EVERY SPOKEN WORD
80 min read · 15,953 words- 0:00 – 2:09
Introduction
- SGSarah Guo
(instrumental music) . Hi, listeners. Welcome to No Priors. Today we have Ivan Zhao, co-founder and CEO of Notion, the beloved productivity application for notes, tasks, and knowledge base. They recently launched an AI Q&A interface as well as a calendar application. We're super excited to have Ivan. Thanks for being here again, Ivan. Uh, we are gonna start with, uh, the hardest question, which is what is Notion?
- IZIvan Zhao
Notion's always pretty hard to define because it can do so many different things. Um, but that's also our goal. We wanna give people one tool that they can do their most work with. For a personal user, that means all your personal notes, all the planning for a trip or for your wedding. For business, for enterprise, for company, that means all your documents, all your tasks, all your issues, calendaring, knowledge base in one tool. Um, the reason we wanna do that because there's just so much fragmentation in the market today. We wish like... It wouldn't be nice as one place to do your most work. And the, our approach here is rather than try to cram all different use cases into one product, uh, what are the underlying software building blocks? What are the Legos that power those use cases? Can we give users those Legos so they can be creative with software themself? They can create and tinker their perfect workflow for their personal life or for their company. And none of this is new, by the way. Like people back in the '80s, even '70s, tried this kind of building blocks approach to software. We're just trying to take a modern spin with cloud work and with AI, uh, to what it's like to break the prison of application-based software.
- SGSarah Guo
It's dramatic to think we've been living in a prison of SaaS fragmentation for the last two decades. But I, I do think it's actually, um, uh, you know, surprising to hear a point of view that is so obvious, which is like of course we want one tool where the data was interconnected.
- IZIvan Zhao
Mm-hmm.
- SGSarah Guo
Why do you think people, um... Why do you think more people don't try that, to have unified tools and unify data underneath?
- 2:09 – 5:39
AI and Computing literacy
- SGSarah Guo
- IZIvan Zhao
I think people try it for different angles. Like even fairly recently, there's this thing called no-code, right? No-code is like coming from this kind of like power user, developer angle of, "Wouldn't it be nice everybody can modify this underlying software they use every day?" That's one angle. The, it wasn't coming from the angle of the, the knowledge and data wants to be in one place, right? And language models sort of give another angle is the underlying knowledge and embedding space wants to be one place. Wouldn't it be nice in one place, right? And the macro is also coming from the, uh, the budget place. Wouldn't it be nice rather than pay for five different vendors and all C-based business just to pay one vendor and save some money? So there have different angles from different times. Um, I will say we are more come from this kind of computing and medium and literacy angle. Like you and me go through school to learn how to read and write, uh, you know, English and Chinese. We've spent years to do that. We all know how to do that. The world, the same MacBook for most people are, are very rigid. It's, uh, more like a machine to do typewriting or watching YouTube. Uh, not much more beyond that. It's not very creative, right? Um, wouldn't it be more nice that more people can use their software more creatively, right? Because there's a separation be- between people who can make software and people who use software, in... That's why SF's rent is so expensive because we're the modern-day Detroit or Manchester, right? We're the factory of the world. Um, Notions largely come from that angle, which is the original angle. We were inspired by our c- early computing pioneer. They thought about that angle, right? They thought about computing could just be like literacy one day everybody can do it. I, I guess they didn't expect AI might make that even give a really interesting twist to it because now language model AI can not only do create software, but also do a lot of thinking working for you so the future is pretty interesting, yeah.
- SGSarah Guo
So for someone who thinks on, you know, um, span of like decades of, you know, what should computing look like and what were, um, what were the most ambitious plans for personal computing, you know, uh, three, four decades ago, like what are you most excited about seeing from AI broadly over the next decade?
- IZIvan Zhao
I think three, four decades a bit too long. If AGI happen that time, like computing might not be necessary. For this decade, I think a one sleeper category is the, the drag, the embedding space. The decades might be too long. I would say like the n- the next year or two. Now the language model can understand what you put into a computer. Understanding. So rather than you do the organization to make your retrieval, retrieve the understanding more easily, machine can do that better than anybody else can, right? So before that, uh, we use keyword-based search where you find your coworker who remember that, that cue. Where does that information sit? Now, uh, just ask Notion AI and you get that in seconds. So that's one I'm personally really excited about. I think not enough people talk about it. And of course the other one is like the agent, the workflow side that, that has a lot of buzz already
- 5:39 – 8:43
Building the Notion AI team
- IZIvan Zhao
so that's interesting too.
- SGSarah Guo
You and Simon, uh, set bet the company on AI and are, you know, have real conviction, and as you are building out the team like-
- EGElad Gil
Um, what does the talent look like you, like have or need to make, um, Notion-
- IZIvan Zhao
Yeah.
- EGElad Gil
... an AI first company?
- IZIvan Zhao
And I kinda argue, you, you folks are one of the earliest adopters of AI at scale as an application.
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
So part of the question in some sense is you've built so many interesting things, like what are the people that you now need to sort of build the next level of stuff, in addition to what you already have on the team? In the early days, kinda just brute force. Simon's really good (laughs) . He had built a lot of things and learned really quickly, right? I would say Notion's a company where largely people interested in interface and design, a lot of full stack, front end heavy folks, and back end people who scale. We have somewhat a small team of search, but we don't have too m- too many ML folks, almost nothing. And in at least my learning, our learning in the past year or so building for AI is okay, you ML folks are important. It's kinda like you no longer do a dis- deterministic thing that you can see how it works. It's almost like a, I don't bake, but feels like baking, right? Uh-huh. Yeah. (laughs)
- EGElad Gil
(laughs)
- IZIvan Zhao
Yeah. That you have to like do something, get the thing, ingredients ready, run through, rinse, press a button, and wait for a while and see does it come up?
- EGElad Gil
Right.
- IZIvan Zhao
Where it's a different way, a different sense of patience, and different type of personality to do that well. Um, a lot of massaging, a lot of preparing. Um-
- EGElad Gil
Some of my friends call that, uh, probabilistic software engineering.
- IZIvan Zhao
Kinda like that. Right.
- EGElad Gil
They think it's, it's morphed into this sort of stochastic world.
- IZIvan Zhao
Mm-hmm.
- EGElad Gil
Or at least partially stochastic.
- IZIvan Zhao
Yeah. So one is like maybe gardening feels like that way. I don't garden either. Uh, so that category of people are to, to me is pretty necessary. The other category is like people who are curious and learn really fast, right? It's like, okay, um, like the group of prompt eng- like, language, prompt engineer, language model sort of make everybody like a, a real time machine learn-
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
... learning engineer. You just prompt right and then you can get your stuff, right?
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
And there's a lot of trick and techniques, uh, um, and how does that plug into user interface? Uh, I think this category of people called AI engineer or something. There's a terminology for 'em. They tend to be pretty young. They tend to be like, we have someone like under drinking age-
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
... working on Notion. Uh, they, they fit into that bucket.
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
And I think both seems to work quite well. Um, we don't have too many researchers at Notion. That's another one I think, uh, um, will be important, but we're fundamentally sit in the application layer, so it's more about applied side of things.
- EGElad Gil
Makes sense. So more manipulating the models and making sure you can scale them to, like user outcomes and things like that.
- IZIvan Zhao
Mm-hmm.
- EGElad Gil
Is that right?
- IZIvan Zhao
And another part is like the scaling part, right? How do you scale to like tens of millions, 100 million users? It's a bit, it's a problem on its own too. It's beyond just a demo on Twitter, right? It's tricky, yeah.
- 8:43 – 12:09
Notion as an application company
- IZIvan Zhao
Yeah.
- EGElad Gil
So, you have often said that, uh, Notion is less a productivity company than an application building company. How do you think about the initial use case and like how- what makes you believe people want to build more applications?
- IZIvan Zhao
I don't think people wanna build more applications. What got me started in Notion, got us started in Notion, it's, um, last year college, I read a paper by, uh, one of the computing pioneers, Douglas Engelbart. Uh, he talk about his papers named Augmenting Human Intellect. So every day we use software today very much like application where you go into one ac- application and do one thing. But for that generation of computing people in their 60, 70, 80s, computers are a lot more, software are a lot more malleable. You can actually tinker and modify, right? Smalltalk, you can go into that and change how the operating system work on the fly. Um, that really inspire me. It's like today, people with software are so rigid, can we create a new breed of software that people can modify, can, uh, can change and customize and bring back some original ethos of those early computing pioneers? That's why we started Notion. Um, the hard lesson for us is that, like you mentioned, most people don't wanna create software. They don't wake up saying, "Hey, I want to create my perfect project management tool or my project, perfect knowledge base." They, their boss ask for something, they just have to get their work done, right? Um, uh, so the, in some sense, our learning and pivot is instead of giving people those, um, software building tool, we have to package the software building blocks together as ready to use templates, as ready to use use cases that people can adopt really quickly.
- EGElad Gil
So you were one of the earliest adopters of AI in terms of, uh, application with any real scale. And I think it's impressive how quickly Notion ended up starting to work in this area.
- IZIvan Zhao
Mm-hmm.
- EGElad Gil
How do you think about how that impacts different aspects of what you've built and what you're building going forward? And how does that impact that vision of saying, "Okay, we have this, um, this effective platform that allows people to both interact with, uh, documents or core use cases in simple ways, add things like calendar, but then also go in very interesting directions in terms of both the set of applications and templates they can use"?
- IZIvan Zhao
Yeah, I think we're lucky. Like I mentioned, we're not trying to build specific use cases, right? We're trying to build the Lego bricks that power those use cases. Um, what are those Lego bricks? Text editing is, uh, one fundamental Lego bricks. Most software have that piece. Relational database, a table is one fundamental Lego bricks, right? Um, different form of permission, commenting. So we've been spend five plus years building those Lego bricks and feels like, boom, AI just jumps in almost like a brand new car engine and can power those Lego bricks in brand new ways. Uh, so it feels very lucky in that way. And that, because we've been building those Lego bricks and refining those allow us to ship features, uh, plugging with AI really quickly. We're one of the early years want to launch, uh, AI writing for productivity software at scale, because we've been spend years building a text editor. Uh, we can do, uh, AI powered database table features really quickly because we've been building relational databases. Um, we've been building a knowledge base for a long time, so we launch AI Q&A really quickly, fairly quickly. Uh, the rack system on top of Notion because we have those Lego bricks. Uh, so in some sense, kinda like...... at just the right moment, right time for us, right?
- SGSarah Guo
How did you, um, begin
- 12:09 – 14:53
Prioritizing AI investment
- SGSarah Guo
to, like, resource and prioritize this effort? Because you're like, "Ah, magic. We have this engine, it applies for our Lego bricks," um, and then you started shipping pretty quickly. But I think there are a lot of organizations right now trying to figure out what to do with AI, and so, you know, in terms of, like, designing the features, prioritizing that effort versus everything you're already doing in a rapidly growing software company.
- IZIvan Zhao
Yeah. I think I had the conviction. My co-founder, Simon, actually, they all had the conviction. Um, funny, 'cause we all live in the mission, right? And OpenAI initially is, uh, still is in the mission. And, uh, some of them are friends, especially Simon's friends work at OpenAI. I remember we go to their office, they ... During the Dota days, they were like, "What is this company doing?"
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
"Kinda interesting." And Simon, uh, and some people at Notion saw early demos of GPT. It's like, "What is this thing?" Spitting out text, s- sometimes gibberish, sometimes useful. Um, I personally, I have to admit, I slept on it. Um, on GPT-3, even saw what GPT-3 feels like. "What is this thing useful for?" ex- like, yes, for marketing, for content writing, to create a first draft. Didn't really click for me. Um, personally for me was, uh, fortunate enough to saw early preview GPT-4, and that's like, "Oh, wow, this thing can think. It can reason. It can know how to do things, has this a little bit-
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
... workflow power to it." That's the big aha for me, and like, it just give me, it personally gave me so much conviction, like, this is gonna change everything. If you think about what knowledge work is in, right? Why do we use software? Fundamentally, SAS, software, this is all where all in the same information, people, paper pushing activity, right? It's like, a piece of paper come in front of you, a human, uh, if, like, change a couple bits, push to another human. Language model can do some form of this now. So, uh, so that just, like, gave me the conviction, like, this is gonna completely change everything we do with a computer. And after that, we sort of just bet the company on it, uh, like, we're lucky now to have those Lego bricks, and then what count, which Lego bricks can works well with AI, which doesn't. We're still trying to figure that out. Who inside the company are good with this technology? We have search, but it's not like we don't have a lot of ML folks, so need to hire more ML folks, need to get people inside a company to have similar convictions so we can move in the same direction. It's quite interesting. It's kinda like ... So it must have felt, dinosaur feels like when asteroid hit the Earth, and it's like, "What do they do with it if-"
- EGElad Gil
Yeah, yeah. There's a lotta change coming for sure.
- IZIvan Zhao
It's a lotta change.
- EGElad Gil
Yeah.
- 14:53 – 17:46
The rapid evolution cycle of AI development
- EGElad Gil
What do you think is missing from the capability side? Because to your point, I think a lot of people weren't really thinking about AI too much until ChatGPT and GPT-4 came out, and there was a period of time where 3.0 and then 3.5, and you started seeing the capabilities incrementing up, and entirely new businesses are suddenly able, enabled with each sort of step with the next GPT level model. You know, GPT-5, or to your point, RAG adds a lot of capabilities.
- IZIvan Zhao
Mm-hmm.
- EGElad Gil
What are the biggest missing gaps for you to take full advantage of this technology? Is it future reasoning? Is it better, uh, thinking and knowledge? Like, what's the-
- IZIvan Zhao
Yeah, I think all but the above. To me, it feels like technology is funda- we're in the tech business. Technology is fundamentally about trade-offs, right? It's like, the plastic can do things that wood cannot do.
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
And we discover plastic, and then we figure out new things. We can boil water like this. Before, you cannot boil water with a wood table, right? So-
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
... um, all of a sudden, we have this thing called language model. They have a new characteristic that deterministic software, uh, cannot do in the past, and w- and we don't really know how it is made fully, so every month, every week, if you're on Twitter, people discover new techniques to get more out of this.
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
And for companies, entrepreneurs, they're also making trade-off, discover what, how the market react to this capabilities of this new wood, this new language model. And so it's a constant evolutional cycle happening really, really fast right now, right? Um, I think if, with that mindset, uh, what are the dimensions? Um, on a technical side, on the l- technology side itself, yes, the model gets c- larger context windows, more reasoning-
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
... better speed, smaller footprint. Um, those are all great. Like, like, for Notion, uh, to power the workflows that we really need, like, we learn, like, GPT-4 is smart, uh, a, a cloud too is smart. We need that intelligence to do reasoning, where for text summarization, cheap, fast is better, right?
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
That's the technology side. And in my opinion, there's so much about human behavior as well, just like inertia in, in our personal behavior, a company's risk tolerance. Uh, that's slowly evolving as well, right? Like, like, what is Steve Jobs always talk about, "You cannot make something too new. You have to be largely the same and change one thing and two things," right? Virgil Abloh, the Off-White guys, like, 3% difference.
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
Just push the boundaries so people can accept it, but still also new, right? To me, it feels like language model power application are kinda in the same phase. If it's too different, people are like, "What do they do with it?" Right? It's such alien behavior. You have to ... The RAG is pretty nice because largely it existing behavior
- 17:46 – 20:00
Notion Q&A
- IZIvan Zhao
but better output, right? So-
- SGSarah Guo
Can you describe the, um, the AI Q&A product for, uh, people who haven't experienced it?
- IZIvan Zhao
Um, right. Essentially everything you put in Notion, Notion help you remember. Right? And this not just apply to Notion. It pretty apply to most RAG systems. Like, why do we use computer? We need to store things and need to recall things.Before language model and RAG, the recall largely happened based on keywords, right? You had the keyword has to be precise, or there's some lexical tech tricks that you can, like, recall easily. Um, uh, im- precisely. What RAG happens, language model can actually understand what you're putting into there, so you no longer need to organize your information in Notion. Whatever you throw in there, you can find it later. What that means is, for a person or for a company or for a team, you can have perfect memory. And not only have perfect memory, the- the right piece of information, if we design our software right, can push to the right person at the right time, right? That's probably more than 50% knowledge work, right? We're still perfecting the system. I think we're- we're one of the first on the market that apply a scale. We still have a- some- somewhat a waiting list, because it's hard to do this at scale still. Um, but for a company, for a team, before search is, uh, one of our weaker point, but with RAG, you completely change that. It changed how I use Notion. I can just ask a question to Notion, like, um, "How large..." When are we moving out of the SF office to a new office? And is someone wro- in the company wrote in some documents. I don't have to ping five different- three different people to find the answers. If it's in Notion, it will find it for me, right? Um, everyday engineers, designers, operation people just keep asking each other on Slack or in email such question. Each question is 10 minutes writing the answers-
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
... 20 minutes to find answers, and there's the delay in the middle. With Notion Q&A, you can completely cut that, right, into, I think, seconds. We're just at the beginning of what RAG can do for work.
- EGElad Gil
It's pretty amazing. I feel like RAG and embeddings are very under-discussed or underappreciated in some sense, relative to the impact that they really seem to be having or starting to have,
- 20:00 – 22:43
Workflow and AI for calendars
- EGElad Gil
and I think Notion, um, Q&A is a great example of that. I guess the other thing that you folks just launched is calendaring. And if you can't talk about it, or if- if there's nothing to talk about, that's fine too. (laughs) But, um, I feel like one of the really interesting things that people are talking increasingly about is agents and sort of the agentic world, and there's a lot of capabilities missing to really make those valuable. But in the can- context of a calendaring application, you could think of all sorts of ways that having AI act on your behalf or help understand things can be incredibly valuable. And so I was just curious how you think about the application of AI relative to calendar versus, you know, some of the core, um, information-related things that you just talked about.
- IZIvan Zhao
Maybe we can group AI stuff into, like, at least in my mental model, it's the RAG, the knowledge, information retrieval is one bucket.
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
Knowledge bucket. Then there's this workflow bucket, right?
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
You used the word agent. Uh, um, that's- that's in that bucket. Calendar's somewhat in that bucket.
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
Um, why do we need to meet? Bec- uh, why do we need calendar? Because we need to meet and we need to schedule time. We need to-
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
... figure out, exchange some kinda bits between my brain to your brain, right? Um, can that bit- can that exchange be done by a language model? Maybe.
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
And can the meeting time be done by- scheduling be done? That's like a baby step, right? Um, and most things we do has this kinda time dimension to it. Uh, can language model help us shuffling our schedule?
- EGElad Gil
Yeah, it feels like there's also that information retrieval piece of it, because, you know, if my calendar auto-populated everything I needed to know about the meeting or the people attending or other things, that's incredibly valuable as a user of a calendar. And so I just feel like there's a lot of these things that kinda tie in together, both in terms of the coordination, which you mentioned-
- IZIvan Zhao
Mm-hmm.
- EGElad Gil
... and the workflow, and then separate from that, there's just-
- IZIvan Zhao
Yeah, I think the-
- EGElad Gil
... what, what do I know about this person?
- IZIvan Zhao
... the calendar part, it's the simpler part of the workflow. Like, the Holy Grail is kinda, like, can just the agents robust do our knowledge work for us, right?
- EGElad Gil
Mm-hmm.
- SGSarah Guo
It's really interesting framing that I didn't have before of a bunch of the work you're doing at Notion actually eating into, like, communication, right? It's- it's sort of obvious in retrospect, but, like, if you look at what you describe of like, oh, like, am I really gonna Slack back and forth about this thing, about, you know, when we're moving, if I can just know?
- IZIvan Zhao
Mm-hmm.
- SGSarah Guo
I'm in Notion, help me know. Or with calendaring, like, you know, the most intelligent version of it is like, "Well, do I need to have that meeting or can you tell me what Ivan was gonna tell me?" (laughs)
- IZIvan Zhao
I know. Like, why do you need to communicate? Because there's something, the work cannot be done asynchronously or by the software itself, right? Then that's why you talk.
- SGSarah Guo
Mm-hmm.
- IZIvan Zhao
Uh, yeah, it's kinda interesting. Maybe it's an interesting question, like, are we gonna communicate more or less with language model? I probably feel it's probably less. Uh, uh, the agent side essentially bet on language model does
- 22:43 – 24:36
Moving past the need for organization
- IZIvan Zhao
the communication.
- SGSarah Guo
One question I have for you, just going back to, like, the implication of RAG and, like, you can be my brain and do my organization for me.
- IZIvan Zhao
Mm-hmm.
- SGSarah Guo
Um, like, what if my brain is really disorganized? Like, do you- do you think that, uh, this changes the amount of work people should do in systems like Notion at input, right? Like, you know, should I be designing my knowledge base in the same structured way?
- IZIvan Zhao
Oh.
- SGSarah Guo
Or can I just dump it all in stream of consciousness in the future?
- IZIvan Zhao
Dump it.
- SGSarah Guo
Yeah.
- IZIvan Zhao
I think organization might be- we might be moving away from the organization wor- world. Why do you need to organize because you can retrieve? Uh, why do you have index? Like, index initially are file cabinets and the little index are sitting on top of so you can find things quickly, right? And they're indexed based on certain, uh, names or c- certain dimensions. But embedding and RAG sort of, it- it just, you have a semantically connection of all the things you throw into this file bag and you can find- bring it out however you like. So I think we might be moving past the need for organization. Uh, that's really liberating. That means on my phone, imagine this experience. I'll have a new idea, or where I see a whiteboard behind me, I just take a picture or write something, dump it, and Notion gonna organize for you, right? So that's- then that's become my perfect memory to start, later could be my perfect assistant to help me do something with this knowledge. And that's the vision we're moving towards. Right. That's super exciting to me.
- SGSarah Guo
So, you are, um ... This is a question from, uh, my friend, Al, who I've been a long time friends with.
- IZIvan Zhao
Oh, okay.
- SGSarah Guo
You know, you are a student of history. You mentioned Doug Engelbart earlier. I know, um, you think about, you know, the transition in terms of like, Alan Kay and what he did in terms of simplifying many of those concepts for like, a broader audience around computing.
- 24:36 – 30:14
History of SaaS doesn’t repeat, it rhymes
- SGSarah Guo
And my almost question was, what lessons in history do you take that, um, inform your point of view of like, how to treat AI strategy with Notion now? Like, from a prior revolution in computing, you know, how, how does that help you decide what to do?
- IZIvan Zhao
Mm. A lot is intuition. I think understanding history give you a sense of ... History doesn't repeat itself, but rhymes. So like, okay, which phase are we in? Um, I personally think we're sort of in this kind of bundling phase. Like, um, um, who said this? Like, there's only two way to do business, bundling, unbundling, right? And, uh, actually, there, the, during the break, I was reading a, a Chinese novel, uh, k- Romance of the Three Kingdoms.
- SGSarah Guo
Mm-hmm.
- IZIvan Zhao
And the opening sentence, uh, for that is, uh, "The empire long divided must unite, long united must divide. That's as, uh, always been." Business is the same way too, right? Um, we're in the bundling phase. I will say that SaaS, it's sort of this unbundling fragmentation phase. If we trace back to SaaS, why is SaaS happening? In the late two thou- mid-2000s, before that, everything's running on Microsoft. That was like a bundling phase. Early days of PC, there's so many different applications. The first version of the, the WordStar, WordPerfect, different text editors, dBASE, different database software. The f- the funny fact of dBASE is like, they start with dBASE II because there's so many, uh, company go bust, that it sounds like if they start with dBASE II, people has more credibility. It feels like this, this product has been around for a while. So that's the '80s. '90s was this kind of bundling phase because Microsoft has OS layer underlying it, and t- the SaaS is because the web becomes good enough to run software, right? And then so, then we have this unbundling phase, a fragmentation phase, and then on to with the, you know, last 10, 15 years, it easier. Really, the money's cheap, easy to create company. There's l- there's so much. P- th- th- too much now, feels like. There's like, information is so fragmented. And now, the new technology happening is AI language model, and if you build more with it or just think more with it, language model wants the information to be one place, wants the endpoints to be connected so it's easier to ... It's, uh, hard enough to look at like, current version of language model to, to do what you want, but imagine on top with different endpoint. That's even harder, right? So, and so we're in the bundling phase because the macro, but we're also in the bundling phase because language model, I believe, wants the things to be together.
- EGElad Gil
I think that makes sense. I also feel like we're in the bundling phase because the nature of how founders think about their business has shifted.
- IZIvan Zhao
How so?
- EGElad Gil
Um, I think that, uh, it's interesting because I remember, I don't know, 10 years ago, I used to argue with people about, "Oh, you should really buy other companies or integrate, or sort of pull all these things together," and in consumer, that actually happened, right? Like, Facebook bought Instagram and WhatsApp and other things, and they effectively created like, a bundle of social products that they could cross use in different ways for distribution and other things.
- IZIvan Zhao
Mm-hmm.
- EGElad Gil
But I feel like what happened is we had a series of highly technical founders 'cause we shifted in the Facebook era from Sheryl becoming CEO to COO, and you went from business-centric f- uh, CEOs in the '90s in some cases, although there was people like Bill Gates who learned and adopted as technologists, to very technology and product-driven founders who often thought, "No matter what product I build, that it always has to be better, and so I, I can't just think of distribution as my wedge, I need to think of every product as being superior, and so I'm not gonna build certain things." And now I feel like people are both building great products as well as bundling them, but also they're much more aggressive about saying, "It can be 80% as good, it can be 50% as good, but I'm gonna have a bundle," and that's HubSpot and that's Rippling. And they, they have very high quality to their products. It's just they realize they don't need every single edge case in every feature as long as they're able to cross sell.
- IZIvan Zhao
Yeah. I think the YC school's philosophy of build one thing, use internet to find the distribution-
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
... that was I think overlapped quite a bit with the rise of internet, right? And, and feels like there's a value to create on the other dimension, which is, like you mentioned, it doesn't have to be as good, 90% as good.
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
But because the synergy of things just make a lot easier, a lot cheaper, less tabs to be open in your browser.
- EGElad Gil
Yeah, it's all integrated. You have the information flow or def- or the system of record for whatever thing that you're dealing with.
- IZIvan Zhao
Yeah.
- SGSarah Guo
I think a lot of people also just, um, perhaps lack that, uh, sort of historical context, right? If you look at the strategy of companies like Oracle, right? It was very much for a decade and a half like, a dominant, at least commercially, attitude of like, "Okay, we're gonna buy the second-best product in this additional software category we wanna be in, and then sell the heck out of it." Worked great, right, actually because, um, it was very hard for customers to deploy these sort of things, or their disadvantages to everything being attached to a single database at some point. And I, I do think there is some analogy to, as you said, language models because having things in the same embedding space is very useful.
- IZIvan Zhao
Very useful, yeah. I think there's bundling of distribution and bundling of information.
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
The, what you're describing to me is more of Microsoft, with more like, bundling of distribution. Language model wants the bundling of data, bundling information.
- EGElad Gil
Mm-hmm.
- SGSarah Guo
So I, I remember hearing from Dylan at Figma early on that there was, um, one crazy user who was in the product like, 14 hours a day. It was you-
- IZIvan Zhao
Mm-hmm.
- SGSarah Guo
... um, early on in the, uh, Notion journey, um, being really design obsessed. I think the company has a reputation for that.
- 30:14 – 34:26
Design at Notion
- SGSarah Guo
Do you think of Notion as like, a design-centric company, and, um, is it important? How do you scale it?
- IZIvan Zhao
Mm-hmm.I think it depends on what you mean by design. Design is, to us, at least to me, it's less about how it looks as how the system plug together, right? And then in that case, the tradeoff you make, do you centralize that thing or do you decentralize that thing? Um, certain company work well, or certain business or product work well being decentralized. Uh, like operation-heavy company could work that way. And Notion, like I mentioned, we're sort of in the bundling business. Our value provided having this one information space, one workspace for people to do all different kind of things, so things need to be, work well together. It's almost building Notion, it feels like building an operating system, building a programming language, right?
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
You don't, you don't-
- EGElad Gil
You don't want that ............................
- IZIvan Zhao
... you don't farm out-
- EGElad Gil
Yeah.
- IZIvan Zhao
... to, like, 50 people to design a programming language. Usually programming language are done by one person. So that means the design here is very much centralized energy. It's kind of like Apple, how they build OS integrated with their hardware, and right? Like, what is the Apple for software? Doesn't quite exist today, right? It truly doesn't quite exist. So that's what I'm interesting, what we're interested in. Uh, so in that case means to build a good product, a good customer user experience, we need to think things more horizontally, more holistically. That means the decision-making tend to be centralized in our design team or in, like, tends to be centralized, right? So, um, less like ... more so, more Apple-like, less Amazon-like.
- EGElad Gil
It's funny 'cause when I first met you, it was just you starting Notion, and it was before you brought on Simon. And you talked about things that way even then, and I felt that one of the reasons I was lucky enough to invest, or, you know, I, I came on board was because you had such a cohesive view of how you wanted to build software, and you had such a cohesive design aesthetic. And it was your mocks, but it was also how you were dressed and how that reflected into the product. I felt like it was extremely striking, you know? (laughs) Like, you're one of a very small number of people I've ever seen where that design aesthetic has just kind of permeated everything in a very cohesive way. And so that's one of the things that got me excited at the time. I was like, "Wow, this is capturing a, uh, aesthetic that could be an incredible product platform." But you also talked about things even then, I remember, in terms of like, "Okay, what's the, what's the cohesive Apple-like thing that you can do for software and things?"
- IZIvan Zhao
Yeah. I think-
- EGElad Gil
It's kind of amazing to see that consistent thread. So I was just stricken while you were talking, you know, (laughs) by that, yeah.
- IZIvan Zhao
Thank you, yeah. Like, I studied cognitive system, cognitive science, which is kind of just like a degree for everything in some sense. It's like a little bit of philosophy, little bit linguistic, computer science. I learned how to code when I was a kid, and I did a lot of art and also in school. So, like, try not to... Like, there's so many things you can steal from all the different places, right?
- EGElad Gil
Yeah.
- IZIvan Zhao
And it's like the boundaries are sort of manmade. And then, then w- or in Notion with most our designer can code. Majority of our design, 80% of designer can code 'cause the moment you can... as a designer or as engineer, you can code or you can design, you can make really interesting trade-offs, right? For, at the end of the day, technology is, at least in my opinion, is about trade-offs. What kind of tradeoff you can make that unlock new user behaviors that's valuable? But if you can do more things, you can more, make more interesting tradeoff that other people cannot make. As a designer, if you can code, you know how to change your design to make easier to build. As engineer, if you can design, you can, like, do the same thing, almost like squeeze the air bubble to whichever direction is easier to squeeze it, right? Um, and therefore, I think being a more holistic help, at least Notion as a company energy, we're trying to be holistic. It also help our, keep our company team very small.
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
Like, we're usually one of the smallest relative to our business scale, because people can think, can do more, can be more holistic, and people enjoy that too because they can do more things, right? It doesn't feels like they have one role, they have to be doing that repeatedly. Has a lot of different benefits, but it's much harder to find such people.
- 34:26 – 36:52
Notion office design
- EGElad Gil
Mm-hmm. I have an important question here.
- IZIvan Zhao
Please.
- EGElad Gil
So if I think about the first office, and this may or may not be, uh, true still, um, Notion is a no shoes, was a no shoes place. Did this contribute to the company energy?
- IZIvan Zhao
I'm Asian, so when you go home, you take off your shoes. Our first office, no shoes. It was, uh, last us to 10-ish people. Second office, 20-ish people, no shoes. Third office, no shoes. Uh, it's actually has heated floors, so even better.
- EGElad Gil
Oh, very nice.
- IZIvan Zhao
Right? It's always all in the mission. Uh, the fourth office, we try to do no shoes. It was still in the mission. Um, I think I made a wrong choice in the rug. The rug kind of hard. When you st- it's, it's a hemp-based rug, so when you step on it without shoes, with socks only, it's just, oh, it hurts. So we decided not to do no shoes at fourth office, and, uh, so far has been stuck that way.
- EGElad Gil
The applied intuition has socks and, uh, slippers at the front. So that way if you need the padding... (laughs)
- IZIvan Zhao
I know, but where do you then...
- EGElad Gil
... or the slippers.
- IZIvan Zhao
Then the question is like, where do you store the slippers?
- EGElad Gil
Yeah, yeah, it's a lot of work.
- IZIvan Zhao
It becomes stinky. It's like 100...
- EGElad Gil
And it is the aesthetic.
- IZIvan Zhao
Yes. If you come to our office, it's like we're still try to be, um, not corporate. It's like that we're trying to use the furniture that people use for homes. I'm pretty picky about what kind of furniture is in the office. Like, ideally design classic that last 50 plus years, so inspire us to build software that way, right? So they make, they also make trade-offs. People who design a chair make trade-off. Th- they made really interesting trade-off for us to solve certain problems.
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
If you know the history of it. So we try to, in the office, use good software, uh, good chairs, good lighting, so.
- EGElad Gil
This is back to the aesthetic point that I made earlier. I actually have felt that in the offices as well. There's that ongoing cohesion. Even the music, I remember, I think it was in the second office, was pl- it was always jazz in the background. (laughs) And I just felt like it all kind of was this consistent vibe, you know? So it's pretty cool. Within Notion, are you using a singular, um, underlying...... LLM, or are you at this point using multiple different things for different use cases? You mentioned sort of the high level reasoning versus the fast, cheap sort of synthesis we try everything. OpenAir, Entropic, uh, are the high end model. Uh, we want reasoning, which is why we work with a high end model, right? Uh, yeah. Mm-hmm.
- IZIvan Zhao
It's kinda like-
- EGElad Gil
Yeah.
- IZIvan Zhao
... everybody building different flavors of this.
- 36:52 – 38:30
How RAG will change the future
- IZIvan Zhao
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
So.
- EGElad Gil
Yeah, makes sense. And then as you look at, um, it feels like with Notion, there's a set of core sort of templates or use cases.
- IZIvan Zhao
Mm-hmm.
- EGElad Gil
Um, you know, there's things around project management, there's other types of almost like applications that people have built to use. There's knowledge base related stuff, there's the things that you mentioned. Um, are any of those you feel differentially impacted in terms of how you think about future AI roadmap or things that, you know, will really change the game dramatically in terms of some of these areas?
- IZIvan Zhao
Yeah, I would say RAG changed all the knowledge side fundamentally.
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
Um, you no longer need to organize. So in Notion, one of thing people love is the left sidebar, right?
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
The left sidebar, you can organize your knowledge base, organize your personal workspace. Maybe the future doesn't have to have that.
- EGElad Gil
Mm.
- IZIvan Zhao
Like what it's like to, um, you know, not fall into your own innovator dilemma to double down that UX paradigm-
- EGElad Gil
Yeah.
- IZIvan Zhao
... but just having a notion that you can just dump things and retrieve, right? That's knowledge side.
- EGElad Gil
That's actually really interesting at a high level to think that, um, everything sort of moves through a form of search over time, right?
- IZIvan Zhao
Move over to search over time?
- EGElad Gil
Yeah. Like you're kinda losing organ- you don't need to self organize information anymore-
- IZIvan Zhao
No.
- EGElad Gil
... in this new world. You can just create a mechanism to interrogate it.
- IZIvan Zhao
Yeah, at least like you- you don't organize your brain, you just dump into it, then you wake up, "Oh, I remember that thing," right?
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
So like-
- EGElad Gil
Yeah, it's interesting, yeah.
- IZIvan Zhao
Oh, some people do, like the Art of Loki, The Art of Memory is actually org- visualize your brain, but for most people it just works without any organization.
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
Magically, right?
- EGElad Gil
Yeah.
- IZIvan Zhao
What it's like for software. We're getting there, right?
- EGElad Gil
Mm-hmm. Yeah, that's cool.
- 38:30 – 42:07
Building our the software in the Notionscape
- IZIvan Zhao
- SGSarah Guo
Are there, um, areas of, uh, software more broadly that you think are outside of Notion's scope that you think are gonna change a great deal from AI?
- IZIvan Zhao
Wow. Um, in some sense it's kinda a race. There is like the, we're in the bun- Notion's in the bundling business. We are, um, we are, uh, we're in the bundling and front office business. Front office, my definition, our definition is, what's happening in fr- like imagining a 1960s office, right? What's in 1960 office? On your desk you have a notepad, you write down something, you maybe have a typewriter. Um, then you have your binders on the left and right, just that's essentially... The notepad is your documents, your notes in Notion, your binders of things are like your Wiki knowledge base in Notion, and behind you will be the file cabinets, that's your relational database in Notion, right? And you have a little push cart to put things into there. Then there's the back office, where it's like the librarians organize all the things, uh, that's Snowflake, right?
- EGElad Gil
Mm-hmm.
- IZIvan Zhao
That's the back in the days IBM. We don't touch that. Uh, we largely touch... Our strength, like I mentioned, is software interface, UI, UX, which is largely what's in front of the human. We're trying to bundle in this one space. At the same time, there's also largely back office powered use cases. They tend to be verticalized specific to healthcare, specific to some kinda workflows.
- SGSarah Guo
Manufacturing. There's-
- IZIvan Zhao
And that's very specific but it's very essential to store somewhere and that vertical integrate that use cases. Uh, that could be AIFI too, and people in fact, we see this in law, we see this in a bunch, very specialized thing that people have that domain knowledge and trying to figure out how do you, uh, instead of human shuffling this, let language model help a lot of that, right? Um, the front office type of things, it's kinda open ended. The back office power things tend to be specific. So I think it will be a race, but the market is just so large and it doesn't, it's not zero sum necessarily.
- SGSarah Guo
Maybe you can talk about the market as you see it for Notion. So, you know, when you guys began, I think early adopter startups were the first to, um, get onto Notion for knowledge base.
- IZIvan Zhao
Mm-hmm.
- SGSarah Guo
Um, you're a much bigger company now. We're also in a different macro where, um, you know, startup budgets are, uh, less robust. Like how do you think about the enterprise and helping the enterprise adopt AI or, or you know, do knowledge management?
- IZIvan Zhao
Yeah. We're still very, uh, early, uh, not at scale yet. Um, I would say, uh, bundling and transoming do a lot of good things. One is you don't have to jump between different tabs to do things. Second is save your cost, right? Like w- we save a lot of customers bill for their project management tool, uh, a- around their issue tracking tool and, and that's, enterprise really care about this. It's very CFO friendly today, um, with this macro. So yeah, bundling has many good benefits besides money, besides information there as opposed to money.
- EGElad Gil
Well, Ivan, I mean, this conversation has been so many interesting topics. Thanks so much for joining us today.
- IZIvan Zhao
Thank you. Yeah.
- SGSarah Guo
Great to see you.
- IZIvan Zhao
Good to see you.
- SGSarah Guo
Find us on Twitter @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: 42:07
Install uListen for AI-powered chat & search across the full episode — Get Full Transcript
Transcript of episode HW5Yeqn9eV4
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