No PriorsFrom Coder to Manager: Navigating the Shift to Agentic Engineering with Notion Co-Founder Simon Last
EVERY SPOKEN WORD
30 min read · 6,429 words- 0:00 – 0:05
Cold Open
- SGSarah Guo
[upbeat music]
- 0:05 – 0:26
Simon Last Introduction
- SGSarah Guo
Hi, listeners. Welcome back to No Priors. Today, I'm here with Simon Last, co-founder at Notion. We talk about their new vision for Notion in the AI age as a platform for humans and agents to collaborate, how the engineering and product org at Notion is changing, and these new tools for thought. Welcome, Simon. Hey, Simon. Thanks for doing this.
- SLSimon Last
Yeah, of course. Yeah, it's really fun to be here.
- SGSarah Guo
Notion's
- 0:26 – 4:10
Genesis of Notion AI
- SGSarah Guo
at scale, amazing platform, lots of users. You did start quite a while ago. I think of Notion as one of the companies that has really, like, raced AI quite aggressively. I was told you first got your hands on GPT-4, uh, at a company off-site in Mexico. Um, is that true? What is the origin story of, like, starting to work on this stuff?
- SLSimon Last
Yeah, I think... Yeah, that year, that was twenty twenty-two. Um, I, I've been watching, you know, what's going on. In general, I've just been, like, super curious about the technology and fascinated to, to try everything and think about, like, like how we can apply it. It wasn't until I played with GPT-4 that it, it became really, really real. So, you know, we-- When we got access to it, it wa- it was sort of like a, a proto-ChatGPT-like interface. Um, and, uh, my co-founder Ivan and I both, both got access, and it was just immediately clear, like, I would say two big things. One is that it was just pretty smart. It c- it, it could follow reasonably complicated instructions. It could write things for you. It could edit things. And, and the second big thing was that, uh, uh, the scope of its knowledge was extremely interesting. Uh, super, super deep, like, um, and, and broad world knowledge. When we played with it, it became just instantly clear to both of us, like, okay, the, the time is now to start th- thinking about how to apply this. It's only gonna get better.
- SGSarah Guo
We were talking about Mexico, GPT-4. You guys saw it was, like, clearly the time. Did you start with, like, a particular vision of, like, what you should obviously be able to do with AI and Notion? Or did you start pulling people from different teams or recruiting people and say, like, "Let's experiment"? How did you begin?
- SLSimon Last
I think we immediately had a long-term and a short-term vision. I would say the, the, uh... I'll start with the short-term one. The, the thing that was immediately obvious was, oh, it could be, like, a writing assistant.
- SGSarah Guo
Mm-hmm.
- SLSimon Last
Um, so it, it could be in your document. You can, like, select some text, have it rewrite it. You can have it write text for you, maybe look something up, and then, uh, you know, give you, like, like, sources or more information. So that was the thing that we immediately, like, like, got to work on, and, you know, we sort of started a tiger team around it, and then we were able to launch it in, like, two, three months after that. And then the long-term vision that we immediately had was like, "Oh, the thing that looks like it may be possible is more of, like, a general assistant." So what if you could just give it all the tools inside Notion that a human would have, be able to, like, create its own databases, query, manipulate them, create documents, edit them, uh, and sort of weave all of these things together to do, like, a longer range task. And so we, we sort of, uh, immediately started on both. The, the short-term one, we were able to ship very quickly, and then the long-term one didn't really work yet, and so that took much longer to get working.
- SGSarah Guo
Are there, like, specific... First launch of the AI-specific Notion features and products was when? Last year.
- SLSimon Last
No, it was, uh, it was, uh, February twenty twenty-three is when we-
- SGSarah Guo
Oh, okay.
- SLSimon Last
-it launched. Yeah.
- SGSarah Guo
My timelines are wrong. Um, are there, like, a few, um, specific learnings or breakthrough moments you think since beginning to release that are interesting?
- SLSimon Last
Yeah. I mean, there's been... It's, it's, it's been a slog over many years or over all the years at this point with-
- SGSarah Guo
Yeah.
- SLSimon Last
With, with many, many learnings. I, I would say, yeah, I mean, just to give you a timeline of the arc of what, what we shipped is, you know... So the first thing was our, our writing system. Uh, we called it AI Writer. Um, that's the first thing we launched. Uh, it was easiest to get working 'cause it's, like, single-step task, rewriting, editing text. Uh, there's no, like, retrieval aspect. It was just, like, raw access of the model to write te- uh, uh, to write the text.
- SGSarah Guo
Mm-hmm.
- SLSimon Last
The next, uh, the next big thing that w- uh, that we, uh, immediately started working on was, uh, a Q&A, doing a semantic index of the entire workspace and then letting you ask a question and, and it can give you an answer that's, that's grounded in the sources. That was also immediately obvious to us that that'd be super useful, and so we started work on that. That one we launched in, I think it was October twenty twenty-three. So we started a beta before then, but then that wa-
- 4:10 – 7:16
Challenge of Semantic Indexing and Retrieval
- SLSimon Last
our, our GA was in October. That was just a, a much bigger effort to get working, obviously. We weren't just, like, like, plugging in the LLM. It was actually doing this, like, real-time updating index. Um-
- SGSarah Guo
Right.
- SLSimon Last
We had to get much more serious about the, the evals and the quality there as well. The, the Q&A has been a, a, a multi-year journey. Basically, what, what we did is immed- uh, as soon as we got the Notion index working, it, it was obvious that, okay, we should in-index everything else as well. And, and so we indexed, like, Slack and Google Drive, and, and we're launching new ones, uh, on a, on a regular cadence and, and, and now we have a, uh, I would say fairly complete index.
- SGSarah Guo
One could argue that those are, like, very difficult problems that, uh, you know, those products natively have not solved perfectly yet. So how did you think about taking that on? I don't know if that's, like, an offensive thing to other product teams, but, like, it's not working yet. [chuckles]
- SLSimon Last
Yeah. It's, it's kind of true. Yeah. This has been-
- SGSarah Guo
Yeah.
- SLSimon Last
-s-something we talk about a lot because it's like, you know, it's like, almost like, what right do we even have to do this? But, but it turns out that most of the companies are pretty bad at making their indexes somehow. Uh, it's honestly kind of baffled us a little bit.
- SGSarah Guo
Right.
- SLSimon Last
But I, I think my take after dealing with all of this and, you know, working with the teams to try to get it working is there's a little bit of just AI-pilled savviness that's pretty important. And then, uh, and then I think most of it is honestly just, like, a bit of a, like, like, craft and attention to detail. I think, like, in, in, in particular with this, like, indexing retrieval stuff, in order to really get it working, you, you have to be quite empirical and iterative and actually be, like, like, trying queries. Like, you know-
- SGSarah Guo
Mm-hmm.
- SLSimon Last
You know, like each, each, uh, data source is a little bit special. Like, you know, you can't just apply a one-size-fits-all to, like, like, querying Slack versus querying Google Drive, let's say. They're, they're completely different kinds of information. And we found that there's just a little bit of, like, like, craft and love that's going to it in terms of, like, actually trying a bunch of different queries, actually using it every day, and constantly iterating and rethinking and, and, and tuning how the retrieval works.
- SGSarah Guo
How did you, um, think about the diversity of how people organize their workspaces and just their-- I mean, even Notion is not-- use of it is not homogenous, right? Like, I'm probably part of fifteen workspaces as an investor, and so I look at them, and I'm like-
- SLSimon Last
Yeah.
- SGSarah Guo
"Well, mine's a mess, and these people are really organized," and the workflow is reflected in how their Notion works.
- SLSimon Last
Yeah, totally. I would say, I mean, the interesting thing is that with embeddings, it almost doesn't matter as much-
- SGSarah Guo
Mm-hmm.
- SLSimon Last
Anymore. The, the AI doesn't really care what the, what the, the tree structure is. For example, the-- all, all the AI cares about is that there's a snippet of text that has the, the context you need, and then it can retrieve it. And so actually we kind of advise people now, like, "Don't worry as much about organization."
- SGSarah Guo
Mm-hmm.
- SLSimon Last
"Just, just, just find a way to get it all piped in and like, like thrown in there."
- SGSarah Guo
You still make decisions that could change performance quite a bit, like chunking strategy or whatever.
- SLSimon Last
Yeah, yeah.
- SGSarah Guo
Right. Yeah.
- SLSimon Last
That's super important. But, but that's sort of not... That's sort of, uh, uh, transparent to the user and sort of in-independent of their, their particular method of organizing things.
- SGSarah Guo
Mm-hmm. It just seems like still a difficult technical challenge given-
- SLSimon Last
Oh, yeah.
- SGSarah Guo
How different the content bases are.
- SLSimon Last
Yeah, yeah. Yeah, I think, yeah, that took a lot of iteration. Yeah. The chunk sizing, how retrieval works, the different, like, steps in the pipeline of retrieval. Um, yeah. There's a lot
- 7:16 – 8:12
The Six-Month Rewrite Cycle
- SLSimon Last
of iteration on that.
- SGSarah Guo
Ivan said I should ask you, um, how many times you've rebuilt r- Notion and rebuilt your harnesses.
- SLSimon Last
Yeah. Yeah, it's kind of a running joke almost. I mean, we, we, we rewrite our AI harness probably every six months or so, and it... and the time to rewrite has kind of been, been decreasing just 'cause, I think, like, like progress has been accelerating. I think this is honestly a, a, a really key thing and something that a lot of companies get wrong is just, like, doing one thing and then just, like, like, sticking with it. You really do have to keenly aware of what the current state of the model is and the technology is, and then designing the harness, the system, and the product deeply around that. And it basically means you have to rewrite it every six months. And, um, I find it pretty fun. It's part of the process. Um, you know, you get to, you get to restart and, and, and, and rethink it. You know, we're working on a, a-- we're about to release a new version of our harness like in the next week or two. Uh, and then, and then we're already thinking about the one after that as well.
- 8:12 – 9:44
Notion’s Coding Agent Era
- SGSarah Guo
I, I think that leads to a, a set of questions I had for you on just like how does Notion as an engineering and product and research organization work now that you have the power of, um, coding agents as well? Because I imagine, like, your willingness to rewrite the harness goes up dramatically, like, agents are gonna help me do it.
- SLSimon Last
Yeah, that's extremely true. Yeah, I mean, I... Yeah, it's been, it's been really fun to, to use the coding agents. I, I think the ambition of what I even consider building has, has, has gone up a lot.
- SGSarah Guo
What do you think has most dramatically changed in how you think about how, um, engineering and product should work at Notion over the last two, three years?
- SLSimon Last
Oof. Yeah, I mean, it's, it's definitely changed multiple times. I mean, in terms of the coding agents, we kind of went through multiple eras. There was kinda like the tab autocomplete era, and then we, and then we got into sort of inserting, rewriting some code. Uh, but, but it wasn't really until the, the agents started working, I w- I would say like early last year, uh, we, we started to adopt the agents. Like, I started using CloudCode, I think, around April of last year. That was a huge unlock. Like, I would say the, the, the big shift there is that, you know, you can really push on getting these agents to end-to-end, you know, implement and, and verify and maintain stuff, but it, but it requires pretty significant thought in terms of how you architect things and what is the verification loop. Um, but, but, but the upshot is I think if you do it well, you can be much more ambitious about what you're building and also make it much more robust than you could've done, uh, with, with, with humans writing it. And then the flip side is if you do it badly, it's all slop.
- 9:44 – 12:49
Impact on Team Dynamics
- SGSarah Guo
Does that change your lens of like what teams should look like at Notion? Like size, seniority, anything like that?
- SLSimon Last
Yeah. I mean, I would say, I mean, the fundamental effect is that, you know, everyone's individual impact in terms of their output can be much higher, um, and your output increasingly depends on your ability and willingness to use the tools. I, I think that's the fundamental thing that's happening. And then like, like how does that play out? I think... I don't think we've seen that much impact on the, the team size really. I think we, we like to work in like, uh, uh, smallish tiger teams, uh, for the most part. Um, I think if you can make teams small, it's almost always better. That was true before, and I think it's still true. Uh, maybe increasingly a little bit, but, but, but not that much. I, I think, yeah, the main thing is to just like, like, really harness the tools.
- SGSarah Guo
Do you think something different happens to the median engineer in an organization versus the 10X engineer or the engineer 10X more willing to use the tools?
- SLSimon Last
Yeah, I think the, the gap is bigger. You can be like 100 or 1000X engineer if you're using the tools right now. I think, I think the, the gap is much bigger. Like the, the, the minimum bar has not changed, but the maximum bar has, has extremely increased. One impact it's had internally, I would say, is like broadly things feel like a little bit more messy and chaotic-
- SGSarah Guo
Yeah. [chuckles]
- SLSimon Last
I would say. Like, but I kinda love that. I mean, it's like there's, there's more proto- there's way more prototypes, uh, you know. People are... Like, uh, uh, for example, our, uh, uh, design team made an en- made an entire, uh, uh, Git repo, uh, they call it, uh, the design playground, and it's essentially like a simplified Notion, uh, with a bunch of like UI primitives-
- SGSarah Guo
Mm-hmm.
- SLSimon Last
In it. And they've made it like really sophisticated. You know, it, it, it has like an agent in there and like, um, and it's, it's pretty cool 'cause it allows them, uh, all the designers can, can spin up like super high fidelity prototypes-
- SGSarah Guo
Mm-hmm.
- SLSimon Last
Really quickly. And so it's no longer like, like pointing at a mock and being like, like, you know, like, like, "How will this look like?" They'll give you like a URL to a prototype that's, that's, that's been deployed. And that sort of thing is true all the way up and down the stack, you know, for all of engineering, just like a little more chaotic, more stuff happening. Um, all the PRs are more ambitious.
- SGSarah Guo
Do you draw a line somewhere about like stuff that is more dangerous to touch or sensitive? Like, ah, there's could be risk of data loss over here or, and not?Or is it kind of you look at it all as it's fair game?
- SLSimon Last
We still do reviews on all the pull requests. And, and I would say-- And, you know, all the pull requests are now written by agents. They're often like larger and, and, and more complex. That's like the worst part. But the better part is that they're often like a, a much better tested, and we can demand sort of a m- a much better testing for the things that merit it. I never produce a PR that like hasn't been like fully end-to-end tested anymore. And so it's like you can get to a pretty high degree of confidence that, that it works, but it requires like you're not just vibe coding by, by saying the thing you want. You're sort of thinking carefully about like, what is the thing I'm tr-- like what is the change I'm trying to make? And like, and, and how can it be verified, and how can it be deployed safely? And then enlisting the agent to
- 12:49 – 15:39
Launching Custom Agents
- SLSimon Last
help you with that process.
- SGSarah Guo
When you think about where you said the general assistant like doesn't quite exist yet, um, what's the-- what do you imagine Notion's agent- agents being able to do like over the next year or two that are still unblocked-- they're still blocked by either capability or your harness work?
- SLSimon Last
We struggled for a few years to build an agent. Um, and you know, it always like, like sort of works, but then, you know, wasn't that useful. Largely just it, it was too early. So we, you know, we, we tried to, to build an agent, I would say actually three or four times. And then, uh, uh, we finally launched it, uh, last fall, so like last August, September. Um, so the-- If you use Notion AI now, it's like the full agent that has access to everything in Notion pretty much. Um, so that, that, that totally works. I would say like the-- a lot of the original vision that we had totally works now. Um, and it, you know, it's like, like fully shipped. Last August, August or September, uh, uh, we shipped our personal agent. Uh, so it's pretty much every user in Notion has an agent, and it basically, it has access to all the, all, all the things that the user has access to. So, you know, it can create a database for you, it can update things, create documents, it can search, search the web, do research. And then the second big thing, uh, that we just launched, um, uh, last week actually was, uh, uh, custom agents. So you can basically, you can create a new custom agent, give it a name, and unlike the personal agent, uh, by default, it doesn't have access to anything. Uh, so you have to grant it access. Uh, but, but then once you do, it can actually run autonomously in the background. So for example, you can give it access to its own database, uh, to file tasks, let's say, and then you can attach it, uh, to a Slack channel, and then it will start responding to people in Slack and, uh, uh, filing tasks. That's, that's, that's one use case. Another one is maybe you could, um, you could give it access to a database of like weekly reports and then, and then let it search the web or search your workspace. And so it's sort of a custom agent sort of represents some work or job, some, some knowledge work task that you want to be done autonomously. One thing I'm really excited about this going forward is, is, um, we want it to be extremely good at sort of bootstrapping its own capabilities, basically from an initial kernel, allowing it to basically bootstrap itself to do anything, right? So even for example, maybe, uh, uh, building an integration that we don't support yet, deploying that, and then, uh, and then using it.
- SGSarah Guo
So you imagine that Notion agents are actually the broader definition of agent, where like writing code is a tool it's got access to.
- SLSimon Last
I think it's pretty key.
- SGSarah Guo
Yeah.
- SLSimon Last
I think it's pretty key. Yeah. I think I, I think of coding agents as like the kernel of AGI. AGI will be a coding agent. Um, and, and, and, and code is just a really, really useful, uh, a primitive for representing like a deterministic logic. The thing that's really exciting about it, um, applying it to, to a knowledge work agent, is that it can bootstrap a capability. You know, so yeah, like I said, if an integration doesn't exist, it can build it. Um, if, if it needs to, uh, you know, connect itself to a new data source, it can do that.
- 15:39 – 17:33
Notion as the ‘Switzerland’ for Models
- SGSarah Guo
Given you have a, you know, Notion is at scale but is operating in a landscape of productivity and platform players that are at even more scale, right? Um, many of these will end up with their own agents. Lots of people from the labs, the Microsoft world are trying to integrate other data sources. So you have this like [chuckles] cross attempt to integrate and index. Like, how do you think that plays out? Like, w-what do you, what do you imagine that Notion agents are best at or what they have the right to go do?
- SLSimon Last
If you look at the landscape, like I, I would sort of say there's the labs, and then there's maybe the, the, the, the software platforms, and then there's maybe like infrastructure. In terms of the labs, you know, we see ourselves as kind of like the, the Switzerland for models. We think and our customers, they, you know, they don't wanna be locked into a certain, certain labs model. They're always, uh, releasing new versions. Any given month, one is better than the other. Um, so we wanna be a, a, a place where basically you can, you can easily get access to all the best models, um, at any time, and you can easily switch around.
- SGSarah Guo
Do you think open source plays into that as well?
- SLSimon Last
Yeah. Yeah. Absolutely. I think, uh, the open source models are actually getting really good. There's like four different Chinese models now that are-
- SGSarah Guo
Competing.
- SLSimon Last
That are quite good.
- SGSarah Guo
Yeah.
- SLSimon Last
We actually just, uh, released one of them in our agent, uh, last week, and, and we're gonna do all four for sure. Um, they're, they're actually quite good, and they're, and, and they're way cheaper than the, the frontier models. So I think there's, there's a lot of use cases where, where, where you'd want that. We wanna give that as an option. In terms of like the other... You know, so, you know, we think of our role as sort of at, at taking all the best models that we can, creating, uh, really high-quality state-of-the-art agent implementations where, where people can easily and conveniently get access to them, and then making sort of a collaborative workspace that is r-really good for, for humans and for the agents, uh, to, to, to coordinate on. I think it's, it's something that's, that's very needed in the world, and we're just trying to do it in a really tasteful,
- 17:33 – 20:09
Designing APIs for Agent Customers
- SLSimon Last
well-executed way.
- SGSarah Guo
You were describing you need the index to make the agents good. Um, you give the agents access to the tools that we humans have in Notion. How do you think about, um, the structure of Notion and like where it's like useful or even like not useful or relevant for agents, like blocks and databases and such?
- SLSimon Last
It's all still pretty useful. Um, e-extremely useful. It, uh, there, there's been a ch- a, a challenge to sort of, you know, we wanna make it really convenient for the agent. I think that's-That's a new thing that, that didn't exist. You know, in the, in the past, it was convenient for humans, and then we also made APIs convenient for humans writing code-
- SGSarah Guo
Mm-hmm
- SLSimon Last
... to use our API. Uh, so we essentially have a new customer, which is the agent. At first, that was definitely a problem. You know, so for example, like our, our API, uh, uses this, this crazy JSON format for blocks that, uh, by default is like crazy verbose and like, like horrible for the agent. But we basically took on that challenge and, um, designed, uh, just really convenient APIs, uh, for the agent. Uh, we created sort of a markdown dialect that, um, looks like the default normal markdown, but it's sort of enhanced with, uh, all the Notion blocks. Um, and the models are really good at it. It works really well. Uh, so, so that's how it reads and writes the pages. And then, uh, for databases, uh, we, we, uh, use, um, SQLite. Um, so, so basically it's, uh, gets to speak in SQLite, which also works really well. So the default thing did not work really well, uh, but then we just like, like took that on as an engineering challenge and, and I would say now we have like extremely convenient APIs that the agents are, are really naturally good at.
- SGSarah Guo
How did you, uh, understand or figure out what would make the API better for agents?
- SLSimon Last
That's a good question. Yeah, I would say it's a, it's a combination of just trying things. It's, it's, it's very empirical.
- SGSarah Guo
Mm-hmm.
- SLSimon Last
So, so we're just playing around and like, like noticing, "Oh, it's not very good at that. Oh, that's way too many tokens. How can we make this smaller?" And then a little bit of just like, like first principles thinking of like, you know, what is it the models are being trained on? And what's, what's in their prior? What do they know? And what do, what do we think it would naturally be good at? And, and like, like how does the agent loop work? And like, what, what would be the convenient, efficient pattern for, for accessing these things? Um, and so... And then just, you know, a, a lot of playing around.
- SGSarah Guo
I hear user research where the user is actually agent, and then, you know, ongoing eval.
- SLSimon Last
Yeah. I mean user... Yeah, you just chat with it. [chuckles] The user's always there. It's ready to talk to you.
- SGSarah Guo
Yeah, I s- actually, that is wonderful, where you have infinite access to it.
- SLSimon Last
You have infinite access to it, yeah. And, and you can, you can script and scale
- 20:09 – 24:48
Simon’s Personal Agentic Workflows
- SLSimon Last
the access as well. [chuckles]
- SGSarah Guo
I assume you have... Actually, I know you do, because you walked in, you're like, "Hey, I need to get access to Wi-Fi. I need power. We can't block the agents while we're doing this." Um, what do you have running right now? Tell me about your setup.
- SLSimon Last
I'm working on a new prototype, and so I have a couple agents. I'm working on that. Um, and then, yeah, my setup these days is just, um, either Claw Code or Codex. I like the, the, the CLI tools. Um, they're, they're, they're super simple and, like, work pretty well. I'm, I'm pretty comfortable in the CLI, so... And then, yeah, my, my-
- SGSarah Guo
You don't need my generated game-
- SLSimon Last
[chuckles]
- SGSarah Guo
... just CLI commands.
- SLSimon Last
It's, it's, it's a very cool idea. Um, I would say, yeah, my, my, my whole goal these days is essentially to just have as many running as possible and to run them all the time. And, you know, so for example, like every night before I go to bed, I'm, I'm like, "Okay, I, it-"
- SGSarah Guo
Let's go, guys.
- SLSimon Last
Yeah. Basically, what I have to do is make sure that I've given it enough stuff that by the time I wake up in the morning, it, it, it will still not be done.
- SGSarah Guo
Mm-hmm.
- SLSimon Last
And so I've, I've maximized-
- SGSarah Guo
That's victory.
- SLSimon Last
Yeah.
- SGSarah Guo
Yes.
- SLSimon Last
That's, that's victory.
- SGSarah Guo
Yeah.
- SLSimon Last
So yeah, like I've, I've, I've done that, I would say, last, last five nights pretty well. My personal record is that I've had a, a coding agent running for I think it was 13 days straight, uh, without stopping, and just, just basically working through like tasks.
- SGSarah Guo
Well, well prompted. Yes. I, I admit to having woken up in the middle of the night at least multiple times this week and just being like, "Are you still going?" [chuckles]
- SLSimon Last
Yeah, I know. Yeah. It's, it, it's kind of nerve-wracking. I, I always like... There's always like, I'll, I'll check it one last time before bed and just really make sure that it's still spinning.
- SGSarah Guo
What about on the Notion agent side? Like, do you have a workflow there that is core to daily work?
- SLSimon Last
Yeah, I mean, I mean, I, I use our personal agent all the time, so it's... It, it has all the context about, about our company and everything that's going on, you know? So like, uh, for example, last night I was asking it about, um, how the custom agents launch was, was going and like, like, like what the, what the signals we're getting from it. Uh, we're super useful for that. And then for... I, I have many custom agents that are, that are running. Uh, my, my, my personal favorite is I have a email triage agent, so it has access to all of my work and personal emails. Um, and it just, uh, wakes up every day and just archives all the stuff I don't need to see. I train it over time to, uh, to, to learn my preferences.
- SGSarah Guo
Do you actually label data for it?
- SLSimon Last
It's pretty easy to do this, actually. So all you have to do is you make the agent, and then you give it access to your email, and then you, you can make a blank page that's like its memory.
- SGSarah Guo
Mm-hmm.
- SLSimon Last
And you let it edit that page. And then you just say, "Okay, now go and, uh, look at my emails and then interview me. Ask me which things," you know. So sort of it will like propose things that it thinks it should archive.
- SGSarah Guo
Mm-hmm.
- SLSimon Last
And then you can kind of correct it, and then we'll use that to essentially generate like a list of rules about like, like what it thinks are correct or not. And so for the first couple of days, I was sort of like, like, like, uh, correcting it on things. After a couple of weeks or so, I, I j- I, I dropped the approval entirely and it, it just automatically archives all the things I don't need to see now.
- SGSarah Guo
Wow. A lot of trust.
- SLSimon Last
It's super useful.
- SGSarah Guo
Yeah.
- 24:48 – 27:28
Notion: Tool for Work is Now A Tool for Agents
- SLSimon Last
of times, you kind of trust that it's working and then
- SGSarah Guo
Is there anything you do internally at Notion to, um, make sure non-technical teams have the intuition for how to build agents or how to, like, express that productivity too?
- SLSimon Last
Yeah, that's a great question. I, I, I mean, we do, uh, sort of workshops and hackathons pretty frequently. So like a, a, for example, like a month ago, I did a, I did a hackathon with, uh, the, the people team and sort of, sort of got them. The, uh, the people team's been amazing. They're actually one of the, the highest adopters of custom agents.
- SGSarah Guo
Cool. [chuckles]
- SLSimon Last
You know, they do all these kind of workflows in like Slack and Notion, kind of like, like manual work like that. And, um, and yeah, I would say, yeah, like, like people are super excited to, to try it and sort of like, like maybe just need a, like a little bit of a push in terms of intuition and like, like getting them started. Um, but then honestly, I've been super impressed. Like, y- I, I think the concept is like kind of intuitive, sort of like, like once you get g- once you get past sort of a little bit of the technical barrier of like what is a prompt and like what is the agent and how does it get triggered and woken up and like, like how does that even work? But then once you sort of get past that, I think it's actually a very human-like interface.
- SGSarah Guo
Yeah. Maybe the, maybe the biggest barrier is actually just getting people to try and assuming it's gonna work at all, right?
- SLSimon Last
Yeah. Yeah.
- SGSarah Guo
You and I, Ivan, originally met on the internet Tools for Thought community. Um, it feels like, you know, the tools we have for thinking are very different now. Has your like core conception of Notion changed over the last few years because of all the AI stuff? Like what, what is the, what, what thinking does the tool do for you? Should agents do for you? What do you get to do?
- SLSimon Last
Yeah, I mean, it's, I would say changed quite a lot. I mean, broadly speaking, before a-AI, our, our, our, our goal was to create the best tool for humans to directly perform their work.
- SGSarah Guo
Mm-hmm.
- SLSimon Last
And then now the goal is to create the best tool for humans to manage agents to do the work for them.
- SGSarah Guo
That's a big shift. Okay.
- SLSimon Last
That's a pretty big shift.
- SGSarah Guo
Yeah.
- SLSimon Last
Uh, it's, it's pretty fundamental. Um, but it, it turns out that you need most of the same primitives. Uh, you actually-- all the primitives that we built are actually still extremely useful. It's, it's more that we just needed some, some new primitives like, like representing what is an agent and, you know, how does it interact with your pages and databases. But, you know, you still need the same primitives. You still need a document. It's an unstructured way to w- you know, to write stuff. Uh, agents love to write markdown documents, so-
- SGSarah Guo
Yeah
- SLSimon Last
... it's still very relevant, and you still need a database. It's, um, you still need structured data. You know, i-if you're working with your, your swarm of like a hundred background coding agents, you don't wanna have a hundred chat threads, uh, you want a Kanban board. It's, you know, the same as before.
- SGSarah Guo
Makes sense. You still need the,
- 27:28 – 29:00
How Building Has Changed for Simon
- SGSarah Guo
uh, the coordination structure. What is one thing that just because you're ahead of the, on this stuff and then trying to figure out how to bring, you know, Notion and then users along with you, what is something that's really changed about, um, how you personally like build even in the last six months?
- SLSimon Last
I mean, it's completely changed. I haven't written code since like last summer. I don't type code anymore.
- SGSarah Guo
Mm-hmm.
- SLSimon Last
Yeah. It's, it's, it's completely shifted. I mean, we went from humans type all the code, uh, to like, we're still typing, but we like tab complete to sort of like, we talk to the agent and it sort of does little tasks for us, but we are still in the outer loop, and then now it's more like I, I design a end-to-end task that involves like making some change and end-to-end verifying it, and then I'm just the, the outer, you know, the outer verifier sort of like, like double-checking at the very end that it, that it's correct, and if it's going off the rails, kind of like, like monitoring it. Um, so it's a com- it's a complete shift. It's, you know, I'm, I'm now like the agent manager instead of the coder.
- SGSarah Guo
Amazing. Well, um, thanks Simon. This has been a super great discussion about how we're all gonna become agent managers and, uh, uh, hopefully in Notion.
- SLSimon Last
Cool. Yeah. [upbeat music]
- SGSarah Guo
Find us on Twitter at NoPriorsPod. Subscribe to our YouTube channel if you wanna see our faces. Follow the show on Apple Podcasts, Spotify, or wherever you listen. That way you get a new episode every week. And sign up for emails or find transcripts for every episode at no-priors.com.
Episode duration: 29:02
Install uListen for AI-powered chat & search across the full episode — Get Full Transcript
Transcript of episode 1dYThQgOyZU
Get more out of YouTube videos.
High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.
Add to Chrome