This Week in Startups - Using AI to diagnose Dementia with Vistim Labs James Hamet + Jason LIVE at iConnections | E1757

Episode Date: June 6, 2023

This Week in Startups is presented by: Coda. A new doc that brings words, tables and teams together. All your valuable data, plans, objectives, and strategies in one place. Go to https://coda.io/twist to get a $1,000 credit! Miro. Working remotely doesn’t mean you need to feel disconnected from your team. Miro is an online whiteboard that brings teams together - anytime, anywhere. Go to https://miro.com/startups to sign up for a FREE account with unlimited team members. VEED makes it super easy for anyone (yes, you) to create great video. Filled with amazing features like templates, auto subtitles, text formatting, auto-resizing, a full suite of AI tools, and much more, VEED gives you the tools to engage your audience on any platform. Head to VEED.io to start creating incredible video content in minutes. * Today’s show: CEO and Founder of Vistim Labs James Hamet joins Jason, as they discuss the current roadblocks with diagnosing dementia (1:18), before James breaks down how they use AI to detect signs of neurological disease (23:29). Then, Jason sits down with Kaitlin Malin at SALT iConnections New York, to discuss the keys to angel investing and AI (40:52). Follow James: https://www.linkedin.com/in/jchamet/ Check Out Vistim Labs: https://vistimlabs.com/ * Time stamps: (0:00) James joins Jason (1:18) The faults with how we diagnose Dementia today (6:46) Visitim Labs’ product (9:25) Coda - Get a $1,000 startup credit at https://coda.io/twist (10:46) Test trials and patient demographics (12:53) Vistim’s go-to-market strategy (18:24) Miro - Sign up for a free account at https://miro.com/startups (20:32) James shares a video demoing Vistim Labs’ capabilities (23:29) James’ previous startup, Neurable (27:02) Catching Alzhimers’ early, habits that prevent neurological disease and the four horseman (39:23) Veed - Sign up and engage your audience on any platform at https://www.veed.io (40:52) Jason’s sits down with Kaitlin Malin at SALT iConnections New York (46:03) Raising a 506-C fund’ (52:13) Jason’s keys to becoming a successful angel (58:27) Jason’s thoughts on AI * Read LAUNCH Fund 4 Deal Memo & Apply for Funding Buy ANGEL Great recent interviews: Brian Chesky, Aaron Levie, Sophia Amoruso, Reid Hoffman, Frank Slootman, Billy McFarland, PrayingForExits, Jenny Lefcourt Check out Jason’s suite of newsletters: https://substack.com/@calacanis * Follow Jason: Twitter: https://twitter.com/jason Instagram: https://www.instagram.com/jason LinkedIn: https://www.linkedin.com/in/jasoncalacanis * Follow TWiST: Substack: https://twistartups.substack.com Twitter: https://twitter.com/TWiStartups YouTube: https://www.youtube.com/thisweekin * Subscribe to the Founder University Podcast: https://www.founder.university/podcast

Now, idea people are fine, but the truth is there's a couple of hundred people in the room, and last night we all had 20 or 30 ideas while we were sleeping. They're called dreams. Some of them might have been nightmares. So collectively, we all have 500 ideas. Of those ideas, in the 500, there's $3 billion ideas in there. Ideas mean nothing. All that matters is execution.

This weekend startups is brought to you by Coda is the all-in-one doc for teams. If you've got a stack of niche workflow tools, or if you're buried in docs and spreadsheets, Coda is the dock that brings it all together. Get a $1,000 startup credit by signing up at coda.io slash twist. Miro helps take ideas from in your head

to out there in the world with its ability to democratize collaboration and input. Sign up for free at mero.com slash startups. And Veed makes it super easy for anyone, yes you, to create great video. filled with amazing features like templates, auto subtitles, text formatting, auto resizing, a full suite of AI tools, and much more, VED gives you the tools to engage your audience on any platform. Head to Veed.io to start creating incredible video content in minutes.

Right, everybody, we've been focused on the amazing progress that AI is making, primarily for business and generative art. So we're seeing it affect things like advertising, copywriting, press releases, reviews, strategy documents. But AI is also

having a tremendous impact in health care. And one of the leading diseases that sadly many of us have had to deal with in our families

or maybe even personally, sadly, is Alzheimer's. Early detection is a key piece of this. And my guest today is working on using AI and a bit of hardware

to track, treat, detect this really pernicious and horrible disease. His name is James Amit, and he is with Vistam Labs. Welcome to the program, James. Jason, thank you for having me. It's a pleasure to be here.

It's great to have you. Tell us a little bit about what you're attempting to do, and then let's look at the hardware here, and we'll sportscast it for those of you listening, and just, yeah, Tell us about what you're trying to solve for here and how it's going. Thank you.

Yeah, I really appreciate the opportunity. So what I'm working on specifically is, I believe it's the biggest barrier to effective neurological care. And that is the diagnostic piece of the puzzle. Because today, so many patients, you know, they come to the clinic, they want to get help. But the existing tests that we have are individually inconclusive, they're costly, they're invasive. And what this means usually for a patient is that they have to pay a lot of money out of pocket.

They get results that don't necessarily help them. And it takes a very long time for them to actually get the right diagnosis. For example, Alzheimer's disease. But in dementia, there are actually 105 different cases of different types of dementia. And so figuring out which test is the right one to run, that is a problem that AI is able to solve. And the way that we solve it is in a very unique way. we've come up with a type of brain scan that is able to look at how your brain processes visual information. And from there, we can actually figure out in advance how you'll perform on a PET scan, on a spinal tap,

on a variety of these tests that you would normally pay big money and a lot of time to get. The average American, it's going to take them, you know, about 18 months lead time to get a PET scan. And we think that what we do is very cool. You can get results in the same day. So right now, in order to test if you have Alzheimer's or to get early detection of it, there are physical tests and kind of test a doctor could administer like your speech or eye movements coordination, that kind of stuff. And then if those don't go well, they see signs there.

My understanding is you do this brain imaging, right? It's not an MRI. It's PET. positron emission tomography. Yeah. So you basically got it. The world of today is you walk in as a patient and your neurologist is going to ask you questions, right? It's kind of like a physical evaluation, a physical test.

They don't actually necessarily listen to your speech or look at your eye movements. It's more basic than that, as wild as that sounds. It might be, hey, can you count backwards from 100? do you know what country you live in? Can you draw this shape on this new piece of paper? It's pretty wild because of how subjective it is. And another problem, of course, is that if you are advancing in stages with dementia, you may still be able to count. You may still know where you live, right? I mean, these are questions that I hope I don't get wrong early on into the disease, because that really is a sign of late stage. So the world of today,

we wait for the symptoms to get so severe that the person can't even count numbers anymore. And that's when we say, okay, let's do a PET scan. It makes sense to do a PET scan. Very rarely will you do a PET scan on a patient who says, hey, you know, I can't remember people's phone numbers. I can't remember my garage door code. You know, maybe I'm like forgetting my phone and I normally always know where it is. Little tiny things.

There's no solution for that type of patient. And PET scans are not cheap. and they require a bit of time to do, where you have to, you take some sort of, I guess you get an injection or you digest some sort of a dye and then it does the scan.

But we're talking about to do it for your brain, what does it cost $5, 10 grand, right? Like, this is not cheap. It's a range. It's between 5 and 10. Yeah, yeah. I mean, having... But by the way, it's also inconclusive

because you could get... It's inconclusive. Of course. You know, none of these tests that exist today are definitive. In order to know that you have the disease, one of the reasons why they wait to do the PET scan is because they need to show symptoms. They need to show that your cognition has changed. And so one of the problems that we believe we're solving is we have a test that measures your cognitive ability. It measures, you know, amyloid proteins in your brain,

just like a PET scan would. And it measures signs of neurodegeneration, you know, physical volume changes in the brain. And you need really all three in order to say, this is Alzheimer's or something like this. So there has to be a better way. And you have a hardware product plus software to do this? Yeah. So our product is entirely AI, entirely software. We do leverage equipment that's already in the clinic.

We use what's called DEG, Electroencephalography. It's not our proprietary device. We're not pushing any single unit. The neurologist already has these devices, usually for epilepsy and for sleep monitoring. All that we ask is that the patient wear that device that the physician already has, and then watch our video. And what our AI does is it looks at how the person's brain interprets the pictures. And we can see what types of pictures, what types of activities, animations, the person's brain struggles to interpret. And that's how we know what type of damage it is, where the damage is, how severe it is as well.

It's pretty amazing. Some people, from what I'm seeing, become blind to certain types of images based on their brain damage. Oh, really? Wow. That's nuts. It's not really surprising either. Most of your brain is visual cortex, and if you have any one of these diseases, your cortex gets damaged physically. So some of those links are broken. EGs are super easy. You put these on your head. It's like a bunch of you know, diodes or something and it just measures brain activity. Am I correct? Yeah, yeah. But there it is. We're showing it here. So this is a device that you don't make. This is a standard device. And they use this for sleeping, right? If you have narcolepsy or you can't sleep, you wear these and they test

your sleeping patterns. Yeah. And this, by the way, this is not a device that you might see in the clinic. This is a little bit more of like a consumer product, just so you know. But yeah, it's exactly that. You have these little tiny metal pieces and then they touch your head, just like what was shown in the picture. And typically with a medical EEG, those will be positions all across your head. It looks like a swim cap. Yeah. It looks, yeah.

And in that one, it looks like an octopus or a swim cap. And it's just basically, you can get the electroactivity in a brain. And then, so you have a video of this at work, you will give them a series of tests, record the brain activity, and then the AI does what?

And maybe we could show this a little bit. Yeah. Yeah. And so what the AI does is it looks at all those electrodes and how they correlate with, with what is being shown on the screen across all the patients in our clinical database. So we've used our method. Yeah, that's how the AI works. It has to have data. That's the training data. That's the training data. Everybody knows the startup market is really hard right now.

And you can't be messing around with expensive software or software that doesn't work. Stop wasting your time. Stop wasting your precious resources, dollars and your team's time, stitching together five different subpar solutions. No, Coda is going to help you do more with less. in Coda, your team can work on entire projects from start to finish. And you have everything you need all in one place. You know what happens when your work is spread out across a bunch of different tools in people's email. Productivity suffers, coordination suffers, communication suffers, in this market speed, efficiency,

and cohesion is what keeps your startup alive. And Coda is that one document that pulls it all together. You can see an instance of me using Coda. Just go to JTrading.com. And you'll see this gorgeous Coda page that we made. And what is it? It's all the trades I made here on the podcast talking about. I think this is doing pretty good right now.

I'm up like 16% versus the S&P of 5%. I'm 3x the S&P. But what's really nice is it does all this automatic in a table with all these calculations. You can really program Coda like nothing else in the world. And we're using it for Founder University now. So if you're ready to join this efficiency revolution with Coda, you can take advantage of this special limited time offer.

It's just for startup. Sign up today at coda.io slash twist and get $1,000 in startup credit on your first statement. that's coda.io slash twist to sign up and get a startup credit for $1,000 coda.io slash twist. How many people have you done this one? Yeah, so we've done this with over 150 people, you know, internationally as well. We did this at Columbia University. We also did this in Germany as well at LMU Munich, which is a very reputable university.

And we collected this data with, you know, almost 160 patients. we have amyloid activity, sorry, amyloid protein levels from pet. We also have tau protein levels from spinal tap and a variety of other endpoints. And so what our AI has done is it's trained itself on their EEG to use new EEG to figure out what those endpoints would look like for that new person. And of those 160 odd people who you've had do this, did some of them have certain degrees of Alzheimer's? and that's how you benchmarked it. And some of them were just young, healthy people who had none. Yeah.

So to accurately describe our demographics, I need to say that we didn't use the product on anyone younger than 50. Okay. So we don't have, you know, healthy young people, but we do have what's called healthy elders. They exist. Okay. That would be me, 52. I'm healthy and I'm an elder.

Yes. And hopefully, you know, you'll stay healthy too. Yeah. And so in our population that we studied, we have these great healthy people who are staying fit, who are highly educated. We have other people who are at the early stages of cognitive decline. This is what we call subjective cognitive impairment. This is where the patient believes that they're suffering from a disease, but there's no clinical evidence that would corroborate that claim. We also have patients who are considered mild cognitive impairment.

this is where there is evidence. For example, maybe they can't count backwards from 100. Maybe they can't redraw the picture, right? So there's real symptomatic evidence that's hard to refute, but we don't know what the cause is. And then we also have diagnosed Alzheimer's patients. And so those are the four cohorts in our population. And so where are you at in terms of this product being in the market? Because is it fall under a medical device?

It's diagnostic. Is this something that's sold to consumers who can then take the information they collect from it like they might with their Apple Watch or Fitbit or their aura and consult their doctor because it's quite a path from what I understand. I guess it's the FDA that who does medical devices.

It regulates it. So how does this work as a startup? Because I've heard ones, you know, people who have different technologies say, oh, we're doing this for quantified self so we can get it into the field early and then people can use it to gain information as they're allowed to do, but it's not for diagnostics or for like a medical treatment.

So how does this all fall into place as a go-to market strategy, if you will? So as an entrepreneur and a serial entrepreneur who's also been in the neurotech fields with other businesses, I do know this field pretty well. So maybe I can move faster than most. But I will say that we will be in the market this year with clinics. and we're looking right now at having our launch in September. We already have three clinics that we'll be launching to. And the reason that I can move this quickly is that our product is actually exempt from FDA approval. And that's because of our ability to measure cognitive performance of people.

Because we're doing that and because it's clinical cognitive performance, we are classified as a cognitive assessment aid, which creates a special product code, which exempts us from the advanced FDA approvals. What this means for our product is that we are not claiming to diagnose Alzheimer's with our product. Our product is simply an AI engine. It produces the endpoints that the doctor is interested in getting anyway, like the pet results and the MRI results. We'll provide similar results.

But we're not claiming that it's Alzheimer's. Now, in a future iteration of the product with 510K approval, we may, but it's not necessary for our go-to-market. Got it. So this is the, I mean, this is quite reasonable. Do we have a reasonable framework here in the United States for allowing innovation like you're doing and then also protecting people from like claims? It seems reasonable. Explain it to us like, you know, for people who are not in the industry, how it works getting from one stage to the other. Yeah. So I will say that not all products will have this opportunity. Not all products will be recognized as FDA exempt. It's all. based on the intended use of the product. And so you have to, when you're planning, your go-to-market strategy, you have to investigate what those exemption codes are and see if maybe your product does fit with one.

I am not an expert. I have experience, but I'm not an expert. We've hired a company called RQM Plus. They've provided us with regulatory advice and guidance. And I guess this is a little plug for them. But they help us make sure that we are not making these decisions in a box and in a silo. And so for any aspiring entrepreneurs out there listening to this, I would recommend the same thing. You know, find a consultant, a regulatory consultant who can help you.

You can do your own research. What you need at least, though, is to have a third party, just double check your research, right? Lend their experience, confirm with what you're saying. And so the place that we're at right now is we've identified this product code. that as long as we are providing cognitive performance indicators, then we are able to be a cognitive assessment aid. And when I say cognitive performance indicators, I don't mean a subjective metric.

I mean specifically, we are predicting with AI, we are predicting specific cognitive evaluation scores. So the FCSRT test, you know, if you're a neuropsychologist or neuropsychiatrist, you might recognize what that is, the free and cute recall test, it's a memory test. we are able to predict someone's performance for taking that test. And so that is what really is allowing us to take advantage of this code.

And yeah, we're not making big claims. We're not saying these are the results of a PET scan, right? That would be dishonest. But what we are saying is we're saying, if you took a PET scan, this is the type of result you should expect to see. If you take an MRI, this is the type of result you expect to see. And there's nothing wrong with that. Could there be products out there that,

take advantage of this. Yes, I believe so. The challenge that I would put to those companies is good luck. Because selling to doctors and hospitals is really, really hard. They're going to want to see the evidence, not just, okay, it's FDA exempts. Awesome. FDA approval, exemption, whatever, it is, that's enough to sell legally, but you still have to sell to your market. and they have their own important criteria for making their buying decision. And clinical data is one of them. Yeah, and these are broken up into different classes. And, like, you know, there's like band-aids.

And, you know, then there's like taking your temperature or like a pregnancy test. And, you know, then you got heart vowels, you know, being implanted by a surgeon. And I think the FDA does a pretty good job. Ours is just software, too. So. Yeah. So, yeah, it's probably, I don't know if it's class one or two or where, falls in all this. I'm no expert, like I said. But you'll be able to get that 510K at some point

to, you know, as you go, but people can just use it for informational purposes now. Founders always ask me and my team for pitch deck punchups. Okay, this is constantly what we're asked to do is help people with their pitches. Well, I've got some great news for you today. We just worked with the team at Miro, that awesome whiteboarding software I've been telling you about to create an amazing pitch deck template for founders, which you can see right now. Now, if you're watching the video, if you want to check it out, head to mero.com slash mirroverse and search for pitch deck. MIRO.com slash mirroroverse.

This will help bring your pitch deck from zero to basically VC ready. Go check it out right now. And if you are a hybrid or fully remote team, Mero is incredibly useful. It's like an old-fashioned, in-person whiteboarding session when everybody's brainstorming. But this one is distributed and it's asynchronous. So you can do it when you have time and you don't need to be in the same room. Miro lets you brainstorm amazing ideas, collaborate on projects, and more from anywhere in the world. When you think Miro, think zero to one but faster.

And Mero is so much more than a simple digital whiteboard your team can collaborate on planning, research, design, and feedback cycles. Faster inputs equal faster outcomes and velocity is how startups win. So here is your call to action. To access our new Miroverse template, sign up today for a free Miro account at Miro.com slash startup. That's MIRO.com slash startups to sign up for free. And then have you had any great results so far in the market or in this testing? Yeah, yeah. So we haven't deployed our product with clinics yet because the product will be launched this year.

But then we will start to collect that data and get their feedback. And we believe that we'll be profitable starting in Q1 of next year. So it's a very quick turnaround. It's very nice that we have an opportunity to make profit even before that. 510K opportunity. And another thing we want to do is expand into other disease states. I mentioned that there are 105 types of dementia. There's also Parkinson's, traumatic brain injuries. So these are areas that we want to, that we want to help and that we know we can. I do have a video that I can share just to give a little bit more color. Yeah. And so the video is here. I turned

off the sound. Yeah, let's see it. So this is, there we go. I paused it. Yeah. This is what a medical EEG device might look like. You can see that it's a little bit more sci-fi, a little bit more like a swim-cap. It's pretty cool. Look, in strange days, if you know that. A great science fiction film at the turn of the century. But yeah, it's like a skull cap. That's got a bunch of webs on it and it records your brain activity.

Yeah. Yeah. And so what we do is we put this device onto the patient. The clinician, you know, loads the software onto the computer. And in a quick instant, we'll have the patient's actually wear it. Where is this part?

What does that device cost? Here we go. So the patient here is wearing the EEG, and they're watching some of our video. Got it. So then sitting there with that cap on, the swim cap, like you said, and the screen changes,

and then a report comes out. Yeah, and this is an example of one of our reports. So you can see, you know, activity within the brain. You can see where we expect distributions of amyloid. You can actually see these are continuums for those different endpoints. We, you know, create a range of where we expect the patient to fall. We also have on the bottom here, this is progression data, how we expect those endpoints to change in the future. I think that

there might be a close up that's a little bit better to see. How much does that device cost that goes on people's heads? I'm curious. Ah, here we go. So that, that was just slightly zoomed in here. So you can see here just some of our end points. So you asked me about the cost. So we're not selling the hardware since that's not part of our business. I've seen prices range between, you know, 5,000 on the low end and on the high end, 40,000. Just for that system for the head to record brain activity. Yes.

And it's not just for the system itself. It's also, it comes with an amplifier, which is a box that does all the, you know, amplification. and signal collection and all that. It's a big, pretty serious piece of hardware. For clinical systems that you'll see in, you know, the regular neurologist office, a better, more accurate range is $10,000 to $15,000. Amazing. The $40,000, the most expensive one I've ever seen was a university research equipment

with, I think it was 248 electrodes. The one I showed on the video is about 60 electrodes. And so there is no consumer version of that yet in the world. but just going off script here on a little bit of a tangent, are there other uses for that thing? At some point, are we all going to have those

as like brain interfaces for our computer and be able to think of a question, have chat GPT-17 answer it, or play a video game with it, or drive a car, or be, in other words, like Professor X.

So I'm actually a pretty cool guy. You probably know about NeurLink, right? Yeah, sure. You may have heard about a company called Neurable. Normal was my last company and we used EEG to, like you said, drive a car, play video games, control things like phone calls and music. And so, yeah, I mean, I built that technology and that's where I get my expertise in EEG specifically.

Got it. How did that company wind up? Does it actually work? Can it work? Yeah, yeah. No, the company is very successful. It's based in Boston. it's actually continued to raise money and grow without me,

which is honestly the dream of any entrepreneur. So I get to say that something has outlived me in a sense, at least outlived my exit. And so, yeah, it works very, very well. I'm very proud of the team that's still working on it. I think that we've raised over $10 million to date with that business. And yeah, we were able to do some amazing projects with some pretty large companies. I don't know what I'm allowed to share, but we used it in pro sports situations.

We used it to control virtual reality video games, wheelchairs, cars. There's some precedent to this with, you know, there's some precedent to using these things to help with people who are optimizing for sports, basketball players, whatever, reaction time, ability to shoot the puck or kick the ball or whatever. This is like a known science, yeah? Yeah, I mean, I got a publication in Nature's scientific reports because I invented a type of way to use this technology to measure how tired somebody was. I could use EEG to figure out how much sleep the person had the night before and when they'll fall asleep next. And when I say next, I mean, you know, are they going to fall asleep while driving? Is it going to be in the next 30 minutes? I was actually taking that technology to market following Nourable when COVID started and the transportation industry crashed.

And so I sort of left that one on the shelf. What would it cost to run your test if like, you know, getting your doctor to check for Alzheimer's, you know, and getting that, you know, special scan. You know, you're talking about 510K. What do you think you can get it down to this early warning system and how often should people be tested then? Yeah, that's a great question. I believe that every American should be tested from the age of 45 onwards. I would like to suggest an annual test. I believe that, you know, one test every five years would be the minimum that I would do myself.

Less than that, you know, you're probably going to miss it. You're going to miss the first signs because our technology can pick up the signs of disease up to 10 years before clinical diagnosis. And so you would want to catch that 10-year window as soon as you can. So I would recommend, you know, between once every five years and once every year, the critical question that you're going to ask me is, well, what's the price, right? Because that determines how often I want to do it. This is less than $1,000 and it's almost entirely covered by insurance. For people with Medicare, Medicaid, it's fully covered. For people who don't have Medicare,

Medicaid, it's about $100 that they're paying out of pocket. Otherwise, it's completely it's a completely covered procedure. Well, and if you catch Alzheimer's early, I guess, if people are getting tested for this every year from the age of 45 on and you catch it early in one patient, I'm guessing the drugs per month for people with Alzheimer's is probably going to be close or the first couple of months on that drug. It could be the cost or, you know, the treatment for getting the prevention, yeah, or something in that range. So if you don't catch Alzheimer's early, if you try to treat it late, that's when the treatments become very expensive. We have like, you know, Atahelm and Aducanamab and some other treatments that are coming to market now. And it's been a, it's had a very big backlash because these drugs, some of them cost about $100,000 a year. That's a ridiculously expensive treatment.

Meanwhile, on the flip side, you know, if you're taking preventative medication, if you're taking medication, if you're taking medication, that slows symptomatic decline like galantamine and denazapil and there are a couple others. Not a physician, by the way, but just sharing some names of some drugs that I've heard from physicians have worked. You know, these drugs are, you know, $100 or less. You're saving, yeah, 99% of your spend with preventative care. And by the way, I should also share an opinion, but it's an opinion that has some substantiated evidence in literature,

that these are preventable diseases. The effect of the genetic factors for someone to develop something like Alzheimer's, it's there, but it's not the largest one. The largest correlates that we've seen for people who develop these diseases, you know, it's comorbidities, it's insomnia,

it's dietary, it's behavioral. One of the things that we see, and it's hard to explain, but people who isolate themselves from society, who have more hermit-like existences. They're the ones who are the most prone for developing dementia, and it's speculated that the social engagements that they have with other people

actually keep their brain functioning longer. So I believe that there are behavioral solutions as well as pharmaceuticals. Wow. So we can prevent Alzheimer's through a number of just behavioral changes and lifestyle changes, exercising, healthy lifestyles, balanced diet, all that kind of stuff. And then there's something in these where mental stimulation or social mental stimulation, perhaps, putting aside the comorbidities of like, I think it's cardiovascular is the main comorbidity or diabetes. Diabetes is a big one. Yeah.

There's actually, there are some scientists who describe Alzheimer's as diabetes type three. Ah, Alzheimer's, that's interesting. And this is a thing about, and this is really important for entrepreneurs, getting good sleep. If you don't get good sleep, that is a precursor to Alzheimer's. If you have your sleep interrupted constantly. So if you're in a noisy city and you got garbage trucks in your backyard, like I did during my 20s and I had to get like, you know, like a noise machine and headphones to just block out like the New York City garbage trucks and ambulances that were constantly going by. that kind of stuff getting woken up all the time, that can be a contributing factor,

according to, you know, the research, right? Yes. Yeah, again. Isn't that wild?

It's a significant factor. I, I don't have the statistics on me right now, but I believe that it's a 30% increase or more of likelihood to develop Alzheimer's. Yeah. All right. And so hopefully. And it's based on how much sleep you don't have.

Right. So if you're trying to be a hero and get less sleep, you could have less cognitive function. in your final years on the planet. I know like young people like to make tradeoffs. But I got to tell you, the sleep thing, it's not much of a tradeoff because if you get eight hours of sleep or six hours of sleep, that two hours you're going to gain, quote unquote, is going to cost you because your productivity is not going to be as good to you.

You might as well try to get seven, eight. I usually get six, but I try for seven eight. I'm going to see what as you. Yeah. You get the six, but you try for seven eight? Is that thing? Yeah, yeah, yeah. I have really worked on it.

You know, the bull, sleeping with two of bold. logs will get you, uh, yeah, that, that is a bit of a challenge. Sometimes it works in your favor, sometimes it doesn't. Well, you know, one, one interesting thought experiment that, you know, the audience can think about is, uh, if it were advantageous to sleep less, then how come every single animal sleeps? Right. Darwin would have made the adaptation. Exactly. Yes. And it's not, and we don't sleep less than other animals. You know, we sleep quite a bit. Um, the very intelligent animals sleep more, it's likely that sleep plays a restorative function for our brain, a critical and unskippable restorative function. Yeah, I mean, all the sports teams now are tracking,

I know a lot of NBA teams and there's different software companies that do this and they are tracking the sleep of their players as the primary goal. I mean, diet right after that, but it's sleep and diet that they stress. And those things are obviously all related. But yeah, Sleep, so critical to performance, especially in elite athletes. What do you think, just broadly in AI while I have you here, what else do you think are going to be, you know, I think Peter Attila calls them like the four horsemen of death and, you know, obviously diabetes is one, heart, lungs and you get brain function. I'm not sure what they all are, but these, you know, diseases, what else is going to fall?

What other things could we see, cancer obviously? What are the things can we see fall in the future, you know, through AI? Because you must have contemporaries who are working on stuff. Are you talking about like what other ailments we can detect early? Detect early and or prevent. Yeah, yeah. So truly, I believe that most of these disorders that we face are preventable. And I think that that's because of what we've seen now with epigenetics,

with the effect of stress and diet on the, human body and its ability to turn on and turn off genes. That is mind-blowing to me. And so I believe that symptom-free living is possible, not just in dementia, but yeah, across these other endpoints. There is a lot of really great work happening right now in diabetes. There's even some great work in reversing diabetes. I think that diabetes, and I think that a lot of people would agree with me, that it can be

prevented and it can be successfully and effectively prevented as well. if you can track that the person is developing the condition early. One of the challenges that we have today is that we're still looking at endpoints that we don't fully understand. We're looking at cholesterol, blood pressure, things that may be indicative of a problem, but things that don't necessarily tell you what the problem is. And so that's the power and opportunity of AI, that we can take all these endpoints that we traditionally look at to distinguish these diseases

and we can have the AI essentially figure out this hyper parameter, this hyper endpoint, if you're familiar with that term, a meta endpoint that helps us know what this is. Because you have your blood glucose level and you also have your age, you got your weight, blood pressure, family history. I mean, there's so many things that come into it. And what you're saying is you throw all that data into an algorithm, do some learning on it.

it might spit back out like, hey, this person needs to get this under control. Yes. Before diabetes shows up. Well, this is like amazing. What I would like to do if I had enough time and actually after my success with this business, it's to go into some of these more vascular problems. And stroke, I think, is another key area where we need more preventative measures. I believe strokes are preventable.

And that's an area where if you don't prevent it, you have huge consequences. So prevention is the only option. In the case of stroke, truly, it's not like you can necessarily treat it. Sometimes stroke will kill the patient. So what I would like to have is a warning system. Can I, as a patient, go to see my doctor in sci-fi, right? They would sit on the bed and get the whole scan and, okay, in five years,

you're going to have your first tumor. That's what I want. Wow. Yeah. That's what I want. And that's the future that I hope that I'm building, starting with the brain. Yeah. Because then you can change the future if you know it's coming, right? If you know it's coming and you know what?

A lot of doctors tell people, hey, listen, you got to lose some weight and you got to stop smoking and people just disregard it and they had a stroke. Well, and you've been speaking a lot about like, yeah, this like performance type medicine, how powerful is it to have that feedback loop? Yeah. Okay. I go and I change my diet for a week. I come back.

Doctor, is it helping me? When the doctor says, I don't know. That's when the patient's like, all right, well, I'm going to go have that cheeseburger later. But if a doctor does another scan and says, hey, look, yeah, you've reduced your odds by 5%. Is that motivating? I mean, super motivated. This is why I think weighing yourself every day is such a good practice. This is why I think the sleep, you know, the aura and the Apple watch Fit Fitbit, the ate sleep, all this stuff that helps people get a little bit of metrics, Neutrasents and different glucose levels. If you start getting that stuff, you're like,

maybe I shouldn't have that bowl of cereal at midnight because my glucose is going to spike and then I'm not going to sleep while you get the double whammy, right? That was what I was doing. That was so pernicious in why I gained so much weight just because I was eating like a bowl of cereal or two at midnight. Make my stomach do backflips, wake me up in the middle of the night,

and my spike in my sugar. Terrible. In the interest of the health of all listeners, ice cream at night is very bad for, you know, your Alzheimer's future. Yeah, ice cream is so great. But yeah, if you eat it within two hours of bedtime, that's a problem. I'm not entirely sure why, but as our brains age, we crave sugar more, especially in the evening. And that has a very strong link to Alzheimer's.

I really believe in this power of like just a little bit of activity. So I just try to walk 20 minutes after dinner or if I eat something, if I'm going to reward myself with a little bit of ice cream, I just tell people, I'll get ice cream. We got to walk to the gelato store and we got to walk back. So it's a mile or a half mile. And I'm like, let's walk to the one further away and get the ice cream. And then we'll walk back to the car. Very simple techniques.

Like just park your car five blocks away from the ice cream store and then walk for your ice cream. Now you got a half mile walk in. I agree. I believe, you know, I've dieted before. I care a lot about my health. I think that it's easier to adopt healthy behaviors like that, like walking more, exercising more than it is to cut yourself off.

I mean, sometimes you want ice cream. There's nothing wrong with that. But you should get your 20 minutes of exercise a day. And that's been pretty well documented. 20 minutes a day. Life changing. That is the best thing that you can do for your body if you're not doing it already. All right, everybody.

Great job on the company and continued success. Are you hiring? And where can people learn more about the company if they want to? Yeah, Jason. Thank you so much. So I can be found on LinkedIn. You know, also our website, vestamlabs.com. You can email me.

I'm James at Vistamlabs.com. Please don't spam. Do you need a headhunter? Are we hiring people? Not at the moments. We just closed around. We just brought in some new team members, which is awesome. We will be raising later this year.

I think we will be growing our sales team. Great. Awesome. There are people who want to help on that end. If you know clinicians, What I would ask is, if you know clinicians who are working in dementia, Alzheimer's, and they're interested in learning more, please do connect us because the only way that I can create a good product is by speaking with more doctors.

Love it. You got to talk to those customers. All right. And we'll see you all next time on this weekend startups. Bye bye. Listen, I'm doing six video podcasts a week. Think that through. Oh, my God.

This weekend start a five days a week for your listening and visual enjoyment. And the video is becoming such a big part of this because we like to share. show AI tools. And then of course, you know, people love to watch all in the video there because it's dynamic. You see us laughing or rolling our eyes. And I can tell you, when you have this much video, it becomes super expensive. It is at least 10, I would say 20 times more expensive to do a video podcast than an audio one. And you need powerful tools to do this. But, you know, how many audio only clips do you see going viral? Zero. Zero point zero. Audio does not go viral. Videos go viral all day long. I know that because we'll put out 10 clips and all of a sudden, boom, one gets 100,000

views. So you need to start leveling up your clips game. You need to check out Veed to do that. Veed is a web-based video editor. That makes it easy for anyone to create great video content and the best part, no editing skills are required. Veed has every editing feature you'll ever need, including automatic subtitles. This is saving our producers a ton of time. You can remove background noise. You can auto resize, scroll the different platforms. They have a suite of AI tools that makes editing faster and easier than ever, and they have hundreds of plug-and-play templates for you to hit the ground running. Veed lets you do it all, V-E-E-E-D. VED, indeed, without having to spend hours learning, complex, editing software, or paying third

parties, start engaging your audience on any platform. Head to VED.I.O and start creating professional quality videos in minutes. That's VEED.I.O to sign up today. So I have to say, it's a bit intimidating interviewing the world's greatest moderator. You know, I get that a lot. But having had a couple conversations with you before this, I think we'll be okay. Yeah, I think we'll be. You're a good conversationalist.

So for our industry, unless you've been living fully under a rock, you may know that Jason is one of the founders of the top 10 All In podcast. And he's also in the midst of planning what's quickly becoming the hottest event in the venture community this fall, the All In Summit. That's crazy. It's, I mean, seriously, if you get a chance to talk to him, ask him about it. But you're also the host of another podcast this week in startups. Yeah, I'm doing that for 12 years and 1,700 episodes.

That's impressive. Congratulations. Thank you. It was one of the original podcasts when podcasts were on your iPod. And for half the room, an iPod is a device before the iPhone. That only played music and you would hook it up. And at night, it would download podcasts from blogs. and then load it up and make an album on your iPod that said, podcasts.

And then you would use your scroll wheel to pick it. And basically, when I found out about podcasting, I was like, this is a great excuse to just talk to people instead of having lunch and then publish it, and a couple of hundred people would listen to it. And then when the iPhone came out and got very popular, you can download it to your iPhone, it was pretty amazing. And just last week,

I had Brian Chesky from Airbnb, Reed Hoffman from LinkedIn, and Aaron Levy from Box, all of them on to talk about AI, which is having a spectacular impact on the technology industry and efficiency, productivity. And if,

it's just wonderful to have these conversations and then have millions of people get to share them. I just got back from the UAE and I was in Abu Dhabi and Dubai and the number of people listening to the podcasts around the world is truly stunning and all in and this week in startups. It's like people think I'm their best friend because I've been in their ears, you know, while they're working out or going for a walk. Having Friday night dinners, we heard earlier. Yes, somebody was telling us that their husband and wife, like their big thing is to put the kids to bed on Friday night, open a bottle of wine.

And I'm like, yes, where's this going? Super fans. And they're like, and then we watch you. And I'm like, wow, I never got that one before. Whatever, whatever. So they all in and chill. Yes, absolutely. But why don't we start at the beginning?

Tell us a bit about Sequoia and how you first got into venture investing. Yeah, so I started my career here in New York. I had a magazine called Silicon Alley Reporter in the 90s. There's probably three or four people in the room who read it back when zines and magazines were a big deal. And I was a journalist for a while. Then I started a couple of companies. I sold one of them to AOL. And then Sequoia backed one of my companies.

And I had introduced Sequoia to three of my friends in startups. One of them was a poker, online poker. One of them was a micro blogging service. And the other one was an electric car. And I said, these three startups are going to go places. and they passed on investing on all three. And those three companies were obviously Tesla, Zinga, and Twitter.

And I was an angel investing at the time, so I didn't invest either. But they came to me and said, hey, can we give you some money and make you a Sequoia scout? And I said, what's that? And they said, it's the newest thing. Rulov, who's our newest partner. I said, oh, I know, Rulof.

You worked with my friends at PayPal. He said, yeah. He's our new partner. He was just there for a year or two And he said, I'll, we'll give you money. You invest it and we'll split it 50-50. And I said, don't you guys, like as venture capitalists get two and 20? Isn't that how it works?

And they said, yeah, but we had 30 because we're Sequoia. But it's never going to amount too much. It's just 25 and 50K checks. So in the first seven investments, I did Uber, Thumbtack, and Data Stacks. And that Uber investment went 4,000X. not 4,000 percent, but, you know, well over $100 million. And so then I started my own venture funds, fund one and two, were $10 million each.

I raised them around the poker table with my friends in Silicon Valley. The third one was $44 million, which I just finished investing. And that's the first one where I had institutional money, Green Spring and some other folks. And then I'm raising launch fund four. And so when you invited me to come, I was like, Well, that makes sense. I'll come and talk. But I'm doing something very different, which is a public fundraise, which is called 506C. And I believe that's going to change the venture capital industry and democratized venture capital. Forever. Can you explain 501C? Sure. The audience for those. I don't know how familiar. How many people in the room are venture capitalists or have a fund or work at a venture fund? Raise your hand nice and high. Okay, great. How many people are limited partners or invest in those type of funds? Great. I'll see you after. It's actually really interesting. That was 50-50 exactly.

I know. It was actually a pretty good balance. So, you know, when you talk to a venture capitalist, when I was a journalist, I'd say, oh, so you're raising another fund? And they'd be like, turn the microphone off. And I'd be like, what? And they're like, we're in a quiet period. We cannot talk about raising funding. Please don't mention that in your story. Because if you do, the SEC will reset and I have to stop raising and then wait six months and get put in a penalty box. And I was like, oh, okay. And then I talked to the lawyers, and they said, yeah, you can't do that unless you're five, I said, okay, 506B, 506C, what's the difference?

They said, one's public, what's private? I was like, well, I have 700,000 followers on Twitter, and 50 million people listen to my podcast, and I don't know any LPs except for my friends, so, yeah, I want to do the public. They're like, don't do it. And I'm like, why not? What's the cost of doing 506C?

And they said, the cost is you have to certify each of the accredited investors. I'm like, how do you certify an accredited investor? and they're like, oh, they're accountant or their CPA or their money manager or attorney send you a note. Caitlin is an accredited investor. And I'm like, okay, what else? And they're like, that's it. I'm like, well, what happens in 506B? They're like, oh, people can just check a box.

I'm accredited. I'm like, oh, so they can lie? They're like, well, I mean, basically, they can lie. I was like, hmm, okay, so we'll just do this certification. And now it's become a little bit of a thing. my friend Sophia Amarosa, who did nasty gal, and she's doing hers publicly, so she tweets, I'm raising a venture fund.

What that does is you get to meet people who you wouldn't have met. And every time I meet someone, they're like, oh, are you raising a fund? I'm like, no, I just finished, but talk to me in three years. And so now I took it to the next level, and at launch, our website is launch.co.

And if you go to launch.com slash memo, as in deal memo right now. I just wrote the deal memo. Here's how we're going to invest. This next fund will probably be $100 million. Here's how we're going to invest it in early stage startups. And what that does is it creates a certain amount of intellectual honesty and rigor for me as a fund manager. Now listen, I've already made the money I need to make in my life,

but I really enjoy investing in early stage startups. So it's a passion of mine. And I like to put my idea out there. I like to say, hey, here's not my annual letter looking backwards. Here's my forward looking. And I don't know if you've sat through this, LPs, and like you watch these pitch decks. I spent like the last three, four months with designers on this pitch deck. And I'm like, all we're doing is talking about fonts and images.

And this has nothing to do with what I do every day. It's incredibly frustrating. I'm $10,000 into this design process. and it's stupid. And then I send the deck to people and they don't remember anything. And I'm a writer, so I'm going to write my thesis.

I'm going to write in crystal clear language as concisely as possible and then I'm going to edit it and edit it and edit it. And I send this deal memo out now and people actually understand what we're doing. So why is it controversial that you put it out there publicly?

You're one of the first? No one's done it. before? I don't think anybody's done it before. I think people maybe are very proprietary about their strategy. What I've learned is if I put my strategy out there, people smarter than me or haters, and I have a lot of those because I got lucky and I hit like three or four lottery tickets, which people really hate. Like if you are the third investor in Uber, people are like, you, Jason, Kevin. You got lucky. I'm like, what about the other nine times? Yeah. You're like, that was luck too.

No, but what I find is people get very aggressive to tell me what I'm wrong. And I just say, they say like, you're fat, you're old. Here's a really something insightful. By the way, you're fat and old. And then I'm like, okay, I'm just going to forget those two other insults. The internet. And then I'm like, but this nugget of truth is really good. Like, this is a leak in my game.

This is how I could be better. And so I just like to take that all in. And I don't fill out in competition with anybody. Objectively, people would say that my fund competes with Y Combinator. And it is true. We're both in Silicon Valley. We both put $100,000 into companies for six or seven percent. But Y Combinator accepts 1.4% of their applicants.

They get 20,000 each class, so 40,000 a year. My firm's much smaller. We're only 18 people. We get 15,000 applications per year. We invest in a much smaller number of companies. 100, 150 a year versus their 500, and we accept 1.x percent. So if you put the two of us together and you assume that the people are applying to both programs, there's 50,000 people applying and 3% get accepted or so, or 2%, maybe. It's not like anybody who's an early-stage investor

could actually tell you there's a difference between the top 2%, the 8% or the 16%. The truth is, when you have a program like we have with our accelerator or Y Combinator, it's like being Harvard or Stanford, smart people apply. Harvard and Stanford do not make those people smart. They take smart people, they don't make them dumb, and on the margins, they make them a little bit smarter. I feel like we were talking about this earlier, and I think where it is most applicable to our industry, too, is just the idea of distilling down, like you're interviewing, you're meeting so many different people, and everyone out here, if you're not looking to raise money as a fund, looking to find that next golden nugget.

So if you had to distill down your process, what would you say is the key to success in terms of taking in all of those applications and really, you know, standardizing it? Yeah, I wrote a book called Angel about this, but I've since refined it, and I'm happy to tell you exactly the playbook. It turns out that when you have two or three founders in an early-stage startup company, you have less of those companies go out of business. than when you have one. For an obvious reason.

Typically, one of the co-founders will quit. They'll go get a job. They find out they don't like it. So you go from three to two or two to one. So you have redundancy. It's like a spare tire. And they also will push each other. The next day is you want to have builders.

You need to have builders on the court team. If you have three idea people, then they're just having to hire people and they're outsourcing. So we give them money and they just give it to an outside dev shop. none of the knowledge is inside the company. So you want to builders on the team.

Could be a developer, could be a designer, product manager, etc. Now, idea people are fine. But the truth is, there's a couple hundred people in the room. And last night, we all had 20 or 30 ideas while we were sleeping. They're called dreams. Some of them might have been nightmares. So collectively, we all have 500 ideas. Of those ideas in the 500, there's $3 billion ideas in their.

Ideas mean nothing. All that matters is execution. At the end of the day, if you look at a startup company, it's about a team that builds a product that delights customers. And all the great successes I've had in investing share incredibly driven, resilient, insane teams that were obsessed with products and who listened to their customers and delighted them. And everybody in capital allocation and founders,

they want to talk about the total addressable market, They want to talk about go-to market strategies, competitive landscapes, matrix. All of that stuff is super important, like when you're analyzing Netflix versus Disney Plus, or Salesforce versus HubSpot. But in the first couple of years, it's a much more simple process. And there are very granular things we look at. One of them is called product velocity. So of these 15,000 people who apply, we meet with 3,000 a year.

We do a 20 or 30 minute phone call in Zoom. We record that phone call. We take that phone call. We transcribe it with AI. We take the AI transcript and we summarize it. And then while I'm here in New York, you know, going down to Mineta Tavern to have a burger alone at the bar, while on the way, somebody on my team will text me, this company is sick. You've got to hear this. Boom.

I put it on my phone. I put it on 1.5 speed and I listen to the founder pitch. We now have all of that in a database. and then we will email the founder every three to six months, an automated message, hey, you met with us on this date, and we're hoping to get an update on the business. And by the way, when you give us an update on the business,

can you rate our firm? How likely are you to recommend launch? And sometimes people will put a three or a four or a five, because if you don't fund somebody, an entrepreneur is going to look at you and say, well, you're dumb. You didn't fund us. So we're going to give you a five.

And then we say, well, why did you give us a five? And it says, I don't feel like you understood our business. And then I email them. Like if it's under a five, they're like, call in J-Cal. I come in, I email them. I understand you met with one of my associates. It didn't go perfectly. We didn't understand your business.

I apologize about that. I would love to jump on a call with you and make sure we fully understand your vision. And it's that kind of hustle that we put into what we do. Most venture capitalists are narcissists and lazy and fat. I don't mean fat like as in I'm fat or I was fat. But I lost 40 pounds, so I'm feeling pretty good about it. And so they take off like two or three months in the summer, two or three months, and they do like two investments a year, all these management fees.

They're just in it for the wrong reasons. Like I would say three out of five, like 60, 70% of them are just living high on the hog on fees. And they don't work hard. And in our company, we are a service company. We're not the stars. We are like the Amman Hotel, is what I tell my team. Have you ever stayed in one of these or what I'm talking about? No, but I've been told I should say it in Amman Hotel.

I mean, it is insane. I stayed at the Amman Hotel in Tokyo and it was life-changing because the level of service and anticipation was incredible. And so we see ourselves as a service provider to these founders. We provide them with capital. We provide them with introductions in a network. And then on the margins, some advice. The great founders generally don't need much.

And then I act personally as a friend to them when they're having tough times because it's really hard. And so we take the work incredibly seriously. We meet with a lot of people. We invest in a lot of people. And then we track their progress. If they double or triple revenue, we put more money in until we hit 15%. And so now we're starting to build 10 to 15% positions in this incredible funnel. 50 million people watch the podcast every year, 15,000 apply.

We invest in every fund about four or five hundred names, and then we try to get to 15% ownership in the top 10. There's probably 20 great companies created every year in terms of unicorns, and then there's probably three or four of them become decacorns, 10 billion or more or 100 billion. There's probably one a year. And so it's a numbers game. Yeah.

And you don't have to hit all of them. You just have to hit one. Well, the beauty of the number, I mean, what you said about your process is that it's really, it showcases how AI is completely changing product. It's wild. You know, I feel like I'm using it daily. I'm telling all of my team members to use it daily. I know I've heard you say that on your podcast.

You're interviewing people regularly on AI. Which companies do you think will benefit most in terms of margins or just in, you know, industry in terms of productivity and the use of AI? So, AI, this is the fastest pace I've ever seen in my 30 years in the industry. if you look at the other platform shifting changes, the internet, cloud computing, mobile, I think this will trump all of them. This is the crescendo moment of all of that innovation. And it's moving at a pace that's faster than I've ever seen.

The innovations in mobile or cloud computing would happen every couple of months. The innovations happening in AI are happening every couple of days. And the reason is the people who are building AI are using AI to build AI, which is creating a loop of speed that is truly unprecedented. I asked my three guests last week, Brian Chesky from Airbnb, Aaron from Box, and Reid Hoffman from LinkedIn, how much faster their teams would be? And they said, oh, it depends on position, you know, it could be 50% faster, 20% faster. And I said in 2023, they all came to the conclusion that their teams would be 30 to 40% faster at their jobs, information workers. Now, for programmers, they're going 50%, 60%, 70% faster.

Not the best ones, but the weaker ones, because programming is a limited set of words. When you type, it knows what you're talking about, and it's going to fill it in. It's a constrained data set. And so they're doing great. Salespeople are doing great. Writers are doing great. Accountants, lawyers, all of that. And so what I predict will happen is there'll be one group of people

which embrace this and use this technology every day. They will get 30 to 40% faster at their jobs every year, which means every two years they will be twice as good at their job. I think there'll be a group of people who will not embrace this, just like we all know people who wouldn't. use PCs or Microsoft Office or email or the internet, and they retired and they were worked out of the workforce. You know, it might have taken 10 years back then for somebody to, you know,

everybody had that like two guys in the office who refused to have a PC on their desk and their assistants printed out the emails. They were old school. Took them 10 years to leave and they kind of got pushed out. This is going to happen in 10 months. If you're not using these tools, you'll be out of a job in 10 months, I predict, if you're working on a computer for most of your day. And it's not meant to be hyperbolic. I think it's going to make every company

two or three times more efficient and capital efficient. So for startups, that means one set. And for big companies, it means another set of complications. But it's happening, whether you like it or not. The number one thing you can do, how many people used chat GPT in the last 48 hours? Raise your hand, hi.

Great. It's half the audience used it in the last 48 hours. And Google Bard is pretty impressive as well. I mean, this stuff is moving at a fast clip. And for startup companies, you know, it's going to drop the cost of starting a company by 50%. For the large companies, I think Google, Facebook, Tesla, maybe not Tesla because of the factories, Uber, etc., I think they will be on hiring freeze indefinitely. they're never going to hire more people.

We can be sitting here five years from now and the number of employees at Google will be the same. So just let that sink in for a moment. It could even be less. And their revenue will grow 20% a year. So the revenue is going to triple every three or four years and their cost is going to stay the exact same. How is it possible?

Well, because they're getting 30% more efficient every year, they're not going to add people. So right now, those large companies can have 500,000 to $2 million a year in revenue per employee, they're going to be doing four, five, six, maybe even $10 million in revenue per employee. And that's how the world's going to bifurcate very quickly. And you're either going to be part of that massive, massive earnings that are going to explode at these big companies, or you're going to be left behind. Now, the good news is the problems that exist in the world, the number of problems, is enormous.

And so this isn't going to be cataclysmic for humanity in the short or midterm. I can't tell you 10 or 20 years from now when we have AGI, you know, general artificial intelligence, what's going to happen? But let's just say the next decade, you know, I'm pretty convinced that what's going to happen is, you know, this app that was too small or this piece of software that wasn't working. funding is going to be done by two people and get to five or ten million in revenue. So it's going to be amazing for humanity. Everybody's going to become a superhero.

So, you know, if you weren't a good writer or you weren't a good designer, you're going to be better than the great designers are today. So just let that sink in. How many people in here are developers? Who codes? Raise your hand really high. Or is a data scientist, either of those two. Okay, it's like five or six people.

It's a small percentage in this room. It's okay. Everybody in this room will be a different. data scientist and a developer within the next 12 months. Because you all, you, everybody in this room knows the programming language of the future. It's called English. And you're fluent in it, if you're here in all likelihood and listening to me, or you're doing a great job faking that you understand what I'm saying. There are people working on U.X where you can talk to and say, I want an app that looks like Uber, but more colorful.

and I want to log in screen that lets you put your phone number in and let you authenticate with your Google or Twitter account. And it designs it. And you're like, okay, you've seen it. Crazy, yeah. Yeah. And then there's another group of people

who are working on software where you say, make me an app like Uber, where you can track a cab with GPS and you can log in and it writes the code. And those two companies have had on the podcast, like multiple companies that do this, I've taken pitches from a dozen of them. And I'm like, can you just merge these two companies

and just publish the... code. So that's coming and it's going to make everybody a data scientist. You're going to be able to upload a data set and say, tell me what trends you see in this data set. And it's going to give you trends. You say, show me what charts should I make of this data? And it's just going to give you the charts. So the idea of going to somebody in your company and asking for those charts is going to be silly. So everybody's going to be a data scientist. Everybody's going to be a world class designer. Everybody's going to be a developer. Let that sink in and imagine a world where everybody in this room could create an app this weekend.

Everybody in this room could create a beautiful brochure and a design and a logo, and everybody could write, you know, beautiful copy. It's literally months away. Well, Jason, thank you so much. If you haven't listened to his Brian Chesky podcast, you definitely should. It's one of my favorites.

It looks like Ron's waving his hand. He has no voice, so I don't know how he's going to be able to ask a question. He wants to ask a question, okay. He's asking somebody else. to ask a question. This is all happening in real time. Byrne Bratton, celebrity for my connections. Okay, one question.

I like it. We'll take one question from the boss. Suspense here. Wow. I'd like to get a guy who I could whisper to and yell out questions. I want that guy. All right, Byrne, let's hear it.

So early stage, which I'm in, will be able to get that 10 or 15%. I think people doing the serious C and D rounds. It's a great question. The people doing the C&D rounds, the late stage, they might be a little more challenged because startups will not need as much capital. And so I think the venture tourists who came into venture over the last couple of years and thought capital was a mode and they could drop $100 million or $200 million into a startup,

the founders may not be down with that anymore or they might just want to sell you some shares. So yeah, venture capital should never have gotten as big as it was. It is a boutique industry that requires small amount of capitals, and small amount of capitals create constraint, and constraint makes for great art. Founders should not have five years of capital. Artists should not have unlimited canvases and paint. Filmmakers should not have unlimited film stock

and an army of actors. You need constraint. You need the pressure that creates the diamond. And that's why the last three, four, five years in Silicon Valley were so wasteful and so destructive. and so fattening and made people lazy. And now that it's over, oh, my job is so wonderful

because everybody starting a company is dogged and hardcore, and they don't want a lot of money. They want to just get a chance and they want to work hard. And we don't have all the tourists, both on the investor side, and we don't have the actors

pretending to be founders. Everybody thought they were a Jedi night the last five years. And now they're all dead. and all their arms have been cut off. Luke Skywalker and Obi-Wan Kenobi, there are 1% of the population. Founders are 1% of the population, the great ones.

And we funded another 20%. They should have never been funded. And now we're back to basics in Silicon Valley, fund less companies, more focus, less free shit in the office, less unlimited vacation, less nonsense, and just more team. product, customer.

That's it. That's the flywheel. Everything else is nonsense and a distraction. And, you know, it's the greatest job in the world what I do. And I really appreciate you inviting me, Ron, as well. Thank you so much for being. And it's been great to meet everybody. Thank you for having me.

Great. Thanks so much, Jason.