that. Understand the, the criticism. But in my mind, um, if you kind of fast-forward months, I guess, in, uh, in AI years, um, the implementation of, and we're already starting to see this, and we'll talk about this, um, uh, different kinds of agents, right?

So I think in addition to having a, a usable definition—and I like yours, by the way, and we should talk more about that, uh, that you just came out with. The classification of agents actually is also useful, which is, is it autonomous or non-autonomous?

Is it, uh... Does it have a deterministic workflow? Does it have a non-deterministic flow- workflow? Is it working synchronously? Is it working asynchronously? Uh, then you have the different kinda, uh, interaction modes. Is it a chat agent, kinda like a customer support agent would be, you're having this kind of back and forth?

Is it a workflow agent that just does a discrete number of steps? Um, so there's all these different flavors of agents. So if I were to draw it in a Venn diagram, I would draw a big circle that says, "This is agents," and then have a bunch of circles, some overlapping, um, because they're not, uh, mutually ex- um, exclusive.

And so I think that's what's interesting, and we're seeing, uh, development along a, a bunch of different paths, right? So if you look at the first implementation of, like, uh, agent frameworks, uh, you look at, uh, you know, BabyAGI and, um, AutoGBT, I think it was.

Not AutoGen. That's the Microsoft one. They were way ahead of their time because they assumed this level of reasoning and execution and planning capability That just, just did not exist, right? So it was an interesting experiment-- thought experiment, um, which is what it was.

Uh, even, uh, you know, the guy that, uh, I'm an investor in Yohei's, uh, Yohei's fund that did, um, BabyAGI. It wasn't ready, but, like, it was a sign of what was to come.

Mm-hmm.

And so the question then is, like, when, when is it ready? Um, and so, you know, lots of people talk about, like, the state-of-the-art when it comes to agents. I'm a pragmatist, so I think of, like, the state of the practical.

It's like, okay, well, what can I actually build, uh, that has commercial value, uh, or solves actually some discrete problem, uh, with some, you know, kind of some baseline of, uh, kind of repeatability or, um, you know, verifiability.

So, yeah.

There was a, a lot and very, very interesting-- I'm not irritated by it at all.

Okay.

I think as, as you know, I take a sort of anthropological view or linguistics view. In ling- in linguistics, you don't want to be prescriptive, you wanna be descriptive.

Yep.

So you're a goals guy.

Yep.

That's, that's the key word in your thing.

Yep.

And other people have other definitions that might involve, like, delegated trust-

Yeah

... or non-deterministic work, LLM in the loop-

Yep

... all that stuff.

Yep.

The other thing I, I was also thinking about, like, just co- the comment on BabyAGI or OGPT. Yeah, in that piece that you just read, we we- I, I was able to go through our backlog-

Yep

... and just kind of track the, the winter of agents and then the summer now.

Yep.

And it, it really, it really-- You can-- We can tell the whole story, like, as, as an oral history just following that-

Yeah

... thread. And it's really just, like, I think-- Uh, I tried to explain the, the why now, right? Like, I had there's better models, of course. There's better tool use, uh, with, like, they're, they're, they're just more reliable.

Yep.

Better tools with MCP and all that stuff.

Yep.

And I'm sure you have opinions on that too.

I do.

Business model shift, which you like a lot.

Yep.

I just, I just heard you talk about RaaS with, uh, MFM guys.

Yep.

Cost is dropping a lot.

Yep.

Uh, inference is getting faster. There's mo- model diversity, which I think is a subtle point. It, it means that, like, you have different models with different perspective. You don't get stuck in a-- the, the basin of a performance of a single model.

Sure.

You can just get out of it by just switching models.

Yep.

Multi-agent research and RL fine-tuning. So I just wanted to let you respond to, like, any of that.

Yeah, a couple things. Um, connecting the dots on the kinda the definition side of it, so we'll get the irritation out of the way, um, completely. I have one more even more irritating leap on the-

Yeah

... agent definition thing.

Please.

Um, so here's the way I think about it. By the way, the, the kind of word agent, I looked it up at, like, the English dictionary definition-

The old school agent, yeah

... is when you have someone or something that does something on your behalf, like a travel agent or a real estate agent, acts on your behalf as a proxy, uh, which is a nice kinda general definition. Um, so the other direction I'm sorta headed, and it's gonna tie back to tool calling and MCP and things like that, is if you, um...

And I'm not a biologist, uh, by any stretch of the imagination, but we have these, uh, single-celled organisms, right? Like the simplest possible form of what one would call life. But it's still life. It just happens to be single-celled.

And then you can combine cells o- and then cell, cells become specialized over time, and you have much more sophisticated organisms, you know, kinda further down the spectrum. In my mind, at the most fundamental level, you can almost think of, uh, having atomic agents.

What is the simplest possible thing that's an agent that can still be called an agent? What is the equivalent of a kinda single-celled, um, organism? And the reason I think that's useful is right now we're headed down the road, which I think is very exciting, around tool use, right?

Um, that says, okay, the LLMs now can be provided a set of tools, uh, that it calls to accomplish whatever it needs to accomplish in the kinda, uh, furtherance of whatever goal it's trying to get done. And I'm not overly bothered by it, but if you think about it, if you just squint a little bit and say, well, what if everything was an agent?

And what if tools were actually just atomic agents? Because then it's turtles all the way down, right? Then it's like, oh, well, all that's really happening with tool use is that we have a network of agents that know about each other through something like an MCP and can kinda decompose a particular problem and say, "Oh, I'm gonna delegate this to this set of agents," and why do we need to draw this distinction between tools, which are functions, uh, most of the time, and an actual agent?

Um, and so I'm gonna write this irritating LinkedIn post, uh, you know, proposing this. It's like, okay. Um, and I'm not suggesting we should call even functions and, you know, call them agents, but there is a certain amount of elegance that happens when you say, oh, we can just reduce it down to one primitive, uh, which is an agent that you can combine in, in-

You can build it

... uh, complicated ways to kinda, you know, raise the level of abstraction and accomplish higher order goals. Anyway, that's my-

Yeah, I'd say that's a relative success.

Thank you for coming to my TED Talk on agent definitions.

What-- How do you define the, yeah, minimum viable agent? Do you already have a definition for, like, where you draw the line between a cell and a atom or?

Yeah. So in my mind, it has to, at some level, use AI in order for it to-- otherwise, it's just software, right? It's like, you know, we don't need a-another word for that. And so that's probably where I draw the line.

So then the question w- you know, the counterargument would be, well, if that's true, then lots of tools themselves are actually not agents because they're just doing a database call or a REST API call or whatever it is they're doing, and that does not necessarily qualify them, uh, which is a fair counterargument, and I, I accept that.

It's, like, a good argument. Um, I still like to think about, 'cause we'll talk about multi-agent systems, 'cause I think-- So we've accepted, which I think is true, lots of people have said it, uh, and you've helpfully, uh, combined some of those clips of, uh, really smart people saying, this is the year of agents, and, and I completely agree.

Um, it is the year of agents. But then shortly after that, it's going to be, uh, the year of, uh, multi-agent systems or multi-agent networks. I think that's where it's gonna be headed next year. Um, yeah.

OpenAI's already on that.

Yeah.

My quick philosophical engagement with you on this cell thing is, uh, I often think about kind of the other spectrum, the other end of the cell spectrum. So single cell is life, multi-cell is life, and you clump a bunch of cells together s- in a, in a more complex organism, they become organs-

Mm-hmm. Yeah

... like an eye and a, a liver, whatever.

Yep.

And then obviously we are, we consider our- us ourselves one life form.

Sure.

There's not, like, a lot of lives within me.

Yep.

I'm just one life. And now obviously, like, I don't think people don't really like to anthropomorphize agents and-

Yep

... and AI, but we are extending our, our consciousness and our brain and our functionality out into machines.

Mm-hmm.

I just saw you wore a V.

Yeah.

Which is, uh, you know, it's-

I have a little, little pendant in my pocket, uh, which I-

I got, I got one of these boys. Yeah, I'm testing it all out. You know, gotta, gotta be early adopters. But, like, we want to extend our personal memory into these things-

Yep

... so that we can we be good at the things that we're good at, and, you know, machines are good at and machines are there. So, like, my definition of life is kind of like going outside of my own body now- I don't know if you've ever had, like, reflections on, on that, like how yourself is, like, actually being distributed outside of you.

Uh, yeah, I'm not... I don't, um, fancy myself a philosopher. Uh, but-

Yeah, you went there, so...

Yeah, I d- I did go there. Um- I'm fascinated by kinda graphs and graph theory and networks and have been for a long, long time. Uh, and, and to me, we're sorta all nodes in this kinda larger thing.

It just so happens that we're looking at individual kinda life forms as, as they exist right now. But, um, so the idea is, uh, when you put a podcast out there, there's these little kinda, uh, nodes you're putting out there of like, you know, conceptual ideas.

Once again, you have varying kind of forms of, uh, those little nodes that are up there and are connected in, in varying and sundry ways. And so I just think of myself as being a node in a massive, massive network, uh, and I'm producing more nodes as I, you know, put content or ideas.

Um, and, um, you know, you spend some portion of your life collecting dots, uh, experiences, people, um, and some portion of your life then connecting dots from the ones that you've collected over time, and I've found that really interesting things happen, uh, and you, you really can't know in advance how those dots are necessarily going to connect, um, in the future, and that's, yeah.

So that's my philosophical take.

See Joss.

That's the... Yes, exactly.

See Joss coming back.

Uh, yep.

Do you like graph as a agent abstraction? That's been one of the hot topics with LangGraph and Pydantic and all that.

Um, I do. Uh, the thing I'm more interested in in terms of, uh, use of graphs, and, and there's lots of work happening on that now, is, uh, graph data stores as an alternative in terms of knowledge stores and knowledge graphs.

Knowledge graphs. Yeah.

Yeah.

Uh, because, you know, so I've been in software now, uh, 30-plus years, right? So it's not 10,000 hours, it's like 100,000 hours-

Mm-hmm

... that I've spent doing this stuff. And so I've grew up with, uh, so back in the day, you know, I started on mainframes. Uh, there was a product, uh, called IMS from IBM, which is basically a index database, what we'd call like a key-value store today.

Uh, then we've had relational databases, right? We have tables and columns and foreign key relationships. Uh, we all know that. We have document databases like MongoDB, which is sort of a nested structure keyed by a, a specific index.

Uh, we have vector stores, vector embedding database. And graphs are interesting, uh, for a couple reasons. One is, so it's not classically structured in a relational way with... When, when you say structured database to most people, they're thinking tables and columns and relational database and set theory and all that.

Graphs still have structure, but it's not the tables and column structure. Um, and you could wonder, and people have made this case, uh, that they are a better representation of knowledge for LLMs and for AI generally than other things.

Uh, so that's kinda thing number one conceptually, and that might be true. I think it is possibly true. Uh, and the other thing that I really like about that in the context of, um, kinda data stores for, for RAG is, you know, RAG you say, "Oh, I have a million documents.

I'm gonna build the vector embeddings. I'm gonna come back with the, you know, top X based on the semantic match," and that's fine. All that's very, very useful, but the reality is something gets lost in the chunking process, in the, okay, well those 10...

Yeah, like you don't really get the whole picture, so to speak, and maybe not even the right, uh, kinda set of dimensions on the kind of broader picture. And it, it makes intuitive sense to me, uh, that if we did capture it properly, uh, in a graph form, that maybe that feeding into a, a RAG pipeline, um, will actually yield better results for some use cases.

I, I don't know, but, uh, yeah.

And do you feel like at the core of it there's this difference between imperative and declarative programs? Because if, if you think about HubSpot-

Yeah

... it's like, you know, people and graph kinda goes hand in hand.

Yep.

You know?

Yep.

But I think maybe the software of before was more like primary foreign key-

Yeah

... based relationship-

Yep

... uh, versus now the models can traverse through the graph more easily.

Uh, yes. Uh, so I, I like that representation. There's something, um, just conceptually elegant about graphs and just in the representation of it. Uh, they're much more discoverable. You can kinda see it. There's observability to it, um, versus kinda embeddings which, uh, you can't really do much with, uh, as a human.

Um, you know, once they're in there, you can't pull stuff back out. But yeah, I, I like that kinda idea of it. And the other thing that's kinda, uh, because I love graphs, I've been long obsessed with PageRank from back in the early days.

Um, and you know, one of the kinda simplest algorithms in terms of coming up, uh, you know, with auth- and everyone's been exposed to, uh, PageRank. But the idea is that... And so I had this other idea for a project, not a company, and I have hundreds of these, um, called NodeRank, is to be able to take the idea of PageRank and apply it to an arbitrary graph, um, that says, okay, I'm gonna define what authority looks like and say, okay, well, that's interesting to me because then if you say, "Oh, I'm gonna take my knowledge store," and maybe this person that contributed some number of chunks to the, um, to the graph, uh, data store has more authority on this particular use case or prompt that's being submitted, uh, than this other one that

may... Or maybe this one was more popular, or maybe this one has... Whatever it is, there should be a way for us to kinda rank nodes in a graph and sort them in, in some, some useful way. Uh, yeah, so...

I, I think that's generally useful for, for anything. I think the, the problem, like, uh, so even though at my conferences GraphRAG is super popular and, uh, people are getting knowledge graph religion, and I would say, like it's getting space, getting traction in two areas.

Mm-hmm.

Conversation memory, and then also just RAG in general, like the, the, the document data source. Most ML practitioners would say that knowledge graph is kind of like a dirty word. Uh, the graph database. People get graph religion.

Yeah.

Everything's a graph, and then they, then they go really hard into it, and then they get a, they get a graph that is too complex to navigate.

Yes.

Um, and so, like, the, the, the simple way to put it is, like, you at, at running HubSpot, you know the, the power of, of graphs, the, the way that Google has pitched them for-

Mm-hmm

... for many years.

Sure.

But, uh, I don't suspect that, uh, know- that HubSpot itself uses a knowledge graph.

No.

Yeah. So, uh, when is it overengineering, basically?

It's a great question. Um, I don't know. So the question now, like in AI land, right, is the do we necessarily need to understand? So right now, uh, LLMs for, for the most part are somewhat black boxes, right?

So we sort of understand how the, you know, the algorithm itself works, but we really don't, uh, know what's going on in there, um, and, and how things come out. So if, uh- A graph data store is able to produce the outcomes we want.

It's like, here's a set of queries I wanna be able to submit, and then it comes out with useful content. Maybe the underlying data store is as opaque as, uh, vector embeddings or something like that, but maybe it's fine.

Maybe we don't necessarily need to understand it to get utility out of it. And so maybe if it's messy, that's okay. Um, that's... it's just another form of lossy compression. Uh, it's just lossy in a way that we just don't completely understand i-in terms of because it's gonna grow organically, uh, and it's not structured.

It's like, "All right, we're just gonna throw a bunch of stuff in there and let the, the equivalent of the embedding algorithm," whatever they're called in, in graph land, um-

So the one with the best results wins?

I think so, yeah. Or it's... This is the practical side of me is like, "Yeah, it's... if it's useful-

I-

" ... we don't necessarily need to understand it."

I have... I mean, I'm happy to push back as long as you want.

Sure, you should.

Uh, it's not practical to evaluate, like, the 10 different options out there 'cause it takes time, it takes people, it takes, you know, resources, right?

Yep.

Set... That's the first thing. Second thing is your evals are typically on small things.

Yep.

And some things only work at scale-

Yep

... like graphs.

Yep, yep. That's actually... Yeah. No, that's fair. And I think this is one of the challenges in terms of implementation of graph databases is that the most common approach that I've seen developers do, I've done it myself, is that, "Oh, I've got a Postgres database or a MySQL or whatever.

I can represent a graph with a very set of-

I know this

... simple set of tables with a parent-child thing-

Yeah, yeah

... or whatever, and that sort of gives me the ability. Uh, why would I need anything more than that?" And the answer is, well, if you don't need anything more than that, you don't th- need anything more than that.

Yeah.

But you... There's a high chance that you're sort of missing out on the actual value that, uh, the graph representation gives you, which is the ability to tra- traverse the graph, uh, efficiently in ways that kind of going through the, uh, traversal in a relational database form, even though structurally you have the data, practically you're not gonna be able to pull it out in, in useful ways.

Uh-

Yeah

... so you wouldn't, like, represent a social graph, uh, in, in using that kind of relational table model. It just wouldn't scale.

Yeah.

It wouldn't work. Uh, yeah.

Uh, I think we wanna move on to MCP.

Yeah.

But I just wanna... Like, just engineering advice.

Yeah.

Uh, obviously you've, you've, you've run... uh, you've, you've had to do a lot of projects and run a lot of teams. Do you have a general rule for over-engineering or eng- you know, engineering ahead of time? You know, like, because people...

We know premature engineering is the root of all evil.

Yep.

But also sometimes you just have to.

Yep.

When do you do it?

Yes, it's a great question. This is a, it's a question as old as time almost, which is what's the, you know, right and wrong levels of abstraction? 'Cause that's effectively what, uh, we're answering when we're trying to do engineering.

I tend to be a pragmatist, right? So here's the thing. Um, lots of times doing something the right way has only a marginal increased cost. In those cases, just do it the right way, and this is what makes a, a great engineer or a good engineer better than a, a not so great one.

It's like, okay, all things being equal, if it's gonna take you, you know, roughly, you know, close to constant time anyway, might as well do it the right way. Like, so do things well. Then the question is, okay, well, am I building a framework?

Is this a reusable library? To what degree? Uh, what am I anticipating in terms of what's gonna need to change in this thing, uh, you know, along what dimension? And then I think like a business person in some ways, like, what's the return on calories, right?

So, uh- And, and you look at, um-

Energy, yeah

... the expected value of... It's like, okay, here are the five possible things that could happen. Uh, try to assign probabilities. Like, okay, well, if there's a 50% chance that we're gonna go down this particular path at someday, like, or one of these five things is going to happen, and it costs you 10% more to engineer for that, it's...

basically it's something that yields, uh, kind of interest, compounding value, um, as you get closer to the time of n- of needing that versus having to take on debt, which is when you under-engineer it, you're taking on debt that you're gonna have to pay off when you do get to that eventuality where, uh, something happens.

One thing as a pragmatist, uh, w- So I would rather under-engineer something than over-engineer it if I were gonna err on the side of something, and here's the reason, is that when you under-engineer it, uh, yes, you take on tech debt, uh, but the interest rate is relatively known and payoff is very, very possible, right?

Which is, oh, I took a shortcut here as a result of which now this thing that should have taken me a week is now going to take me four weeks. Fine. But if that particular thing that you thought might happen never actually...

you never have that use case transpire or just doesn't... it's like, well, you just saved yourself time, right? And that has value because you were able to do other things instead of, uh, kind of slightly over-engineering it or way over-engineering it.

But there's no perfect answer. It's an art form in terms of, uh... and yeah. We'll, we'll bring kind of the layers of abstraction back on the code generation conversation, which we'll, uh, I think have later on, but-

I was gonna ask, we, we can just jump ahead quickly.

Yeah.

Like, as you think about vibe coding and all that-

Yeah

... how does the percentage of potential usefulness change when... I feel like with over-engineering a lot of times it's, like, the investment in syntax.

Yeah.

It's less about the investment in, like, architects acting.

Yep.

Yeah. How does that change your calculus?

A couple things, right? One is, um... So when, you know, going back to that kind of ROI or, uh, return on calories kind of calculus or heuristic, you think through. It's like, okay, well, what is it going to cost me to put this layer of abstraction above the code that I'm writing now, uh, in anticipating kind of future needs?

If the cost of fixing, uh, or doing under-engineering right now, uh, will trend towards zero, that says, "Okay, well, I don't have to get it right right now because even if I get it wrong, I'll run the thing for six hours instead of 60 minutes or whatever.

It doesn't really matter," right? Like, because that's gonna trend towards zero, the a- the ability to refactor code. Um, and because we're gonna-

Yeah

... not that long from now we're gonna have, you know, large code bases be able to exist, uh, you know, as, as context, uh, for a code generation or a code refactoring, uh, model. So I think it's going to make it, um, th- make the case for under-engineering, uh, even stronger, which is why take on that cost?

You just pay the interest when you get there. It's not, um, just go on with your life, vibe code it, and, uh, come back when you need to.

Yeah. Sometimes I feel like there's no decision-making in some things. Like, uh, today I built a autosave for, like, our internal notes platform.

Sure.

And I literally just asked on Cursor-

Yeah

... can you add autosave?

Yeah.

I don't know if it's over or under-engineer.

Yep.

I just vibe coded it.

Yep.

And I feel like at some point we're gonna get to the point where the models kind of decide where the right line is.

Yeah. But this is where the, like, the, in, in my mind, the danger is, right? So there's two sides to this. One is, uh, the cost of kind of development and coding and things like that, stuff that, you know, we talk about.

But then, like in your example, you know, one of the risks that we have is that because adding a feature, uh, like a save or whatever the feature might be to a product, as that price tends towards zero, are we going to be less discriminate about what features we add as a result-

Mm-hmm

... of making more product, products more complicated, which has a negative impact on the user, a neg- negative impact on the business. Um, and so that's the thing I worry about. If it starts to become too easy, are we going to be- Too promiscuous in our, uh, kinda extension, adding product extensions and things like that.

It's like, ah, why not-

Yeah

... add X, Y, Z, whatever. Back then, it was like, "Oh, we only have so many engineering hours or story points," or however you measure things. Uh, that at least kept us in check a little bit. Uh-

Yeah, and in over-engineering, you're like, yeah, it's kinda like you're putting that on yourself.

Yeah.

Like now it's like the models don't understand that if they add too much complexity, it's gonna come back to bite them later.

Yep.

So they just do whatever-

Yep

... they wanna do. Yeah, and I'm curious where in the workflow that's gonna be, where it's like, hey, this is like the amount of complexity in over-engineering you can do before you gotta ask me-

Yeah

... if we should actually do it-

Yep

... versus, like, do something else.

I think that... So, you know, we've already, it's like we're leaving this, uh, in the code generation world, this kinda compressed, um, cycle time, right? It's like, okay, we went from auto-complete, uh, in the GitHub Copilot to like, oh, finish this particular thing and hit Tab, to a, oh, I sorta know your file or whatever, I can write out a full function to you, to now I can, like, hold a bunch of the context in my head, uh, so we can do app generation, which we have now with Lovable and Bolt and Replit Agent and other things.

So then the question is, okay, well, where does it naturally go from here? So we're gonna generate products. Makes sense. We might be able to generate platforms. It's still I want a platform for ERP that does this, whatever, and that includes the APIs, includes the product, and the UI, and all the things that make for a platform.

There's no- nothing that says we would stop. Like, okay, can you generate entire software companies someday, right? Uh, f- with the platform and the monetization and the go-to-market and the whatever. And, you know, that, that's interesting to me in terms of, uh, you know, what...

when you take it to almost ludicrous levels of- ... uh, abstract. It's like, okay.

Turn it to 11.

Yeah.

You mentioned vibe coding, so I have to, uh... This is a blog post I haven't written, but I'm kind of exploring it.

Yeah.

Is the junior engineer dead?

I don't think so. I think what will happen is that the junior engineer will be able to... If all they're bringing to the table, uh, is the fact that they are a junior engineer- ... uh, then, then, yes, they're likely dead.

But hopefully, if they can communicate with carbon-based life forms, they can interact with product, if they're willing to talk to customers, they can take their kinda basic understanding of engineering and how, uh, kinda software works, I think that has value.

So I have a 14-year-old right now-

Yeah

... that's taking Python programming class. Uh, and some people ask me, it's like, "Why is he learning coding?" And-

Yeah, we ask this a lot

... and my answer is, is because it's not about the syntax, it's not about the coding. What he's learning is, like, the fundamental thing of, like, how things work, and there's value in that. I think there's going to be timeless value in systems thinking and, uh, abstractions and what that means, uh, in whether functions manifested as math, which he's gonna get exposed to regardless or...

There are some core primitives to the universe, I think, uh, that the more you understand them, uh, those are what, what I would kinda think of as, like, really large dots in your life that will have a higher gravita- gravitational pull and value to them that you'll then be able to cl- uh, so I want him to collect those dots, and he's not resisting, so it's like, okay.

Um-

Yeah

... while he's still listening to me, I'm gonna have him do things that I think will be useful.

Yeah. You know, part of one of the pitches that I evaluated for AI engineer-

Yeah

... as a term, is that maybe the traditional interview path or career path of software engineer goes away, which is because what, what's the point of LeetCode?

Yeah. Yeah.

And, uh, you know, it, it actually matters more that you know how to work with AI-

Yep

... and to implement the things that you want.

Yep. That's one of the, like, interesting things that's happened with generative AI. Um, you know, you go from machine learning and the models and just that underlying form of it, which was, like, true engineering, right? Like the actual, uh, what I call real engineering.

I don't- ... think of myself as a real engineer actually. I'm a developer. Um, but now with generative AI, we call it AI, uh, and it's obviously got its roots in, in machine learning, but it just feels like fundamentally different to me.

Like, you have the vibe. It's like, okay, well, this is just a whole different approach to software development, to s- uh, so many different things. And so I'm wondering now, it's like, an AI engineer is like, if you were, like, to draw the Venn diagram, it's interesting because the cross between, like, AI things, uh, generative AI and what the tools are capable, what the models do, and this whole new kind of body of knowledge that we're still building out, it's still very young, intersected with, uh, kinda classic engineering, uh, software engineering.

Yeah. I, I, I just described the overlap as it separates out eventually until it's its own thing.

Yeah.

But y- it's-

Yeah

... uh, starting out as a software thing.

Yeah.

That makes sense. So to close the vibe coding loop-

Sure. Yeah

... uh, the other big hype now is MCPs.

Yeah.

Obviously, I would say Cloud Desktop and Cursor-

Yeah

... are like the two main drivers of-

Yeah

... MCP usage.

Yep.

I would say my favorite, the Sentry MCP. I can pull in errors-

Yep. Yeah

... and then you can just put the context in.

Never used it. Okay.

In Cursor. How do you think about that abstraction layer? Does it feel almost too magical in a way? Do you think it's like you get enough? Because you don't really see how the server itself is then kinda-

Yep

... like repackaging the information for you.

Yep. Uh, I, I think, uh, MCP as a standard, um, is one of the better things that's happened in the world of AI because a standard needed to exist, and absent a standard, there was a set of things that just weren't possible.

Now, we can argue whether it's the best possible manifestation of a standard or not. Does it do too much? Does it do too, too little? I, I get that, but it's like just, like, simple enough to both be useful and understandable and adoptable by mere mortals, right?

It's not overly complicated. Uh, you know, uh, a reasonable engineer can put a, stand up an MCP, um, uh, server relatively easily. The thing that has me excited about it is like, uh, so I'm a big believer in, um, multi-agent systems, and so this going back to our, uh, kinda this idea of an atomic agent.

Uh, so imagine the MCP server, like obviously it calls tools, but the way I think about it... So I'm working on, uh, my current passion project is Agent.ai. Um, and we'll talk more about the... I think we should 'cause I think it's interesting.

Not to promote the project at all, but, uh, there's some interesting ideas in there. One of which is around we're gonna need a mechanism for, uh, if agents are going to collaborate and be able to delegate, um, there's gonna need to be some form of discovery, uh, and we're gonna need some standard way.

It's like, okay, well, I just need to know what this thing over here is capable of. We're gonna need a registry, which Anthropic's working on. I'm sure others will and have been, uh, doing, uh, directories of, of, uh...

And there's gonna be a standard around that too, so how do you build out a, a directory of MCP servers in a standard way? I think that's gonna unlock so many things just because... And we're already starting to see it.

So I think- MCP or something like it is going to be the ma- next major unlock because it allows, uh, systems that don't know about each other, don't need to. It just dec- it's that kinda, uh, decoupling of, like, Sentry and whatever tools someone else was building, and it's not just about, uh, you know, Claude Desktop or things like...

But even on the client side, I think we're gonna see very interesting consumers of MCP, uh, MCP clients versus just the chat body kinda things, uh, like, you know, Claude Desktop and, uh, and Cursor and things like that.

But yeah, I'm very excited about, uh, MCP in that general direction.

I think the, the typical cynical developer take, it's like we have OpenAPI.

Yeah.

What's the new thing? Uh, I don't know if you have a... Do, uh, do you have a quick, uh, MCP versus everything else?

Yeah, so it's, so... I like OpenAI, right? So just a descriptive thing-

No, uh, it's OpenAPI

... OpenAPI. Yes, that's what I meant. Uh, it's, so it's, it's basically a, a self-documenting thing. We can do machine generated, lots of things from that output. It's a structured definition of an API. I get that, love it.

Um, but MCPs, uh, sorta are kinda use case specific. They're perfect for exactly what we're trying to use them for around LLMs in terms of discovery. It's like, okay, I don't necessarily need to know kinda all this detail.

And so right now we have, uh, we'll talk more about, like, MCP, uh, server implementations, but-

We will?

I think it-

Oh, okay.

I, I, I don't know. Um, maybe we won't. Um, but-

We should

... uh, at least this is in my head.

Yeah, yeah. This is, let's go there. Let's go there

... it's like a background processor. But, um-

Yeah

... I do think MCP, uh, as a pro- adds value above OpenAPI. Um, it's, yeah, uh, just because it solves this particular thing. And if we had come to the world, which we have, like, it's like, hey, we already have OpenAPI.

It's like if that were good enough for the universe, the universe would've adopted it already. There's a reason why MCP is taking off. It's because marginally adds something that was missing before and doesn't go too far, and so that's why the kinda rate of adoption, you, you folks have written about this and talked about it.

Uh-

Yeah, why MCP won.

Why... Yeah. And, and, and it won because the universe decided that- ... this was useful. That's why. And maybe it gets-

Yeah

... supplanted by something else, and maybe we discover, ah, maybe OpenAPI was good enough the whole, the whole time. I, I doubt that, uh-

Yeah

... but we'll see.

It, it... The, the meta lesson, this is... I mean, he's a, he's an investor in com- DevTools companies. I work in developer experience at DevRel in com- in DevTools companies.

Yep.

Everyone wants to own the standard.

Yeah.

Uh, I'm sure you guys have tried to, have launched your own standards. Actually, it's, it's HubSpot known for a standard, you know, obviously inbound marketing, but is there a standard or protocol-

No

... that you ever tried to push?

Uh, no. And, and there's a reason for this-

Yeah

... is that, uh,... And I don't mean, mean to mean, uh-

No, no

... speak for the people of HubSpot, uh, but I personally-

You kinda do.

... am not s- am not smart enough. That's not the, uh, like I, I think I have a-

You're smart

... uh, not enough for that. I'm much better off understanding the standards that are out there, and I'm more on the composability side. Let's, like, take the pieces of technology that exist out there, combine them in creative, unique ways.

Um, and I like to consume standards. I don't like to... And that's not that-

Not make them

... I don't like to create them. I just don't think I have the, both the raw wattage, uh, or the credibility. It's like, okay, well, who the heck is Dharmesh, uh, and why should we adopt a standard he created?

HubSpot, right? Yeah.

Uh.

Yeah, I mean, there are people who don't monetize standards. Uh, like OpenTelemetry-

Yeah

... was, is a big standard, and LightStep never capitalized on that. So, so no.

S- okay. So if I were to do a standard, there's two things that have been in my head in the past.

Ooh.

Uh, one around, um, a very, very basic one around... I don't even have the domain for... I have a domain for everything. Um, um, for open marketing, um, 'cause the, the issue we had in the... HubSpot grew up in the marketing space.

There we go.

There was no standard around data formats and things like that. It doesn't go anywhere. But the other one, um, and I did not mean to go here, but I'm gonna go here. Um, it's called OpenGraph. I know the term was already taken, but it hasn't been used for, like, 15 years now for, um, for its original purpose.

But what I think should exist in the world is right now our information, all of us, uh, nodes are in, uh, the social graph at Meta or the professional graph at LinkedIn, both of which are actually, uh, relatively closed in actually very annoying ways.

Um, like very, very closed, right? Uh, especially LinkedIn. And I personally believe that, um, if it's my data, and if I would get utility out of it being open, I should be able to make my data open or publish it in whatever forms that I choose, as long as I have control over it.

It's opt-in. Uh, so the idea is around OpenGraph that says here's a standard, here's a way to publish it, and I should go to open... I should be able to go to opengraph.org/dharmesh.json and get it back and it, it's like, "Here's your stuff," right?

And I can choose along the way, and people can write to it, can, and I can approve, and there can be an entire system. Uh, and I, and if I were to do that, I would do it as a, uh, like a public benefit, uh, non-profity kinda thing as, just as a, a contribution to society.

I wouldn't, uh, try to commercialize that. Yeah.

Have you looked at At Proto?

What's that?

At Proto.

At who?

The, it's the protocol behind Blue Sky.

Okay.

My good friend Dan Abramov, who was the, uh, face of React for many, many years, uh, now works there and, you know, he, he actually did a talk that I can send you which basically kinda tries to articulate what you just said.

But he does, he loves doing these, like, really great analogies, which I think you'll like, which is, like, you know, a lot of our data is behind a handle, behind a domain.

Yep.

So he's like, "All right, what if we flip that? What if it was, like, our handle and then the domain?"

Yep.

So, um, and, and that's really, like, the... Your data should belong to you.

Yep.

And-

Yeah

... I should not have to wait 30 days for my Twitter data to export.

Yep, totally. And you should be able to at least, yeah, be able to automate it or do, like, yes, I should be able to plug it into an agentic thing. Um-

Yeah

... yes. I, I think we're, um... 'Cause so much of our data is, uh, locked up.

I think the trick here isn't the, that standard.

Yeah.

It is getting the normies to care.

Yeah.

'Cause normies don't care.

That's true. But building on that, normies don't care. Um, so, you know, privacy is a really hot topic and an easy word to use, but it's not a binary thing. Like, there are use cases where, and we make these choices all the time, that I will trade not all privacy, but I will trade some privacy for some productivity gain or some benefit to me that says, "Oh, I don't care about that particular data being online if it gives me this in return," or, "I don't mind sharing this information with this, uh, company if I'm getting, you know, this in return."

But that sh- sorta should be my option.

I think now with computer use you can actually automate some of the exports.

Yes, yes.

Like, uh, something we've been doing, um-

Oh, okay

... internally is, like, everybody exports their LinkedIn connections-

Yep

... and then internally we kinda merge them together to see how we can connect our companies-

Yeah

... to customers or things like that. Um-

And n- not to pick on LinkedIn, but, uh, since we're talking about it, but they Are-- feel strongly enough on the, you know, do not take LinkedIn data, that they will block even browser use kind of things or whatever.

They go to great, great lengths, uh, even to see patterns of usage, uh, that says, "Oh, there's no way you could have, you know, gotten that particular thing or whatever without..." And it's, um... So it's... There's-

Wasn't there a Supreme Court case that they lost?

Yeah. So the one they lost was around, uh, someone that was scraping public data that was on the, uh, on the public internet, and that particular company had not signed any terms of service or whatever. It's like, "Oh, I'm just taking data that's on."

There was no, uh... And so that's, that's why they won. But now the, you know, the question is around can LinkedIn, and I think they can. Like, when you use, as a user, you use LinkedIn, uh, you are signing up for their terms of service, and if they say, "Well, this kind of use of your LinkedIn account, that violates our terms of service," they can shut your account down, right?

They can. Uh, and they... Yeah, so, you know, we don't need to make this a- ... a discussion, uh... By the way, I l- I love the company, don't get me wrong. I'm an avid user of the product, uh, you know, got, um-

Yeah, I mean, you got over a million followers on LinkedIn, I think so.

Yeah, I do and I've-

Um, they should listen to you

... and I've known people there for a long, long time, right? That I have lots of respect, and I understand even where the mindset originally came from of this kind of members first approach to, uh, you know, uh, privacy first.

I, I sort of get that, but sometimes you sort of have to wonder. It's like, okay, well, that was fifteen, twenty years ago. There's likely some controlled ways to expose some data on some members' behalf and not just completely, uh, be a binary.

It's like, "No, thou shalt not have the data." Um, but-

Oh, just pay for Sales Navigator.

So- Right.

Before we move to the next layer-

Sure

... of abstraction, anything else on MCP you mentioned, um-

Let's move back, and then I'll, I'll tie it back to MCPs. Um, so I think we opened this with Agent.ai. Okay, so I'll start with here's my kind of running thesis, is that as AI and agents evolve, um, which they're doing very, very quickly, we're going to look at them, uh, more and more...

I don't like to anthropomorphize. We'll talk about why this is not that. Less as just like raw tools, uh, and more like teammates. They'll still be software. They should self-disclose as being software. Uh, I'm totally cool with that.

But I think what's going to happen is that, uh, in the same way you might collaborate with a team member on Slack or Teams or whatever you use, you can imagine a series of agents that do specific things, just like a team member might do, that you can delegate things to, you can collaborate, you can say, "Can you take a look at this?

Uh, can you proofread that? Can you try this?" You can u- uh, whatever it happens to be. So I think it is, uh, I will go so far as to say it's inevitable that we're going to have hybrid teams someday.

And what I mean by hybrid teams, so back in the day, hybrid teams were, oh, well, you have some full-time employees and some, uh, contractors. Then it was like hybrid teams are some people that are in the office and some that are remote.

That's the kind of form of hybrid. The next form of hybrid is like the carbon-based life forms and agents and AI and some form of, uh, form of software. So let's say we temporarily stipulate that I'm right about that, uh, over some time horizon, that eventually we're gonna have these, uh, kind of digitally hybrid, uh, teams.

So if that's true, then the question you sort of ask yourself is that then what needs to exist in order for us to get the full value of that new model? It's like, okay, well, you sort of need to...

It's like, okay, well, how do I, if I'm building a digital team, like how do I... Just in the same way if I'm interviewing, uh, for an engineer or a designer or a PM, whatever, it's like, well, that's why we have professional networks, right?

It's like, oh, they have a presence on some, uh, likely LinkedIn. I can go through that semi-structured, structured form, um, and I can see the experience or whatever, uh, you know, self-disclosed. But okay, well, agents are going to need that someday.

Uh, and so I'm like, okay, well, this seems like a thread that's, uh, worth pulling on that says, okay, so I create... So Agent.ai, um, is, is out there, um, and it's-

LinkedIn for agents.

Uh, it's LinkedIn for agents. It's a professional network for agents. And the more I pull on that thread, it's like, okay, well, if that's true, like what happens, right? It's like, oh, well, they have a profile just like anyone else, just like a human would.

It's going to be a graph underneath, just like a professional network would be. It's just that... And you can have its, uh, you know, connections and follows, and agents should be able to post. That's maybe how they do release notes.

Like, "Oh, I have this new version." Whatever they decide to post, um, it should just be able to behave as a node on the network of a professional network. And as it turns out, the more I think about that and pull on that thread, the more and more things, like start to make sense to me.

So it may be more than, uh, just a pure professional network. So, um, so my original thought was, okay, well, it's a professional network, and agents as they exist out there, which I think there's gonna be more and more of, will kind of exist on this network, um, and have their profile.

But then, and this is always dangerous, I'm like, okay, I want to see a world where thousands of agents are out there, uh, in order for the... Uh, because those digital employees, the digital, uh, workers don't exist yet, uh, in any meaningful way.

And so then I'm like, oh, can I make that easier, uh, for like... And so I have, as one does, it's like, "Oh, I'll build a low-code platform for building agents. How hard could that be?" Right? Like very hard as it turns out.

Uh, but, uh, it's been fun. So now Agent.ai has one point three million users. Three thousand people have actually, you know, built some variation of an agent, sometimes, uh, sometimes for just for their own personal productivity, uh, about a thousand of which have been published.

Uh, and the reason this comes back to MCP for me... So imagine, uh, that and other networks, but I'll just, uh, since I know Agent.ai. So we... Right now, uh, we have an MCP server for Agent.ai, uh, that exposes all the internally built agents that we have that do, like super useful things.

Like, you know, I have access to Twitter API that way, like I can subsidize the cost, and I can say, "You know, if you're looking to build something for social media, we'll build like these kinds of things with a single API key," and it's all completely free right now.

I'm funding it. Um, that's a useful way for a developer to say, "Oh, I have this idea. I don't have to worry about OpenAI. I don't have to worry about now, you know, this particular model is better. It has access to all the models with one key," and we proxy it kind of behind the scenes, um, and then expose it.

So then we get this kind of community effect, right, that says, "Oh, well, someone else may have built an agent to do X." Like, I have an agent right now that I built for myself to do domain valuation for website domains because I'm obsessed with domains, right?

And like, there's no efficient market for domains. There's no Zillow for domains right now that tells you, "Oh, here are what houses in your neighborhood sold for." It's like- Well, why doesn't that exist? We should be able to solve that problem, and yes, you're still guessing.

Fine. Uh, there should be some simple heuristic, so I built that. It's like, okay, well, let me go look for past transaction. You say, "Okay, I'm gonna type in agent.ai, agent.com, whatever domain. Uh, what's it actually worth? I'm looking at buying it."

It can go and say, "Oh," And which is what it does. It's like, "I'm gonna go look at are there any published, uh, domain transactions recently that are similar, either use the same word, same top level domain," whatever it is, and it comes back with an approximate value, and it comes back with its kinda rationale for why it picked the value end comparable transactions.

"Oh, by the way, this domain sold for, uh, published..." Okay. So le- that agent now, uh, let's say, existed on the web, uh, on Agent.ai, then imagine someone else says, "Oh, you know, I wanna build a brand-building agent, uh, for startups and entrepreneurs to come up with names for their startup."

Like, a common problem every startup's like, "Ah, I don't know what to call it." And so they type in five random words that kinda define whatever their startup is. Uh, and you can do all manner of things, uh, one of which is like, "Oh, well, I need to find the domain name for it.

Uh, what are possible choices?" Now it's like, okay, well, it would be nice to know if there's an aftermarket price for it, if it's, is it listed for sale? Awesome. Then imagine it calling this valuation agent. It's like, okay, well, I wanna find where the arbitrage is, where the agent valuation tool says this thing is worth $25,000.

It's listed on GoDaddy for $5,000. It's close enough. Let's go do that, right? Now, that's a kinda composition use case that in my future state, thousands of agents on the network, all discoverable through something like MCP, and then you as a developer of agents have access to all these kind of Lego building blocks based on what you're trying to solve.

Um, then you blend in orchestration, which is getting better and better with the reasoning models now. Just describe the problem that you have. Now, the next layer that we're all contending with is that how many tools can you actually give an LLM before the LLM breaks?

That number used to be, like, 15 or 20 before, um, you know, you kinda manage, started to vary dramatically, and so that's the thing I'm thinking about now is like, okay, if I wanna expose 1,000 of these agents to a, a given LLM, obviously I can't give it all 1,000.

Is there some intermediate layer that says, "Based on your prompt, I'm gonna make a best guess at which agents might be able to be, uh, helpful for this particular thing"?

Yeah.

So.

Yeah, like RAG for tools.

Yep.

Um, I did-

Right

... build a latent space researcher on Agent.ai.

Okay. Nice.

Um, yeah, that seems like a, you know... Then there's gonna be a latent space scheduler, and then once I schedule-

Yeah, there you go

... the research-

Yep

... you know, and you build all of these things.

Yep. By the way, my apologies for the user experience. Uh, you realize-

No, no, no. It's pretty-

... I'm in here. It's like, yeah

... it, it, it's pretty good. I, I would say-

I think it's a normie-friendly thing.

Yeah, uh-

That's, that's your magic. You know, HubSpot does the same thing.

Yeah, just to, like, quickly run through it, you can basically create all these different steps, and these steps are, like, you know, static versus, like, variable-driven things. How did you decide between this kinda, like, low-code-ish versus doing, you know, low-code with code backend-

Yep

... versus, like, not exposing that at all?

Yep.

Any fun design decisions?

Yeah, and this is, I think, um, I think lots of people are likely sitting in exactly my position right now, coming through the choosing between deterministic... Like, if you're, like, in a business or building, you know, some sort of agentic thing, do you decide to do a der- deterministic thing, uh, or do you go non-deterministic and just let the LLM handle it, right, um, with the reasoning models?

The original idea and the reason I took the low-code stepwise, uh, very deterministic approach, A, the, um, reasoning models did not exist at that time. That's thing number one. Thing number two is if you can get a... If you know in your head what the actual steps are to accomplish whatever goal, why would you leave that to chance?

Right.

There's no upside. There's, uh, literally no upside. Just, just tell me, like, what steps do you need executed? So right now what I'm playing with, um, so one thing we haven't talked about yet, and, and people don't talk about UI in agents.

Mm-hmm.

Uh, right now the primary interaction model, uh... Or they don't talk enough about it. I know, uh, some people have. But it's like, okay, so we're used to the chatbot back and forth. Fine. I get that. I think we're gonna move to a blend of, uh, some of those things are gonna be, uh, synchronous as they are now, but some are gonna be async.

It's just gonna put it in a queue, just like... And this, uh, goes back to my... Man, I talk fast, um, but, um, I have this, uh... I only have one other speed that's even faster, um. So imagine it's like if you're working...

So back to my, oh, this, we're gonna have these hybrid digital teams. Like, you would not go to a coworker and say, "I'm gonna ask you to do this thing," and then sit there-

Right

... and wait for them to go do it. Like, that's not how the world works. So it's nice to be able to just, like, hand something off to someone that's like, okay, well, maybe I expect a response in an hour or a day or something like that, and there's some implicit contract that we have, um, with our coworkers in terms of when things, uh, need to happen.

So the UI around agents. So, so if you look at the output of Agent.ai agents right now, they are the simplest possible manifestation-

Yeah

... of a UI, right? That says, oh, we have inputs of, like, four different types. Like, we've got a dropdown.

Mm-hmm.

We've got multi-select. We... All the things, it's, like, back in HTML, uh- ... the original HTML 1.0 days, right? Like, here are the smallest possible set of, uh, primitives for UI. Uh, and it says, okay, because we need to collect, uh, some information, uh, from the user, and then we go do steps and do things and generate some output in HTML or markup are the, the two, uh, primary examples.

So the thing I've been asking myself, if I keep going down that path, like, so people ask me, I get requests all the time, it's like, "Oh, can you make... The UI is sort of boring. I need to be able to do this," right?

And if I keep pulling on that, it's like, okay, well, now I've built an entire UI builder thing.

Right.

Where does this, like, where does this end? And so I think the right answer, and this is what I'm, uh, gonna be back coding once we, I get done here, is around injecting a code generation UI generation into-

Mm

... the Agent.ai flow, right? As, as a builder, you're like, "Okay, I'm gonna describe the thing that I want," much like you would do in a vibe coding world.

Yep.

But instead of generating the entire app, it's gonna generate the UI that exists at some point, uh, in either that deterministic flow or something like that. It says, oh, here's the thing I'm trying to do. Go generate the UI for me.

And I can go through some iterations, um, and what I think of it as a, um... So it's like I'm gonna generate the code, generate the code, tweak it, and go through this kinda prompt style like we do with vibe coding now.

And at some point I'm gonna be happy with it, and I'm gonna hit Save, and that's gonna become the action in that particular step. It's like a caching of the-

Mm-hmm

... generated code that I can then now f- like, incur any inference time cost. It's just the actual code at that point. Um, so.

Yeah, I invested in a company called E2B, which does Code Sandbox, and they-

Okay

... powered the LM arena web arena.

Okay. Yep.

So it's basically the... Just like you do LMSys, like text to text, uh, they do the same for, like, UI generation.

Yeah. Yep.

So if you're asking a model, how do you do it?

Yep.

But yeah, I think that's kinda where-

Yeah

... this should go.

That's something I'm, I'm really fascinated by. So, you know, the early, um, LLMs, you know, were understandably but laughably bad at simple arithmetic, right? That's the thing like my wife and normies would ask us, like, "You call this AI?

Like, it can't-

Right.

"My son would be like, it's just stupid. It can't even do, like, simple arithmetic." And then, w- like, we've discovered over time that, uh... and there's a reason for this, right? It's like, it's a large... There's, you know, the word language in this is in there for a reason in terms of what it's been trained on.

Uh, it's not meant to do math. But now it's like, okay, well, the fact that it has access to a Python interpreter that I can actually call at runtime, that solves an entire body of problems that it wasn't trained to do, and it's basically a form of delegation.

And so the thought that's kind of rattling around in my head is that that's great, so it's, it's, like, took the arithmetic problem and took it first. Now, like anything that's solvable through a relatively concrete, um, a Python program, it's able to do a bunch of things that it couldn't do before.

Can we get to the same place with UI? I don't know what the future of UI looks like in a agentic AI world, but maybe let the LLM handle, but not in the classic sense. Maybe it generates it on the fly, or maybe we go through some iterations and hit cache or something like that, so it's a little bit more predictable.

Uh, I don't know. But...

Yeah. And especially when is the human supposed to intervene?

Yeah.

So especially if you're composing them-

Yep

... most of them should not have a UI, because then they're just-

Yeah

... web hooking to-

Yep

... to somewhere else. I just wanna touch back. I don't know if you have more comments on this.

I was just gonna ask, when you, you said you go- you're gonna go back to code, what are you coding with? What's your stack?

Yep. Uh-

What's your-

So Python's my language. Uh, I, like, I'm, I'm glad that it won in terms of the AI, uh, language- ... its lingua franca, uh, like it or not.

It's the second-best language for everything.

Uh, by the way, they're, I think, exact like N of one of things that I disagree with Brett Taylor on, uh, when, when he was on, and just generally I'm a massive Brett Taylor fan. Uh, smart, one of my favorite people in tech.

Like, it was, like, a segment in there, he was talking about, like, "Oh, we need a, a, a different language than Python or whatever that is, like, built for-" "... uh, built for AI and built..." It's like, no, Brett, I don't think we do, actually.

It's, it's, it's just fine. Um, it c- it deals with it just fine, just expressive enough, and, a- and it's nice to have a language that we can use as a common denominator across both humans and AI. It's...

It doesn't slow the AI down enough, but it does make it awfully useful for us to also be able to participate in that kind of future world, uh, that we can still be somewhat useful. Anyway, but yeah. So it's, uh, Python, uh, Cursor as my, uh, kinda code gen thing.

Yeah, I w- I would also, uh, mention that I really like your code generation thing. I have another thesis I haven't written up yet about how generative UI has kind of not fulfilled its full potential.

Mm-hmm.

We've seen the Bots and Lovables, and those are great. And then Vercel has a version of generative UI that is basically a function calling pre-made components, and there's something in between where you should be able to generate the UI that you want and-

Yep

... pin it and stick to it, and that becomes your form.

Yep.

Your... Yeah. And, and so the, the way I put it is, um, you know, I think that the two form factors of agents that have seen a lot of product market fit recently has been deep research-

Mm-hmm

... and the AI builders, like the Bot- Lovables.

Yep.

I think there's some version of this where you generate the UI, but you, you sort of generate the Mad Libs and fill in the blanks-

Yeah

... forms.

Yep.

And then you f- you, you keep that stable-

Yep

... and the deep research just fills that in.

Yeah. Yep.

And that's it.

I like that. Yeah. Um, so I, I, I, I love those, uh, kinda simple, uh, simple implementations and kinda abstractions. But then if you look at the kinda, I'll say almost, like, the polar opposite of that. So, so right now most of the UIs that you and I think about or conceive or even examples are based on the primitives and the vocabulary that we have for UI right now.

It's like, oh, we have text boxes, we have check boxes, we have radio buttons, we have pull-downs, we have nav, we have clicks, touches, swipes, now voice, whatever it is. The set of primitives that exist right now, we will combine them in, uh, in interesting ways.

But where I th- AI is going to be headed on, I think, on the UI front is the same place it's headed on the science front, that originally it's like, oh, well, based on the things that we know right now, it'll sort of combine them.

But we're, like, right at the cusp of it being able to do actual novel research. So maybe a future version of AI comes up with a new set of primitives that actually work better for human-computer interaction than things that we've done in the past, right?

It's like I don't think it's, it ended with the, uh, the checkbox, radio button, and dropdown list, right? I think there's life beyond that. Uh, yeah.

I know we're gonna move to business models after, but when you talked about hybrid teams, one way we talk to folks about it is, like, you had offshoring... You had onshoring, which is like, you know, move to cheaper place in the country, then offshoring, you know, it's like AI shoring.

Yep.

You're kinda moving some roles to AI.

That's a thing people say, AI shoring?

Yeah.

That's the first I've ever heard of that.

Ooh.

Yeah.

Yeah. Um-

I don't know, man.

But I, I think to me the most interesting thing about the professional networks is, like, with people, you have limited availability to evaluate a person.

Yep.

So you have to use previous signal-

Yep

... as kinda like a evaluation thing. With agents, theoretically, you can have kinda like proof of work.

Yeah.

You know? You can run simulations and, like, evaluate them in that way.

Yep.

How do you think about that when running, building Agent.ai even?

Yep.

It's like, you know, instead of just choosing one-

Yep

... I could, like, literally just run across all of them-

Yeah

... and figure out which one is gonna work best. Um-

I, I'm a big believer. So, uh, under the covers when you build an, um, a... Because the primitives are so simple, you have some set of inputs. We know that what the variables are. Every agent that's on Agent.ai automatically has a REST API that's callable in exactly the way you would, uh, you'd expect.

Automatically shows up in the MC- MCP server, so you're able to invoke it in whatever form you, uh, decide to. And so my expectation is that in this future state, whether it's a human hiring, uh, an agent to, to do a particular task or evaluating a set of five agents to do a particular task and, uh, picking the best one for their particular use case, we should be able to automate that.

It's like I just wanna try it, um, and there should be a policy that the publisher builder of the agent has that says, "Okay, well, I'm gonna let you call me 50 times, 100 times before you have to pay," or something like that.

Uh, we should have effectively, like, an audit trail. Like, okay, this agent has been called this many times. We also have, uh, kinda human ratings and reviews right now, and we have tens of thousands of reviews of the existing agents on Agent.ai.

Average is, like, 4.1 out of five stars. And all those things are nice signals to be able to have. But the, the kinda callable, uh, verifiable kinda thing, I think is super useful. Like, if I can just call, uh, an A...

Give me an API that says, "Here are five agents," and it solves this particular problem for me. If I have, like, a simple eval, I think that'd be so powerful. I, I wish I had that for humans, honestly.

That would be so cool. Uh-

Yeah, because I mean, w- when I was running engineering teams, people would try and come out with these rubrics-

Yep

... you know, when hiring-

Yeah

... and it's like

They're not really helpful, but you just kinda need some ground truth.

Yeah.

But I feel like now, say you wanna hire, yeah, a AI software engineer.

Yep.

You can literally generate like 15, 20 examples of like your actual issues-

Yep

... in your organization-

Yep

... both from a people perspective of like collaboration and like actual code generation-

Yep

... and just pay for it to run it.

Yeah.

Like today we do take-home projects and we pay people.

Sure, yeah.

Like this should be kind of the same thing.

Yeah.

It's like I'll just run you. But I feel like people are not investing in their own evals-

Yes

... as much internally.

They're not. I mean, that's the present company included, right? Um. Like everyone talks about evals. Everyone accepts the fact that we should be doing more with evals. I won't say nobody, but almost nobody actually does. Uh, that's the...

And yeah, and it's a topic for a whole other day. I'm not-

Yeah.

Yeah.

It's funny, 'cause, I mean, 'cause obviously HubSpot is famous for launching graders of things.

Yes.

You'd be perfect for it.

Yeah.

Somehow. I don't know. I agree on evals, by the way. Uh, I mean, I just force myself to be the human in the loop or-

Yep

... you know, someone I work with, and, and that's okay. But obviously the scalable thing needs to be done. Just a fun fact on or question on AI, Agent.ai, you famously-- you've already talked about the Chat.com acquisition-

Mm-hmm. Yeah

... and all that, and that was around the time of custom GPTs-

Yes

... and the GPT store launching.

Yes.

And I definitely feel Agent.ai is a kind of the GPT score, but not taken seriously.

Yep.

Do you feel OpenAI if, like they woke up one day and they were like, "Agent.ai is a thing, like we should just reinvest in GPT store." Is that a fear sort of?

Um, I think that's an... It's not, won't be Agent.ai, uh, driven. It'll-- It's an inevitability that OpenAI... I, I don't have any insider information.

Yeah.

I'm an investor, but uh, no insider information. Is because it makes too much sense, um- ... but for them not to-- Like, and they, they've taken multiple passes at it, right? They did the plugins back in the day-

Mm

... then the custom GPTs and, and the GPT store. Because, you know, being the platform that they are, I think it's inevitable, um, that they will ultimately come up with, um... And they already have custom G- You know, it, it's gonna happen.

I'm not-- You know, on the list of things I promised myself I would never do is compete with Sam Altman ever, uh, not intentionally anyway. Um, and-

But here you are.

But yeah, here I am. But I'm not really, right? It's not, um, not really. It's, uh-

I mean, it's free, so like, whatever. But you know, at some point-

Yeah. But I mean, he's-

... if it's actually valuable

... he's-- they're solving a much, much bigger problem. I'm like a small, tiny rounding error in the universe. Um, but the reason that compelled me to actually create it in the first place, 'cause I knew custom GPTs existed, and I did have this, uh, rule in my head that, you know, don't compete with Sam.

He's literally like at the top of my list of people not to compete with. He's so good. Um, but the thing that I needed in terms of for my own personal use, which is how Agent.ai got started, 'cause I was building a bunch of what I call solo software, things for my own personal productivity gain, and I found myself doing more and more kinda LM-driven stuff because it sh- was better that way as, you know, AI, AI sort of showed up in those, uh, solo projects a, a bunch.

And so the thing I needed was an underlying framework to kinda build these things, and high on the list was, I wanna be able to, uh, straddle models. Because certain steps in the thing is like, oh look, for this particular, uh, thing, uh, involves writing, so maybe I wanna use Claude.

For this particular thing, maybe I wanna do this. Even around image generation, different types of, uh, whether it has text or doesn't have text or whatever, and I wanna be able to mix and match. And my sense is that whether it's OpenAI or Anthropic or whatever, they're likely going to have an affinity for their own models, right?

Which makes sense for them. But, like, I can sorta be, for my own purposes and for our user base, uh, a little bit of the Switzerland. It's like we don't think there's like one model to rule them all based on your use case.

You're gonna wanna mix and match and maybe even change them out, maybe even test them, uh, back to the kinda eval idea. It's like I have this agentic workflow, and here's the thing we've, uh, been playing with recently, 'cause we have enough users now where they, like the LM, and I look at the bills, uh, and it's like, oh, like I'm spending real money now.

Um- And, and this is just human nature, right? Uh, it's not just, uh, normies, but it's like, so you have this, you know, dropdown of all the models, um, that you can say which model do you wanna use in, in your Agent.ai agent.

And as it turns out, people pick the largest number. So they will pick-

Yeah. Yeah

... uh, GPT 4.5-

Oh, 3

... or whatever, whatever it is, right?

Yeah.

It's like it's, uh-

Oh my God, you're doing 4.5?

Yes.

Ouch.

Yes.

Yeah.

But the thing I've promised myself, um, is we will support all of them regardless of what it costs. And like, once again, I see this as a, just a research thing, uh, you know, benefit to humanity and inference costs are going down, at least I, so I tell myself late at night, uh- ...

uh, so I can sleep. Um, so they pick the, uh, the highest numbered one, and so we have an option in there right now that says, and which is the first option, it's like, let the system pick for me.

Auto optimize.

Uh, auto-

Yeah.

As it turns out, people don't do that. They just pick the, the G because they don't trust it yet, which is fine. They shouldn't trust it, uh, completely. But one thing we discovered is that, um, if we backchannel it, and this is the thing we're testing with, is that, oh, if I can just run the exact same agent, uh, that gets run a thousand times, and we'll do it on our own internal agents first, and if the ratings and reviews, because we're getting human evals all the time on these agents, we can get a dramatic multiple orders of magnitude reduction by going to a lower model with literally, like, no change in the quality of the output, right?

It's like, which makes sense because so many of the things we're doing doesn't require the most powerful model. Um, and it's actually bad because there is higher latency. It's not just a cost thing. But, um, so anyway, like in that kinda future state, I think we're gonna have model routing and a whole body of, and people working on that problem too.

It's like, uh, help me pick the best model at runtime. Um.

Would you buy or build model routing?

I buy everything that I can buy. I, I, I don't wanna, I don't- Literally I don't, I, uh, I don't wanna build anything, uh, if I don't have to.

One of the most impressive examples of this, I think, was our Chai AI conversation, which I think about a lot. He views himself explicitly as a marketplace. You are kind of a marketplace, but he has a third angle, which is the model providers.

Mm-hmm.

And he lets them compete. And I think that sort of tri three-way marketplace-

Yeah

... maybe makes a lot of sense. Like, I don't know why every AI company isn't built that way.

It's a good point, actually. Yeah. Um, it makes sense. I, uh, on my list of things I'm super passionate about, I'm very passionate about, uh, efficient markets or, uh, and/or extremely irritated by inefficient markets. And so efficient markets for the normies listening are markets that exist where every possible-- Uh, efficient markets are the ones that every transaction that should occur actually does.

That's an efficient market. That, that should happen. And so then why do inefficient markets exist? Well, maybe the buyer and seller don't know each... about each other. Uh, maybe there's not enough of a trust mechanism. There's no way to actually price it or come up with fair market value, fair, uh, fair pricing.

And as you kind of knock those dominoes down, you know, the market becomes more and more... And lots of, uh, latent value exists as a result of inefficiency, and whoever removes those inefficiencies for, like, high-value markets makes a lot of money.

That's been, uh, proven time and time again. This is one of those examples of there's an inefficiency right now because we are, like, either over- uh, using over-models or whatever. Let's just reduce that to an efficient mark- the, the right model should be matched up with the u- right use case for the right price, uh, and then we'll-

Very-

Yeah

... very interesting. You ever looked into DSPy?

I have looked at it, not deeply enough, though.

It's supposed to be as far as I think the, the, the, the only, like, evals first framework.

Yep.

Right? And if, if evals are so important... And by the way, the, the, the relationship between this and all that is DSPy would also help you optimize your models-

Yep

... uh, because you did the evals first.

Yep.

I wonder why it's not, not as popular, you know? Um, but I mean, it, it is growing in traction, I would say. We're, we're keeping an eye on it.

Let's talk about business models. Obviously, you have kinda two work as a service and results as a service.

Yep.

I'm curious how you divide the, the two.

Yeah. Uh, so work as a service is... So we, we know about software as a service, right? So I'm licensing software that's delivered to me as a service. That's been around for decades now, um, so we understand that.

But the consumer of that service is generally a, a human, uh, that's doing the actual work, uh, whichever software you're buying. Work as a service is the software is actually doing the work, whatever that work happens to be, and so that's work as a service.

So if I come up with kind of discrete use cases, whether it's kinda classification or a legal contract review or whatever, the software's actually doing the thing. Uh, results as a service is you're actually charging for the outcome, not actually the work, right?

That says, okay, instead of saying, "I'm gonna pay you X amount of dollars to review a legal contract," or for this amount of time or number of uses or something like that, I'm gonna actually pay you for the actual result, um, which is...

So my take on this in, like, in the industry or parts of the industry are super excited about this kind of results as a service or outcomes-based pricing, and I think the reason for that, I think we're over-indexing on it, and the reason we're over-indexing on it is the most popular use case on the kinda agent side right now is, like, customer support.

Uh, well-documented. A lot of the providers that have, uh, you know, agents, uh, for customer support do it on a number of tickets resolved at times, you know, X dollars, uh, you know, per ticket. And the reason that that makes a lot of sense is that, uh, the customer support departments and teams sorta already have a sense for what a ticket, uh, costs to resolve through their kinda current, uh, current way.

And so you can come up with an approximation, um, for, A, what the kinda economic value is. There's also a at least, uh, semi-objective measure for what an acceptable, uh, resolution or outcome is, right? Like you can say, "Oh, well, we measured the net promoter score or a cSAT, uh, for tickets or whatever.

As long as the customers... 90% of the tickets were handled the way the customer was happy," that's whatever your kinda line is. As long as the AI is able to kinda replicate that same SLA, and it's like, okay, well, it's the same.

They're fungible, uh, one versus the other. I think the reason we're over-indexed, though, is that there are not that many use cases that have those two dimensions to them, that are objectively measurable, and that there's a known economic value that's constant.

Like customer support tickets, because they're handled by humans, makes sense, and humans have a, kind of a, a discrete cost, and especially in retail, which is where this originally got started in B2C companies that have a high volume of customer support tickets as they're distributing across.

A ticket is roughly worth the same because it takes the same amount of time for most humans to do that kind of level one, uh, tier one support. But in other things, the value per, uh, outcome can vary dramatically, literally by orders of magnitude, in terms of what the thing is actually worth.

That's kinda thing number one. Thing number two is how do you objectively measure? So let's say you're going to do a, uh, a logo creator as a service based on results, right? And that's a completely opposite subjective thing or whatever.

And so, okay, well, it may take me 100 iterations. It may take me five iterations. The quality of the output is actually not completely under my control. It's not up to the software. It could be you have weird taste, or you didn't-

Right. Yeah, yeah

... describe what you're looking for enough or whatever. It's like it was just not a solvable problem, and design kinda qualitative, subjective disciplines deal with this all the time. How do you make for a happy customer? There's a reason why they have, "Oh, we'll go through five iterations," but, you know, our output is we're gonna charge you $5,000 or $500, whatever it is, for this logo.

But that's hard, right? To kinda do at scale, so.

Just a, a relatable anecdote. Uh, we... Our podcast, actually, we just, uh, got a new logo.

Yeah.

And we did a 99designs for it.

Yep.

And there are so many designers who are working really hard.

Yeah.

But, like, I just didn't know what I wanted.

Yeah.

So they were like... I was like, "Just too bad." Like, I, I know... Like you, you seem great, but-

Yep

... you know.

Yep. Yeah. And that's another example of a, a market made efficient, right? It's like I've been a 99designs user and customer for dozen plus years now. Um-

It's fantastic.

Yeah.

And so many designers. Like this doesn't cost them that much for them to do.

Yeah. Yeah.

It's worth a lot to us. We can't design for shit.

Totally.

Um, yeah.

Yep. By the way, pro tip on 99designs, um, is that on the margin, you're better off kinda committing to paying the designer-

Yes

... that you're gonna pick a winner. Whether you like it or not-

Yeah

... doesn't really matter, uh, and that gets higher participation, which... And the... You're still gonna get a bunch of crap that happens. You get a, bunch of noise in it, uh, but the kinda quality outcome is often a function of the number of iterations, uh, and, and, and logo design is one of those examples.

If you can... If you had to choose between 200 logos versus 20 logos, chances are closer that you're gonna find something you like.

Yeah. Uh, for those interested-

It's worth that delta

... I have a blog post on my reflections on the 99design thing, and that's on, that's one of those. They, they give an estimate of, uh, how many designs you get.

Yep.

And I think that the modifier for, like, we will pay you-

Yep

... we'll pay somebody, and maybe it's you-

Yep

... is, like, 30 to 60, but actually it's 200.

Yep.

So it's underpriced.

Yep.

Do you think some markets are just fundamentally gonna move to more results-driven business models?

Probably, and I don't n-understand enough markets well enough to know. But if we had to kind of s-sort, order, rank them, there's likely some dimension along which we could sort that. It's like, oh, these kinds of businesses is their objective measure of, uh, of kind of truth or, uh, the outcome.

Um, is there a way to kind of price it, um, in terms of the... So low variance or variability on, uh, on the value per outcome. If those things are true, whatever industries that is true in, customer support, uh, is an example, but there's likely lots of other examples where, uh, those two, two things are true.

But then the thing I wonder, though, is that from the customer's perspective, would they rather actually pay, uh, for work as a service versus an actual, um... It's like maybe the way they think about it is, that's sort of my arbitrage opportunity, like in the, I can get work done for X, but the value is actually Y.

Why would I want that delta to be squoze out by the kind of provider of the software if I have a choice? I don't know. Um, that's-

Oh, I mean, okay, I... Attribution.

Yeah.

Like, there's, you know, there's 18 things that go into that, and you're one of them. So, like, you know, it's, uh, it's hard to tell.

Oh. Yes, it is.

So, yeah. By the way, have you seen... Uh, obviously you're in this industry, not exactly HubSpot's exact part of the market, but what have you seen in attribution that is interesting, you know? Because that, that, that directly ties into work as service versus results.

Yeah. Not enough because we are so, um, as a world, as an in- just pick your thing, so behind on track... And this is why I think Web3, uh-

Oh, boy

... in the way that it was meant to be done-

Yeah

... is going to make a comeback because fundamental principles of that makes sense. I think what happened in that world was kind of bunch of crypto bros and grifters and NFT stuff or whatever that was loosely related, like there was no re-actual...

But the idea of a, uh, of a blockchain, of a trackable thing, of, you know, being able to fractionalize digital assets, uh, attribution, having an audit log, a published thing that's verifiable, all those primitives make sense, right? Like, and, and maybe there's, you know, a limited, but it's not zero, set of use cases where the kinda what we would now call, like, the inference cost or the, the overhead, the tax for, uh, storing data on the blockchain.

It has a... And there's certainly a tax to it. It doesn't make sense for all things, but it makes sense for some things, uh, for sure. Uh, so... But we just don't have, like, attribution in any meaningful way, I, I don't think.

Um-

Isn't it sad that it's so important and-

I know

... no answer?

Yep. It's partly comes down to incentives. Uh, so the people that actually have the data or parts of the data from which attribution could be calculated or, uh, derived don't really have the incentives to make that data available.

Um, so even something as simple like on, uh, like the PPC side, right, on the Google Search thing, uh, which, you know, that's sort of my world or has been. We have less data now than we did back in the day in terms of like click-throughs and things like that before Google would actually send you, "Here are the keywords people typed."

And, you know, years ago they, you know, they even took that away. So it's hard to kind of really connect the dots back on things, and we're seeing that across, it's not just PPC, but just all sorts of things.

They took that away from Search Console.

What's that?

They, they- Their Search Console has that.

Y- yes.

And they took that away.

Search Console has that, but your website, if you go to Google Analytics, you can connect it back to Google Search Console.

I see. I see.

Yeah. Yeah. So, um-

Uh, okay. All right. Yeah. Well, it's a known thing.

Um-

You don't have to make it a rant about Google.

What about software engineering? Do you think it will stay as like a work as a service, or do you think... I think most companies hire a lot of engineers-

Yep

... but they don't really know what to do with them or, like, they don't really use them productively.

Yeah.

And I think now they're kinda hitting this like, you know, crisis where it's like, okay, I don't know what I will price an agent-

Yeah

... because I don't really know what my people are doing anyway.

Yeah.

Like, uh, how do you think that changes?

I, I think, um, so I'm actually bullish on engineers in terms of their kind of long-term economic value. Um, not despite all the movements in code gen and all the things that we're, you know, already seeing, but because of it.

Uh, because what's gonna happen as a result of AI, and people have talked about this, um, in, um, even other disciplines, we're gonna be able to solve many more problems. The semi-math guy in me is like, okay, so we always say, "Oh, well now, you know, agents are gonna be doing code or whatever, and so there's gonna be a million software enginee- uh, you know, virtual digital, you know, software engineers out there.

And so the value per engineer is gonna go down because I'm just in that, in that same mix. I as an engineer." What they don't recognize is that it's not just about the denominator, there's a numerator as well, which is what's the total economic value that's possible.

And I would argue that's growing faster than the kind of denominator is that the actual economic value that's possible as a result of software than what engineers, uh, can produce, you know, with the tools that they will have at hand.

Um, so I think the value of an engineer actually goes up. They're gonna have the power tools. They're gonna be able to solve a larger base of problems that are gonna need to be solved. Um-

Yeah. It feels to me like it'll stay as, like, work as a service.

I, I, I think so.

You're paying per work. I don't think there's, like, a way to-

And there, there will be a set of, um, engineers that... And we see this all the time, you know. There are, uh, like in the media industry, you have people that are kinda writers, but then you have freelancers that, you know, you know, write articles or write however they manifest their kind of creative talent, and both make sense, right?

There's like the work for hire. There's also the kinda outcome-based or like, "I produce this thing." And maybe they, some of those engineers actually produce agents, so they put it in a marketplace like Agent.ai someday, and that's how they make their millions.

Right.

Um, yeah.

Any other thoughts just on agents? We got a lot of like misc things-

Miscellaneous

... that we wanna talk to you about.

I think we covered a lot of territory. So I'm, uh, excited about agents. My kind of, uh, message to the world would be, don't be scared. I know it's scary. Uh- ... easy for me to say as a tech- uh, techno optimist, but learn it.

Even if you're a normie, even if you're not an engineer, if you're not an AI person, you don't think of yourself as an AI person, use the tools. I don't care what role you have right now, where you are in the workforce, uh, it will be useful to you.

Um, and start to get, uh, get to know agents. Use them, build them.

And I, I think my message for engineers is always like there's more to go. Like we're still ear- in the early days of figuring out what an agent's, uh, stack looks like.

Yeah.

And, uh, I want to p- push people towards agents with memory.

Yes.

Right? Agents with planning.

Oh, we have to talk about memory. We gotta talk about memory.

Let's go.

Yeah, let's do it.

'Cause I think that's the, uh, that's the next, in my mind, the next frontier is actual long-term memory, both for agents, uh, and then for agentic networks in a trustable, verifiable, I won't say privacy first, but, uh, privacy-oriented way.

I have an issue with the, uh, the term privacy first, um, 'cause a lot of times we say privacy first when we don't really mean that. Like privacy first means I value that above all things. Doesn't matter what we're talking about, and that's just not true- ...

not for any, uh, for any human. Um-

Anything that wants to be used.

Um, so it's-- But, uh, but so memory is an interesting thing, right? So the thing I'm working on right, right now, uh, lots of things, uh, in play in, in Agent.ai is around implementation, uh, of memory, and there are, uh, great projects out there, Memzero being one of them.

But the thing that's interesting for me, right, is and so we see this in, uh, ChatGPT and other things right now where it does have the, the notion of a longer-term memory can pull things back into, into context, um, as needed.

The thing I'm fascinated by is, uh, cross-agent memory. So if I'm an agent, uh, builder right now, uh, it's like, okay, here are the things that I sorta know or I, I learned from the user, um, in terms of pulling out the, uh, I'll call them knowledge nuggets for lack of a better term, and that's great.

But then when the next agent builder comes out and it's the same user, shouldn't all the things that Agent One learned about me, if it's gonna be useful for Agent Two, as long as I opt into it, it's like, yeah, I don't care.

Those things... In fact, I would find it awfully annoying to s- tell Agent Two and Agent N and Agent N plus one all the same things I've already told it, uh, because it, it should know. Like, the system should know.

And this is part of the reason why I'm, like, a believer in these kinda networks of agents and shared state is that that user utility gets created as a result of having shared, shared memory. Not just we should solve the memory problem, uh, for an independent agent, but then we should also be able to share, um, share that context, share that memory across agents, and that's part of the value prop for Agent.ai is like, okay, when you're building, it's like so we've got, you know, and, uh, one, you know, whatever million users, uh, and we're gonna have growing memory about all of them.

So instead of you going off on your own thing and building an agent out as this kind of, uh, disconnected node in the, in the universe or whatever, here's the value for building on, on the network or on the platform, ours or, ours or someone else's because more, there's more u- uh, user value that gets created.

It's more utility. Yeah. How do you think about auth for that? Because part of memory is like- Ooh. ... selective memory. So take like scheduling. Yep. I want you to have access-- If I have a scheduling agent, you should be able to access the events you're a part of- Yep ...

and, like, what times I have available. Yep. But it shouldn't tell you about other events on my calendar. Like, what's that layer like? I have so many thoughts on this. This is the-- And the, like, the opportunity out there, like solving these kind of fundamental, like, this is going to need to exist, right?

So right now, uh, the closest approximation we have, um, is, is auth, uh, OAuth 2.0, right? Yeah. Um, and everyone has this, like, okay, approve. Scopes, yep. And it's a very, very coarse, uh, set of scopes, right? Like, based on the, the provider of the, um, the, uh, OAuth server, be it Google, whoever it is, HubSpot, doesn't matter.

It's like, oh, I pick a set of scopes, and they could have defined the scopes to be super granular. Right. Fine. Uh, but that's sorta up to them. But that is going to move so slowly, right? So for instance, the use case I have right now, like, I use email for everything.

I use it as a, um, like an event and data bus for my life, right? And why-- I mean that, like, literally. It's like I'm like anything that I do, if there's a way to kinda get that into email, 'cause I know it's an open protocol, right?

It's like, okay, I will be able to get to that data in useful ways, uh, and this is before, so I have three million that I've built a vector store off of that is, that solve my own, uh, personal use cases.

So I'll give you the example, but obviously I'm not gonna build my, all my own software for everything. But if a startup comes along and says, "Dharmesh, can you make your email inbox available in exchange for these things?"

I'm like, "Hell, no." Yeah. Like, that's the literally my kinda like everything. Like, my life is in here, right? Um... So you need to sh-share subsets. Yes. And so I think there's a, and maybe this is not the actual implementation, but imagine if someone said, "Okay, I have a trusted intermediary for that first trust however defined," that says, "Okay, I'm gonna OAuth into this thing," uh, and it gets to control it.

I can say in natural language, "I only wanna pass email to this provider, uh, where the label is one of X or that's within the last thing and no more than 50 emails in a day or whatever," so I don't have them dumping the entire three million, uh, you know, backlog.

Whatever controls I wanna put on it, it's unlikely that the, all the OAuth, um, server side right now, the Googles, even the big ones, small ones, doesn't really matter, are gonna do that. But this is an opportunity for someone, and they're gonna need to get to some scale, build some level of trust that says, "Okay, I'm gonna hand over the keys to this intermediary."

Yeah. But then, uh, it opens up a bunch of utility because it gives me control, uh, more fine, fine-grained control. Um... Yeah, I'd say LangChain has, has an interesting one. There are a bunch of people who has tried to track, crack AI email.

Every single one of them who's tried has pivoted away. Yep. And I'm waiting for Superhuman to do it. Yep. Uh, I don't know why they haven't, but you know, some point. They have some cool AI stuff. Yeah. But you need-- I, I think the pace needs to increase.

But I think this goes back to, like, Open Graph. Yeah. Right? Yeah. Which is like I, I think Google is not incentivized to build better scopes. Nope. And, like, they're just not gonna do it. Nope. So, um- We can't even get, like, we haven't been able to get semantic search out of Google for, like, still- No, totally.

You know, just now they made the announcement this week. What do you mean? Semantic search? In Gmail. Oh, I see. So, okay, so they have all the-- they have my three million emails. Why don't they have a vector store where I can...

The, just, like, basic crap, right? Yeah. Actually, that's, that's really bad. It's not, you know, they're indexing the entire internet, uh- Yeah. ... in real time. Like, I th- I don't think my email is that big a deal, but...

Yeah. My standard thing on memory is it sounds like you are using, uh- Memzero ... Memzero. I am. There's also Mem- MemGPT, now Letta, uh, which give a workshop at my conference. There's Zepp, which uses a graph database- Yep ...

just kinda open source. Yep. Kinda interesting. And LangMem from LangGraph- Yep ... which I would high-highlight. Also, like, it's re-really interesting this developing philosophy that's, that people seem to be agreeing on an, on a hierarchy of memories. Mm-hmm.

From semantic memory to episodic memory to, I think just overall sort of background processing. Like, we have independently reinvented that AI should sleep- Yep ... to, uh, to do the deep REM, uh, processing of memories. Yep. It's kind of interesting.

Yep. Yeah, that is. It's-- The other-- I mean, just on the notion of memory and hierarchies, um, so, you know, I talked about, uh, the memory we're working on right now is at the, at the user level, and it's cross-agent, right?

Um... Yeah. But the other kinda one step up would be, so ma- once again, uh, going back to these kinda hybrid, uh, digital teams, is that, uh, you can imagine to say, oh, well, my team, uh, has this kinda shared tea-- I don't want shared with the world or other thing, other word.

Like, this set of agents across this group of people, I wanna have shared state like we would have in a Slack channel or, or something like that, and that should sorta exist as an option, right? Yeah. Um, and the platforms, uh- Yeah ...

should provide that. And, uh, the, the B folks I should also mention have mentioned that they're, they're working on that- Okay ... as well. So imagine being able to share, you know, selective conversations with people. Like, that's nice.

Uh, Limitless has, I guess, voice-based shielding.

Yeah.

Uh, that I don't think there's act-

I'm an investor in that too, by the way.

Oh, really?

Just-

Oh, yeah.

So full... Uh, okay. I'm trying to think about all the things I've said. Uh, invest in OpenAI, Perplexity- ... LangGraph-

Superhuman

... Crew AI, Limitless. Uh, a bunch of them. So if I've said anything, by the way, I have no insider knowledge.

Yeah, that's fine.

I have no in-

Yeah, yeah.

I'm not trying to plug or pitch or anything like that.

No, no, no, no. I get-- I think it's understood. We're, we're often like, you know, if you have skin in the game, you're probably invested or-

Yep

... you know, may, may or may not. I'm not an investor in B, but I'm just a friend.

Yep.

And, uh, I, I think you should be able to free, speak freely of your opinions regardless.

Yeah.

Okay. We have some, um, miscellaneous questions that may be zooming out from Agent.ai.

Sure.

First of all, you mentioned this, and I have to ask, you have hun- you know, so many AI projects you'll never get to.

Yep.

Uh, what's one or two that you want other people to work on?

Oh, wow. Um-

Just drop some from your-

I want other people-

... read list

... to work on.

'Cause you'll never get to it.

Yeah, yeah. What I need to do, um, 'cause I've had this thought before, so I have this... is, like, maybe, like, pick one a week or something like that and give the domain away. Uh, like-

Oh

... I have people submit their-

Dharmesh's giveaway

... kind of one-pager or something like that. It's like, if you can convince me that you have at least enough of an idea, enough, uh, like, willingness to kind of commit to actually doing something, uh-

It's the ones that you keep mentioning, but you've, you've, you've, you haven't gotten to it for whatever reason.

Yep, yep. Um, trying to think. Like, some of them I don't have the underlying business model.

You don't have to.

We're gonna have, we're gonna have to come back to this, maybe do a follow-up episode. I don't, uh... Like, they're just not jumping to mind. Um-

You don't need the business model. Just, just-

Yeah

... okay. Okay.

So I own Scout.ai.

Okay.

Uh, I think that's an interesting, uh... By the way, pretty much all of them, there was an idea at the time. It's like, it was one of those late night, I was like, "Oh, I could do this. Is the domain available?"

And I go grab it. Um, I'm trying to think what else I have on, uh, in the AI space. I have a lot of, like, nonprofit, uh, domain names as well for, like, a nonprofit like OpenGraph. Um, I'm not sure why things are not jumping to my head.

Uh, I, I, I have agent.com, which obviously is tied to Agent.ai. Uh-

Ooh.

That-

That's gonna be big.

That's gonna be big. Uh, I-

Oh my God. That's gonna be like $30, $50 million.

It, it's gonna be big. Um- It's... Yeah. It's gonna be, I think, end, end up being bigger than, uh, chat.com, which, uh-

It has to be

... was 15. Yeah.

Yeah, yeah. It's more work-oriented.

Yep.

That's interesting.

Yeah. Do you wanna talk about the chat.com thing? Um-

I, I-

I would love just the backstory. So did you just call up Sam one day and be like, "I got the domain"?

Yeah.

Did, did they kinda get back to you-

No, I'll, I'll give you-

... knowing that you had it?

It's, uh, it's, it's a good story. Back, uh, in the original ChatGPT days, uh, the first thought I had in my head, which lots of people had in their head, is that OpenAI is going to build a platform, and ChatGPT is actually just a demo app to show off the thing, and there's been precedence for tech companies that have had, uh, you know, uh, demo apps to kind of help normies understand the underlying technology.

And even after the kind of the boost or whatever, so my original thought was, well, someone should actually create, like, an actual real product. And so I'm like, and that product should be called chat.com because GPT is not a consumer-friendly thing at all.

Like, that's an acronym, uh, not pr- doesn't roll off the tongue. And so like, I'll build ChatGPT because that was just a demo app back then. So I, you know, got chat.com. And then as it turns out, ChatGPT is like a real product, and I was at an event here in San Francisco that Sam spoke at, where he launched, uh, uh, plugins, I think it was the, the announcement at that time.

Yeah, the update. Yeah.

Yep. And that's the thing, is like I had sort of suspected, it's like, okay, things sort of be-- Like, there's no way that OpenAI is gonna launch plugins for ChatGPT if they were not thinking of it as an actual platform.

Mm-hmm.

So it's not just about the, uh, GPT APIs.

Yeah.

This is like a real thing. I'm like, crap. Like, this violates the first rule of Dharmesh, which is don't compete with Sam.

Yeah.

Um, I knew when I bought the domain that there was competition for the domain, uh, there were, um, other companies looking to buy it. I, I don't know who they were. I had suspicions. Um, so I bought it, and then I'm like, okay, well, I'll reach out to Sam.

I was like, "Hey, Sam, uh, I happen to have got..." I don't know if, you know, he was or wasn't, uh, kinda in the running or trying to acquire it or not, but I have chat.com. I don't- not looking to make a profit or whatever.

If you want it, you'll obviously do something much better, bigger with it. I don't wanna be in the compete with Sam game, um, effectively is, is what I said. Uh, and so they did want it. Um, and yeah, we struck a deal.

Looks like it's been a very good deal if, uh, the valuations are, you know, to be-

It's-

... to be real.

Yeah. Uh, it's gonna be-

Who knows? Who knows?

It's one of those weird things, like, uh, yeah.

The agent.ai domain evaluator said that late in that space has for between 5 and 15K.

Okay.

So...

Does that feel right-ish?

Well, it's missed, uh, it's, it's missing that it's, uh-

So this is V1 of it. This one does not incorporate the transactional data.

Mm.

I have not published that one yet. Uh, so that's... And because it's also operationally very intensive, uh-

Yeah

... that other one. But anyway.

Uh, we, we actually had it donated by a listener.

Oh, okay.

So I-

Awesome

... don't know what the real cost is.

Yep.

But, uh, it's missing that it's linked to an influencer.

By the way, I also own crew, crew.ai, which I've offered... I'm an investor in, in, uh-

You did?

Yes.

Yes.

I bought that. Uh-

Do you want-

And I've told him that, like, whenever you're ready, you let me know, I'll sell it to you at cost. Uh, yeah, so.

Yeah. I mean, that, that is some value add. Since you ma- buy a lot of domains, what, what are your favorite, uh, domain buying tips apart from have a really good domain broker, which I assume you have?

Uh, no, I actually don't. Uh, I do-

You don't?

I do my own deals. Um-

Oh my God.

Nice.

Um, I have a, like a very cards face up approach to life. Um, so there's... So, you know, some people would tell you, it's like, "Oh, well, if someone they know that it's you're behind the transaction, the, you know, the price is going to go up."

Sure. But it's still, like, willing seller or willing buyer, whatever, doesn't mean I'm gonna have to necessarily pay that price. Uh, it's like, okay. But the upside to it, uh, uh, 'cause I always, you know, reach out as myself when there's a domain out there, um-

And they can look you up.

They can look me up.

Uh, right.

But then I also come off as, like, legit. Like, okay, well, there's very few people are not gonna return my email when I say I'm interested in a domain that they may have for sale, um, or had not considered selling but, you know, would you consider selling?

Uh, so yeah, and some of the, like, uh... So I own, like, some of my favorites, I still own prompt.com by the way. That, that could be a big one. Um, it's, it's... and, but I- Owned, and this is one, uh, I don't regret it.

I, I went into a good... I owned playground.com. And so the original idea behind playground.com was, at the time, uh, OpenAI had their, uh, playground where you could, uh, can play around with the models and things like that, right?

It's like, okay, well, there should be a platform neutral thing. There should be a playground across all the LLMs, then you can... And there are obviously products and, uh, startups that, that do that now. And so that was my original thing.

It's like, oh, there should be playground.com, and you can go test out all the models and play around with them, just like you can with, uh, with OpenAI's, uh, GPT stuff. And then, uh, so Su- Suhail was out there with, uh, with, with Playground, uh, the company, um-

Yeah, he's been on the pod, yeah

... and I think he reached out, might have reached out to me over, over Twitter or something like that. I'm, so we knew of each other. I'd never-- I've still nev- never met him. And, and he asked me whether I would, you know, consider...

And that was a tough one because I'm like, I actually have the business idea already in my head. I think it's a great domain name, uh, and it's like a really simple English word that has, like, relevance in a whole new context now.

But once again, uh, I took, uh, took equity, so it's like I look on the bright side, that's like I... So domains get me into deals that I would never have been able to-

Mm

... likely get into other ways, so yeah.

Yeah.

Yeah. We should securitize your GoDaddy account and just make it a fund.

It's a fund.

And then yeah.

It's basically a fund.

Yeah, um- And by the way, so back to the kind of, uh-

I hope you read those GoDaddy, by the way

... three things, whatever. So I've been vested, uh-

GoDaddy sucks

... I don't know if it's public yet, um, but in a company that's going to treat domains as a fractionalizable, uh, tradable asset. Because that's the, kind of the original NFT in a way, right? It's like, okay, well, if you can-

It is the NFT

... and then if you can make both fractionalizing, but also just the transfer of, like right now it's so painful when you buy a domain. You go through an escrow service, and there's just all this... It's like I just want like instantaneous, like charge me in Bitcoin or credit card, whatever it is, and then I should show up and I should be able to route out the DNS.

Like that should be minutes, not weeks or days. Um, anyway, so-

Yeah

... so.

That's for ENS on Ethereum.

Yep.

That's, is basically the same. They, they, they should bring the-

Yeah, but it needs to be that way for normies

Yeah, exactly.

For normal humans, yeah.

They should bring the, yeah, the ICANN and all of that as a, as its own, its own thing.

I have a question on-

Yeah

... on just that y- y, you know, you keep bringing up the, your Sam Altman rule. One of my favorite, favorite, favorite My First Millions of all time was actually without you there, but talking about you.

Okay.

'Cause, uh, Sean was describing you as a fierce nerd.

Mm-hmm.

Which I, I'm sure you, you, you were there. Uh um, and, uh, I think Sam also is a fierce nerd and, and he is, uh, uh, I was, I was listening to this Jessica Livingston podcast-

Yep

... where what she had him on and described him as a formidable person. I think you're also-

Yeah

... very formidable, and I just wonder what makes you formidable. What makes you a fierce nerd? What, what keeps you this driven?

Yeah. Sam's fiercer and nerdier, just for the record. Um, but I think part of it is just, like, the strength of my conviction, I guess. Like I'm, I'm willing to, like, work harder and grind it out, uh, more than people that are smarter than me, and I'm only slightly stupider than people that are willing to work harder than me, right?

Like I'm just the right mix of, uh, the kind of grind at it, kind of work at it, stick to it for extended periods of time. If I think I'm right, I will latch out, latch on and not let go until I can either, like, prove to myself that it's not, um, so even, like the natural language thing, it's like, you know, took 20 years, but eventually I got to a point where, uh-

Yeah, yeah

... the world caught up and, and it became-

No, I-

... possible. Uh, but yeah, I think, and part of it is, uh, I think, this is partly I think what makes me... Like, I'm a nice guy. Uh, and they're, sometimes they're the most dangerous kind, right? It's like, okay, well, and I, I, I don't make enemies or whatever, but so my advice would be my, this is my take on competition.

I don't think of it as like war. I think of it as, uh, they're opponents, uh, right? And this, it's, it's not war.

You're playing a game.

It's like it's, it's a game, right? And you can, and use whatever analogy. I happen to, uh, play, um, a fair amount of chess. I'm a student of the game. That's partly I think what, uh, makes me effective.

Uh, I'm solving for the long term, uh, so I'm kinda hard to deter. So for those of you out there looking to kinda compete with HubSpot, uh-

Good luck

... no. Uh, I'm gonna be here. Uh, 18 years, I'm gonna be here for another 18 years, so. But not that you shouldn't do it. It's a big market. Uh, I'm not trying to sway anyone, but...

Yeah. I think, like, something I struggled with, with is conviction.

Yeah.

You said you pursue things with conviction, but, like, you start out not knowing anything.

Yeah.

And so how do you develop con- conviction when there's, you, you find it along the way or you, you stumble along the way, then you lose conviction, and, and then you stop working on it? You know-

Yeah

... like how do you keep going?

The way I've sort of approached it is that, um, so I don't generally tend to have conviction around a solution or a product. I have conviction around a problem, uh, that says, "This is an actual, real problem that needs to be solved."

And I may have an idea for how to be solved, uh, you know, right now, and that I may be c- get dissuaded. It's like, ah, I'm not smart enough, technology's not good enough, whatever the, you know, constraints are.

But it's the problem I have conviction around. It's like, oh, that problem still hasn't gone away. Uh, so, uh, like I sort of file it away in the back of my brain, and I will revisit. It's like, okay, well, you know, the kind of board changes, uh, I don't know if changes really fast now with AI, like things that weren't possible before are now possible.

So you kinda go back through your roster of things that you believe or believed and say, "Maybe now, uh, now is the time. Maybe then wasn't the time." Uh, but I'm a big believer in kind of a- attaching yourself passionately, uh, with conviction to problems that matter, um, that...

And there are some that are just too highfalutin for me that I'm not gonna ever be able to kinda take on. I, I have the humility to recognize that.

Yeah. I feel like I need a, um, updated founder's version of a serenity prayer. Like give me the confidence to, like, do what I think I, I'm capable of, but like not to overestimate myself, you know?

Yeah, yep.

Uh, you know, anyway, uh, when you say board changes, how do you keep up on AI?

A lot of YouTube as it turns out.

Really?

Uh, yeah, a lot. Um-

Okay. Fireship?

I don't know what Fireship is.

It's a current meme right now. Whenever OpenAI drops something, you know, they love this like live streams of, of stuff-

Yep

... from, on the OpenAI channel. The top comment is always, "I will wait for the Fireship video."

Okay. Okay.

Because Fireship just summarizes their thing in five minutes.

No, I, so my kind of MO, so I... By the way, I, I keep very weird hours. Uh, so my average go to bed time, uh, is roughly 2:00 AM.

Oh, boy.

But I do get-

A few hours

... average seven, seven and a half hours in. Uh-

Yeah. Good. That's great

... I don't use alarm clocks 'cause I don't, I don't, uh, have meetings, uh, uh, in the morning at all, uh, or try not to at least. Uh, so my late night thing is, uh, is I'll watch probably like a couple of hours of YouTube videos, often in the background while I'm coding.

Um, just kind of-

That's how you've seen our talks.

I have, yeah.

Yeah.

I've seen, yeah.

Okay.

Yep. Um, and so I... There's so much good material out there, and the, and the thing I love about kind of YouTube, and this... By the way, in terms of like use cases and things, agents that should exist that, uh, don't yet, I would love to...

Technology exists now to build this, is to be able to take a YouTube video of like a talk of, let's say, on Latent space, or not on Latent, but on the, um, AI engineer event, and say, "Just pull the slides out for me, uh, 'cause I want to put it into a deck-"

Yeah

... "for, you know, use or whatever."

Yeah.

Or some form of, uh, kind of distillation or translation into a different-

Oh, I see

... a different format.

Pull the slides. Got it.

Yep. Pull the slides out of a video. Um, so I think that's interesting. I've... Yeah. So by the way, on the kind of A- Agent.ai thing, like one of the commonly used, uh, actions, uh, primitives that we have is the ability to kind of get a transcript from a video, and that seems like such a trivial thing or whatever, but it's like, like if you don't know how to do it programmatically or whatever, if you're just a normie, it's like, "Okay, well, I know it's there, but I can copy and paste it, but like how do I actually like get to the, the transcript for a You- ...

And then, uh, getting to the transcript and then being able to encode it and say, I can actually, uh, give you timestamps." So if you have a use case that says, "Oh, I want to know exactly when this was.

I want to create an aggregate video clip." This was the actual original, um, agent that I built for my wife, that she wanted to pull multiple clips together without using video editing software, 'cause she wanted to have this, uh, aggregate thing, uh, she's on the nonprofit side, to like send to a friend.

Uh, anyway.

There are video understanding models that have come out from Meta-

Yep

... but the easiest one by far is gonna be Gemini. They just launched YouTube support.

Yep.

So, um, they're doing good work over there.

By the way, in terms of like the coolest thing AI-wise recently, I'll say the last, uh, week to 10 days, has been the new, um, image model, Gemini Flash experimental, whatever they call it, uh, because it lets you effectively do editing.

Yeah.

Um, the... And just... And so, you know, my son is doing a eighth grade research project on AI image generation, right? So he's kind of gone deep on, uh, Stable Diffusion and the algorithms and things like that. I don't know much about it, but one thing I do know, I know enough about Stable Diffusion to know why editing is like near impossible, that you can't recreate it because it's like you can't go back that way.

It's gonna be a different thing, because it's sort of spinning the roulette wheel another time the next time you try to, you know, a similar prompt. And so the fact that they were able to pull it off, it's still, it's still a very much a, a V1 because, you know, if you...

I know I've, I-

There's-

One of the test case like, oh, take the HubSpot logo and replace the O, which is like this kind of sprocket, with a donut. And it will do it, but it won't size it to the degree that it will actually fit into the actual thing.

It's like, okay, um, but-

Yeah

... but that's where it's headed, I can see.

So do you know the backstory behind that one?

No.

Uh, Mostaf, Mostafa, who was part of... So they had image generation in Llama 3.

Okay.

Uh, lawyers didn't approve it. Mostafa quit Meta and joined Gemini and did-

Yep

... and shifted. Uh, and it is rumored, and that's all I can say, is that they got rid of Diffusion. They-

Yep

... they, they did autoregressive-

Yep

... image generation. And I think it, it's been interesting, these two worlds colliding, because Diffusion was really about the images and autoregressive was really about languages, and people were kind of seeing like how are they gonna merge. And on the Midjourney side, David Holz was very much betting on text diffusion being-

Yep

... uh, being their path forward. Uh, but it seems like the autoregressive paradigm is one. Like Next Token is-

And Suhal and Playground are doing like exceptional work on that kind of domain of-

Autoregressive?

... uh, I don't know if it's autoregressive, but around kind of image editing and not just-

Yeah

... the kind of text to image, and actually building like a UI for like a-

Yeah

... Photoshop kind of thing for actual generation of images versus, uh-

Yeah

... just doing text to-

It is fascinating

... text to image.

I thought Diffusion was kind of dead. Like there wasn't that much... It was just like bigger models-

Yep

... you know, higher detail, and now autoregressive come along and now like the whole field is open.

Yeah.

Um, and I think like if there was any real threat to like Photoshop or Canva, it's this thing.

Yeah.

Just to wrap up the conversation, you have a great post called Sorry, Must Pass, which if I did the math right, you first wrote in 2007?

Yep.

The first version.

That sounds about right.

And then you re-updated it post-COVID. You mentioned you made a lot of changes to your schedule and your life based on the pandemic. How do you make decisions today, you know, in the, the... Has anything changed like since you...

Because you updated this in 2022.

Yep.

And I think now we're kind of like, you know, five years removed from COVID and all of that. I'm curious if you made any changes.

Yeah. So that, so that post, Sorry, Must Pass, was... The issue that happened, um, is my schedule just and life just got overwhelmed, right? It's like just, I just, uh, too many kind of dots and connections and, and I love interacting with, uh, new people online.

I love ideas. I love startups. There's lot... But as it turns out, uh, every time you say yes to anything, uh, you are by definition saying no to something else. Um, this, uh, you know, despite my best e- you know, attempts to change the laws of the universe, uh, I have not been able to do that.

So that post was a reaction to that, because what would happen for me, uh, would be when I did say no, I would feel this guilt. Because it's like, okay, well, whatever happened to me, it's like, oh, can you spend 15 minutes and just review this startup idea or whatever.

It's like, uh... And sometimes it would like be someone that was, you know-

A friend

... second degree removed, like intro-

Yeah

... through a friend or something like that, and I felt, uh, you know, real guilt. And so this was a very kind of honest, vulnerable, here's what's going on in my life. So, so this is not a judgment on you at all, what- uh, whatever your project or whatever your thing you're working on, but I have sort of come to this realization that I just can't do it.

So I'm sorry, but I, I... So my default thing right now, and lots of people will disagree with this kind of default position, is that I have to pass. Because unless, and, and Derek Sivers, uh, said this really well, it's like either a hell yes or it's a no, right?

So, and I'm gonna... There's gonna be a limited number of the, the hell yeses, um, that I'm gonna be able to kind of inject into my life. Um- So yeah, that, and that's, of all the blog posts I've ever written, that has been the most useful for me.

So I'd, um... And so, and I send it, and I still send it out personally, right? I don't have a auto- I don't automate my, uh, email responses at all yet. Um, don't do automated social media posts. Um, but yeah, that one's been very...

And I, uh, so I encourage everyone, wherever your line happens to be, I think this, um, lots of people have this guilt issue, and that's one of the most unproductive emotions, uh, in, in human psychology is, like, no good comes from guilt, not really, unless you're like a sociopath or something like that.

Um, maybe you need, um... Anyway, you don't, you don't need more guilt. Uh, yeah.

I, I would also say, so I, um, I would just encourage people to blog more.

Yeah.

Because a lot of times people want, like, to pick your brain.

Yeah.

And then they ask you the same five questions that everyone else has asked.

Yep.

So if you blogged it, then you can just, "Here."

Yeah.

Like

So one of the things I'm, I'm working on, uh, and there are startups that are working on this as well, uh, but I started before them, is like a dharmesh.ai, right?

Yeah.

That just captures... And it's interesting, so that's one of the agents, um, on, on Agent.ai, uh, on the underlying platform.

Oh, there... There's a Dharmesh agent?

It's, it's out there. It's dharmesh.ai, yeah.

Nice.

It's, uh, it's pure tech space, no video, no audio right now. Um, but uh, the, the thing that's, like, I found it useful in terms of just the how, how do I give it knowledge? So I have a, kind of a private email address, 'cause a lot of the interactions that I will have, or if I do answer questions, because I...

The other thing I, by the way, I don't do any phone calls, like, at all. Even, like-

No Zooms

... like, at all.

Yeah.

I mean, I'll get on Zooms with teams, but no one-on-one meetings, no one-on-one, uh, it just doesn't scale. So I've moved as much as possible to an async world. It's like I will, as long as I can con- control the schedule, like, I will take 20 minutes and write a thoughtful response.

But I reserve the right, uh, anonymously with no attribution, to kind of share that, uh, either with my model or with the world, um, you know, through a blog post or something. But it's been, like, useful because, uh, now that I have that kind of email backlog, I can go back and say, "Okay, I'm gonna try to answer this question."

Go through the vector store. Um, and it's shockingly good. Uh-

Okay

... and I'm still irritated that, uh, Gmail doesn't do that out of the box. It's like, they're Google. Um, I think it's co- it's gotta be coming now. It's their... I think they're finally-

Uh-

Uh, the giant has been woken up. I think they're, uh, their kind of-

It's a very-

... clock speed has gotten faster now

... y- you know, it's one of the biggest giants in the world, ever.

Yeah.

So yeah. When I first told Alessio, uh, you know, you were one of our dream guests-

Wow

... I never, I never expect- actually expected to book you because of Sorry Must Pass.

Oh, yeah.

'Cause so we were just like, "Ah, let's send an email," and like, "He'll say no, and we'll move on with our day." Uh, so I just have to say like, uh, uh, we're very honored that you-

Oh, I'm-

... just spent some time with us

... thrilled to be here. A- a-

Yeah

... huge fan. A first time, first time guest.

Yeah, yeah.

But, uh, yeah.

And-

Thank you for all that you do for the, for the community. I, I, I speak for a lot of them. You guys taught me a lot of, uh, what I think I know. It's, uh, yeah.

Appreciate it. Yeah. I mean, uh, I am explicitly inspired by, by in, um, by HubSpot.

Oh, thank you.

Inbound marketing, uh, I think is a stroke of genius, and like, the AI engineering is explicitly modeled after that. So, uh, like, you created your own industry, uh, you know, subsection of an industry that became a huge thing because you got the trend right.

Yep.

And that's what AI engineering is supposed to be-

Yep

... if, if we get it right.

Yeah.

Um, how do we screw this up?

How do we square what up?

How, how do I screw this up? How do we screw AI engineering up?

Oh, um-

If, you know.

Yeah. The, so the, the common failure modes, right, is, um... So the original thing that makes inbound marketing work, the kind of kernel of the idea, was to kind of, uh, to solve for the customer, solve for the audience, solve for the other side.

Uh, because the thing that was, you know, broken about marketing was marketing was a very self-centered, "I have this budget, I'm gonna blast you and interrupt your life and in- interrupt your day and, because I want you to buy this thing from me," right?

And inbound marketing was the exact opposite. It's like, use whatever limited budget you have and put something useful in the world that your target customer, uh, whoever it happens to be, will find valuable. Uh, um, anyway. So the, the common failure mode is that you lose that, uh, and I don't think you will, but it is very, very common, right?

It's like, ah, like now I'm just gonna like turn the crank and squeeze it just a little bit. Or like it's, uh, but you, you... The ri- reason I think, uh, folks like me, uh, you know, appreciate that community so much is you, you have that genuine want to act.

And there's nothing wrong with making money. There's nothing wrong with having spot. None of that. But at the, at the core of it, it's like we wanna lift the overall level of awareness for this group of people and create value and create goodness in the world.

Um-

Mm

... I think if you hold onto that, over the fullness of time, uh, the market becomes more efficient and rewards that, uh, that generosity. Uh, that's my kind of fundamental life belief.

Okay.

So I think you guys are doing really well. Continue doing well.

Thank you for, thank you for your help and support.

Yeah, my pleasure.

Yeah. And just to wrap in very Dharmesh fashion, you have a URL for the Sorry Must Pass-

Ooh. Yeah, I do

... blog, which is sorrymustpass.org.

Yes.

So yeah, the, I thought that was a-

Yep

... good nugget. Um, yeah, thanks so much for coming on.

Oh, thanks. Thanks for having me.