He Was Promoted 3x in 5 Years at Google (Bilawal Sidhu)

so you have actually given a TED Talk on AR and VR. I really consider you one of the thought leaders in spatial computing. Hasn't it mostly been hype so far? Yes, we're a couple of years out, but this is the closest we've ever been, particularly thanks to generative AI. People have been talking about this notion of walking around with your glasses and being able to ask any question about anything you see in the world and understanding the backstory of a space or place that you care about.

All the experts I've talked to, like yourself, have a lot of praise for the UX and the interaction that the Vision Pro created, yet it's basically a flop. It's like the Tesla Roadster, right? How many of those actually got sold? Would you consider it to be a commercial success? Was it more of like a status symbol thing? I wouldn't call it a flop as much as a really high-end developer kit that was positioned as a sort of high-end device that the media consumption nerd would want to buy.

There are plenty of people who get no promotions in five years, like Google. How did you manage to get three promotions in five years? I always did the risky thing and I followed the craziest ideas. When people said, "Oh no, you should work on this area; this is far more strategically valuable," I was like, "No, I'm going to go work on the thing that I have a lifelong deep interest in."

Lavel Sadu, I've been following you on X for a few years now. We've been interacting on X. I can't wait to dive into this conversation. Thanks for being here. Thank you for having me. Likewise, big fan, longtime listener, first-time dialer.

So you have actually given a TED Talk on AR and VR. You have a YouTube channel that's blown up in this space. You have a TikTok with over 1 million subscribers that people should check out. I really consider you one of the thought leaders in spatial computing. What I have to ask is, hasn't it mostly been hype so far?

Ah yes, the perennial question: how far is AR/VR out? It always feels like it's two to three years out and kind of has been for the last decade. It's particularly funny that I say that, like, you know, I've got my Snapchat AR glasses right over here, which is supposed to be a glimpse into the future, but is this kind of a dev kit right now?

Really, I think the best way to think about this space is like that classic Peter Thiel quote: the world of atoms is so much harder than the world of bits. I mean, we've seen such staggering advancements in the software world. But yeah, I mean, I've spent the entirety of my professional career effectively working in AR/VR since 2013, and every couple of years it felt like, "Oh gosh, mobile VR is the unlock that we need. Just once Gear VR gets out at scale, people will be watching passive VR media, and that's going to change everything."

No, no, no. Actually, it's all about mobile AR. It's the phone in your pocket. You got to meet people where they are. And then all the advanced headsets along the way. I will say on the headset side, Meta has basically been like, you know, carrying the entire industry while everyone else sort of slowed down their investments.

Of course, we're seeing that change. Apple, of course, with the Apple Vision Pro. We're going to very soon see something from Google. They've been teasing out their XR platform with the work that they're doing with Samsung. So I'm sure there's exciting stuff ahead. But even Meta, the clear leader, I think they have like 70-something percent of the market. It's like on the order of millions of headsets. That is just not a significant enough install base to make a difference, right?

To create a thriving developer community, to attract even some of the bigger brands and the bigger IP holders to create content. And so that's why you're seeing the biggest players subsidizing this type of content, like you're seeing with Apple TV and the Star stuff.

I hate to say we're a couple of years out, but this is the closest we've ever been, particularly thanks to generative AI and sort of the multimodal capabilities that were always missing in the earlier iterations of the vision for the metaverse and spatial computing.

Say more about that because it almost sounds like a tongue twister of trends when you say generative AI is helping AR and VR progress along. What is the role? What are the areas kind of within the value chain where it's helping?

This episode is brought to you by Adio, the next-generation CRM. Imagine a world where your CRM is powerful, easily configured, and actually intuitive. Adio makes that a reality. As I've profiled in the Product Growth newsletter, setting up Adio takes less than a minute, and within seconds of syncing your email and calendar, you'll see your relationships in one place, enriched with valuable data, including things like relationship strength.

Adio lets you build happier style automations, create powerful reports, and works for any go-to-market motion, from PLG to sales. Adio is designed for the next era of companies like yours, and companies like yours shouldn't have to deal with inflexible, one-size-fits-all CRMs. Join industry leaders like 11 Labs, Replicate, and Modo. Head to adio.com/productgrowth, and you'll get 15% off your first year. That's adio.com/productgrowth.

This episode is brought to you by WorkOS. If you're building a SaaS app, at some point your customers will start asking for enterprise features like SAML authentication and SCM provisioning. That's where WorkOS comes in, making it fast and painless to add enterprise features to your app. Their APIs are easy to understand so that you can ship quickly and get back to building other features.

Today, hundreds of companies are already powered by WorkOS, including ones you probably know, like Vercel, Webflow, and Loom. WorkOS also recently acquired Warrant, the fine-grained authorization service. Warrant's product is based on a groundbreaking authorization system called Zanzibar, which was originally designed for Google to power Google Docs and YouTube. This enables fast authorization checks at enormous scale while maintaining a flexible model that can be adapted to even the most complex use cases.

If you're currently looking to build RBAC access control or other enterprise features like single sign-on, SCM, or user management, you should consider WorkOS. It's a drop-in replacement for Auth0 and supports up to 1 million monthly active users for free. Check it out at workos.com/AAS to learn more. That's workos.com/AAS.

Absolutely. So, I mean, AI has been used in augmented and virtual reality for a while, right? Particularly computer vision. That's how these systems figure out where you are in 3D space, like precisely locating you, understanding what's in your environment, and letting you interact with computing using hands, for example, or other forms of gestural inputs, whether it's controllers, using hands, eye gaze, etc.

So we've had that mix of sort of classic AI and some ML-based approaches. But I think the big paradigm shift that was missing was just natural voice and being able to talk to the computer and give it context on what you're seeing and respond contextually. As it's become very trendy to say, I think that's the part where, in my head, there's this vision of sort of like a Jarvis-like AI system.

Step one of it is, yes, I need to be able to do touch gestures. And oh my God, Apple nailed it to the point where, gosh, sometimes when I take the headset off, I forgot I've taken it off, and I'll find myself pinching stuff on the walls. But I still can't talk to it and give it, like, just ask it questions about what I'm seeing.

You see instantiations of this on mobile, something like Google Lens. Apple has started rolling this out in their Photos application. But once you can take that context of your video stream, your audio, throw that into these massive context windows that these models have, suddenly you start seeing something that feels like a helpful assistant that has full context not only into your prior digital life but also your physical surroundings.

That, to me, is just super exciting, and I think we're on the cusp of being able to do these things that would have been way, way, way harder in the future. Just to close it out with one concrete example in AR, people in science fiction have been talking about this notion of walking around with your glasses and being able to ask any question about anything you see in the world and understanding the backstory of a space or place that you care about.

That was a traditional content authoring problem, like two to three years ago, pre-large language models. Now you can start creating this type of content procedurally on demand without having to create a bunch of this offline curated authored content. So I think Google's Project Jarvis is a great example or a glimpse of what is to come.

Of course, Meta is doing very similar things on the other end of the spectrum, taking their Ray-Ban glasses and putting the smallest, teeniest version of Llama on there so you can ask some basic questions. But you'll see those trends converge for sure. And that's why I think generative AI is going to be huge.

Not just on the perception side, though, also on the creation side for these types of real-world experiences, but also 3D. If you think about the metaverse, like we go back to the Matthew Ball instantiation of like super extensive definition of what is the metaverse, and you break it down to the spatial 3D spiritual successor of the 2D internet that we know and love.

Well, hey, the 3D world is massive, and it's empty, and it's one hell of a canvas. But it's not much use if you don't have the means to populate it. Creating 3D content has been so hard. So the level of democratization we've seen on the video side, some of the 3D scanning stuff, and the convergence that's going to happen therein, we finally have the means to populate the so-called metaverse as well.

And so that has me very bullish about the future of spatial computing, especially as it gets a boost with all of these generative AI technologies. All the experts I've talked to, like yourself, have a lot of praise for the UX and the interaction that the Vision Pro created, yet it's basically a flop. Why?

I think it's like the Tesla Roadster, right? How many of those actually got sold? Would you consider it to be a commercial success? Was it more of like a status symbol thing? You know, the first thing I'll say is I agree there is not that much to do on that device yet, especially because they're seeding a new ecosystem from scratch compared to Meta, where you've got the Quest ecosystem, but you can also tap into Steam VR, which has been thriving for the last almost decade now.

So I wouldn't call it a flop as much as a really high-end developer kit that was positioned as a sort of high-end device that the media consumption nerd would want to buy. Oh, it's like your connected living room. Oh, you're thinking of buying a 65 or 75-inch OLED TV? Why don't you get this thing instead? You could watch 3D content, and gosh, it's amazing for things like that where they are subsidizing and doing strategic deals.

For example, I had a chance to watch the first Dune in 3D before I watched Dune 2 in cinema, and it was fantastic. Amazing media consumption device, amazing second screen experience, but all the apps are kind of not there yet.

I think this reminds me a lot of the first iPad and the first iPhone, where when the first iPad came out, people were like, "Yo, this is just like a larger iPhone. There's not much to do. It's like, you know, kind of blown-up iOS apps," until people really started taking advantage of the form factor.

I think we'll get there, but where I think Apple did mess up, because I don't love everything about it, is this thing is freaking heavy. Even for somebody like me, I have pretty much all the VR headsets, other than perhaps some of the Chinese ones. I need to do a third-party mod so I can actually wear it in a fashion that it's comfortable.

A lot of people do that, which is weird because they clearly went with sort of form over function. I don't know, maybe that's why all the rumors about the next instantiation are like a cheaper, lighter-weight, you know, maybe more plasticky device that doesn't feel so heavy and lopsided on your head.

But this is classic Apple. They're dropping sort of like the crème de la crème, you know, Tesla Roadster, and they're going to work towards the Model 3. I have no doubt that they'll get there, and these things that look like bulky VR headsets will keep getting thinner and thinner until they look like a pair of ski goggles.

You compared it to the iPad. I feel like the iPad is something that actually sold. You mentioned Meta. I feel like they've actually sold some headsets in this space. I believe that both the Quest and the Ray-Ban glasses have outsold the Vision Pro. So tell me more about that. What is Meta getting right that Apple is getting wrong?

It's the price, to be honest. Let's be real. Meta is basically subsidizing this thing and giving it away, and they're seeding that ecosystem. It's like a brilliant strategy. I feel like Quest 2, particularly, and now Quest 3, reached this threshold where, you know, the biggest barrier to playing with VR systems was, "Oh, you needed a PC to really take advantage of it."

You didn't really have standalone. I think Quest 2 reached that point where there were enough interesting applications on the market on the social side, on the first-person shooter side, on the media consumption side. Even apps like YouTube VR that are glaringly absent from the Apple Vision Pro were available.

So it became this really cheap entry point into this beautiful ecosystem. You have this experience where kids that are in VR chat or Rec Room and are playing on iOS devices see their friends jump in with VR. They're like, "I want to get that thing." So that network effect has also been fantastic for them.

Yeah, they've got millions of devices, maybe even more than 10 million total, and that's the largest install base, right? So they're crushing it. So that's on the VR side. The Quest 3 is like a phenomenal device. I actually gave away my Quest Pro recently because I was like, "Why did I even get the Quest Pro?"

The Vision Pro is the best high-end workstation device, and I think in terms of the closest console-like proxy, like an Xbox that you just buy and there's cool stuff to do, no doubt it's the Quest. It's only going to get better because the Quest 3 is like this cheaper variant that has the best of the Quest 3, which is itself phenomenal.

So I expect that to only grow. They seeded the ecosystem. They did a good job of building the stuff that they want to start creating their own install base, having missed the mobile era, as people like to say. But at the same time, they reversed a couple of controversial decisions earlier on where, you know, initially you could not use Steam VR and some of these desktop VR applications.

They wanted to lock it down, so you had to use this Oculus app on your PC through their store. They've reversed that, and I think it's been for the better. You're seeing the same thing on the other end of the spectrum play out, of course, with open-source AI.

But also their glasses, though I would argue, I mean, Zuck on record saying this, the glasses were a surprise for them, right? Bose had had in the markets these sort of like really lightweight Ray-Ban looking glasses with pretty decent audio and microphone inputs, no display.

I think with a combination of the Ray-Ban aesthetic brand, whatever you want to call it, I love them both, and it has effectively replaced AirPods for me. I don't use AirPods when I'm taking calls, when I'm walking around doing convos with ChatGPT or Claude to get my ideas down. That's the device that I use.

But I think even Meta was surprised how well that's done. Just this past week, I was reading some headlines, total rumor mill, about how they may be taking away some resources on the VR side to push more on the AR side. That makes sense. I think they've got a great barbell strategy.

You've got the desktop tether, or sorry, desktop on your face kind of equivalent, you know, standalone VR headset, and then you've got the lightweight, ambient, always-on, take-it-everywhere-with-you experience as well. They've shown something in the middle of that spectrum with Ayan, the $10,000 pair of glasses you can't buy, but is a great tech demo as to where this might be going.

But I think for the time being, you're going to have these two separate device categories, and gosh, Meta's got a strong contender in both. I'm actually curious to see what Apple's going to do for the lightweight form factor. I was expecting that's the route they'd go down first, but no, they built a rather bulky headset, even though they tried to make it seem cool.

I don't know if it fully had the impact that they might have thought it would have. Let's move on from Apple and Meta to Google, your former employer. One of your hottest takes when I was researching for this episode was that you're still bullish on Google.

I think so because, at least for me and the top 5% of new tech users that I daily interact with, search is not something we use at all. ChatGPT, Claude, Perplexity have fully replaced Google. So why do you think Google can compete?

It depends on which angle of the Google business you look at, right? I think certainly there is more competition than ever before on the search front, though I think people underestimate just how much mind share Google has. It's that thing where, like on Twitter, when you go look at those, like, you know, maybe this is like some self-recursive loop that actually doesn't make sense, which is like you go on search trends and you just look at ChatGPT versus Claude.

And it's like ChatGPT is just by and far out ahead of Claude. Who's actually using Claude? I'm using it every day, and like perhaps in our early adopter community on Twitter or LinkedIn or whatever people are, but the vast majority of people aren't.

In the studies I've seen, Gemini actually has great adoption, but that's not why I'm bullish. I'm bullish on Google because they have the most complete models of the physical and the digital world. I think that's going to be a huge asset in this future, whether you want to call it spatial computing, whether you want to call it AR/VR, whether you want to call it ambient computing, whatever the term is.

Whether you want to call it AI assistance to make that assistant computing experience helpful such that it can actually help mediate all your interactions in the digital world and the physical world. I mean, Google's got all the pieces. I'm talking about Google Docs, Calendar, like the G Suite set of products, my email, everything perfectly synchronized.

All my stuff is on Docs already. Photos are where my images are, you know, on and on it goes. They've got YouTube, which is like the second biggest search engine and clearly perhaps the most complete repository of visual content on the freaking internet, right?

Like all sorts of verticals that you might not even think about exist there. So when I think about that data moat, it's insane. Same thing with Maps. This is like my old world, right? Like satellite, aerial, ground-level imagery that's fused together, like far out blows anything Apple's got, blows Microsoft out of the water.

And so I think we're at this phase where everyone's building LLMs, scraping effectively the same data set, call it the public internet. I think as things slow down that we're hearing about, at least on the pre-training side, the same level of returns aren't being garnered with these larger training runs.

Oh, we need to scale on inference time, compute, and other things. Well, even in those factors, Google has an advantage because they got TPUs. Like Apple is training a bunch of their models on their tensor processing units, like on all the T stuff, right?

And so I think the company that has the right consumer products, the right data moats, and the right sort of ecosystem and distribution to get this right to create that truly helpful Jarvis-like assistant, they check all the boxes. Apple does too, but then again, I don't know if they've got the same level of software chops as we're seeing, right?

Maybe the ChatGPT partnership will just take them far enough. Gemini supposedly going to be in mode in there too. But all these things considered, I feel like people forget that what you see about Google is like the tip of the iceberg.

The deeper you go, it goes down to like, you know, undersea cables and last-mile infrastructure to get you that YouTube video at super low latency fashion, which lets Google do things like cloud streaming and Stadia. Now, a bunch of these products do well; some of them fail abysmally. Notebook LM being a great example of like, you know, an amazing moment for Google.

But gosh, that is one hell of an ecosystem to play with. And if you're a startup, gosh, you better pull like, you know, an open AI if you really want to compete. So all those things brought together make me exceedingly bullish on Google still.

I can agree with most of those things, especially like YouTube and Waymo, where I almost feel like, you know, the future earnings of Google are at this point like three years before Instagram took over Facebook, where we could all tell Instagram was going to become the more important part of Meta.

For me, it's like I can already tell like YouTube and Waymo are going to become more important than search because what Google has consistently shown in the air over the last two years is that it has been almost a year behind Claude, which is, you know, three to four months ahead of ChatGPT in terms of user experience.

Like Gemini is a usable product, but I don't know anybody who's extensively used the four major models: Meta, ChatGPT, and Claude, who's using Gemini. The only people I know who are using Gemini are people who haven't really tried the others extensively.

I don't know. I mean, I agree with you on the Waymo point and I also, same thing for AlphaFold as an innovation, right? Like that's getting, there's a subsidiary company that's set up around, you know, basically drug discovery now.

I mean, I don't think we priced in the impact of all of these offshoots that can come out of, you know, sort of the Alphabet portfolio. But personally, Gemini is like maybe not the best. Like it's not my daily AI assistant on my phone. That's definitely Claude, ChatGPT as well.

Sometimes, though, you know, it kind of varies. If I'm using like advanced voice mode, I'll go to ChatGPT. But me and a lot of creators, especially for our podcast on the Teddy AI show, we're using Gemini extensively in AI Studio because of the 2 million token context window, the fact that it is multimodal, and you can drop in like the MP3s of all the recordings.

You can drop in like a bunch of producer notes, my own notes, and kind of get this synthesis across modalities. It's kind of wild. And for the YouTube stuff too, like short-form video, hey, what's like the best hook to pull? I want you to like watch this video, and it gives you really, really good feedback and answers.

It's funny, AI Studio is technically this sort of dev tool that's for you to try stuff before you use it, and I find myself using it a lot because I can upload multiple audio files, video files, a bunch of images, and that's the way I tend to use it, is for summarization and querying over very large data sets that I can't do with something like Claude.

I have to keep starting a new goddamn conversation periodically. Yeah, yeah, Claude is unusable from the client right now. They just cap it way too early.

Okay, so for specific use cases, it seems like Google is ahead. I do buy your point that the ecosystem is pretty interesting. We know that TPU performance of TPU data centers is a lot behind Nvidia data centers, but they could eventually catch up to an Nvidia data center.

They obviously do have that last-mile infrastructure, so I see your point. I see the market cap there. But I have to agree with the overall trends you've seen in market caps where Nvidia and Microsoft are kind of whipping Google right now. They're crushing it.

I mean, like who's really making money in AI? It's like Nvidia, right? Everyone's got to pay the Jensen tax. Yeah, you may use TPUs for like inferencing, like your inference workloads, but you're probably going to do your pre-training on like, you know, an Nvidia cluster, right?

And so I think, I mean, they're absolutely crushing it. Microsoft, on the other hand, it's interesting. I've got some friends that are doing cool things there, and I chat with them periodically. But personally, I don't use Copilot, and I don't know anyone who actually does too.

It's like, and there's of course the whole funny thing we could talk about here where like the Salesforce CEO has just been like hammering Microsoft Copilot in his tweets. I don't know if you've seen this; it's hilarious.

Of course, he's probably got some competing product offering to plug there. But yeah, it's like Microsoft is seemingly making all the right moves, but in terms of actual product stuff, like remember like two years ago, when we probably first connected, everyone was like, "Google's dead. Bing's going to be the thing."

Who is going to Bing? Like, no one. Meanwhile, the relationship with OpenAI has become like kind of, you know, there's like a bit of more of a friend-enemy dynamic than there perhaps was in the past.

And I mean, back to Nvidia, the fact is any company that wants to be successful, back to just like the scale that you need to win right now, at least at the platform and infrastructure layer, is like you need to partner with a hyperscaler.

So just this past week, I saw another AWS partnership with Anthropic, and I was like, "Oh, finally, maybe we'll get more quota so we can like actually talk to Claude every day." Meanwhile, Microsoft, despite all their epic wins in this new division, I don't know.

I find Copilot to be like a shittier version of advanced voice mode. I want to like it, but maybe it's like, I don't know, maybe it's super useful for people that are on Teams and Outlook all day.

It's like the equivalent of if you're in the Google ecosystem and you're using Gemini. Maybe it's good enough. Maybe 80% as good as the frontier model is fine if it just meets you where you are.

And you certainly are seeing this with things like Photoshop and Adobe, right? Where you've got generative fill in Photoshop. It is nowhere near the state-of-the-art in painting that any other model, even like the old stable diffusion models, offer. But it's in Photoshop, so you're going to use it, right?

So yeah, those are my thoughts. I expect more from Microsoft despite their strategy because on the product experience side, there's nothing that's called out to me. Now GitHub Copilot, I need to go use it since the update.

But yeah, I don't know. Maybe it's gotten better than Cursor now. But yeah, I think where Microsoft mainly won is by buying a lot of OpenAI.

Let's move on to product management, which most of the listeners of this podcast are. Where I want to start is you actually landed a PM role at Google, which is very, very rare for people to just move into product management as kind of like your main title at Google.

Usually, when someone asks me how do I break into Google, I tell them, "Yeah, you know, it's a five-year journey. You need to work at the companies that Google hires from." And DEO isn't one of those, but you moved from DEO, so I need to hear more about this story. How did you land the PM role at Google?

The throwback story, let's do it. Yeah, I mean, I got very lucky with DEO too. I ended up joining their Emerging Technologies Group, and this is right when all these consulting companies were buying up digital agencies.

This is like 2013. I graduated college, studied computer science and business. I was in a business fraternity, and so like what everyone was excited about was the groupthink of like you go into finance or you go into consulting.

At that time, I needed H1B sponsorship. I'm proud to be a citizen now, but back then, I had to be very safe, so I wasn't going and doing film and working in 3D animation. That just wasn't an option, even though I minored in that.

Given I had duct-taped together a weird disciplinary skill set, a creative technologist, if you want to call it that, I was a Google Glass Explorer. I had already started prototyping with that stuff. I started playing with the earliest instantiations of 360 cameras.

I started doing a bunch of visual FX workflows to figure out what AR experiences look like, and that just let me do the funnest things at DEO, working on IoT projects, drone projects, like, I don't know, a field service expert on an oil rig going through a maintenance checklist.

All the stuff, by the way, that's still the things that have high ROI in AR/VR. Then when the DK2 happened, I really got into consumer VR and game engines and 360 video. We produced a video for DEO at the time, like a training simulation video essentially using the earliest 3D 360 cameras.

Doing that and doing soul-searching, it came up to about two years. I got promoted quickly, had reasons to stay, but I was like, "Holy crap, this is the next revolution." Of course, it ended up being a longer journey than that because everything didn't blow up and pop off after two years.

But I had conviction this is the right direction, and suddenly all the other stuff that I had learned reconciled all these different facets of me, especially my passion for visual effects in 3D. I think of AR as basically like the visual effects pipeline running in real-time on the phone in your pocket or the headset on your face.

That gave me an opportunity to join Google in 2017 originally as a technical producer. So that latter, they hired us on, they had no idea they were building out this new function called Daydream, now called Labs.

How do we even hire for these types of people who know technical stuff but can do production stuff too? They were like, "Oh, TSC ladder," so they tossed me in there. I just started doing a bunch of cool stuff that was inherently within the cracks of all the other functions.

You obviously had product and engineering teams, you had some design teams, you had BD teams, and I ended up being sort of like the bridge and glue between them all, like really close to the users in terms of taking these camera systems and using them, I don't know, in production, like Teen Choice Awards, Coachella, Elton John, working with the NFLs of the world, but also being savvy enough about the underlying tech to shape the roadmap for what the engineering team should build next.

I remember, you know, one of my buddies now, like shout out Ryan Cassidy, hit me up basically being like, "Yo dog, have you ever thought about being a PM? Because what you're doing is like the product manager role."

Up until that point, I hadn't even considered it. It always struck me sort of as like a boring sitting in meetings, writing docs all the time job. Surprise, surprise, there's definitely a lot of that.

It was at this nexus position where, at least in Google, and this varies based on the company that you're in, engineering PMs kind of run the show and have a lot of autonomy and influence. So yeah, I basically did a rotation. I had a chance to take one of the things I was working on anyway.

It's like, "Can I just PM this part?" The hiring managers and all those folks were super cool with it, or I should say the manager of that was super cool with it. I did that process, and then I went through the conversion interview.

So I got the same feedback that you are giving people, which was sort of like, "Hey, this is a multi-year process. You got to do it." I was just kind of like in 2018 antsy, and I just wanted to make the shift happen quick.

I'm going to do a rotation, a six-month rotation. I'm going to ship something that demonstrates reps, and then at least at that time, you still had to go through the full loop of PM interviews to transition over.

Gosh, it was definitely several months of like, as you well know, the Java product management is very different than how you interview for product management. So I watched a lot of the videos, read all the various books that you can imagine, and I just kind of bulldozed through it.

I'm glad I did. In fact, I wish I had done it sooner. So when I give people advice, and I've helped a bunch of people convert ladders at Google since then and mentored them, it's like don't wait for that promo. Don't worry about this thing. Just move on the ladder that you care about and do it as quickly as possible because after then, you know, the world is your oyster.

So that was my take on kind of ending up in this weird sort of, you know, like one-size-fits-all ladder. Later, it was shared with me that some of the highest transfers to PM at Google are from this TSC ladder, which I found very interesting.

It makes sense. You get a bunch of people with some domain expertise and a smattering of things, and you make it happen. So after that, I was hooked on the product thing, and that's what I did for the rest of my career there.

Yeah, I love this route. I now have a new sort of case study to talk to people about, which is essentially join Google in an area where it sounds like you're a domain expert. So you don't necessarily need to join in the PM ladder, which tends to be a bit haughty-taughty about having needed to have been a PM before.

You can join at a role that's kind of a virtual PM, which a lot of technical producer roles are. A lot of technical program management roles are. I mean, PM and project managers, it's not even clear that they're that different at big companies these days.

I feel like there's a, and even your producer point, like in the gaming and entertainment world, that producer role is so interesting. And even I just interviewed the director of DARPA. The program manager role is just very similar.

In aerospace and deep tech or whatever the trendy thing to call it now, there's this sort of bridge. I don't know how you think about the PM role because there are so many PM styles, like the colors of the rainbow effectively.

But I think getting close to what you're doing, yeah, I would like an area where you have domain expertise. So key because that's also where then the engineers want to work with you. Because they're like, "Oh, this person actually knows what they're talking about."

And like, actually, what are your thoughts on this? And so that becomes a very good, you're just able to make better product decisions if you understand the space better rather than approaching everything from kind of a generic MBA framework-oriented approach, which I think, you know, it's like a lot of PMing has turned into, unfortunately, these days.

Yeah, that's exactly right. Basically, there was a trend when I first joined product management in 2008 where everyone was an industry expert. By 2018, there was, it's like, you know, it's like fashion too. There's just these cyclical trends. You go from skinny to loose.

It's like industry expertise to no expertise. Currently, the trend is towards industry expertise again for PMs. It's almost like a pendulum. You can just swing too far in one direction.

So what I land like firmly in the middle, like you want some industry experience. At least more industry experience is better than none. But you do see terrific PMs who switch industries as well.

So don't let industry expertise silo your career and not allow you to achieve whatever you want to do. And I'd also say, like, you can always go back. I think a lot of people make it seem like an irreversible one-way door.

It's funny, like a lot of the advice I got was, you know, there was a time earlier around 2019 where I almost left the company to join a very large social media company, I'll say. And, you know, the feedback I got basically was like, "No matter what you do, having this ladder label, whatever, is going to be so prestigious that you might as well stick around and get your promo or something like that."

I did end up doing that for other reasons, but I think there's so many people that can go back to being an engineer or like I did the PM thing. I didn't like it. Now you can go be a designer.

Most of the stuff I did right out of college, despite having a CS and business degree, was like product design at DEO, and that made me useful there because nobody else freaking knew how to do that. Everyone was like a wizard in Microsoft products, but they didn't have this other skill set.

But that still helped me as a product manager too, like, you know, later on in my career to have those design chops and do it. I think when a lot of people think about like, "Oh, this role," it's like this title, and like people overthink it.

They're so much easier to jump around now. If your goal is, "Hey, I'm trying to be on the fastest path to, I don't know, like product VP or something like that," maybe you can't take as many detours.

But I like to say surrender to serendipity, smell the roses, chase your passions because when you're chasing your passions and not chasing a job ladder title or something like that, then that's all a means to an end.

And then the end is totally worth it, even when you have to deal with all the, I would say, the less glamorous parts about product management, which people don't talk about as much.

What are those less glamorous parts for people who don't really understand what it's like to be a PM at Google? I would say, like, I describe working at Google to being like you're doing a tour on like the Star Destroyer, you know, like the tanker ship. Choose your freaking analogy.

It's massive. There's a huge chain of command. There are so many divisions. Everything is like hierarchical to the max. Yet the waves that you leave in your wake, if you can get the tanker ship to turn in the direction you believe in, you want, you're convinced is right for the business, there is nothing more satisfactory.

But to do that, it may depend on like, depends on what kind of team you're in. You know, I spent four years on the Google Maps team. Getting the Maps team to rethink how they view 3D in Maps was like sort of like an exercise in applying the innovator's dilemma and the innovator's solution, kind of dealing with organizational change, finding out all the very respected people and getting them to change their minds.

That's not easy. It takes time. The visual analogy, it always struck me as sort of like you're a tugboat. As a PM, you're sort of a tugboat, and you're moving, steering this massive ship in the direction that you want.

But you need to align with other tugboats and like all orchestrate and coordinate together to go in a certain direction. So I think, like Alex Kosi famously called it his deck, like hilariously leaked the slime mold deck on Google culture, I think nails it, which is sort of like the coordination headwind is just really high.

So you're going to have to have a deep Rolodex. You're going to have to manage upward, downward, across, influence without authority. But again, like if you have, if you're passionate, if you have a vision, you know what you're talking about, and you can make reasonable, like, you know, data-oriented or just like product sense case for what needs to be done, the way it needs to be done, people are very accommodative.

Because I find most people in these large companies are either like, you know, like there's a bunch of folks. There's this disproportionate number of folks that want to change stuff and get things done. So you got to find them and partner with them.

But then there's a lot of other folks that are sort of like what Elon might call the manager class now with the Doge thing, where it's like, "Oh, I got into the country club. My sole job now is to not rock the boat."

Oh, crazy PM with crazy idea that will rock the boat? No, thank you. I think you got to learn to take, like, you can't just know is not an answer, and you got to try again and again and work.

You know, dare I say, apply the 48 Laws of Power to go figure out how to get stuff done. But those are the downsides. It's a lot about moving hearts, minds, organizations, like hearts, minds, people's visions, processes to get stuff done.

A significantly smaller amount that's focused on the actual insight, doing the actual product work and getting stuff done. I think that's why larger companies are, you have to measure your wins in the order of certainly quarters, maybe even years, to have substantial impact.

Whereas as a startup, the opposite of the tanker ship is the speedboat. Oh, like we're going over here today. Oh, boom, let's go do it. No owner approval processes, no having to ask five teams for permission where they have some dependency on your stack that was made 15 years ago.

All this type of stuff you don't have to worry about, right? But again, the magnitude of the waves you can make are much smaller than the tanker ship. So I don't think it's glamorous.

You know, my style has always sort of been like the freaking servant leader kind of style a bit. And I don't know, it's like as long as you don't make it about you, but you're able to go bad for people and bad for what you believe in, you'd be surprised how much you can get that done in big tech.

Because I think like one other thing I hear a lot is, "Oh, it's impossible to go do anything zero to one or new in these tech companies." It's just not true. There's a lot of people doing it. They're just not going to follow the playbook that everyone gives you.

Perhaps by design, because the path well-trodden, well, that's how you end up adding another button to your Google Cloud AI product. I was in Google Cloud. I think we had some overlap. I was there from 17 to 19.

I'm curious, what was the work environment like, at least in Cloud? I think people would be surprised, but you know, the engineers, it was rare to find them in the office at least after 4:30. I think marketing evacuated by 5. PMs generally there till 6:30, often pulling virtual all-nighters before product reviews, up until 4:00 a.m., going home for like a two-hour nap, and then coming back to get grilled by executives about something they hadn't even prepared about.

I'm curious what it was like in the Maps team, and I believe you also spent some time in the spatial computing team, so I'm curious to hear about both of those. Sure, yeah, I could start in chronological order. The AR/VR team, that was Clay's Daydream team, which then evolved into the Labs team, as we were discussing.

That was very much like cherry-picked people from all the various product areas of Google that wanted to go build this next computing platform. It was very much like, I would say, very PM heavy. There were a lot of PMs, like, and yeah, like almost a PM-to-engineer ratio.

I don't even know if I should share this, so yeah, so like I probably won't share it, but it was not great. There were just like way too many PMs, and there were like kind of the product surfaces were interesting, but there were only a few of them that were like successful external-facing products.

So Google Lens, as an example, or YouTube VR, as an example, or ARCore as the AR platform on Android. These were like interesting things, and there were a bunch of other really interesting R&D products and projects happening, and I kind of loved that because this is what I love about the tanker ship model too.

It's like basically it's like all these big companies, like whether it's Google or Meta, is like, "Yo, we just got to catch all the Pokémon." So for any given field, let's just find all the foremost experts in all of this stuff, right?

At the time, Paul Debevec, who did a bunch of like the light stage stuff, was under Shriram's AI team. Shriram now is the VP of AR/VR/XR at Google. Really smart guy who did the Microsoft Holoportation demo.

If you go type in Microsoft Holoportation, this amazing video of him doing a volumetric Skype with his daughter in the early 2010s, it was just like mind-blowing stuff. And so these cats were around back then running the research agenda. Paul was on these folks' team, and I loved it because there was so much cross-pollination all the time in all these domains.

There was a lot of sharing. There was a lot of team spirit, if you want to call that. And of course, this is all like squarely in the office. So when I transitioned to PM there, even before then, getting access to all these people was easy, especially if you had interesting ideas.

"Oh, we're going to go do something with this big YouTuber or this big brand. Can we get some of your research stuff in there as a launch conduit for them?" But then when actually being a PM, the way I structured my engagements was like I would always take one external-facing launch that had impact, like an AR feature, and then on the other side, I was like PMing all the volumetric capture and the 3D capture stuff on the research side.

So helping them make sure their next phases of their research agenda got greenlit. That was the part like PMs didn't want to touch. Too nerdy. What's the product impact? Blah, blah, blah, blah, blah.

But yo, I had my perf launch impact story locked down, working with the actual engineers that are going to ship stuff. But then on the other hand, getting to play with the people that were really looking at the tech that then turned into neural radiance fields and 3D G-splatting, all the 3D stuff you hear about today.

In fact, a couple of those folks just launched a paper earlier this week that's making some waves in itself. So I loved the vibe. It was amazing. I would say as a PM, it was still kind of tough, like given that there were so many other folks trying to find their bit of the map.

To launch anything interesting, you have to work with another established product area, right? So if like Lens, it had to be you're working with Search. If it's YouTube VR, you're working with YouTube proper. Even if you're doing ARCore, you're working with Android.

Now some of these things have been consolidated in the future, but I loved that experience. It very much felt like, you know, kind of Skunk Works type vibes.

Contrast that with Maps. So after doing two years of that, I had an opportunity to keep working on stuff, actually fly glasses there, which would have been an interesting, very long trajectory to get to what Google may or may not be announcing hopefully in the next couple of months.

Maps was like a timeless mission where Google, where Daydream and Labs was like, "Well, what are we trying to do? Is it about AR? Is it VR? What is the future?" There was a lot of pathfinding for that North Star.

Whereas what I loved about Maps was like, we've got the most timeless mission in the world. It's like help people find things and get stuff done in the real world. In the world that I worked in, our job is to map the world. Next year, we'll be mapping the world. The year after that, we'll be mapping the world.

And a decade later, we'll still be mapping the world. There was just a much larger, more established team, a lot more, by the way, I don't know, you look like you want to jump in here with something if you want to feel free to.

Okay, yeah, I mean, like a lot of lifers, like the people, you know, I read this book called "Never Lost Again," which was written by Bill Kild, who's now marketing director for the Niantic folks.

This is like John Hanke and Brian McClendon, the co-founders of Google Earth's predecessor, Keyhole, sort of what that turned into. I read that book, and it was like, early on, it was like getting to meet all those heroes.

I was like, just like, I'm like, these are the people that have done all these crazy firsts and all have very fascinating backgrounds from the private sector, from the public sector. You wouldn't imagine the breadth and depth of stuff that was happening there.

And yeah, man, but it was very different in the sense that there was like a huge consumer team that was solely responsible for consumer products and services. There was a huge enterprise team that did all the enterprise stuff, and then there was a huge platform team that built and maintained the map of the world.

And of course, a huge operations team, which was responsible for, I mean, like one of the draws to the job to me was like the hiring manager at the time was like, "Yo dog, do you want to go make glasses that may be launched, or do you want to like move satellites around in the sky?"

And I was like, "Huh, move satellites around in the sky? What are we talking about?" It's like literally there's teams that are orchestrating tasking for satellites. There's like so many people know Google has Street View cars and trekkers, but very few people know Google owns its own fleet of airplanes.

It licenses a bunch of imagery too, so it's just like the world's most immaculate sensor collection, you know, kind of operation and mapping operation on the planet. To go from like doing AR for fun whimsical things, oh, like, you know, so-and-so launched some cool things with it, awesome, to like, "Yeah, YouTube VR videos," and like, "Oh, we can scan Simone Biles for the Olympics."

How cool is that? For how cool is that to, "Oh, we're creating the next generation 3D map of the world, and by the way, do you want to be the product lead for that effort?" It was a no-brainer. I loved it.

Yeah, it was like getting a PhD in maps over the next four years while getting to ship a bunch of cool stuff. Okay, the Maps team, I feel like, has very puzzling product decision-making processes.

So I need to hear more about how this vision, strategy, and roadmap is created because just like as a user, I think if anything, I feel like the Maps product has taken some features from Waze and made them worse.

So like yesterday, my wife searched for gas stations along the route, and it chose some place called gas station an hour away. The performance in low data is pathetic compared to when the iPhone came out in 2008 when like we had no data whatsoever, and it would just download the map.

Now downloading a map is insane. I live in, like, for some reason, a low data area, and if I leave my house, it takes me like 30 seconds to download a route to a place versus my Kia, which no one would think is like a good maps.

It just pulls it up right away. So I was curious, how are they making decisions there? How are they prioritizing stuff? It doesn't seem like an Apple-style product culture where quality is always maintained on the things they're launching.

Ooh, that is quite a scathing commentary. You know, I know exactly the PM who'd be responsible for this, and you should maybe bring him on the show and talk to him about it because I think, like, I mean, quality is huge. It's very data-driven, especially for things like this.

That said, it is so incredibly hard to build a product experience that scales to Maps to two billion users now and works in all the regions the way it should be, you know, the way you'd expect.

So there's a lot of nuances that, you know, perhaps don't get addressed when you're trying to build for such a massive canvas. My personal gripes with Maps and some of the things that, you know, I would like to see change certainly are like there is certainly a very cluttered UI.

It is extremely cluttered. A lot of unification has happened. The stuff that's closest to the world that I care about is like the way I view it is like we've solved getting you to your latte in your gas station.

That's a solved problem, whether it's Maps, you know, or Apple Maps or whatever it is that you end up using, right? You can do that. The things that only Google can do, like especially given this visual spatial model of the world, is where a lot of my interest was.

But even in that world, as I went through looking through some of the older product decisions, I found that a lot of, you know, I won't single out Maps for this, but a lot of billion-user product surfaces, there is a path where, you know, essentially there are high impression surfaces in the products, and you can plop whatever freaking feature into those high impression surfaces, and you're surprised people are going to click on it, and you're going to get engagement.

So a part of me feels like, especially with Search and Maps, because Search gets a lot of the similar criticisms of like sort of bloat in different senses, is like I think it's a problem of incentives where it's so easy. It's like, "Well, what have you shipped?" You know, on like, "What have your engineers shipped?"

And what is your, like, the director of that team? What have y'all like as a portfolio shipped? And you got to have these wins. So you take all these roof shots, going back to the Alex Kosi, you know, roof shots versus like moonshot thing.

Because of all the coordination headwind, because like, holy crap, we need to turn out, like, I don't know, let me give you an example of like landmark-based navigation, where like I want navigation to happen where it's not based on, "In 300 meters, turn right," but take a right at the Denny's or whatever.

But to do it in a fashion that works everywhere is so much effort. How do you measure the data quality for that? And like the data quality efforts and like launch quality efforts were immaculate.

I've seen so many things get rolled back because the initial signs weren't good. As you're doing your percent ramp, you have to do experiments. So like despite all of that, I think for these established companies that, let's say, hit some degree of feature completeness where every other feature is a vitamin, it's no longer a painkiller.

You do have this culture of like rotating features that just aren't all that to get people on the promo train, and then you deprecate that feature, and some other feature pops up, and on and on it goes.

That was, I would say, one of the aspects of that product management, especially at big companies, that I dislike the most because that's kind of the game that a lot of people are playing, and that's what people encourage other people to play too because it's such a well-designed playbook.

It's like, "Oh, let's manufacture a metric that may not even matter," and then move it. And yeah, you moved a metric, but it didn't matter to what end rather than moving metrics that matter.

So yeah, those are my thoughts on it. I think Google has, as a product, has gotten a lot better. Their rendering stack is way snappier now than it was previously.

It was getting very cloying, the rendering stack across web, native, iOS. From what I can tell, at least externally looking in now, Google Maps is a lot more popular on iOS than it used to be.

And the features like, you know, live view, so you can like search what you see with a camera, and immersive view, so you can preview a place, are really exciting and popular, and people love them.

I still think if somebody went through the Maps UX and like removed 25% of the buttons, it would be a better product experience. I still think that there's still, but that's like Google product design aesthetic versus other products as well.

So I've certainly seen even on the design side some very talented designers go back and forth. At some point, maybe we close by saying it runs into this weird problem of like Google Earth had this issue too, but like Photoshop or Premiere in the sense that you built this product.

There are so many people used to working a certain way. How do you not alienate those users while still making it attractive to the next generation that say might be like, "Well, I get Apple Maps as default on my phone. Why should I download Google Maps?"

And I think that's very, very hard to do. Yeah, to meet your kind of incumbent user base, keep them happy, but still make substantial changes for the longevity and growth of a product that bring new users to the party as well.

And so, you know, that's my take on Maps and some of the struggles with established applications. But hey, look at them. They still just crossed the two billion landmarks just a couple weeks ago, and so there's no, which I was like super excited to see because as Pachra Ste down, they mentioned immersive view in there too, and I was like, "Oh cool, like in Pachra's legacy, he views that as like, you know, one of the steps in the right direction."

But hey man, if I'm being honest, there are many more steps to take, and I'm, as I mentioned, I'm rooting for Google to take those steps.

So I heard two big main reasons there. The first: the perils of performance review processes for product managers. The second: needing to acquire new customers competing against things like Apple Maps.

So sticking on the first, you know, it feels like you really nailed the Google performance review process. Three promotions in five years. There are plenty of people who get no promotions in five years at Google and are perfectly happy with that and are even considered as doing well.

So talk to us about that. How did you manage to get three promotions in five years? The first thing was like not aiming to get promoted. That was the first goal because every decision I made at the time, like even like going back to leaving AR/VR and coming to Maps, so many people at the time were like, like, you know, like these directors and VPs that I respect that are like, "Hey, you've got a great network here. Why would you reset and start from scratch and do all those things?"

What I found is that no matter what I was doing, even within Maps, I followed a similar thing where like, you know, I started off on the platform side PMing all the photogrammetry, like 60 engineers that were building and maintaining this map of the world and creating all sorts of like the imagery that gets published into Maps, Earth, the stuff for flood forecasting, the stuff that powers navigation directions, the stuff that powers AR and 3D.

At some point, I wanted to keep moving. I don't know, like I keep going back to this recent DARPA interview because I found so many parallels with how they view product development with sort of how I've implicitly viewed it.

Not knowing is like when you go to a team or you go to an area, there are crazy ideas you walk in with, but typically in two years, a year and a half, even like all the strategic surprise, like all the novel insights get sucked out of you, and then you need something new to like create the next set of insights.

Because you don't want to end up being, or at least I never ended up wanting to be sort of like the person in the country club I was describing, which is like, "Oh, I did my things. You know, I've got my lapel of medals and awards, and now my job is to say no to as many things as possible and maintain the status quo."

Because there's more that can go wrong by like people trying new things than by maintaining, you know, just steady as we go. So I always did the risky thing, and I followed the craziest ideas.

When people were like, "Oh no, you should work on this area; this is far more strategically valuable," I was like, "No, I'm going to go work on the thing that I have a lifelong deep interest in." Oh, by the way, here's a bunch of other people that are also interested in this, and that became the goal.

Then I've also been very lucky that as I was doing those things, my managers have always been so supportive to be like, "Yo, you're actually crushing it. Do we want to try this? It may not happen, but why don't we try it?"

A theme in my life, especially now as a creator too, is like, man, very rarely do I reach out about stuff. Everything is kind of organic and inbound. What I find is like if you're putting out, you know, if you're putting out the message of like what it is that you want to do, you find the band that, like, you know, marry men and women that want to join you on this journey.

But you'll also find like all the accolades and the validation follows after. So, you know, that's what I'd say as a starting point.

I will caveat this answer saying it's like, as you all know and many others would know, the promotion process is a gauntlet. I think one of the other things that helped a ton was just like I never let my level define who I could and couldn't talk to.

I think there were a lot of people that felt, even when I started out, like I remember like my manager at the time was like, "What? You went and got coffee with the VP? How'd you get that meeting?" I was like, "I emailed them. Like, what do you mean?"

So I think a lot of people put themselves in this box, and what I found that in many times in my career, being good at like just sharing, like you don't have to like make this again. If your goal isn't promotion, you're doing cool stuff you care about.

You want to talk about cool stuff. You want to get them to unblock certain problems that they're best positioned to unblock after, I don't know, you've tried like, you know, fangling senior directors to align on a thing, which, oh my God, those are the meetings I don't miss, though. They are freaking hilarious and funny.

You want to just go find a more expedient solution. When you go do that, like those folks remember, and like you'd be surprised. They'll be down to help you out when performance season comes out and give you that peer review or whatever.

And then that, especially for like the more senior promotions, that kind of visibility, are you messaging your stuff? Does the organization view it as important? Just all happens organically because you're not trying to manufacture the perf packet.

You're trying to do the thing that you're doing, and all these other things then result in that happening. That said, if I could go back and have had Claude to write those promo packets, oh my God, life would have been easier.

Those were a painful few days every single time, and I always did it like very last minute. And yeah, that's my journey. It's like, I don't know if it's like a great lesson or example for anyone to be honest, but if there's one lesson I would leave people with is like, please don't chase being a PM.

Don't chase the title, the promotion. Chase cool products and experiences that you want to work on and launch, and ideally cool people you want to do it with. Because also when you're in the trenches, you know that L7, L8 eng lead that you're in the trenches with is going to be there to bat for you because like that's just how it works.

And then it doesn't, it's not even a question that you're asking for their help versus, you know, the other game that a lot of people play. And I'd rather not even articulate it.

I'll draw out two lessons for people here. One is, and I noticed this about the content that you have too. A lot of your content, especially the stuff that has gone really viral, is what I would characterize as very unselfish.

It wasn't like super intentional, engineered for virality. It's almost like cool experiments with technology or cool art that you were creating. So it seems like you throughout your career at Google and now your career as a creator, you're really good at kind of getting passionate about some sort of technology and becoming this creative builder around using it and actually showcasing its potential.

And I think that's very, very hard to do. So I want to talk about that in a second. But I also want to draw this kind of second lesson that I think is really interesting about how you do things, which is this is, I think, the definition actually of agency, which is like not being blocked by your level.

That's probably the thing that I think probably hurt me the most in my time at Google is I was exactly that person that you just described, where I paid so much attention to their level. So if I was X level, I would only communicate directly with people X plus or minus one.

I never thought to just like plus two, just like willy-nilly, just grab a chat with somebody like that. And I think that there's a certain amount of confidence, but also driven out of the first lesson of your passion for technology that allows you and enables you to make such choices.

So what do you think it is about you that is driving that? How can people learn from your passion about technology and apply that in their own life?

It's a really great insight, this agency point. Somebody else brought it up too. I was catching up with my old business school teacher who had randomly reached out on LinkedIn. Reconnects you have, it's wild.

I don't really have a way to describe it other than like this is what I love about being a creator. It's like it's so permissionless. I don't need anyone's permission to share whatever the hell I want online.

I think even a lot of creatives, I interact a lot more with the creative side of the world now, and they have this thing where it's like, "Well, I'm going to take my pitch, and I'm going to like go pitch it to this new emerging AI studio, and one day let me do it."

Then I'll, homie, just go freaking do it. But I don't know what it is. I think it's like conviction or belief in like the thing needs to be done.

Like many times, it just could be like, "Hey, we have these assets. We could be doing this. Why aren't we? Can we talk about this?" I'd ask questions like this, and usually, especially when I'd reach out to senior people, like, "Well, let me tell you there's this long backstory, and let me tell you about the time we launched 3D in Maps in, like, you know, 2008 or whatever and what happened after that."

And I'll get the full background, and that stuff is so helpful for me because I think in order to be effective and have high agency in organizations, A, you need to know who the influencers and decision-makers are, and they're not always the highest-paid person in the room.

Which is very, like, oftentimes it's like who the highest-paid person in the room delegates or defers to for certain decisions. In GEO, and I like, it's a lot of these area tech leads. These are usually like principal engineer plus type people that are just the domain experts.

Knowing the map, the chessboard, or whatever you want to call it is one piece of it. But then getting out as much institutional knowledge from them because the moment then you have that subsequent convo and you're not anchoring it in like, "Here's me, naive new PM that's like excited to change the world like every other PM," and then gives up, they see your persistence because you're trying again and again.

And B, you're refining, like you're coming with context that they didn't expect you to have, and that immediately builds trust and credibility. Then people are like, "Yo, you kind of get the old Maps. You also get where we're going. Help."

Then you can be one of those people that leadership looks to to help those transitions happen. I think right now is such an interesting time if you do want to be in product and doing these type of things to help, no matter what vertical you're in, connect the past to the present and the present to the unknown and bring people along for that ride.

That's kind of how you articulated some of my creation stuff too. And yeah, man, like the way I'd like to describe my content, it's like imagine if we were just grabbing coffee and I was showing you something cool. That's the vibe I'm trying to go for, to show something to you.

I've tried some of the, "Oh my God, like XYZ thing is dead, and here are 10 examples" thing, and some of them have gone viral, but I always feel very icky after I've done them, even though it's a part of playing the game.

Maybe I need to go back to doing a couple more listicle threads, but that's the vibe I try to go for. It's like I'm trying to leave something. It's funny, I was catching up with this really cool guy who works at Google Labs, used to work at YouTube.

Shout out Tron, you should have him on the pod. I think he'd be great, by the way. He was basically like, when you're so, everyone is so stuck in their corner of things.

So the other lesson I'd pull out is if you have the broader view, if you can build a map of the space, then you will undoubtedly have insights versus the people that are stuck in their little corners of whatever the hell.

And so I don't know, man. I've been trying to articulate when I was talking to Tron, like what is my vibe as a creator? What am I trying to be? I think I'm still figuring it out, trying to reconcile these three facets of my personality, which I can go into the three dogs in me, as I like to call it.

But really the answer I come up with lately is like I'm some weird combination of Marques Brownlee's accessibility, Ben Thompson's deep strategic insight, and Corridor Digital's mad science VFX experiments.

Those are sort of like the three facets I end up going to. I try to connect dots between them where I can, but it's usually very conversational. I won't try to patronize people, but I leave a nugget in there for the people in the audience to get something out of it too, which is what I appreciate about MKBHD.

I love reading Ben Thompson's stuff, but I've shared it in the past with, you know, Gen Z folks and APMs, etc., and they're like, "What? This is like, I can't get past the first page. It's too dense," right?

And then the same thing with experimentation. It's one thing to play with the tools and do the tutorials. It's another thing to play with four primitives that people aren't putting together in an interesting way, and the likelihood of something interesting happening goes up based on that.

And that doesn't matter if you're building videos or content like me or software too because, again, we have all these Lego bricks, and they haven't been put together in the most obvious combination.

To go full circle back to the Google thing, it's the same thing at these big companies. You'd be surprised when you go in and take a stock at what's out there in Meta or Amazon or Google. There are all these primitives sitting on the pantry shelf that have not been used.

So hey, if you got creative ways to use that to move the ball forward, I think, you know, at some point, the mission is bigger than you, and that inherently gives you agency because you are convinced this thing needs to exist in the world.

An area you talked about in another podcast that we haven't hit on yet but I think was an important part of your performance reviews, you tell me if not, was when you said especially on teams where design resources were light, I got to flex into the technical side.

Tell me more about that. I've done like, you know, the various types of product roles I've done: emerging platform like ARCore type stuff, research PM. I've done platform PM, right?

Where you're like, you own a platform of data products and services that run consumer applications, enterprise applications, and other stuff too. And then consumer itself.

What I found is when I was in that platform or research mode PM, usually like, you know, like the design VP ain't doing out a bunch of designers for the research teams or the platform teams.

Like, "Oh, you build data products and services. The consumer teams have resources to build around." And so that was like the closest, like, it gave me the superpower where like I knew the tech like the back of my hand.

I've been doing this since I was like 11. All the stuff from graphics beautifully transitioned over to computer vision. I got a full speedrun on computer vision and AR/VR by the time I got to platform PMing on Maps.

Yo, like, "Oh, here's like a cool new data product that we made." I was like, "Well, guess what? I'm going to take this into Maya and whip up this crazy cool prototype. I'm going to take this off-the-shelf 3D reconstruction software, run it through our in-house pipeline to do some semantic stuff, throw it into this third process."

Other products that I ended up helping shape partnerships with those products too, you know, basically let you go from like, "Huh, this is kind of a neat idea," and come to the meeting the next week with like, "By the way, here's what I whipped up over the weekend."

Everyone's like, "Whoa!" Trust isn't just earned; it's demanded. Whether you're a startup founder navigating your first audit or a seasoned professional scaling your GRC program, proving your commitment to security has never been more critical or more complex.

That's where Vanta comes in. Businesses use Vanta to establish trust by automating compliance needs across over 35 frameworks like SOC 2 and ISO 271. Centralized security workflows complete questionnaires up to five times faster and proactively manage vendor risk.

Vanta can help you start or scale your security program by connecting you with auditors and experts to conduct your audit and set up your security program quickly. Plus, with automation and AI throughout the platform, Vanta gives you time back so you can focus on building your company.

Join over 9,000 global companies like Atlassian, Cora, and Factory who use Vanta to manage risk and prove security in real-time. For a limited time, my listeners get $1,000 off Vanta at vanta.com/AAS. That's Vanta.com for $1,000 off.

Imagination and my reframe on that, on product visualization or product design, is how can we show people a glimpse of the practical thing that we're trying to build? Whether we're building it for real or creating a pseudo-vision video, I think it's the same sort of thinking that you need to go through. Many people, when they make a vision video, are too untethered from reality; it's not exciting enough.

There have been these vision videos, some that I made myself, and sometimes we'd work with a team or even get agencies to help us out. A lot of companies do this for new tech development, and that was just a huge asset for me. I could go from understanding the tech to seeing the product potential and then to visualizing the user experience. If you can have a version of the PM, UX, and tech in your head, I think that's a superpower.

Now, I find it so cool because I ended up studying computer science and business, but I did a bunch of VFX and product design. I learned After Effects, 3D, and all this stuff. You don't need to do that anymore. If you have tools like VZero, Cursor, and Figma AI, you can go from an idea to a very concrete thing to react to.

A lot of people like to say prototypes over PRDs. I like to say, why not both? I think a lot of PMs will find that sounds like wizardry to them. They're so far from that world; they live purely in meetings and documents. They're not actually connecting tech together or building designs. How can PMs adapt that mindset and move towards it?

I think the first thing is, if you're not working in a domain that you're passionate about, it's tough. I get it, though. I have friends who view their jobs and are still PMs at Google, very comfortable. One of them just moved to an IC L8 role because now Google has IC directors, and it's just doing all this cool stuff, not all the management stuff. There are ways to carve that future out, but you can't do it if you're not passionate.

You have to have those superpowers and insights. Connecting the dots is crucial. On the weekend, even though you're not getting paid for it, you're looking up stuff for your own self-interest and curiosity. I lurked on Twitter for so long, basically until the start of last year when I started posting. When we started DMing at the beginning of 2023, I never took that platform seriously.

You've been crushing it, by the way. Hats off to you for both Twitter and LinkedIn. I've seen your rise. I would often see you mentioned on Blind too, and it was kind of funny. It's really cool, and now I know the guy.

Back to answering your question: if it's just a job for you, your mind in the shower or on your walk isn't going to be making those connections. I like to view it as finding a problem space that's worth loading and keeping in your context window for a long time. If it's not, there are other ways you can optimize that. Maybe you're building something on the side. Focus your energy on that side piece versus your day job.

But it's also not that hard. You could go into MidJourney, and especially Ideogram. I feel like PMs are sleeping on it for some reason. Ideogram and Google Imagery folks got sick of the tanker ship and got on the speedboat. A16Z did a startup that does phenomenal text rendition. If you're doing mocks for UI/UX design, you need to take a look at Ideogram. You can do similar things in MidJourney.

I used to love doing this even at summits, even in 2022. This was when there were internal models at Google. I just used the internal image models when we were debating things over some summits. We only had a certain amount of resources, and we could only fund this thing or that thing. You had to go into full influence mode at that point.

These artifacts are so useful because, again, not only do they help people visualize, but they help people come to the same page and then build on the thing with you. You could do that; VZero is super easy. My buddy Matt Wolf, one of the top YouTube creators, is going crazy with VZero and Cursor stuff.

If you're listening to this and thinking it sounds out of reach, I'm telling you, you're one weekend crank session away from having Ideogram, Cursor, and VZero be invaluable assets to help you do your product job better. So much of what we're doing right now is just that the tools are already there. People are just not taking advantage of them and applying them to their specific domains.

People are applying them to the stuff that gets broadly circulated, which is fine, but think of your contextual knowledge and the problems you're solving. Try to use these tools as part of that product development lifecycle. You'd be surprised how easy it is to get in.

I think ChatGPT is 100% awareness, like you mentioned with Google search trends. I kept asking my audiences on my India tour how many of you guys are using Claude. It was something like 30-40%. Ideogram, VZero, and Cursor among the PM audience would have been one hand if I had asked them.

Maybe we can walk tool by tool through them. How do people get started with the tool, and what use case should they be using for it? Let's start with Ideogram. I view that as you're writing your two-pager on what product you're building or feature you're building. Maybe you're even as deep as into your PRD, or you're thinking about the strategy of a space, or you just want to visualize an idea.

What I find is it's very useful to use Claude, ChatGPT, or your LLM of choice to write your prompts for you. You might even provide some context, like visual references. You can upload five images into Claude, for example. Here's a slide that I have in mind that stills my idea. Here are a couple of other competitive apps that I'm looking at. Help me write a bunch of prompts to explore a design for your product feature, whatever it is.

Then just throw that into Ideogram and see what happens. You can use the various remixing tools, the in-painting. They've got a canvas editor now where you can do all of this in a Figma-style canvas and select specific parts. You can get even more into a flow state without having to jump into Photoshop or wherever.

That's kind of how I view Ideogram personally. How I use it right now is for packaging for videos, which is also important for product folks. What does a landing page look like for this thing? Is this something that people would actually click on? How do you test that? Go mock up a landing page, and then you can toss that into VZero or Replit Agent or whatever you're using and turn that into a real thing.

VZero is really good for UI with code. It's not going to build the full thing for you. Replit Agent is cool in that it'll handle the backend and all that stuff for you too. If you want to get to a deployable app with just a couple of prompts, Replit Agent is a great starting point.

Now, if you code or want to get into coding, Cursor is the way to go locally on desktop for sure. It's super easy, and there are a bunch of guides out there. I particularly point folks towards McKay Wrigley, who does some great writing on it. My buddy Riley Brown has also started doing some good stuff on it.

Riley's journey is particularly interesting because he calls himself a "senior software composer." That's the thing he's developing. I was like, I don't know, man, that just sounds like a product designer, maybe like a design savant to me. But I think proof is in the pudding. You've got a bunch of these folks who had never written a lick of code, were making TikTok videos, doing marketing, and are now building out web applications for themselves to facilitate conversations at their workplace.

The bar is even lower. That's how I think about these tools. One more thing I'd say is I find it very helpful to use Perplexity for research. I'm actually doing this right now for a company that I'm advising. Building context on the space, doing competitive analysis, rather than going to your product strategy team and waiting for them to take a month to get back to you with a "here's what we think about the landscape."

No, you should build your own map of the landscape. The tools to do that are so much easier now. I would find myself previously, when I was on my personal device, just sharing links from iOS to my work email. "Read this, read this, this is the thing." At some point, when I wasn't sitting in a long set of meetings, I would start building some loose connective tissue around the background research.

You build a map of the space too. That's really helpful context. Going back to Gemini, especially these larger models, or even giving it to Gemini to start, say, "Hey, can you distill this down into 2,000 words?" Then tossing that into Claude as context when you're doing your prompt iteration for what to pull out of VZero or what to pull out of Ideogram.

Those are some ideas. There are so many other ways to use these tools, but hopefully, this will give you a concrete starting point. So, Ideogram is going to get you your mockups. It sounds like VZero is going to get you your prototypes, and Cursor would get you an MVP. Is that a fair mapping?

For the full-stack PM, that's how you think about it. Traditionally, full-stack PMs only exist in startups, but you kind of exploded that by bringing the full-stack PM model to Google. I feel like that must have been one of the key factors in your promotions.

How do you manage your time at Google? I feel like I was just overwhelmed with meetings and PRDs. How did you find time for these extra full-stack activities? It's funny. Shout out to the guy who told me I should be a PM, Ryan Cassidy. When I was asking him for advice on leaving Google, he was sort of like, "You don't actually have one gig; you have two gigs, and then you're doing a third gig on the side."

I think I was always a bit of a workaholic in the sense that I get bored very easily. I always needed to be doing at least three high-priority things. Then I could have three or four things on the back burner that I pop up to stack when needed. That diversity of having three high-priority things with different groups of people and different levels of risk made it always fun for me.

I don't have a better answer than I was a workaholic. I absolutely worked 12, 13, 14-hour days a lot. I think that was a shock to a lot of people, especially in Maps, where in Daydream and the Labs team, there was a lot of culture of people with a startup mentality. In Maps, everyone was like, "Yo, we Gucci, man. We got a billion users; it's all right."

Whatever you're thinking of doing isn't going to make a huge splash. I think that almost made it easy for me because so many people were like, "Oh, I couldn't possibly do something ambitious enough that would be interesting as a geo-level priority." Meanwhile, I found this very good balance.

You know how people say you should have a hobby that keeps you in shape, makes you money, and blah blah blah? I feel like there were two sides to my coin while I was at Google. There was the "Hey, I'm a product manager at Google building stuff, connecting the physical and digital worlds," and then my personal version of that was "I'm blending reality and imagination with visual effects."

Even when I moved on to Maps and we were doing some cool AR stuff, like launching the AR Core Geospatial API, it was just freaking awesome. Gorillas used it to turn downtown Manhattan into a music video and do a bunch of other cool stuff.

I would be so deep on a set of facets of this problem during the week that when the weekend came around, it was like, "Oh yes, Saturday morning, I'm going to get my cappuccino in." I'm a bit of a cigar aficionado—not the best habit—but I was telling you that nicotine-caffeine combo.

Saturday morning, I would have an idea for a video that I wanted to make using one or two new tools that came up. Even though I moved on from the AR/VR stuff, I was still playing with all these tools. I got into early access for TikTok's Effect House. I think I'm still in the top 1% of AR creators on TikTok, which is kind of wild.

It was like I was recharging while still learning things that applied to the day job. I was learning about real-time computer vision with all the AR stuff I was doing on the weekend. So when I went during the week and we were doing large-scale, world-scale offline computer vision, there were still a lot of insights to be gleaned from that.

That's kind of how I viewed it. I lived and breathed this stuff. I don't think I took work-life balance as a priority, and I didn't need to because I just enjoyed it. One thing I will say that helped me be much higher leverage is that I got much better at saying no to meetings and skipping a bunch of crap.

I think every PM goes through this phase where, especially if you're doing a bunch of initiatives, you'll always have touchpoints. I've been fortunate to work with some stellar PGMs, but other times, there are just sinks for the sake of sinks. Figuring out when to pull out of those has been huge.

The last thing I'll say is that morning times were sacred. That's when I did most of my product thinking. I usually had to talk to the engineering team in Zurich, and before that, I would do my important thinking for the day. I tried reading "Make Time" and "4-Hour Work Week," and to be honest, there were still many days where I had a litany of meetings that never ended.

I stacked them all together, and it's probably not healthy. PMs are probably the only job function that will do eight meetings back to back, eating lunch in the meeting too. In retrospect, a lot of that was unhealthy, and I think I could have done a better job of spacing things out.

But gosh, man, I was just so passionate about what we were doing that it was easy for me to do that. The weekend was just a treat to flex this adjacent but still relevant side of the problem. During the week, I'd be awful at creating content. That's why I love shorts.

For the first, I would say especially, I started making content in 2016, but when I was balancing both worlds, especially during COVID, short-form video was the right content format for me. I could do something interesting enough in a couple of hours, get it out there, and feel that full satisfaction versus trying to do what Mark Rober did.

He was working at Apple and doing these long-form videos on the side. God bless him; maybe he wasn't a PM. It would have been harder if he was. Those are some thoughts. I don't have anything crazy other than I just love this, and I can't not think about it.

Even when the workday is done, at night, I'm scrolling Twitter. Those are the things I'm looking for, sharing, and searching. All our B2B PMs might feel left out, but the real thing here is to find some love about what you're building.

Even if you're building for some persona you'll never be, you can't feel like you can't get passionate about your product as passionate as others. I think what I love about your work, especially if people haven't seen your shorts, is that a lot of the shorts are true AR shorts.

Some will be in the real world, and then there'll be some crazy rendering inside. They're like creative work products, the type of thing I would expect coming out of a product designer. It also feels like you didn't take the product management role as narrowly as some people do. You blended a lot of design and engineering into it, which I think is a glimpse into the future of PM.

That's a really good observation. I feel like that's a glimpse into a lot of job functions that are going to turn into this sort of generalist-specialist, if you want to call it that. You're maybe the conductor of the symphony, but you know how to play a couple of instruments in the orchestra or have played some of the instruments.

They could be domain expertise too, not just skills. Even the B2B folks, this is the part of Twitter that I find funny. People are giving the B2B SaaS startup founders a lot of crap because they're so exuberant and boisterous about what they're doing and how they're changing the world.

I think more people in established and startup companies doing B2B stuff need to be, dare I say, as delusional about what they're building as some of these celebrity personality startup founders. There’s so much to get excited about.

People look at video intelligence as this lame thing where you upload content to a CMS, and it writes captions for you. There's another way to think about how it fundamentally reshapes the world and lets you convert video into structured data for so many different applications.

There are ways you can get excited about it. Just find that spark of excitement. You can't always relate to the target persona. Many times, I couldn't, especially on Maps, because you're thinking of a generation. If I was building something for me, I'd be working at Niantic and building out their scan app.

For me, that would be the ultra-nerd type person. There needs to be whatever facet of it is. Either you love working with the people, you love the tech stack or the associated technologies, or you love the industry set of problems.

Those are probably the three that come to mind. I think there are too many people that just run life on autopilot. I think that's where the temptation comes in. One of the weirdest things about this experience is that there are these people I looked up to early in my career at Google.

Some of them I'm still friends with, like you said, who didn't move as fast. When I would dig into why that happened, it was either a lack of agency or a bad luck component. Some people can take on a project. If you worked on Stadia, you could have been the best PM, but that project got axed.

I'm sorry for all those folks. Even there, the folks that crushed it moved on. They were entrepreneurial enough to say, "You know what? I'm going to go build this exact thing at Magic Leap, and then I'm going to go be a VP at Nvidia doing the exact thing."

There are people that just... I don't have a good way to say this. If you don't move towards something you're passionate about, you're going to be far more likely to default to behaviors like, "Let me just figure out how to game my next promo."

You're going to default into that. I think there are some insights to learn from how to play the processes and the game right. But for me, if you're not excited about the thing, getting out of bed in the morning, then why are you even doing it?

Why not find something else where you could do it better? I think life would be a lot better if you did it that way. Or, like I said earlier, don't worry about putting your day job into maintenance mode and find that spark somewhere else.

If you don't have that spark inside and outside work, I think it's going to be tough. The other group that might feel a little left out is the parents who feel like outside of work, they have no time. But I think there too, it's more about what I noticed about your work.

Like you said earlier, it was the anti-promotion review culture. You weren't like, "Oh, I could get a promo here." You went in and worked in areas you were excited about. It's about finding that excitement and joy.

Totally. It's also about going to a place where other people think it's crazy and foolish to try things. You'd be surprised if you go to these incumbent organizations how many sacred cows there are.

What I mean by sacred cows is things that people view to be truths that aren't actually true. They're just like, "Well, we tried that back in whatever." There are so many answers for things like this everywhere, and it's all based on scar tissue and the things that have happened.

I think going to a new place, as I mentioned earlier, is like once the strategic surprise has been sucked out of you, go do something else that builds new insights. Then you go to a new landscape and try building new things. It's a lot of fun.

I think people get comfortable too easily. We're hardwired from an evolutionary perspective to chase security. It's like, "Well, I know this. I've worked with this team. The VPs and those directors know me already. It'll be easier."

It's so easy to make that case. It's funny talking about parents. I remember especially moving to Maps. I remember talking to my mother at the time about this decision. She was like, "No, B, what's wrong with you? You have to stay."

But then I look at her career. She's a retired bureaucrat who worked for the Indian government. That's all she knew. In this world now, you've got a bunch of these large players that are like tanker ships and a bunch of speedboats.

The market's getting tougher or whatever, but again, if you're passionate, that's why you'll stand out, and you'll get to work on something fun. Learning for the sake of learning and cultivating passions is so much easier.

Even for parents, the fact that your research is now, after you put your kids to bed, sitting on your MacBook Air or your phone on Perplexity, typing in a couple of queries and getting an answer that would have required you to do two or three hours of research is insane.

There are so many interesting high-leverage opportunities to discover knowledge and thus discover what you're passionate about, where you will stand out. My consistent advice to people is if people aren't vying for telling your manager, "I want this guy as a PM," that should be your goal.

The only way you can do that is if you can level with what actually exists and demonstrate agency for getting stuff done. You don't always have to be a visionary. To be clear, there are plenty of successful PMs that are pure operators.

They're really good at recognizing an opportunity. They may not have even come up with the insight, but they can rev on it. Those folks are very successful in large companies too.

Those are some thoughts. I almost think that's the dominant archetype in these large companies these days: the smooth operator. You took the kind of more disruptive approach, which I think is almost like a barbell strategy. Either you're going to flame out, or you'll succeed a lot like you succeeded.

That's totally right. I think there are so few people that take that route that I don't think we have enough of an inclination of what happens if people purely chase. It also goes back to how you measure success.

If your success is, "I want to get to X level as quickly as possible," and that is the focus because your day job is, "I want to make a living for my family," then you draw meaning, excitement, and passion from something else. I think that's a perfectly fine route.

That just takes me back to all my friends who decided to go into finance. All these IIT grad types that go into being a quant or doing high-frequency trading and stuff like that. They're not deeply passionate about it.

Okay, maybe they nerd out about repurposing some naval satellite dishes to get the latency for their trades down even further. There's that part, but largely it's a means to an end. Then they're buying all the guitars possible, going to all the concerts possible, taking all the lessons.

There are other ways you can find meaning, so it just depends. If you can find that Ikigai, where it's both, I always love the Ikigai framework. It's really hard to find.

If you're looking for a job, toss a screenshot of that into Claude with a bunch of your conversations in context and ask Claude to interview you to find things like that. You'd be surprised at the stuff you can do.

That's the way I want to live my life. It's all got to somehow fit together. I will say that's a detriment of mine too; it doesn't all just have upsides.

So PMs are taking your advice. They're looking internally for that PM job. If they found it, I know you mentioned you've advised some people on these types of internal transfers and transitions. How do you go about manufacturing it? What are the step-by-step things you should be doing?

Great question. One thing I'll say is that Claude is getting hot lately. I don't know if you've had Logan Kilpatrick on the show. He's a PM in Cloud, but he owns AI Studio, and that's putting him in a very interesting position.

Back when I was a platform PM, I'd have to go find folks like Logan to get my APIs launched. Go find the Google Maps platform enterprise PM or whatever. I know for a fact that these roles might exist within the product area you're already in.

First off, figure out the map of where you're at. The next thing is, certainly at Google, and I'm sure the culture is very similar at Meta, little coffee catch-ups are great. But don't say, "Pick your brain." Every time I get that "pick your brain" message, I just have an internal eye roll.

My brain feels like someone just doesn't have much respect for the brain. Beyond getting triggered by the terminology, this idea of a coffee where you get to understand is great. You can search literally everything at Google.

There's an internal search engine called Moma. You can look up all the other projects, the full map of who reports to who, what the teams are, and how many headcount there are. A few of these things have gotten closed off for more sensitive projects, but for the vast majority, the world is still an open oyster, at least so I hear.

Take advantage of that. Go one or two ways: either double opt-in mutual intro. I've always been a big fan of that. You have somebody connect you. "Hey, so-and-so is great. You all should grab coffee." Boom, usually, you get something scheduled.

Or just cold message somebody and ask them for their thoughts. From there, do a mix of people who are ICs working on the team itself or in the area so you get a better understanding of what's on the ground.

Then also hit up hiring managers, the folks managing those people. Even if there isn't something right this second, something may open up in the future where you're uniquely positioned for it.

You've probably heard this, and I think it's so true: the best jobs never even make it to the job description phase. They're like an email or a short memo of what people need to staff this thing. Very rarely does it go up into the internal job portal at Google, and certainly into LinkedIn and some of the external-facing stuff.

It's very different for engineering, though. I feel like in engineering, the more interesting roles are kind of like this. Maybe the manager roles are like this, but for product, so much of it is less about those jobs to begin with.

The ones that make it out there end up being sort of milk toast, bland, framework-autobot-regurgitator PMs, which I don't like at all. Those are some pieces of advice. Just do some scoping using the internal search, talk to some folks, and position yourself for something exciting.

Many times, you'll have a hiring manager that's cool with you doing a 20% project where you end up actually launching something with that team. That's a great way to build a relationship.

That's how I retained my relationship when I moved over to Maps from AR Core. I owned a bunch of 3D stuff, including the visual positioning system. I made the pitch for launching VPS as a consumer-facing product.

I ran into all the same stuff: "We've tried this before. Here's why we didn't launch it last time," etc. Safety concerns, security concerns. Yes, you can hack through them one by one, but a huge part of making that possible was building a partnership with the AR Core team, Clay's team.

I had a really good ally there in my PM counterpart. We were basically joined at the hip to make this stuff happen on both sides. He handled the AR Core distribution, getting exec alignment over there, while I figured out all the map stuff, the enterprise stuff, the safety stuff, the data products, and the scale.

What are the next feature sets? That kept that open line of dialogue. I talked to Eric till date. I just met him about a month ago to talk about the latest that's happening.

There are so many ways to take a dip in the pond, survey the landscape, take a dip in the pond, and then make a more proactive step. The last thing I'll say, because this is cool, is that a lot of people do this externally. They'll get five external offers, but they won't do the same thing internally.

This always blows my mind. If there are two teams within the same PA that want you, in addition to whatever your current role is, and then there are two other teams where you could go here or there, you might end up carving out a more favorable situation.

You're not going to negotiate tremendously on comp internally, but you're certainly going to be able to negotiate tremendously on scope. As you move up the PM ladder, a lot of it is about how much scope you're responsible for.

Those are all such underused things that people just don't do. When they do, I get hit up a bunch, being like, "Hey, I looked at your stuff in M, the stuff you did there, and the stuff you're doing now. How do I do this stuff?"

I'm like, "Well, are you excited about it?" If there's no excitement, they just read that spatial computing is going to be a thing, and it seems kind of cool. They're thinking about it as one thing in addition to the four other areas they are interested in.

A lot of this just keeps going back to that thing: why product manage a product that you don't want to use or are convinced needs to exist? Sorry if that's not the most practical advice, but that's how I genuinely feel.

When you're doing this outreach and reaching out to folks, maybe the foremost expert, I remember this conversation. I was putting out this deck on my thoughts on all the latest 3D reconstruction research and comparing it with other stuff.

This guy responded, Richard Zisy, who wrote the book on computer vision that all the PhDs read. People from a bunch of my entertaining man were like, "Did you see Richard responded to your email on that group chat?"

That resulted in a great conversation with Richard and some insights that we incorporated into the product we were building. But that wasn't from a place of, "Let me go pick Richard's brain on cool things we can do." It was from a place of, "Here's my opinion on things, but you're the OG. You probably have a lot more insights here. What do you think?"

I think that's such an interesting curiosity hook for people to help you too. If I could make it practical, what comes to mind is the same way you think about packaging a tweet or piece of content so that people will engage with it.

You have to think about the same thing when you're reaching out to these folks. Usually, they'll still give you the time of day, but I'm telling you, the meeting will be so much more interesting if you don't start with, "I don't know anything about AR/VR, but I'm really keen to learn about it."

Or, "I don't know anything about 3D maps, but I tried using this one app. What can you tell me about it?" It becomes so boring versus, "Hey, I've actually gone second. It's like 0-1. You got to do your 0-1 thinking time, put some inference time, compute going two to three levels deeper, then reach out to those folks."

You'll have a more interesting conversation, and if they vibe with you, trust me, you will be in their head. Those neurons will be like, "There's this person I talked to that would be perfect for this thing." I've seen that happen so many times, not just for roles but with cross-team partnerships.

You go talk to that hotels or travel team that's swamped in a bunch of other stuff. They didn't have time, but you got them sufficiently excited that when their next summit came up, they were like, "Yo, we need something for under 24 users that's visual forward."

I remember a guy that I talked to.

Let's fast forward from product management to where I feel like you're one of the few people fully exploring the space. You've said generative AI is super cool, but I'm continually blown away by the work happening in procedural 3D modeling. Can you walk us through that?

Yeah, I mean, generative AI has this slot machine kind of stochastic chaos to it. You type in an image prompt, and you'll get a different image every time. You can lock down the seed, and okay, you're kind of getting the same image now, but you don't have a very effective means of going from the idea in your head to something on a canvas with fine-grain control.

What I mean by that is, okay, you made a MidJourney generation. Let's take the exact shot I'm looking at. You've got the mic, the bokeh with the acoustic guitars in the background, the posters—everything looks great.

"Oh, actually, I don't want the guitars there. I wanted something else. I want there to be an LED wall with a map of the world that’s slowly rotating." You have to do manual in-painting and then go back to the slot machine to get something.

Maybe you have something exactly in mind, or you want to change the expression slightly or change your T-shirt. You were in this pose, and now you want to be in another pose. All of that is still solvable in image space.

When you move that over to video, suddenly the problem is multiplied to the max. The most interesting things I've been seeing is that before generative AI and diffusion transformer models were equated with generative AI, you had this thing called procedural 3D modeling, which was also called generative AI.

Architects use this to figure out how to build a building that has certain constraints. I need these types of windows, this type of slanted roof coverage so I can put solar panels on it. You've got these constraints that you're giving the system, allowing it flexibility to figure out the rest.

You exert fine-grain control on the stuff that matters and use generative AI to take it all the way. That's what I become very excited about: this hybrid approach. Think of it like wireframing.

If you're doing wireframes, sometimes you don't want to have a high-fidelity prototype because it gets you thinking about aspects that don't actually matter yet at that stage. The equivalent of wireframing in the 3D world is called gray boxing.

In video game land, whenever people make video games, before you see the final visuals, there's going to be this heavily gray-boxed, low-poly version of everything to get the mechanics right. Does this feel good?

You could use Pixar's storyboarding to nail the story in the storyboarding phase with visuals and audio attached to it before you go make the movie for real. I see this beautiful confluence happening between these technologies.

I had a chance to ask Jensen about this at GTC. He thinks every pixel in the future will be generated and not rendered. I think that's true, but I think underpinning it will be an explicit 3D scene graph.

What do I mean by that? Whether it's an advertisement, let's say I want to author an advertisement for Volvo, and I want that creative contextualized for South Korean demographics versus Indian demographics.

How do I author that content at this higher level of abstraction that almost represents a wireframe or a document object model? I can lean on the creativity of these models, giving them very clear guidance on what to do.

Not just twiddling and iterating on text prompts, being able to do that with 3D is going to be a game changer too. Imagine this on your browser. You'll have a lightweight 3D engine running that keeps track of the world you're in.

The 3D version of a document object model is called a scene graph. Here are the characters, these are the characters in all my generations, these are the environments that I care about, and here are the rules of physics for the world.

You can attach interactivity to things in your scene graph. Awesome, do that. I can move a cost to the left or right exactly where I want. I got everything was a perfect take, but I need to change this one little bitty thing over here.

Now you can do that, but still for that final render, you can lean on the creativity of these generative models to get output that would have taken way longer to produce the traditional way. You'd still be texturing the thing, lighting the thing, and manually animating the thing.

I always find the best of both worlds very exciting. Something about this next level of iterative control is going to come out of this combination of scene graphs, where you have explicit fine-grain control of the elements that matter, with the creativity of generative AI models.

You're seeing the earliest instantiations of that right now, more for product photography. If people want to put their product on Shopify but put a bunch of different backgrounds, you're seeing the simplest version of that.

Let me tell you, people are going to pour gasoline on it, and movie-making is going to look a lot different by the end of next year.

Another technology that I've seen you kind of following the thread of more than others is Nerfs. What's been happening in that space recently?

Totally. Yeah, so Nerfs stand for neural radiance fields. Think of them as a 3D representation that models the complexity of reality. It's a way of making photorealistic 3D objects and scenes.

What's cool about it is the force multiplier and the democratization that's taken place. Going back to the AR/VR team I mentioned, I was PMing this augmented perception team that was doing 3D capture. Back then, we needed these crazy GoPro rigs and cameras on a dome to volumetrically capture a space.

Then you needed to throw this at a Google data center to process. In contrast, one of the techniques that have stood the test of time, if you will, is called photogrammetry. It's the art and science of measuring stuff in the real world.

Basically, taking images and even other sensors like LIDAR and using them to make measurements. Photogrammetry has been around since before computing was a thing. We reached a place in the early 2000s where you could make a good GTA 2-3 looking model.

That's the best way I'd put it. I got a bunch of images of the world, and I can fuse all those images into this cardboard cutout model where all the static structure in the world is perfectly recreated.

It's amazing for skyscrapers, buildings, and concrete structures, but awful for things on the ground and vegetation. If you still look at Google Earth, all the vegetation looks like broccoli trees.

It's not what it actually looks like, but it's a good enough abstraction of what's on the ground. Enter image-based rendering. There are these two separate tracks in computer vision: photogrammetry, where you care about measuring stuff, and the other is sort of like 3D photos on steroids.

You might remember the earliest instantiations of this being PhotoSynth for the real nerds. In Facebook, in 2018, they rolled out this feature where you could take a photo and move around to see a little bit of what's happening.

That line of research continued. We got something called multiplane images that turned into multi-depth map images. But really, in 2021, came out this paper called neural radiance fields.

It took this age-old problem, historically called light fields, and asked how to replicate the entire field of light that represents something in my head. All the light rays that I'm getting, such that if I move my head around in it with a VR headset, everything looks realistic.

The reflections move, the glint, the veneer on your skin, the rim of your glasses—everything works perfectly. That ended up being a really hard problem. LRO tried to build a company around it and failed.

Along comes a bunch of Google researchers and Berkeley research in 2021. They come out with this paper and say, "No, actually, this is way easier than you think." For the first time, we got this representation where the complexity of reality could be represented by the weights of this multi-layer perceptron.

This neural network gets so good versus explicit triangle polygons at representing the complexity. This neural network did a better job, but it still took a lot of compute. Back then, I had to beg for TPUs or wait way too long to play with this stuff.

But then Luma dropped. Luma came out with a mobile app that let you do exactly that. You're capturing with your phone, and now you've got a 3D scan of the spaces, places, and objects you want to capture.

You can reframe it infinitely in both. You want to use it in your other content? You would throw it into a game engine and start doing things like that. But it was still a data center in the cloud.

Fast forward another year, Gaussian splatting came out, which again follows this theme of old school meets new school. They took this really old school concept of Gaussians, a way of representing stuff, and combined it with this OG physics concept called spherical harmonics.

They put it all together to create something that was way faster to train than Nerfs but could render in real-time in your browser. Suddenly, that same idea of radiance fields was pulled out of the neural bit.

They made these non-neural-based radiance fields that are amazing. Music videos are being made with them now. People are embedding them as photography on their product pages.

Even more interestingly, the photogrammetry industry, which had traditionally been researchers playing around with their fun little research papers, is now saying, "Whoa, you're telling me this works with the existing pipeline I already have for drone scanning the world or doing all sorts of 3D reconstruction for Hollywood?"

This is just one cool new artifact I can get out of that that models the complexity of reality and is editable. That's kind of where we're at with radiance fields.

I'll stop by saying I've got a couple of cool videos on this, "3D Gaussian Splatting Explained," on my YouTube channel. You'll be surprised at the various utilitarian and delightful applications of this tech.

But fundamentally, I've been so bullish about this idea of taking observations of the world and turning them into 3D renditions. I think we've solved it for static scenes, and the next paradigm is volumetric animated 4D scenes. I think we're not too far from that future either.

Super cool stuff. People have to check out the videos. We talked about a lot of things that you have expertise in. Of course, you're the world's leading expert in yourself, and that's what I want to talk about next, which is all about your journey.

The first step of the journey that's really interesting that people always want to hear about is you're a PM at Google. For some people, this is their terminal sort of job. This is the thing they always dreamed of and wanted, but you decided to make the leap to creation full-time. How and why?

That decision took me a year to make. It's funny; I was catching up with somebody who left Pixar and is doing a startup. I always ask this question, and it feels like for most people, it takes them six months to a year.

There were several reasons. Personally, I had a bunch of success, and the path ahead for me was to be a manager. I just started managing APMs, and it was kind of fun, but I hated managing, to be honest.

There are many reasons for that. I knew what I needed to do. I stopped learning new things. I could not find another thing within Google that excited me enough where I was like, "Hey, it's time to reinvent myself again. Let's go do this for another two years."

All the exciting stuff was happening externally in generative AI. In 2022, I started playing with MidJourney V2 as it came out. I started playing with all the Instant NGP, the Nerf research. At that point, we were working on a Nerf product to put into Google Maps' immersive view.

The indoor views and immersive view are radiance fields, so there was some connection there. But all this other stuff, like MidJourney and Stable Diffusion dropping, connected with the VFX creator part of me that wasn't getting as much of an itch.

There might have been an alternate world had I not decided to stay in Maps. I could have gone to YouTube and started doing a bunch of AR effect stuff. At that point, mapping the world as a mission is so beautiful.

I feel like it needs more generations of people to go in with energy to keep reimagining it. Don't always jump if you get an opportunity. There are times when sticking it out makes a difference.

Of course, your boy got promoted for sticking it out and pushing that stuff. But this VFX side became the shed aspect of my personality. I make some shorts on the weekend. Yes, adjacently, I'm learning about some AR stuff.

But the moment the generative stuff popped up, there was very little crossover with my day job. Suddenly, all these interesting things were happening between all the LLM stuff. I started playing with GPT-3 before ChatGPT even came out.

I had to try to write PRDs and stuff. It was kind of fun to see how bad it was back then compared to where it is now. The same thing with MidJourney. I made my first video using MidJourney and After Effects.

I thought, "Can I make a video that mixes some of the generative stuff with some of the stuff I usually do in visual effects?" I did the video in a fraction of the time, and it did great. It got a couple of million views.

Then I started making similar tweet threads on Twitter, where I was like, "Let me just try and pit bash some of these things together." What I found was, "Okay, do I want to go to this place where I could stick it out at Google?"

I have to still be on, as you know me. I described myself as a workaholic. I can't not be a workaholic and do the two jobs I was doing at the time. At the same time, I know the way to move upwards is basically playing the 48 Laws of Power.

There's organizational stuff I would have to do to get there. On the other hand, I look over at the playground, and everyone's just having fun with all this stuff. My channel had become big enough by this point where I was starting to get sponsored gigs.

Most of them Google didn't let me do. For example, I got hit up by Amazon, saying, "We're trying to advertise the Boys TV show. Can you make a TikTok video with visual effects? Maybe do something with airplanes?"

I was like, "Oh, this is my bread and butter." I went to create a gut ticket, and some person was like, "Oh, Amazon, we're too deep with them. You can't make this video."

But I did get to do stuff for Autodesk. I think you saw that video, the sponsored video I did with Autodesk. That was a lightbulb moment for me because, like many other people, by that point, I was feeling like, "Oh, you know what? I could stay at Google forever."

That was kind of the mindset when that gig happened. That was maybe one or two months into me thinking about what to do next, with a year taking me to actually decide to leave. I was like, "Wait a second. Autodesk is paying me to fly me to check out cool stuff, and they're also paying me to make this video about my life journey using their tool."

That job kind of thing started happening. That's when I started talking to a bunch of other creators in the space. I talked to somebody who's in an adjacent space to you, my buddy Rom Nan. Shout out to Rom for hooking me up with him.

I talked to Nick Singh, and he was just like, "Dude, you are framing this up." At that point, my plan was to keep building out this creator thing on the side. Maybe I build out a little side team. I figure out how to keep minimal conflict of interest.

I'll keep the Billy FX brand that's external-facing for creators. There's Belaval to do the personal stuff, but that's all Google. I didn't want to even have a personal brand at that time.

Talking to folks like Nick and a bunch of other AR creators, like Da Allen, who was just on the Meta stage doing a bunch of stuff, I realized, "A, it's possible to do this as an indie. B, there are so few people who are actual creative technologists in this field that are doing it. C, I could go back to Big Tech whenever I want."

I talked to some people who were working on AI efforts at Big Tech to keep it generic. They were like, "Dude, I don't even know. I'm scared to leave to do YC because I don't even know if there will be a job."

Again, what are you optimizing for? Is it for the job, or is it for the thing that gives you spark and passion? I remember distinctly going to CES last year. I met up with a YouTube creator I hadn't seen in a while. He has, I don't know, eight million subs or something.

After that convo, all these other convos, I knew what the next year was going to be. We had our roadmap locked in. Something in me just wanted to take the leap.

It's funny; my manager at the time was super nice. She was like, "Yo, dog, you have a lot of PTO. Do you want to take a month off and just try it out? See if you want to do it?"

I was like, "You're right." Maybe this goes to why I had a month of PTO left over. You're at Google; isn't everyone resting and investing? Contrary to popular belief, there are a lot of people hustling.

What percentage of people? I'll let the other folks debate that, especially given the recent Stanford study about 10% of engineers being ghost engineers. But that month was amazing for me because it was the first time my default network of the week was off.

I wasn't in any meetings, and that's when I made some of the craziest things. I don't know if you ever saw me reskinning my mom's living room in a 3D thing. I made this 3D scan in my parents' living room, and then I used ControlNet to restyle it in different styles of decor.

It was just such a simple thing I cranked out in an evening. The next day, I packaged it up, and it blew up. That was my first 1.9 or 2K likes on Twitter. I had like 2,000 followers back then, so I was small.

That gave me some content market fit and confidence. After I lined up a few brand gigs that would cover over, I didn't want to end up in a situation where I leave Google and then I'm not making any revenue that year.

I had visibility into those things, so I dipped, and it was the best decision ever. Maybe there's a world in which I should have left a year sooner. I think I could have been much further along, but then again, hindsight is always 20/20.

I got to ship cool things during that year that I wouldn't have otherwise. That's what my transition was like. If there's something people can take out of this, it's that it wasn't easy. It was hard.

One of the things that's really important to do is be okay with not having it all figured out. Be okay with making yourself vulnerable and being upfront with people. For example, I told my manager at the time, and they immediately assumed I was going to Meta.

That was their default reaction for some reason. Meta does some funky things too, where at one point, they'll email you on your corporate ID. I was like, "What the hell? We talked. You've got my Gmail."

The next day, I got a ping from my VP out of the blue, saying, "Hey, I just want to say we really appreciate the work that y'all are doing." I was like, "Maybe that was a coincidence, but I don't think so."

Putting yourself out there, saying, "I'm not going for a competitor. I'm going to do this creation thing. I'm deeply passionate about it," changed that dynamic too. I was able to talk to so many other people far more openly, even some of the folks that were a little annoyed that I was leaving.

We were kind of on this journey together, and it just made for a seamless transition. The fact that I had been dabbling in the creation stuff for many years prior, and specifically in the AI stuff the year prior, gave me a really good on-ramp.

I already knew Lonus, Matt Wolf, Heather Cooper, and a bunch of these creators by then. That gave me a circle to transition to, and then we all helped each other out as we were doing our versions of this creator journey.

Yeah, that was my very long-winded answer about deciding to leave Google. Maybe this is probably the most in-depth version of that story that I've shared thus far.

How you're doing the creator journey feels a bit more under the covers. It reminds me of Sahil Bloom, where people may not realize how many different agencies he spun up and how many different things he has going.

I feel like you aren't spinning the plates like everybody else. What are you focused on as a creator? You have a YouTube account with an enviable number of subscribers. I don't believe you have uploaded a video recently. What are you focused on as a creator, and how does that play out in terms of the revenue you're earning?

That's a great question, and this is the question I'm going into this winter with. At the end of last year, I wrote a strategy doc for myself, like every PM would. Here's my strategy doc for what I'm going to do this year.

Basically, I've got three dogs in me to set things up for people. There's the product builder in me that wants to build the machine inside the machine. I don't want to just use the machine; I want to build the machine with the people who are worthy and capable of building it.

There's the creator in me, which is a lot of what you see on Twitter and LinkedIn. It's like, "Hey, I'm in the sandbox with Lego pieces. You can put them together in this cool combination. Hey, check this out."

I'm not trying to be James Cameron, and I'm not trying to make a cinematic masterpiece. I'm trying to put together things that otherwise people would not and essentially learn some technology and share my interest.

Interestingly, there's the evangelist. This is one I've been flexing a lot more with my collaboration with TED. I'm doing the podcast, and I just signed a deal with them to be their technology curator.

I'll be curating a bunch of their AI and tech talks at TED 2025 and TED Next, which is really exciting. The goal for that is to put the "T" back into TED. Let's not just be the place where you go and a famous person announces their next book.

We want the crazy PR of mystery, six-sense type talks, where crazy people come up with crazy ideas, and that's the stage to show it on. Evangelism is more of a "God, know." As a creator, you always go into plug mode, but as an evangelist, you're connecting the past to the present and the present to the future or the unknown, helping people come along for the ride.

Over the past year, I've really flexed my personality numbers two and three: the creator and the evangelist. I haven't done as much product stuff. It's interesting; every year, the same thing happens at the end of the year.

When the headcount comes out at various places, people hit you up, saying, "Do you want to just be an influencer, or do you want to be an operator building the future?" I always think about that. Is there another way for me to scratch that product itch without doing a startup and without working at a big tech company?

I don't have the answer yet, but that's one of my goals for next year. The way I've dabbled in that is by scouting for A16Z, which I love doing, and angel investing in some startups. I got preed on Pika, Hedra, Backbone, and a couple of other really cool VFX companies that do amazing iPhone stuff.

A couple of other things are coming out soon, including a defense tech one. I can live vicariously through the founder, but it's only so much you can do. Living vicariously through, "Oh, you got on a call once a month with the founder, and you learned some things," is not quite the right way to do it.

What else is there? The closest thing I found was as part of my monetization stack as a creator. I taught two cohorts of a cohort-based course, like every other AI creator on Maven.

I had an interesting takeaway from that that I intend to double down on next year. I did two cohorts of this generative AI masterclass. The first cohort was a lot of creative directors, Hollywood people, and visual artists who wanted to click by click on everything.

"Hey, let's get into Comfy UI and figure out how to do this in office hours." The second cohort ended up being a product director from YouTube who just moved to the AI team, an engineering director from Netflix, and a group PM from Meta.

All of those folks came to it with a mindset of, "Hey, I want to get savvy about the tech primitives at my disposal so that when all their PMs are advocating to put stuff on the roadmap, they have a conceptual framework for how to think about the opportunities with regards to audio-visual 3D media in generative AI."

That was a lot more fun. I really enjoyed doing that, and it scratched that strategy itch in a more async way. I'm thinking of building a brand that is distinct from myself, not Belaval.

The working title is "Spatial Intelligence." I need to make it better. Something that encapsulates the future of creation and computing. I want to create deeper content that nobody would want to read otherwise but is very hard to find in general.

There are so few people that have actually done work on these intersections and actually shipped. This ends up being the thing that comes up again and again with the brands I've worked with and the founders I work with.

They're like, "Yo, you're a creator, but you can level with me as a product builder." Sometimes filmmakers are using these tools or giving advice that isn't anchored in the practical realities of building a real product or even thinking about what that means for your growth mechanism, which obviously you know a lot about.

I love doing the creator mad science experiment thing. I'm going to keep that going. That segues into a bunch of brand gigs, which I'll keep going. I've started doing speaking recently, which is great.

The TED curation deal I'm excited about. Twitter and LinkedIn are great, but I want to focus on this other side. It's funny; you mentioned this YouTube channel, which is true. My two biggest channels, Billy FX, have almost a million on TikTok and 400,000 on YouTube.

They still keep getting half a million views. That's been one reason I haven't touched it. I keep asking myself if there's something else I should do with it. Is there a world IP? I've occasionally had folks hit me up from the agency world who are trying to build, I hate to say it, but the Pixar of AI or the AI studio of AI.

Whatever the meme is, it's like, "Yo, man, why don't we do what you're doing, but we'll put a rapper around it? It'll be like the back rooms." I've thought about something like that.

Instead of the standalone shorts that you were talking about being in isolation, building this world together is interesting from a tech futurism standpoint. Pivoting it a little more towards the pros and cons of essentially building the panopticon.

All this mapping tech is the panopticon, and I think there's a good balance. If tech is mediating all interactions in the digital and physical world, we need to equally share optimistic views but also the cautionary tales.

There's too much cautionary stuff going on, so I do think about how to put that together. Right now, that's a P2, and that's my fourth bucket that I may not get to unless I have some grand enlightenment over the holidays.

I feel like you're prioritizing your life so much better than the average creator. I think anybody in the world, if they were sitting on a 1 million TikTok account and a 400,000 YouTube account, would be trying to pull all the money out of it.

So how does your revenue break down today? You mentioned speaking, brand deals, scouting, and the podcast. How does it break down?

I would say if I stack the pyramid, the base of it is definitely speaking and brand gigs. That's the predominant way for creators to monetize. The next level up is the deals with TED, these long-term engagements.

The next level up from that is ad hoc stuff that I might do, like teaching a course or passive revenue from YouTube, Twitter, LinkedIn, and all that type of stuff. That's basically how I think about the stack.

Honestly, it's not complicated. The thing is, it's so easy for creators to go and get brand gigs if they wanted to. You don't have to do it in two ways. The ones I do are where there's tremendous brand alignment, and the price is at a point where it makes my money's worth.

I just need to put a week or two into it. For example, doing a Lenovo or Qualcomm engagement is a lot more satisfying for me right now than getting paid a couple grand to do a mid-roll spot in a newsletter.

If I were in a situation like you or some of my other creator friends, this is what I want to build out with spatial intelligence. You have a regular edition that goes out every week, and you've got a lot more readership.

Stacking enough of those annual deals with brands can be a very lucrative business. I'm learning a lot from Rowan Chang, who's doing his newspaper. He's just crushing it.

He focused to the point where everyone's like, "Yo, we're going to do podcasts." Like you're saying, "I'm focused." There are people way more focused than me. It's like, "Dude, I'm going to go do a podcast. I'll teach this course on the side. I'm going to do a bunch of brand gigs. I'm also trying to be a creator in residence."

Oh, by the way, I did that for six months at YouTube last year. That was a lot of fun to do. These types of more opportunistic engagements, but Roland was like, "I'm just going to scale the newsletter."

I think there's a lot to that. For me, the last year has been Twitter and LinkedIn. That's the platform I put the most energy into. What I'm hearing now from people, and again, I could be wrong, I'm updating my map of the world, is that people are saying, "Dude, no, your Substack should be way more valuable. You need to be building it out way more."

So I'm trying.

To think about how to kill, create that like Community Hub that can kill a couple birds with one stone.

And yeah man, like the money's getting to the point where like, why ever do a traditional job unless you really get some super crazy scope?

And um, yeah, barring getting an opportunity where somebody's like, I don't know, literally he like, "Yo, we need you to run like meta reality lab" or something like that, I don't think there's anything that would excite me enough to go root for one team and just keep doing that.

But I got a lot more to do on the monetization stack, let me tell you that.

It's like I'm just really happy and grateful that thanks to X and LinkedIn, I've had a bunch of organic inbound that's just managed to make things work, you know, again and again.

Yeah, so 2024 is almost in the books.

If you were to compare that to year five at Google, which I think over those five years the Google stock something like tripled, so your initial equity grants were pretty large golden handcuffs.

How is the comparison?

Yeah, yeah, well let me tell you this also: I got a pretty fat retention offer in 20120, which, uh, so my comp was great.

I would say like I don't think I'm quite at that comp.

I'm not going to hit that comp this year, maybe like 200k off.

I think next year I could definitely do it.

And this is the other interesting thing: last year, you know, it's like I made like 150k.

You know, this year I'm G to make like 400k.

And so like I have no doubt the next year I could make 700, 800, 900k.

My goal is a million, but let's see if I get to that.

Amazing, thank you so much for all of the insights you've shared.

Personally, I have to say just after this conversation, I feel like I have a different perspective on passion and finding passion in your work.

So thanks for sharing a little bit of your passion with us.

Thank you for having me, it's a pleasure.

Thank you so much for listening.

If you found this valuable, you can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app.

Also, please consider giving us a rating or leaving a review, as that really helps other listeners find out about the podcast.

You can find all the past episodes or learn more about the show at product-growth.com.

See you in the next episode.