Please enjoy this transcript of my interview with David Baszucki (@DavidBaszucki), the founder and CEO of Roblox. TIME named Roblox one of the “100 Most Influential Companies,” and it has been recognized by Fast Company for innovation on their “Most Innovative Companies” and “Most Innovative Companies in Gaming” lists.

Previously, David founded Knowledge Revolution, where he and his brother Greg created Interactive Physics, a leader in educational physics and mechanical-design-simulation software.

Transcripts may contain a few typos. With many episodes lasting 2+ hours, it can be difficult to catch minor errors. Enjoy!

Additional podcast platforms

Listen to this episode on Apple Podcasts, Spotify, Overcast, Podcast Addict, Pocket Casts, Castbox, YouTube Music, Amazon Music, Audible, or on your favorite podcast platform. Watch the interview on YouTube.

DUE TO SOME HEADACHES IN THE PAST, PLEASE NOTE LEGAL CONDITIONS:

Tim Ferriss owns the copyright in and to all content in and transcripts of The Tim Ferriss Show podcast, with all rights reserved, as well as his right of publicity.

WHAT YOU’RE WELCOME TO DO: You are welcome to share the below transcript (up to 500 words but not more) in media articles (e.g., The New York Times, LA Times, The Guardian), on your personal website, in a non-commercial article or blog post (e.g., Medium), and/or on a personal social media account for non-commercial purposes, provided that you include attribution to “The Tim Ferriss Show” and link back to the tim.blog/podcast URL. For the sake of clarity, media outlets with advertising models are permitted to use excerpts from the transcript per the above.

WHAT IS NOT ALLOWED: No one is authorized to copy any portion of the podcast content or use Tim Ferriss’ name, image or likeness for any commercial purpose or use, including without limitation inclusion in any books, e-books, book summaries or synopses, or on a commercial website or social media site (e.g., Facebook, Twitter, Instagram, etc.) that offers or promotes your or another’s products or services. For the sake of clarity, media outlets are permitted to use photos of Tim Ferriss from the media room on tim.blog or (obviously) license photos of Tim Ferriss from Getty Images, etc.

Tim Ferriss: David, so nice to see you. Thank you for taking the time. I’m excited to have this chat with you.

David Baszucki: Hey, Tim, it’s great to be on the show, and when we started chatting before I came on, I had read one of your books literally 10 or 15 years ago and it got me inspired to do kettlebells and I did some this morning.

Tim Ferriss: And I saw photos of your beautiful kettlebell collection. Could you, just because now I can’t not take the bait, how did you jazz up your kettlebells that you ended up sending me a photograph of?

David Baszucki: So just to frame it, I’m pretty sure in your book you made your own travel portable kettlebell with some pipes that you could screw together, which I think — 

Tim Ferriss: From a plumbing shop or a hardware store, that’s right.

David Baszucki: Yeah. So I have five kettlebells and we use them a lot at my gym and have fun. They’re all made of iron and so we took them over to an auto place where they make low riders and do custom paint jobs of sparkle cherry red, sparkle orange, sparkle green. And so there are all these really beautiful automotive sparkle colors and it just makes them a lot more fun.

Tim Ferriss: And a lot of folks perhaps who are coming into this podcast will assume that we connected because of the amazing business and innovation story of Roblox, but that’s not actually how we connected.

David Baszucki: No.

Tim Ferriss: We connected because we have a friend in common, Dominic D’Agostino, some listeners may recognize as, effectively, Mr. Ketone, master of all things exogenous ketone related and an amazing scientist in his own right on a number of different levels.

And I had also had Chris Palmer of Harvard on the show a while back related to something called metabolic psychiatry, and your name and your Baszucki group kept coming up over and over again. And that is the thread that I pulled on, which ultimately connected the two of us. And I think I had mentioned your name, I’d invoked your name several times on the show including on The Random Show, and that’s how we got connected. So maybe as a way of just setting the table for a little bit of the metabolic health discussion, and everybody listening, we will get to Roblox and all of that of course. But this I think is something that will probably strike a chord with a lot of people listening on a lot of dimensions.

So as a way of setting the table, perhaps you could describe early in your son’s freshman year at college, what happened?

David Baszucki: Yeah, thanks Tim. And I’m going to share that my son and my family are comfortable sharing this story and so I feel I have some flexibility. But you could imagine as a parent of a high school student who had just started at UC Berkeley, all of the hopes and dreams of a parent of a student going off to school and a student that’s been very successful in math and science and academics and athletics and how much hope a parent has for that student going off to school. And like any other student, my son Matthew at his freshman year at Berkeley, it brought back memories of when I had started school. He hit it pretty hard. He was in computer science, he was rushing a fraternity, and there was a lot of demands I could see from afar it just seemed on him. The studies, the creativity, what a rush would be at a college like Berkeley and all of that.

And that was a time when he entered what I would call his first manic episode. We got some cryptic texts from him that were very alarming. We got some texts from some of his friends in his fraternity and dorm. I went to get him and he had entered what you would call a manic episode.

And for us, a manic episode is something that no parent has ever seen before, and is really something very surprising and weird and different. And he had entered this and what he had started on was really an eight or a nine-year journey with us over eight or nine years that involved some of the wildest stuff you could ever imagine. It involved him going to the hospital several times. It involved us not quite knowing how to really care for him. It involved trying to navigate the medical system. It involved going to Stanford and having him being locked up on the psych ward, and really started this journey that if we rounded out was only solved through ultimately getting him on a ketogenic diet.

Coincidentally, we ran into another CEO founder who said they got progress on their bipolar with a metabolic diet and a ketogenic diet, and that seemed like the craziest thing I had ever heard about. How is it possible after eight years and hospitalizations and very difficult times and complete disruption and I would say possibly concerned, would our son ever go back to school? Would he ever work? Would he ever integrate? We worked with Dr. Palmer and others and he tried a ketogenic diet, and literally within three weeks or four weeks, we saw progress that we had never seen with any drug or medication. Mind blown, really, and a miracle. And that was really the catalyst of starting our whole adventure down the ketogenic route.

Tim Ferriss: And if we flash forward then a bit to December of 2017, why was December so significant?

David Baszucki: We were now into a bit of a situation where our son had run away, had flushed all his meds, literally streaming on social media, streaming as he had run away. Had subsequently caught a bus, had made his way down to San Diego, had lived in, I think a lifeguard shack in San Diego. We had some monitoring of him. I knew he was full-blown manic at that time, and I tried my first try to come down and pluck him off the street with the help of some police and get him into a hospital.

Given the laws of our situation, I got to San Diego near some of our relatives, called in the cops, but he was pretty convincing. He said, “Hey, I’m free. I don’t want to go to the hospital.” I knew he was completely out of his mind and he ran away, and the police wouldn’t grab him, he just went running away from his dad. And so that was pretty scary.

Following that, we got a report that he had then hitchhiked to Los Angeles from San Diego. He had a phone and a laptop and that was it. And the communication started getting more sparse with him. In retrospect, it’s really scary because I think in retrospect, what I know now, I would’ve flown down there and hired 200 people, rented a hotel room and started searching all of Los Angeles, just go to every Starbucks in Los Angeles. Let’s find this guy. In the moment that I would say for one or two days, didn’t quite figure out what to do except we have a son who’s gone AWOL in Los Angeles.

This was a terrifying thing, a powerless kind of thing, here we were family with all of the resources in the world. And then by some miracle, he texted me from a Starbucks once again full-blown manic episode. And I was just able to work with him and to say, “Hey man, it’d be fun to buy us a latte together. You want to just chill out there for a couple hours? I’ll come down, we’ll buy a latte, it’ll be fun.”

And surprisingly. He said, “Sure, I’ll just chill out here.”

I’m like, “Oh, shit, he’s going to chill out there.” So flight, rental car, like SWAT team stuff.

I’m on the airplane, I’m sitting next to someone and they’re saying, “Hey, why are you going to L.A.?”

I said, “Oh, my son is AWOL with bipolar. I think I have an hour to get to him and to pick him up.”

So I pop into the rental car, just drive flying over there. I get to this Starbucks and there is my son, just a street person of your son. Nothing except a plastic Safeway bag with his laptop and a cell phone and a charger that he’s sleeping in a Starbucks with. And I’m just like, “Oh my gosh.”

And so now I’m thinking, okay, I can’t call the cops because I’ve been through this before. I cannot lose this guy. So my son and I got into this, I’m saying, “Hey, we should go see your relatives in San Diego. That’d be fun.”

He’s like, “Yeah, that’s a good idea. Let’s go see the relatives. You don’t mind if I get some smokes, man, do you?”

I said, “No, man, we need some smokes. Let’s get some smokes.” So we buy cigs, we buy Diet Cokes, we get in the rental car, he’s just smoking, completely manic and we’re going to see a relative.

Now I’m driving to San Diego texting 30 people at the same time, texting my wife, texting his uncle who’s a psychiatrist in Carlsbad. And during this 60-minute drive, I’ve got his uncle lined up as a hot stop so, “Yo, hey, we should go see Uncle Alex.”

“Oh, that’s a great idea. Let’s go see Uncle Alex.”

So I’m able to get Uncle Alex warmed up, and so we come in hot to Uncle Alex who’s a psychiatrist, and yeah, let’s just grab a dinner with Uncle Al. Good, so we’re rolling. So then we’re hanging out, and then we got to figure out how do we get my son into a hospital without the cops coming and having him run away? If he runs away, what are we going to do? So he had had a lot of adventures on the street and his hands were really beat up. I don’t know if he was punching a concrete wall or what he was doing, but his hands were really beat up.

So Uncle Alex, who’s pretty savvy, says, “Matt, man, we should go out, get a good steak dinner, what do you think?”

And Matt, thank goodness said, “Oh yeah, I’m really hungry. Let’s get a big steak dinner.”

And then I look at his hands, I say, “Hey, Matt, man, your hands are pretty beat up. We should just stop and get those checked out on the way to the steak dinner. Just, we’ll pull into the hospital, check out your hands.”

And Matt’s like, “Cool, let’s go do it.”

Oh my gosh. We’re texting, we got the hospital lined up and we pull in and just go into the hospital. They know we’re coming, go into this waiting room, but now we’ve got 30 minutes to keep Matt together in this waiting room so he doesn’t run away. “Let’s go have a smoke outside. Let’s do all of this stuff.”

Finally, after what seems like a lifetime, the doctor comes in, “Yeah, we’re ready.” The doctor’s preflighted just like, “Oh yeah, let me see your hands.” And this doctor, I forget their name, but she was a genius. She’s just like, “Matt, it just seems like maybe you want to take a rest for a day or two, just get off the street.” And wow, that was huge.

Tim Ferriss: And he was open to it.

David Baszucki: Yeah.

Tim Ferriss: Wow, man. Yeah.

David Baszucki: And he could have ran out of there. I think that started the journey of some, what is called insight in some level of insight where a bipolar person has a small inkling that things are not quite right and they want to participate in treatment. And this thing called insight is this very valuable thing that when someone does not have a sliver of it, they will run, they will sleep outside, they will not participate in the journey. That’s what led to ultimately many drugs, many interventions, and finally us finding ketogenic therapy.

Tim Ferriss: Yeah, thank you for the context and the story. I actually read something that your wife wrote or shared on metabolicmind.org and did not know this particular episode, this chapter in your family’s history. And it struck me for a couple of different reasons and I won’t read it all, but if you’d indulge me for a second, I mean, I’ll just read a little portion of this, which complements what you already described, but this is from your wife.

“At 4 a.m. the Friday before Christmas, I lay curled up and crying on my bedroom floor, convinced my son was no longer alive.

“Matt, then 21, had been a star in elementary and high school, but he began experiencing insomnia and panic attacks. After a manic episode led to a hospitalization at age 19, he was diagnosed with bipolar disorder. Two further hospitalizations and 10 different medications failed to provide stability.

“That December of 2017, escalating into mania, Matt had left home, cut off contact, and wandered the streets, eventually taking a bus from the Bay Area to Southern California. From his increasingly alarming Snapchat and Instagram posts, we knew he had no money, and that he’d slept one night behind a dumpster and another in a lifeguard tower. As dawn broke that Friday morning, his social media channels had gone silent and his three sisters hadn’t heard from him in 18 hours.”

This makes me quite emotional as well because I’ve had two or three friends basically follow this exact same pattern with bipolar, and I can only imagine the effect that would have on a family. And for people who also hear you mention metabolic psychiatry or ketogenic therapy or diet, and it might sound like some type of hand wavy panacea because they don’t understand perhaps some of the plausible mechanisms, what is actually happening? Why did this intervention help your son when so many other things had failed, or at least what is your best understanding of that? 

David Baszucki: I want to take a step back, and your audience may be familiar with what a ketogenic diet is, but in the big picture, almost all of us live our lives day-to-day burning glucose, and we have monitors. Too much glucose can lead to insulin resistance and diabetes. And society I think is just learning about glucose and how it brings energy to the body.

Now, interestingly enough, the body has a second way of generating energy called ketones, and most people never touch their ketones. It’s a more primitive way of generating energy. It’s a way of generating energy that we all go into if we don’t eat for a day or two because there’s no more glucose, we go into ketosis. But it’s also a way of generating energy that people who live far up in the north and eat seal blubber all day long aren’t getting any glucose, they’re not getting any carbs. And so they eat that seal blubber and they go into this thing called ketosis.

It’s arguably something that is closer to the way we lived as primitive people than the way we live today. We arguably live today with more carbohydrates in our diet because about 10,000 years ago, we had this thing called the agricultural revolution and the agricultural revolution in this amazing thing. Humans figured out they could generate a lot of food that was support a lot of people. It was an amazing invention, but it was more carbs than what the traditional diet has. Being in ketosis comes from fasting or eating very low carbs and more fat than we would be used to. What happens when someone is in ketosis is you run on ketones, this alternate energy pathway, and you get very consistent energy and very clear energy, especially to your brain.

The thesis would be, I guess my thesis would be, there are a lot of people around, arguably interestingly enough, with bipolar people who have pretty big brains, people who are trying to process a lot, and if they’re not getting consistent energy to their brain, which could happen from glucose spike, glucose crash, glucose spike, glucose crash, one might argue that what people see in bipolar is actually just a little bit of a symptom of not having enough brain energy to their brain. So keto, what we did with Matt and people have been exploring with keto diets for a long time for epilepsy. Also, I like to think when those Aboriginal people would run 300 miles in America a thousand years ago with a bag of pemmican, that was pretty keto food. That was high fat, low-carb food and allows people to go pretty far without crashing.

We put Matt on a keto diet and we worked with a dietician and Dr. Palmer, and this is a diet less than 20 grams of carbs per day, which is almost unfathomable for modern people, right? Because a quarter of a Coke probably has 20 grams of carbs. And also a diet that probably had more fat than protein, which is also something that we’re not used to because I think the last 50 to 100 years, we’ve migrated to less fats in our diet as well, somewhat. So we migrated Matt to what is called a ketogenic diet. And after 20 plus meds and treatments and all of it, we started and he started to see results from this diet, which was an absolute miracle.

Tim Ferriss: How do you make it as easy as possible to follow a ketogenic diet? And I say that as someone who’s spent probably upwards of a year in nutritional ketosis, and I always see the benefits, but I typically come off of it at some point because I find it difficult for compliance depending on travel and various things. But still, for instance, over the next month, I’ll probably spend two to three weeks in nutritional ketosis, and there are a lot of reasons to do that. People should also listen to Dominic D’Agostino or Chris Palmer, but activates anti-cancer pathways. It directly, quite aside from the anti-inflammatory effects and just the ability to starve certain types of cancerous cells of glucose, there’s so many upstream benefits. You develop this mitochondrial and metabolic machinery that has some durability if you’re in ketosis even for say four to six weeks, something like that. How do you make it as easy as possible?

David Baszucki: So this is, it’s hard, right? And another Matt story that we shared publicly is as we learned about ketosis and how you monitor your diet, and at this point Matt had been on eating primarily food that had been measured made by a cook. We know all of the ingredients so we could measure the carbs, the fat, and the protein. We went on a trip to Mexico as a family for a week, and we thought we had it right in the restaurant, but where we were wrong is possibly me, somewhere in the family, we forgot that avocados, even though they have a lot of fat in them, carry some carbs more than we expected. And so we were a couple of days into the Mexico trip, everything was cool. Matt was eating fish, olive oil, butter and some avocados. Those avocados had more carbs than we expected and nudged him out of this, for him, very strict ketone zone.

I think he arguably is good at a ketone level of two or 2.5. We weren’t measuring at the time, we didn’t have a ketone measuring device, and we can talk about that. And so all of a sudden here we are, day three or day four, and Matt’s starting to get some manic symptoms, more trouble sleeping, which is an early warning sign, a little more agitation and things are starting to get a bit edgy and we didn’t know what’s going on. And then we identified, oh my gosh, we’re a little off on the diet.

The next two days, Matt went to just literally small amounts of fish and butter, asking the chef, “Get more butter, more butter for my fish and more olive oil,” and he popped right back in and I just saw the correlation and how tight it was. So you’re exactly right. It’s very hard, especially if we’re going to restaurants or things like that. I am not as strict as Matt. I feel my life doesn’t quite depend on it, but I am also a huge advocate of some of the things you mentioned, sharpness of focus, just body weight control, consistency, energy in the afternoon, all of these good things that come from it. So when I go into a restaurant, the first thing that comes out in a restaurant is the big thing of bread. You have to push that aside, just get that out of my face.

You have to know how to order specifically. Yeah, I’ll take the burger without a bun and some extra mayo and maybe some extra butter and eat the burger with the mayo and the butter and maybe the lettuce and the tomato and all of that. But the bun and the fries are not fully keto aligned as well. Matt is even more strict. He’s gone pretty keto to the point of carnivore as well, which is very heavily animal type products. So I’d say it does take some practice, but I think once you get used to it and you can figure out how to maneuver in a restaurant, I think you’re in pretty good shape. The hardest thing for me would be, say we were to go to a wedding or something and you’re not fully controlling the menu and you’ve got all these friends around you and it’s all of this special food. Sometime in that case it’s kind of hard to do that unless you just don’t eat anything and sometimes you have to figure that out.

Tim Ferriss: And I’ll just speak personally for people who might be curious. I mean, what I’ve ultimately found easiest and Dominic, Dom has some good recommendations related to this because he’s spent a hell of a lot of time in nutritional ketosis, but basically two huge salads a day, intermittent fasting, so I just have two meals somewhere between 2 p.m. and 8 p.m. I have two huge salads, lots of olive oil, with a sliced ribeye on top of it, and it does the trick. You have to be a little careful, like you mentioned, in terms of protein fat ratios. If you consume a ton of protein at one sitting, you can knock yourself out because the liver, through gluconeogenesis, converts all those lovely amino acids or some of them right back into glucose.

David Baszucki: My son knows that and I figured that out as well. Protein is not a complete free lunch. You will pick up that glucose effect.

Tim Ferriss: Yeah, yeah, yeah, totally. Where have you landed in terms of measuring ketone levels? So for people who are wondering, I’ve used the Precision Xtra device, but then you end up having to do a million finger pricks, which is fine. Eventually you start to feel the level. I’m sure Matt, at this point, can tell when he’s at two millimolars or 2.5 millimolars, which is pretty high for people who are wondering, it’s actually pretty challenging for me to get there unless I’m fasting, so he must be very tight on the controls to get to two to 2.5. But once I hit, I know what it feels like on a Precision Xtra for me to get to 1.2, 1.3 millimolars. In terms of mental sharpness, I can feel the click over to that mode. What do you use for measurement now, or what would you recommend people use?

David Baszucki: Yeah, and we should come back to knowing what it feels like because I do feel, and this might sound a little wacky, I can either do an optimism simulation or a very minor, obviously with all due respect to people who are clinically depressed, a minor depression simulator by how far I go into ketosis or how far I glucose crash and I feel I can set both of those moods a little with my diet. Once again, not obviously the level of depression of people who truly suffer it, but touch on the edges of that.

So I, like you, initially tried some of the finger prick stuff and what’s interesting with fingerprint stuff is best practice would be twice a day check your ketones, kind of stuff. For many of your audience, they’ve probably tried CGM, which are continuous glucose monitors now, which are a complement to ketosis, and those are things where you just slap it on the back of your arm, hook it up to your mobile phone, and you get a graph of your glucose level for two weeks that you can look at. That’s a gentle early sign of ketosis if your glucose is just not spiking, but it’s not the full picture. 

What has started coming out now, and it’s surprisingly not available in the United States, are continuous ketone monitors, CKMs, they’re available for sale in Canada. They’ll probably be approved in the next year in the US and, I’ve got to be honest, I’m in a smuggling ring bringing CKMs — 

Tim Ferriss: “It fell off the back of a truck.”

David Baszucki: — into the United States.

Tim Ferriss: I might need to join that WhatsApp group.

David Baszucki: Yeah, I’ll send you a CKM and, just like a CGM, you can watch your ketones 24/7 throughout the day and then really see where they’re at. They’re very close to probably what you would assume, right? You’re probably one, 1.5. Matt’s two, 2.5. 

I also find it hard to really pop up over in that one to 1.5 zone. I’ve got to really push it to get there, but there’s definitely a feeling, there is a feeling. For me the feeling is one of not irrational optimism, but a little bit of a calm optimism that we can do this and things are going to be okay. And I’m excited about challenges that, I would say when I’m glucose crashing, may seem completely untenable and challenges that seem completely untenable, this is impossible to solve. In a moderate level of ketosis, it’s like, “Hey man, it’s chill. You’ve got food, you’ve got shelter, you’re not going to die. You can solve this thing. Let’s go do it. It’s going to be exciting to solve this challenge.” And so I can feel that feeling.

Tim Ferriss: Yeah, I’ll just mention a couple of other things that people might find interesting. Having done a lot of ketosis since 1998 or so, that’s when I first started. I was actually doing a weird, for some people weird, approach to it, a variant called the cyclical ketogenic diet because I was training in athletics, so I would do about 18 hours of, after a glycogen depletion workout, of carbohydrate loading so that you could take advantage of insulin and so on for packing on a little muscle while you’re doing the ketogenic diet.

But the point of that is a few things that I’ve seen repeatedly and it’s N of one, but you do see some of this reported in the literature as well. Number one, I need less sleep. I, actually on average, I would say spend two to three hours less time in bed and I wake up feeling fully refreshed. I wake up and I am awake versus waking up and needing 60 minutes to get up to speed. That is a clear benefit. A weird one that people shouldn’t screw around with too much, but my breath hold times double at least.

David Baszucki: Can I ask, because this scares the crap out of me? What’s your breath hold time?

Tim Ferriss: Oh, my breath hold time. If I did it right now, my breath hold time would probably be 45 seconds. I have a really compromised left lung from being born prematurely. I have something analogous to emphysema in my left lung, but when I did, I’ll give you two examples of breath holds. So one was after doing 10, and these are not breathing exercises that put me at risk of a shallow water blackout, but breathing exercises for 10 to 15 breaths along the lines of Wim Hof and then doing a breath hold on the exhale when at about three millimolars in terms of ketone or BHB concentration. That was like two minutes, 50 seconds. So I went from basically 45 seconds to two minutes 50 and had a friend right next to me who is a witness to this, and I’ve done it many times since.

I don’t think extended breath holds are great for your brain. I did another experiment when I was on day nine of a 10-day water fast and I was probably around five, I want to say four to five millimolars. I was really deep and there’s a point at which you could argue the really high concentrations are perhaps not great for you, but I was on day nine of a 10-day water fast and I did a hard shell hypobaric oxygen treatment where you can get up to 2.4 atmospheres, 2.5 atmospheres of pressure. And I was doing that for other reasons, but I thought to myself, well, let’s see what we could do in this type of environment. So I did a bunch of, let’s just call it Wim Hof breathing, breath hold on an exhale and I stopped at nine minutes because I was just terrified myself. I thought I was going to cause an aneurysm or something. Felt totally fine.

David Baszucki: So heavy ketosis, arguably a little lower metabolic rate. If you’ve just gone nine or 10 days without food, you’re probably a little skinnier, like less stuffed.

Tim Ferriss: A little bit.

David Baszucki: And then you’re saying in that hyperbaric chamber you pushed two or two and a half times as much oxygen into your body? Damn.

Tim Ferriss: Yeah, nine minutes. And that was without feeling any impulse to breathe. I stopped because I was like, you know what? I’ve never gone anywhere close to this long without breathing and I’m not getting paid for this, so let me just stop and call nine minutes a good breath hold. Yeah.

David Baszucki: Can I ask what you think your average sleep is per night?

Tim Ferriss: Oh, my sleep is terrible. It’s really fractured. Sleep for my entire life has been a problem. So I would say that generally I’m spending probably nine hours in bed. I have a latency, so the onset to sleep I would say, and I wear an Oura ring, so I’m able to track some of this. You could debate maybe the accuracy of the algorithms and so on, but around, yeah — there you go. Showing me an aura rig. So I would say 47 minutes to an hour plus for me to get to sleep, that would be an average night. It could be much longer. I probably wake up two to three times per night. And overall, if you look at my sleep score, like the absolute maximum would be around 90. It’s much more frequently. This is on the kind of aura rating scale, but 90 would be an absolutely best night of sleep for me.

It is more typically in this 60 to 70 range. So sleep is a real problem more than almost anything else. I mean there are a few things that contribute to dramatically improved sleep quality. One is zero caffeine, not a little caffeine, zero caffeine, even at very small quantities, I find that my sleep is interrupted. That is a pretty tough pill to swallow or not swallow for me, to be frank. But in addition to exercise and all the levers you would expect, sun exposure, a huge one for me for improving sleep quality, if I get at least 45 minutes of sun exposure. But ketosis just does a lot of heavy lifting on the sleep side of things. And it’s not possible for me to isolate variables here, but I have clinically diagnosed OCD and a lot of ruminative looping. When I am in ketosis, the volume on that goes from a 10 to a two.

So although I haven’t tracked it with an Oura ring, I would suspect my sleep latency is dramatically improved, right? I’m falling asleep a lot faster because my brain isn’t basically saying, oh, finally I have been waiting all day to tell you so much. It’s not that kind of situation quite as much. And then there’s a lot of other things going on, and I would say to folks, not to devalue therapy because I have therapists, I engage with therapy, but if you have some of these fundamental physiological issues, let’s just say with fuel utilization, talk therapy in and of itself is probably not going to fix those things. 

And whether that’s looking at, for instance, I have three relatives with Alzheimer’s disease right now, and I’ve done some experiments with providing them with exogenous ketones, so supplemental liquid ketones in this case that they can drink. And if I give them 25 to 30 milliliters of BHB bonded to 1,3-butanediol, there are some real concerns around 1,3-butanediol. Just to make it clear, I think there could be some real liver toxicity from extended use, but putting that aside for the moment, give this to a relative with Alzheimer’s and within 20 to 30 minutes, longer sentences, verbal acuity, noticeably, very noticeably improved. They’re telling stories instead of giving one answer, one word answers.

David Baszucki: Totally.

Tim Ferriss: It’s incredibly noticeable. And for instance, I slept like dog shit last night. Not to get too technical. But I just had a ketone salt mix prior to this conversation and it’s like I can feel it now. I can tell when the light switches come on. And there’s a reason Alzheimer’s is sometimes referred to as type three diabetes. And furthermore, I mean I don’t want to dox this scientist, you might know who I’m talking to, but I don’t think he’s been public about this, where there’s also — there might be some explanatory power in various types of infections as catalysts for certain types of what we would term psychiatric disorders.

And if your glucose metabolism is compromised in some way— And people listening, we’re not going to talk about ketosis the whole time, but honestly, if this is the only thing you take from this conversation, I think for a lot of you it will be well worth it — I’ve had Lyme disease twice and there’s a bunch of ridiculous woo-woo nonsense around Lyme disease and quite a few infectious diseases. Not everyone has Lyme disease. You might just be depressed. There’s a lot of overlap with symptoms for various syndromes. But growing up on Long Island, I had two absolutely verified cases of Lyme disease and other coinfections. Then you take a bunch of antibiotics and you do kill that infection despite what people might say. Nonetheless, you might experience what people call long Lyme disease, like long and what, in retrospect, I’ve realized that a lot of my longer duration symptoms, I think were probably neuro-inflammation, probably microglia, but who knows?

There could be other aspects to it. What solved my symptoms, that came after Lyme disease, was three weeks of strict ketosis, very strict ketosis. And I formed a bunch of pet theories or hypotheses as to why that might be the case because this was 10 years ago, but only recently — well, let me back up. Had at least four, maybe five friends or their wives who had actual proper documented cases of Lyme, a hundred percent success rate of getting rid of their cognitive symptoms and joint pain with strict ketosis. I’m not saying it’s a cure all, but it’s four for four or five for five at this point.

David Baszucki: That’s amazing.

Tim Ferriss: Yeah, long-winded way of saying, it’s such an accessible intervention, obviously do it with doctor supervision and neither you or I play a doctor on the internet. But man, it’s right there. It’s right there in front of you and accessible.

David Baszucki: I feel the same thing about sleep. And I feel 10 years ago I used to fear insomnia because that could mean a bad day, but I’ve never correlated this with ketosis. But I now know I can have three nights of what might be considered bad sleep, but I’m not as afraid of it. And now if I wake up at 2:00 a.m. I just listen to some interesting thing I want to listen to, a book or a podcast or something, and I fall back to sleep and I actually feel it’s just like a free learning period rather than something to be afraid of. And I no longer have those days, maybe 10 or 15 years ago where I’d just wake up and just go, it’s going to be a terrible day. I’m exhausted.

And then on the talk therapy thing, the way I sometimes think about this is, it’s kind of interesting that we first go to talk therapy rather than what I would call mechanical therapy. Mechanical therapy is what’s up with the machinery in your brain? Your brain is a machine. And talk therapy, with all due respect, in many cases, very, very valuable. But in many cases, if the machinery is not functioning, not getting enough energy, has a core at the molecular level thing going on, talk therapy is not going to do anything. And so I think that the physiological mechanical should always be the first place to go and in many cases, boom, that can take care of quite a bit.

Tim Ferriss: Yeah, yeah, yeah, absolutely. And I would actually go further because there have been many periods in my life where I’ve, I don’t want to say ignored, but for periods of time, developed a blind spot related to the mechanical therapy. Looking at examining the engine and oil levels and so on, instead of just trying to improve my driving technique. 

And I would go further and say that there are times when, and keeping in mind people, please, I do talk therapy regularly, so I’m not knocking it. I view it as necessary but not sufficient. It can actually put you into a precarious position because if you’re doing a lot of talk therapy but making no progress, you can develop almost a learned helplessness and the self-flagellating can get worse where you’re like, well, wait a second, I’m working with this great therapist. They’re giving me all the tools. Why am I of such a fucking failure? I thought I was so smart. Why can’t I figure this out?

And you can end up in this very dangerous, in some cases, situation where you feel like you cannot fix things because you’re unable to use the talk therapy to fix whatever the underlying issue is. Whereas in reality, it could be a purely physiological issue. Or you at least need that base level of physiological support before you can begin to do some of the higher level functioning and reorienting.

Let’s shift gears. You mentioned listening. This is going to sound like it’s out of left field, but it’s sitting right in front of me. So I might as well ask, this is actually from Lifehacker back in the day. Lifehacker.com. And I want you to tell me if this is still true. “I tend to work well,” this is quoting you, “whenever I’m listening to pure ambient noise, like those you’d find on the Atmospheric Calm playlist on Spotify. It’s my go-to playlist if I need a quick boost in creativity or productivity.” Do you still have any playlists that you listen to of that type?

David Baszucki: I still listen to Atmospheric Calm on Spotify. I am unable to function with anything other than ambient noise in the background. And so ambient means no lyrics, no words, no people saying stuff. For me, if I want some chill music, it’s got to be spa music, ambient music. I can’t get distracted by that.

Tim Ferriss: All right, I’m going to try that playlist because you and I are the same in that way. If there are lyrics, forget about it. It’s just generally not going to work for me.  

Let’s talk about Roblox. So there are a million different entry points here. The most obvious one would be to say, David, tell me the genesis story of Roblox. And I do want to hear the genesis story, but I might want to start with one that’s also sitting right in front of me. And maybe we could just start with this. This is under the heading of the future. Why don’t you give people an overview of what Roblox is for those people who have no idea whatsoever? And then the future I have here procedurally generated real-time worlds, aka dreaming in real time, that’s just too attractive for me not to leap into. So what is Roblox? And you could give some stats and figures if that helps give people an idea of the scale and scope of this. And then could you elaborate on the future as I teed it up?

David Baszucki: So hey, the out of the box big picture thing here is humans are just compelled to try to figure out ways to connect and communicate. And we didn’t used to have language and then we figured out language and we could sit around the campfire and communicate and tell stories. And then we tried to communicate at a distance and we had smoke signals or semaphores, and then we figured out writing and the mail system and we had the Pony Express and all of that, but we still wanted to communicate more at a distance. And then we had the telegraph system and then we had the telephone system and we use that a lot. And then we have text. And then in the midst of COVID, all of a sudden video, what we’re doing right now got to be more. And it’s just this core human thing of wanting to connect with people both real time and whatever.

And technically we’re not quite done. Technically, there’s going to be more. And it’s not necessarily dystopian. It could arguably very positive, whether it’s the holodeck we’ve seen on Star Trek or some of those things where maybe instead of a video call, I’m hanging out with my dad and we’re walking around ancient Rome together even though he’s in Carmel and I’m here. Or maybe he feels like he’s right in my office together. And so behind all of Roblox is what I feel is this unstoppable wave of technology that is going to happen. And we have graciously landed in this opportunity to usher it in, initially coming from what seems like a gaming platform, like people playing together, but arguably a platform that, if it’s done well and if it’s done safely and with civility and with scale, can be a very important, not just play, but working platform, communication platform, lonely kid with cancer in the hospital connection platform, lonely kid who’s having a hard time finding their people and finding them digitally platform and maybe even a way to experience music or political rallies even.

So the good news is, it’s good to be in a company with just a big thing happening behind it. The way Roblox presents today is you could think of it as a 3D gaming play platform with about 120 million people on it every day where all of the games, all of the creations, are made by people on the platform, whether it’s a 12-year-old hobbyist, whether it’s a team of 50 people making 10 million dollars a year, where it’s everything in between.

And where, through these user created experiences, about three percent of all the gaming in the world is now starting to happen on Roblox. And gaming is a pretty big market. What’s really beautiful about it is that we see emergent games like you would expect with user generated things where maybe we have a traditional view of what games are. But on Roblox, a top game is Dress to Impress, like a fashion game where you pick clothes out for five minutes and you compete in a fashion show or, what was recently hot, Grow a Garden, where your garden is always growing in the background and you’re tuning it and you’re trying to make it better.

So it’s really a fascinating, interesting journey. I think it started with a combination of great people and just a big vision. It’s an enormous responsibility because there’s probably nine billion hours of people on our platform every month. And at peak times there’s over 40 million people. 

And from day one, we’ve built this as a platform for all ages. So we have nine-year olds on the platform. All of their communication is filtered. They’re not able to share images, but they are able to go play hide and go seek. And we put enormous effort on safety and civility with all these things. And we actually, I think have done something very lucky, is unlike almost every other social platform, you name it, it’s 13 and up. We’ve accepted that we have young people on the platform from day one and really built the infrastructure around that rather than denying that. So it’s a really fun company to run, it’s in a really big interesting market. I think we’re going to see people doing virtual 3D work on the platform.

Tim Ferriss: What do you mean by 3D work? What would be an example of that?

David Baszucki: I just think as, over time, Roblox gets more and more photorealistic and more real-time rather than having a video call. What’s interesting about a video call with 20 people is we see 20 windows and only one or two people can talk at the same time because it gets all confusing. But in a 3D Roblox world, we’re all in the same space and we kind of hear us all at the same time, just like the real world. So I think over time some types of video calls will get replaced with 3D calls. We’ll see music concerts. If you’re not live, you’ll be there in a 3D holodeck type version with your friends dancing and seeing everything around you.

And I do think we’ll ultimately see political rallies where, in addition to the stadium of a hundred thousand people within the rules and the laws of the state where that political rally may be occurring, we may see both a video version, a physical version, and a 3D version where you can be there with your friends and go to that. So I really do think we’re at the start of just used to be the phone and now video is pretty big. And someday 3D is going to be pretty big too.

Tim Ferriss: Yeah.

I can’t remember. Alan Key? I might be getting the attribution wrong.

David Baszucki: Alan Kay?

Tim Ferriss: Kay. Alan Kay. Is that “The best way to predict the future is to create it”? 

David Baszucki: I think the future, in many ways, we sometimes don’t have the hubris to feel many things are inevitable. And so, one other way to predict the future I think is many things are just inevitable. There are enough smart people around that the wheel was inevitable and one could take credit. I invented the wheel, but that wheel was probably going to get invented by thousands of different people anyways. And I think we have a little bit of the same vision at Roblox. We are working on something that is inevitable. We are participating in building it, but I don’t think we would lay claim to being the inventors of it.

Tim Ferriss: Although I’m fascinated by the Incan Empire, it seems like they never landed on the wheel idea. It’s like what they were able to accomplish without the wheels is totally bananas. But that’s — go figure.

David Baszucki: Yeah, that is bananas. That is. But then again, if they would’ve survived, they might’ve come across it.

Tim Ferriss: For sure. Maybe they just ran out of runway. So let me ask you, since a lot of entrepreneurs are listening to this, and I’m sure even if they have not played Roblox, they have heard of Roblox and the numbers you’re providing our mind-boggling numbers in terms of the breadth and size of Roblox and where it’s going. You mentioned the creator community, and I’ve got some numbers in front of me, like Roblox creators earned more than one billion dollars in the past year, and this, in a sense, open development community seems to be key to growth. How early on did you figure that out? Did you try to do things internally for a while and then prototype it and then pour gasoline on it? Or was that just from the very first sort of nascent stages of this product of the company, part of the plan?

David Baszucki: I believe one has to always be innovating in a company like this. And part of building a company like Roblox is those innovations have to be happening year after year. One almost needs a system for innovation. We, along the way, with the way this economic system called the developer system, Roblox was always what is called a user-generated content platform, which means creators are making stuff, people are learning STEM, people are getting excited. Even the ego burst of having three friends play a game can really motivate a young person to get into computer science. And so, initially, you could say Roblox ran on the excitement of having friends see what you’re working on.

We initially had a much more primitive economic system, a club membership thing, like some much older virtual worlds from young people. And we had one of these moments that we’ve had so many in the company where one number is going great and one number is not going so great. And, at this point, we had one number going great, which is user growth and hours growth, but we had this other number going not great, which is revenue. And that was really — 

Tim Ferriss: Can you explain what the membership — what that looked like at the time just so I have an understanding?

David Baszucki: The membership, the early membership model was, if you subscribe for $5 a month, you get some cool stuff, you get some skins for your website, you get more places to build stuff so you could build more. But that was actually a very dangerous revenue model because there should never be any impediment to building and creating. That should just be unlimited and free. So we had arguably a primitive revenue model. We weren’t making building free. And it was like selling a little extra cool thing like extra chrome on your car or something. And that was getting stale and tired.

And what we learned in that is, sometimes, your intuitive, big picture, very difficult thing is the right solution, but we did what many other entrepreneurs would do, which is, like, “Oh, my gosh, we’ve got a revenue problem. We need to forensically diagnose this. Let’s look at the 50 things. What did we change? What did we break? Oh, we can’t find anything we broke. Okay, let’s spend three more months making a list of all the small little tweaks we can do to improve revenue. Oh, let’s stack rank them. Okay, there’s 50 things. Well, okay, let’s do the top 10 of those. Oh, my gosh, those aren’t working.” And then, in the back of our heads, we had been saying there’s one thing that’s really difficult, which is we need a digital economy, virtual currency. We need players to be able to buy Robux. We need them to be able to go into any game and use their Robux. We need the creators to sell things for Robux.

Whatever they could think of, we have to trust that these creators, in a pizza delivery game, they’re going to sell us a scooter for anyone who wants to deliver pizzas faster. We’re going to have to trust that, in Bird Simulator, they’re going to sell the ability to turn into an eagle faster than if you just play. And so, all of a sudden, okay, let’s go do the big strategic thing. This is going to take a while. Let’s just go, go, go, go.

Tim Ferriss: Were there any sources of inspiration who led you to consider the digital economy Robux route? Were there any antecedents or influences?

David Baszucki: You could say Adam Smith. Like, literally, the way our economy works, people have some currency. They buy things. When artists or creators make stuff, people decide what they want to buy. People are very creative in making that. We are literally inspired by the real-world economy, and so we said we have to have kind of a microcosm of the real-world economy. What was exciting is, even though we knew this was going to take two or three months to build, the second we committed, like, “Forget all these fixes, forget all of these little things, we’re going all in at what we think is the big strategic fix to this problem.” It was very relaxing and fun. It was just like we’re — 

Tim Ferriss: Yeah. I can see you getting visibly excited. I’m just wondering, with such a seemingly dramatic move, number one, am I over dramatizing the risks involved in doing it? Is it more like, hey, it seems big, and it is big, but it’s actually not that risky, number one? Number two is how did you decide internally to hit go on that decision?

David Baszucki: I think we were smaller. We had 20 people. We had spent three months trying to find what we broke. We’d spent three months trying the top 10 of the 50 stack rank things and we were like okay. But I think we had hope and conviction that, if we built this right, we would build essentially a system that, if Roblox grows, our economy will grow. And we’re not going to be day in day out trying to fix the economy. If we build this system, we can get back to making Roblox fun. We are always working on safety. We continue to focus on that, focus on those things, because we’re going to build a system that’s going to scale with the system. And, by the way, that did turn out to be true. If we typically double the number of users or the number of hours, we’ll generally double the revenue. The system just — 

Tim Ferriss: Got it. So, in this case, the numbers that were looking good, that you mentioned before, hours played, is now directly correlated with another number which looks good — 

David Baszucki: That’s right.

Tim Ferriss: — which is the revenue — 

David Baszucki: Revenue — 

Tim Ferriss: — as opposed to diverging because one is subscription per month or whatever the interval was. Got it. Okay.

David Baszucki: We wanted to get to a point where revenue equals K times hours. And that was the hope.

Tim Ferriss: What’s K?

David Baszucki: Just a constant. Pick a number.

Tim Ferriss: Oh, okay. Got it.

David Baszucki: Revenue equals some multiple times hours. So then, psychologically, what was exciting is, a few days into this, it’s like pretty relaxed. Right? We’re obviously risking wasting three months on this, and we’re risking that this won’t work. But the opposite side of it is, “Oh, my gosh, we really can imagine this thing working. And if this thing works, it’s really going to work. And this one thing is going to work really, really well.” And it’s a pretty complex feature. We needed digital currency. We needed users to be able to buy digital currency. We needed anyone’s game. In your hide-and-go-seek game, you needed a facility to be able to sell a flashlight for five Roblox. We needed the facility for you to take all of your Roblox and turn them into cash to support your living as creator. We needed a discovery component to see what are the interesting games that you can spend Robux in.

We needed developers to do all of this stuff, so we had to build all of this in parallel. But what was interesting is we had kind of some secret rocket engine behind it and that for the first time ever, as a complement to this feature, you or I could imagine making a living on Roblox because, before, we were hobbyists and, all of a sudden, whoa, if I could make five grand a month with my Roblox game, I might just work on that full time rather than an hour a day. So we felt there would be a secret afterburner here that people would work harder on their Roblox game. So, fast-forward, and we got within a day of shipping and we were, like, “Okay, it’s like we’re kind of hoping this works. I hope this works.” And then like, now, the dev community had heard about it. Everyone’s really excited about it. And the day we shipped it, I would say, within four hours, we knew it’s going to work just like boom.

Tim Ferriss: Was that a revenue metric? Was it just an adoption metric? How did you know it was going to work?

David Baszucki: It’s, oh, my gosh, of the top a hundred creators on Roblox, 22 of them already have Robux features. Oh, my gosh.

Tim Ferriss: Okay, got it, like integrated that quickly.

David Baszucki: Yeah. Oh, my gosh, a bunch of users have already bought Robux in the first four hours. Oh, my gosh, people are spending Roblox in these places. Oh, my gosh. And so, yeah, it was a little example of what we would call doing the hard thing and taking the long view. It’s a little bit like, when the strategic thing is right, everything else follows. And, metaphorically, it’s a little bit like, first and foremost, if you’re having a mental health crisis, work on your body and your machinery maybe before you work on the talk therapy. Like get the strategic things right first.

Tim Ferriss: And when you were about to ship to make this virtual economy a reality, this digital economy, prior to shipping, did you have a framework for how you would decide whether it was pass, fail? In other words, was it time-bound, like, “We’re going to follow X, Y, and Z metrics for four weeks. And if it’s not working, we’re going to roll back to the prior version?” What was the plan B. Or was there nothing like that?

David Baszucki: I feel, this one, we’ve got conviction. I think what actually happened is, in the back of our minds, we all really knew this was the right big way to go. It was just so much work that we constantly kicked the can down the road, tried to fix things, tried to do the small things, and then, finally, when we accepted reality, we got to do the right, big-picture thing, all of a sudden, it’s like, “Okay, we accept the reality. We got to do the strategic thing,” and then it was all good. And then I think we started believing this is going to work.

Tim Ferriss: All right, let me ask a question about this digital economy. So you talked about different creators making a living and many of them are doing much more than making a living on Roblox. Some of them are fantastically successful. And one of my employees sent me his two kids’ favorite games and then Dad’s favorites. One of them is a game you mentioned, Grow a Garden. And you also mentioned Dress to Impress, which shows up for one of them. There are a bunch of other ones.

Well, you know what, just for fun, there are a couple that show up a bunch. The Mimic, 99 Nights in the Forest, and there are many more. But I wanted to talk about Grow a Garden specifically, because Grow a Garden was also the game that came up when I was in a text thread with my friend Kevin Rose. And Kevin’s an amazing entrepreneur, a fantastically successful investor, and he kept sending me these videos, these screen captures or iPhone videos, I don’t know which, of playing Grow a Garden with his little girls. And he just loves the game. He’s got to be one of your biggest spenders, I would have to imagine, based on some of the conversations I’ve had with him.

And he had a question, and I would love to know how you think about this as sort of the Fed and the president and everything involved at the top of the pyramid for this digital economy. His observation was that, when something like Grow a Garden comes out, and I apologize that I haven’t participated in the ecosystem quite so I can’t speak to this, but, once something is really successful, a lot of copycats come out, a lot of clones sort of come out. How do you think about handling that type of situation where people might go to an imitator believing it to be the original, they spend money, et cetera, all of these possible complications, or maybe ensuring that creators are incentivized to put the time into developing their games. If they fear they might be cloned, how do you think of handling a lot of these issues which get handled offline as well, of course, with the USPTO. And you have drug development and certain types of rules and regulations around IP. How do you think about handling that in your digital economy?

David Baszucki: Yeah, so I would say, first, more and more, from an IP copyright standpoint, things that are typically IP and copyrighted like the name Grow a Garden and things like that, or avatars or things like that, the same protections exist on our platform that would exist in any other platform. What is trickier is barring a name or a trademark or a copyright, a form of game play that’s traditionally not been protected. I’m not a lawyer, so I might not be giving the exact right thing, but if I made an experience on Roblox called Water Your Plants a Lot and Create an Amazing Garden, and it kind of works like Grow a Garden, that’s a little hard to protect.

What we do find, and we work a lot on it, is people who make similar type experiences do try to draft on the main experience. They get very creative. We have the ability because, when people search for Grow a Garden or when they look for it and they type that in, we can see a lot about all of the games, the game that matches exactly like Grow a Garden. Or even if someone types in the word “garden”, we know pretty well that there’s one game on Roblox with the name “garden” in it that has 25 million people playing at the same time. And there’s a bunch of other games with the word “garden” in it that have 10 people playing, and so we can be pretty intelligent in really showing, like, “This is Grow a Garden even if you type ‘garden’, and here’s a few small other ones.” But we would not block Dave’s Try to Create a Garden game even if Grow a Garden were there.

Tim Ferriss: What are some of the other best decisions? I’m also going to ask biggest mistakes after this. I’m just going to plant that seed, because I often ask people what are their favorite failures or mistakes. In addition to the implementation of the digital economy, what have been some of the best decisions? Those could be design features. Those could be related to business model. They could be related to org chart of the company and how you thought about that. It could be really anything, but some of the best decisions that, since made, have contributed to the success of Roblox.

David Baszucki: I think one of the best decisions we’ve made is trying to optimize creator revenue over profits actually. And so, when I’m running Roblox and we have our CFO and our board and we have a lot of options and we’re making billions and billions of things flowing through the company, we end up with this really interesting decision where we’re trying to keep our employment cost as efficient as we can. We’re trying to keep the cost of running Roblox as efficient as we can. We’re trying to have really many interesting ways to efficiently purchase Roblox without that costing a lot. And we end up with two final places the cash can go. One is it can go back to the developers or the other is, quote, we can make profit.

And, time and time again, I think we’ve leaned a bit on the direction of “Let’s move back more to the creator community” rather than being a ridiculously profitable company. We do generate cash, and we put some cash in the bank. But, generally, we’re trying to create that creator community. And I think that goes hand in hand with, when we think about designing the product, we have groups that work on our simulation engine, and the user experience, and the economy, and trust and safety. Our economy team’s been pretty gracious in that their goal is to generate revenue. But the real primary goal is to make Roblox engaging and interesting and fun, not at the expense of revenue. And so our economy team has been very gracious in a sense saying, “Yeah, the primary goal is still user engagement, not making money.” So I think that’s been a good decision. We have — 

Tim Ferriss: Now, could you say more about that because people might have — 

David Baszucki: Well, you could imagine if you were — 

Tim Ferriss: — questions about that?

David Baszucki: If you were the head of Roblox economy and you said, “Hey, Dave, for my job, all I want to do is make the most money.” That seems like a logical thing for that economy team. What I would say is, “Well, we want to be careful. We like you, for every feature you build, to make sure that it’s neutral-to-positive on fun as well, and that, if you come up with a wacky feature that people get confused and just spend a lot of money and they’re not quite as happy, but we make more money, that would fail all of our metrics. And so, in a sense, we can make more money, but we really want to move all of the things. We want to move fun. We want to move how much people enjoy Roblox at the same time.

Tim Ferriss: Well, it also seems, tell me if I’m misreading this, like playing the long game in the sense that, if you want to build the largest company possible, if you want to have 10 percent of all gaming content on Roblox, if you want to go beyond that. 

David Baszucki: I would say, subject to constraints, actually, because what our mission is is connect a billion users with optimism and civility. And so there’s some pretty big guardrails around that in that we would not just take the billion daily users. We would take the billion daily users if the average user on our platform might come away with a higher level of civility than if they hadn’t even played. We are actually trying to teach civility at the same time we’re growing the company.

Tim Ferriss: And what I was going to say is not mutually exclusive with that. I was just going to say that you need the game developers to be happy for the long-term interest of everyone involved, including Roblox, right? So you don’t want to kill the golden goose by pulling out as much profit as possible.

David Baszucki: No. We think the more efficiently we run the company, the more of the money flowing through that flows to the developers. It’s a much better long game than just trying to be a hyper-profitable company.

Tim Ferriss: What are some of the missteps or mistakes along the way that have stood out for you?

David Baszucki: Generally, missteps happen either not taking the long view and sometimes trying to do too much rather than doing less better. I’ll share a classic mistake that’s really hard to get one’s head around. Maybe five to eight years ago, there’s a whole category of gaming that was around being in a clan and making points and having rankings and all of that. And those are features that are beautiful within a single game or beautiful within certain games. But, arguably, those are features that aren’t necessarily something we should have been building as a platform. We should have trusted the developers to, like, “You go build your game with that stuff,” rather than us saying, “We want to be in the gaming business rather than the platform business.” And so we spent a lot of time building out some of that kind of clan ranking functionality, clan point stuff, and we threw it all away because it was trying to do too much on the platform.

We’ve done many, many things well. One thing we did really well was Roblox was initially a PC/Macintosh company. And there was a time when people did not believe that 3D stuff should happen on a phone. Phone games were two-dimensional, more puzzle-type thing. And there was a technological leap that, just like with the iPhone, all of a sudden, websites that were used to being these big PC websites, with a pinch and a Zoom, would work on an iPhone, the exact same website. There’s no more of this separate mobile-web stuff. It was a huge innovation. And I think we did it right, thinking that this 3D immersive, metaverse, holodeck stuff will ultimately work on a phone as well. And most Roblox games run on a phone or a big screen at the same time. And that has really, I’d say, turned to pass and worked out really well for us.

Tim Ferriss: Why did you have confidence in that when others didn’t?

David Baszucki: It just seemed inevitable. I think the reason we felt it was inevitable is people were watching movies on their phone at that time. And movies are kind of three-dimensional. And games are starting to approach movies. It just seemed inevitable. But, you are right, that was a little bit of me and a couple other people in the company believed it. But that was a pivotal time when probably 90 percent of the company said, “No way is this one going to work,” but we took the risk on it.

Tim Ferriss: What was the risk?

David Baszucki: Distraction.

Tim Ferriss: Distraction?

David Baszucki: Yeah, like we’re going to spend all this time on this iPhone version of Roblox. But that same kind of stuff taking that big, long leap, the very — there was a really fun, very simple game called Survive the Natural Disaster on Roblox. And early on in the prototyping phase, I had a version running on a very old iPhone. And it was that same feeling as the virtual economy. It’s just, “This is going to work.” Boom. Now, iPhones and Androids are the biggest platforms for Roblox.

Tim Ferriss: Yeah. The inevitability part is worth underscoring. I’ll just share a quick story, which was in 2008 or 2009 when I first met Tobi, CEO of Shopify. This was when they had nine or 12 employees, something like that, and Harley as well. And I ended up becoming an advisor to the company, which was pretty good timing. But the point of the story is that the way I decided to look at that really seriously was, number one, I asked my fans how I should update my first book related to e-commerce specifically. And they mentioned Shopify over and over again, but I took a look at it.

And, to borrow from this wild man, hacker/investor named Pablos Holman, he often will look out and he says, “Okay. Well, what’s going to happen next year is pretty hard, maybe, to predict with any precision. But what might happen in 10 or 20 years can be a little bit easier.” If you ask yourself, “Will there be more or less e-commerce?” I think it’s pretty obvious, right? At that point, more. Will there be more phones? Will there be more broadband connectivity? Yes, and yes, like if you had to choose a yes or no. And pretty soon, when you start to add up a few of those, you come to the conclusion that this seems like an inevitability. It also seems inevitable that someone’s going to win in this space, so why not bet on the horse that my readers have recommended over and over again? And that was the decision process. There was more to it, of course. I mean, Tobi is a genius, and Harley is amazing, but — 

David Baszucki: But if it was Tobi from Shopify, you probably also had some intuition that this is the type of person that can make that happen.

Tim Ferriss: Oh, for sure, yeah. I mean, it can’t just be a good idea. I mean, the team was critically important. And Tobi is one of the best first principles thinkers and systems thinkers — 

David Baszucki: Tobi is a — 

Tim Ferriss: — I’ve ever met in my life.

David Baszucki: — very systems thinker, absolutely.

Tim Ferriss: Yeah. Yeah. But the inevitability part, I sometimes just like to point out to folks, like, yeah, I’ve done a lot of angel investing in this and that, and I’ve had some luck, but you can train yourself to ask some of these questions to narrow down the list of players you might consider just based on these inevitabilities. And if you can’t find an inevitability, it’s like, okay, maybe you just opt out of that particular playing field, whatever that might be. 

Let me ask a couple of questions from one of my employees. This is, again, the father who plays with his kids. We can do just a couple of these. Are there any games that maybe aren’t super popular, but you particularly like? Are there any games that are not sort of the greatest hits that a lot of people listening would know where you’re, like, “Yeah, I am actually a fan of this game or that game?” Any come to mind?

David Baszucki: Yeah. I think what I’m a fan of, and I’m hoping that happens more on Roblox, is more nichey content that super fans would have. So there’s a couple simulations. One is, and I forget the name, but it’s a simulation of a full airline company. And it’s a simulation, it seems unbelievable, but a simulation where you go to the airport, you buy tickets, you wait in the lounge, you get on the plane with everyone else, you take your seat, you go on a flight for an hour, you get served with a flight attendant, and everyone plays a role. You’re either a passenger. You’re a flight attendant. You’re a pilot. You’re an executive in the airline.

That kind of thing really blows my mind just because of the potential for amazing role playing. And, yeah, I think the potential for individuals to, “Hey, there’s a role for everyone.” There’s the first-class passengers. That’s a role. There’s also the baggage handler or whatever. I think that’s really fun. I also like model railroading. And I think I like all the railroad games on Roblox because that — there’s this hobby 30, 40, 50 years ago before we had computers that people would make all these model railroad sets in their Midwestern basements. And they don’t do that anymore. But I like the idea that that hobby can go digital on a platform like Roblox.

Tim Ferriss: How many games are there, roughly, on Roblox?

David Baszucki: The number is almost meaningless because it’s millions and millions and millions, and so what it turns into is more these crazy numbers. Like how many people make more than a million dollars? I think that’s pretty big. How many people can make a living. And that’s thousands and thousands that can make a living. How many make any money at all? We will validate that I’m giving you the right number, but that’s tens, if not hundreds, of thousands. And then how many are making cool stuff to show off with their friends? That’s in the millions. So there’s a lot.

Tim Ferriss: All right, this is another one, again, from my employee. What are things kids can do to be extra safe on Roblox or I guess, by extension that parents could do to ensure their kids are extra safe on Roblox?

David Baszucki: So this is a big one, and I think we are really working to keep kids on Roblox actually. And this might sound counterintuitive, but we’re really working to keep people in a place where text is filtered and we can monitor critical harms and where there’s no image sharing or video sharing. I would say, the backstory of the industry right now is there are a lot of 10-year-olds who have phones. And there’s a lot of 10-year-olds who can install software that is for 13-year-olds. And a lot of that software allows more open communication, unfiltered communication, and sharing pictures. We’ve all heard there’s a lot of things that go on on the internet that are really terrible. When people start sharing images, they can get blackmailed, they can start trying to meet someone in the real world.

So we’re working really hard to just stay on Roblox. And I would say, there are controls for parents as far as if you only want your kid to communicate with the people you pick, we will offer that for you. But I would say, we accept the responsibility and the challenge that not all kids are talking with their parents. And there are so many kids out there who when they’re 10, they get handed an iPhone and they go try a bunch of stuff, and we have to build Roblox in a way that is as safe as possible for those kids.

Tim Ferriss: Got it. What does the future hold, do you think, just for safety precautions or other technological innovations that might just help you to manage the entire system in that way to mitigate risk?

David Baszucki: I think we’re rapidly approaching a position where what is going on is in addition to all of our investigators, and in addition to all of that, the advancements in AI that we’re bringing forward on our platform. And some of the things we’ve committed to make this essentially not just a very, very, very, very, very good system, but a system that we almost take for granted. One thing that we have in the works right now, by the end of this year, using AI, using age estimation, using the camera on everyone’s phone, we’re going to know pretty well the age of everyone on our platform.

And in addition to filtering all text on our platform, and in addition to monitoring for critical harms, we’re going to start clustering people by ages just so unless you happen to know that person who’s a farther age away, we’re not going to let you communicate at all. So I think we’re going to zero in on that. So I do think, over time, both on the communication and the content side as well, we’re going to get to that well beyond very, very, very, very, very, very good type system.

Tim Ferriss: Yeah, how long have you been embracing, how long have you been an AI shop, I suppose is one way to put it? Roblox. How long have you had elements behind the scenes, exactly?

David Baszucki: Behind the scenes, Roblox is a hundreds of different AI model shop. We build them ourselves. We have built all of our own texts, safety models, all of our own voice safety models. A lot of the models we use for figuring out what is a cool game to play. We’ve built our own translation system that is starting to translate from one language to another. We have started to debut some of the more, I wouldn’t say more exciting, but more futuristic stuff, which is 3D creation by AI, so that even if you or I aren’t used to using 3D tools, we can talk about things and have those be created. And we have more coming.

As you said earlier, I think you use something like procedurally generated, real-time dreaming. We have our eyes on that. And I think beyond, I would say, oh, can I use AI to make a game? It’s interesting to think if someday, will AI literally create an evolving game as we’re walking around, almost as if you and I are in a dream world or the Holodeck talking about the Holodeck and just have that fill in around us?

Tim Ferriss: And that’s another thing that seems kind of inevitable, right? Why wouldn’t that happen? It seems like, I don’t know, you’re holding a finger up, so what do you think?

David Baszucki: I think you’re exactly right. I think a long time ago, if we read a comic book, we would see Dick Tracy with a TV set on his watch. And we would just say, that’s completely crazy. We’ll never see a TV set. And now we have a smart watch. And a long time ago we had HAL in 2001, and no one’s ever going to be like that. And arguably, the AI systems today are better than HAL in 2001. So yeah, I think the average consumer, you, or me, or other people, if we have a crazy vision of some future technology, there’s a good chance we’re going to figure out how to build it someday.

Tim Ferriss: Yeah, yeah. The Impossibles are worth questioning, particularly within the realm of AI. I was listening to a professor, well, Fei-Fei Li, I believe her name is.

David Baszucki: Absolutely. I know they’re working on a beautiful, wonderful idea of a company.

Tim Ferriss: They are, and she’s incredible. We were actually at Princeton at the same time. And she had Andrej Karpathy as a student at one point. I’m sure I’m mispronouncing that. But the point is they were generating descriptions from images, and Fei-Fei I think at one point, she talked about this in her TED Talk, was saying, “What if we could generate an image from the description? So go the other direction.” And even at that time, which was not that long ago, if you think about it, Andre was like, “Yeah, no, that’s impossible. That’s not going to happen.” And yet here we are, really just in the blink of an eye.

So a lot of these things, like the holodeck you’re talking about, it’s like, sure you guys are working on this. There are other people working on some version of this. The ability to walk around Rome or something with a friend or to be in their living room with very, very lightweight hardware, it’s like, man, this stuff feels like it’s just around the corner. Who knows? People have said that about fusion for a long time, but I think in this particular case, it’s just so technologically-enabled and the development is so rapid that it’s hard for me to imagine a future without these things. I mean, how far away do you think these things are? The holodeck has come up a couple of times, how far away, if you were a betting man?

David Baszucki: There’s some various things to think about, right? How long till a Hollywood movie is maybe AI generated? Three to five years maybe, crazy. How long before there’s a product where instead of whatever you like, TikTok, Shorts, Reel, Spotlight, whatever short form video product you like, how long before some of that activity rather than you thumb scrolling, you’re watching a continuous video of your dream rather than all of those, who knows? That’s going to be kind of crazy.

I think one of the things we have our eyes on that is technically very difficult. How long could we support a photo realistic music concert with 100,000 people in the same stadium? Where if you and I were on the other side of the stadium and I waved a flag, you could see me and we could have a full simulation of how that concert. We’ve got our eyes on that. I’m not going to make a prediction, but that’s a really big technical lift to build that kind of technology out. But I think that’s going to happen as well. How long before my glasses have full AR overlay and I’m getting a lot of feedback? We’re starting to see early signs of that. So there’s a lot of cool stuff coming down the pipe.

Tim Ferriss: I know a lot of friends who have gone from private to public, and it’s not to say it’s a bad thing. In a lot of ways, it can be an incredible thing, but you have more voices at the table in some respects. And I’m wondering how you think about preserving some of the, or ensuring that some of the constraints and values you put in place early continue even with those additional voices at the table.

So for instance, you mentioned TikTok, question from another friend was around Roblox Moments. And then they said, “Context, Moments has recently released beta short-form video format similar to TikTok.” And then I think back to when you were talking about STEM and some of the games that my friends use with their kids, which relate to critical thinking and so on. And I know nothing about Moments, so you could describe it. But when I hear similar to TikTok, I think, man, well if kids are able to use that instead of something that is STEM-focused, that’s quite a battle for attention. And I’m not sure who wins there, maybe the TikTok analog wins. How do you think about product development moving forward?

David Baszucki: Yeah, so one thing, I think that there’s a separation on intention in that historically side-by-side, the big thing I shared with you about this desire to connect, I think there is also a desire to tell stories and consume stories. And a lot of consuming stories is a bit of a different emotional head space, chilling out, watching a movie, watching a TV show, scrolling through short-form video, things like that. What I think is exciting is people have different mental models of I want to go hang out with people and do stuff together, or I want to be by myself and consume stuff by myself. And that gives me some positivity on building a connection platform because that desire, I think, for people to be with friends and meet and hang out with friends is always going to be there.

On the Moments thing, the reason we created Roblox Moments behind the scenes is there’s a couple things going on. People are looking for ways to find cool Roblox content. And one of the ways to find content is see what your friends are doing or see what other people are doing and jump into that. And so we think it is a gentle way to help people find more interesting stuff. You mentioned STEM on that side. I do feel once again, behind the scenes of everything that’s happened on Roblox from then to now, if we were to tabulate the number of new entrants into computer science, or graphic arts, or economics that had been inspired by Roblox, it probably is in the millions given just that experience that people have had on our platform.

Tim Ferriss: What are some of the challenges of being a public company CEO?

David Baszucki: If your machinery is functioning, you can keep making those, I would say, bigger, more optimistic decisions. I think when we talked earlier about you’re feeling you’re in ketosis or you’re feeling you’re not in it, I think there’s a little bit of a connection to if the machinery is running properly, the decisions tend to drift a little bit more long-term and strategic because you’re not in fight or flight. Whereas if you’re in a glucose crash and you’re a little more fight or flighty, those decisions can tend to lean more tactical and that long-term thing’s going to be — 

Tim Ferriss: Yeah, that’s a good point.

David Baszucki: The long-term thing’s going to be impossible. Everything’s an emergency, let’s just do, do, do, do, do. And so I think trying to stay in the right balance of long-term decision making is a big thing.

Tim Ferriss: So for people who may not be watching the video and perhaps they’ve seen photos of you, but when we had our conversation a while back, this was our first conversation discussing the possibility of doing this podcast together. Biohacking came up as one topic we could explore. Maybe you could speak to some aspects or elements of your self-care routine. Because you obviously take good care of yourself, clearly exercise. What does the regimen look like, what does a week in the life look like in terms of when you are performing at your best, supporting the machinery?

David Baszucki: A lot of it is just trying to have some form of movement every day. As you said, get some sun every day. On the diet side, definitely pretty much two nights ago I had a glass of wine. I could definitely feel it. So try to keep very, very low on the alcohol. I would say, try to stay in moderate ketosis, have the workouts fun and set up, and a balance of those things. It’s not that complicated though. Like you I’m trying to eat in that say, 1:00 to 6:00 window and trying to know generally what pieces I would eat. Trying to more get to bed at 9:00 rather than 11:00 type thing. Pretty simple.

Tim Ferriss: What do your meals look like? I imagine maybe you have some commonalities in meals from day to day.

David Baszucki: Mine are very similar to yours. I would say, low carb everywhere, fair amount of meat, fair amount of eggs, fair amount of butter, coupled with lettuce, veggies, stuff like that.

Tim Ferriss: And what about non-negotiable exercise? Do you have a weight training three times a week? Do you have something else that is sort of non-negotiable?

David Baszucki: CrossFit three times a week, hiking with a weighted vest three or four times a week, pretty simple.

Tim Ferriss: And you’ve got your Oura Ring on. What kind of stuff do you track for yourself? Are you recording these workouts? Are you just doing the workout of the day as prescribed by the Crossfit Gym? What type of stuff are you doing?

David Baszucki: Yeah, I’ve got a trainer that I’ve got a dialogue and my whole thing set up in my garage. So we have a fun little thing going on there. On my Oura Ring, somewhat similar to when you were talking about your sleep score. I don’t look at my Oura thing maybe more than once a month because I actually don’t want to get freaked out by how bad my sleep scores are.

Tim Ferriss: Yeah, good idea. Good idea.

David Baszucki: So I’m like you. I’m in the 60 to 80, 60 to 90 zone on the sleep, and I try on the sleep just not to worry about it.

Tim Ferriss: So when you say you don’t look at it more than once a month, what are you doing once a month? Are you crunching that yourself, just scrolling through the Oura Ring or are you exporting the data and doing something with it?

David Baszucki: I’m just looking at the Oura Ring. What I’m mostly looking for once a month, is that the Oura projected stress score, the Oura projected cardio age. I’m looking for what’s my lowest heart rate while I’m sleeping. I’m looking at HRV, just do a scan of those things, maybe temperature. But nothing too much more than that. I would say, I’ve spent quite a long time wearing a CGM or a CKM to dial in my diet and watching those type things. That would be my biggest recommendation for people is to wear a CGM if you can. I think you can buy them on Amazon right now because it’s pretty interesting. 

At Roblox, we give everyone a CGM. And — 

Tim Ferriss: Which CGM do you give them?

David Baszucki: I think it’s FreeStyle or whatever. I don’t know. Or I think they can buy either one. The other thing we do at Roblox is we have pretty good snacks, but we label all of the snacks at Roblox on two axes. One is the whole food axis, and not all whole foods is necessarily metabolically what I would call metabolically good. Like fresh-squeezed orange juice might be whole foods. We also — 

Tim Ferriss: Hemlock’s a whole food too.

David Baszucki: And then we also put it on another axis, which we just picked up the Good Energy axis. It’s not a strict keto axis, but it’s pretty close to that. And we’ve got the Casey Means book from that and all of that. So every snack at Roblox is either, is it whole energy or not? Is it whole food or not? And what’s surprising is with that, with talking about it at company meetings, with giving out the CGMs, I get all kinds of Slacks from engineers saying, “Oh my gosh, my life’s been changed. I’ve been wearing the CGM. And we used to eat just white giant plates of white rice every dinner. And we’ve read a few books and I’m not eating that. I lost 30 pounds and I feel so sharp. And that’s amazing.” And then I’m like, so actually we’re getting a lot of that from employees in the company.

Tim Ferriss: Yeah, it is remarkable how much stabilizing your glucose metabolism and just how far that goes, touches every aspect of your life. David, just a few more questions then we’ll wrap right up. This is just kind of a couple of rapid fire before we land the plane. Favorite books or books that you’ve gifted or recommended a lot to other people? Are there any books that come to mind?

David Baszucki: One of our board members gave me this book called The Infinite Game, and it’s the original infinite game book. There’s some follow ons from it about how to implement the infinite game, so I forget the original author.

Tim Ferriss: So there’s a book by Simon Sinek, which is The Infinite Game. The original Finite and Infinite Games, I believe, is Carse.

David Baszucki: Yeah, that’s the one, the original. And so that one really got me into thinking fun, play, Roblox is a long game. It’s not a short game. That one I think really got me thinking, so I’m a big fan of that.

Tim Ferriss: Were there any books that had an outsized influence on you as an entrepreneur or a company builder?

David Baszucki: What’s interesting is I’ve never really ever liked any business books ever.

Tim Ferriss: Yeah, I feel you. I feel you.

David Baszucki: The books I was obsessed with in my youth were the books about Magellan, and Captain Cook, and Mutiny on the Bounty, and Joshua Slocum, and just all of these crazy explorers, Amundsen, and Scott, and all of that stuff. For some reason, that was my go-to category.

Tim Ferriss: Probably in some ways a more helpful set of reading than the typical business books where you have complete information in retrospect, whereas these people just forging off into the unknown, are dealing with catastrophe, and challenges, and curveballs at every turn.

David Baszucki: Amundsen versus Scott is such a balance of knowing how to play the long game and being prepared versus not, and it’s a great comparison.

Tim Ferriss: Okay, two last questions. This one is the billboard question. If you could put, metaphorically speaking, a message, a question, a quote, anything on a billboard to get the message to millions or billions of people, anything non-commercial, could be a mantra, something you live by, anything at all, what might you put on that billboard?

David Baszucki: You know what, I’m thinking metaphorically right now, just off our whole conversation and then popping back to Surrealistic Pillow. I don’t know if everyone would get it, but how about feed your head?

Tim Ferriss: Feed your head, yeah. Yeah. Feed your head.

David Baszucki: That’s what ketosis does.

Tim Ferriss: Yeah. Yeah. God, I was kind of on the fence about whether or not to do ketosis before my next trip to Mexico. And I’m thinking it’s time to get into ketosis. David, is there anything else that you’d like to say or point people to? Anything at all? Anything you’d like to ask my audience? Of course, people can find Roblox, R-O-B-L-O-X.com, Roblox. They can find you on X @DavidBaszucki. We’ll link to all of these things as well as YouTube and Baszucki Group. Is there anything else you’d like to say or mention before we wind to a close here?

David Baszucki: No. I want to thank you for having me on, and I think going back to your earlier book, I just want to also think, one other thing I really remember is your mixing of, I forget what kind of oil you were mixing with your coffee really early on.

Tim Ferriss: Oh yeah. Could have been MCT oil, could have been any number of things that I was experimenting with.

David Baszucki: Yeah. And I think you were onto something there because I have my coffee with whole cream, which is zero carb, which I think has some overlap with your MCT oil. So my takeaway would be, if you want to dabble in ketosis, go buy some whole cream for your coffee rather than half and half.

Tim Ferriss: Yes, heavy cream, just to be clear, it’s not half and half. Heavy cream, which sometimes coffee shops will have in the back. It is effectively pure fat and man, is that stuff delicious, also. But back in the day, with kids with epilepsy when it was hard to get them to maybe choke down the butter and this, that, and the other thing, heavy cream, that was the key to the kingdom of feeding your head. David, thanks so much for the time. I really appreciate it. And for folks listening, we’ll link to everything in the show notes at tim.blog/podcast. And until next time, be just a bit kinder than is necessary to others, but also to yourself. And feed the machine, feed your head. It doesn’t always have to be something you think your way out of, sometimes you have to fix the physiology. Thanks for tuning in.