SPEAKER_01: when you sent a hotmail it said sent by hotmail or powered by hotmail yeah it's actually funny

SPEAKER_00: you should say this it was the idea to do that was tim draper's idea 100 and i i not me and and

SPEAKER_04: i just want to give credit to him because it was incredibly cheeky at the time to embed a commercial message involuntarily to everything sent by your customers so you signed up for hotmail it feels

SPEAKER_00: like a normal email account but now every message you sent has this get your free email at hotmail.com

SPEAKER_07: right like a call to action controversial this week in startups is brought to you by perspective ai surveys they never capture what customers are really thinking that's why we use perspective ai real insights straight from your customers and the first two months are on us just go to get perspective dot ai twist dot tech say it without saying it head to get dot tech slash twist or your favorite registrar to get a clean sharp dot tech domain today and northwest registered agent starting your business should be simple with northwest registered agent you can form your entire business identity in just 10 clicks in 10 minutes from llcs to trademarks domains to custom websites they've got you covered get more privacy more options and more done visit northwestregisteredagent.com twist today

SPEAKER_11: all right everybody welcome back to this week in startups you got a treat today one of my favorite

SPEAKER_13: human beings one of the greatest investors in the history of venture capital and one of the smartest most considered individuals uh who i am lucky enough to call a lifelong friend steve jervison is back on the program he is the co-founder of future ventures he's also the j in dfj if you remember that he's done four funds of 200 million dollars each at future ventures he's on his fourth fund these funds are for

SPEAKER_15: really really deep tech really challenging bets that other people aren't willing to make like the bets that steve made on companies that people predicted would fail spectacularly like spacex and tesla

SPEAKER_13: welcome back to this week in startups my guy steve germanson how you doing brother oh my gosh thank

SPEAKER_18: you that was the kindest intro i think i've ever received and very unbecoming of you to actually say

SPEAKER_21: nice things well we are friends and so i break chops and he breaks my chops but now that's all in love

SPEAKER_23: you're a huge nerd uh you know the best oh here we go now we got no but you're a huge node nerd in

SPEAKER_11: the best possible sense of the word you uh as recreation your hobby is sending up rockets so i'll go to dinner with you and you'll have me by your house and you say oh i gotta step outside for a second and launch this rocket um you just love deep tech yeah how'd you get that love of deep tech yeah it goes

SPEAKER_27: back to my earliest memories the first thing i ever bought with my allowance when i was i think five

SPEAKER_29: or six years old was a chemistry book it had a lot of cool pictures of like lab experiments but like literally was my first volitional purchase of my life which wow goes to show how far it runs and then the apple 2 is like this incredible wake up of like wow you can program in computers and i think some people

SPEAKER_31: who've had a taste of computer science just fall in love with it and so i did their own games and

SPEAKER_04: things and so short version is it's been my whole life i've struggled in the first few years to find a career that would really tap into that passion for lifelong learning about technology and where

SPEAKER_31: it's had taking us and what we can do with it um you know i bounced around a bit but when i finally found venture capital 30 years ago it was like wow it's the perfect match for me how did you get

SPEAKER_33: that apple 2 what year was that when did you first see it take me to that moment when you put your

SPEAKER_36: hands on that magical device and started typing oh it was amazing i remember exactly it was in seventh

SPEAKER_29: grade so i was about 13 years old um i was you know socially kind of maladapted i didn't really i

SPEAKER_04: mean this is 1980 no this would have been even before that this would have been around no no you're right around 19 i think it was around 79 though that i got it for some reason it was either 79 or yeah yeah yeah i think it was well maybe like 76 77 i wasn't the very first year that it came out but it was pretty close so let me let me give you a more precise answer no it would have been yeah around 78 79 and um and my my dad worked in the chip industry actually so he made memory chips he made the actual chips that i then manually plugged in to upgrade from 16 to 48k of memory which was like a big advanced killer bit right so yeah not mega but thousands and um my my eyes just lit up so i immediately got into basic and would write simple programs like you know print something go to ten and

SPEAKER_29: your dad bought this for you yeah it was a gift yep i don't remember if it was a birthday or christmas it just showed up and i just could not imagine anything else that i want to play with so i bounced around with legos and fisher technique this other kind of building block set but this when apple two

SPEAKER_31: came around that was it and i used to write games for it i wrote a mastermind game which is like this colored peg kind of game in a um some simple like almost like blast star that elon did like little but

SPEAKER_04: actually my little graphic so graphic sort of shooting them up games a um a text-based adventure game where you know enter words like go north go south you know like zork exactly just like zork bingo uh and i and i loved the games i mean there were oh my gosh i remember actually it's funny i've

SPEAKER_31: been talking to richard garriott a lot lately and over the last few years but there was a good i think 30

SPEAKER_29: year stretch where we didn't see each other but i but i was just a fan of ultima and i was a fan of um was it bill anyway oh shoot i forget the name that was a pinball construction kit and

SPEAKER_56: anyway i remember pinball construction kit yeah you got to make your own pinball game yeah that was

SPEAKER_04: so fun exactly that was incredible like and as a programmer i could tell those people know how to

SPEAKER_29: do things i have not figured out how to do yet with shape tables and flipping two memory registers for video ram like i could not figure out how they did what they did it was like squeezing insane performance out of this pretty simplistic machine in retrospect and that's a big ups to your

SPEAKER_13: dad because it was a thirteen hundred dollar computer at launch but that's in 1977 to 82 dollars was like i think the window the two came out which would be close to seven thousand dollars right now for the base model it was that's a significant purchase for your kid um but it was mind-blowing i got the pc junior my dad bought me the pc junior which was that keyboard yeah it was a it was a chiclet keyboard as you remember and then they subsequently came out with a better one but these were really the introduction of a whole generation to computers before that you had to go to a laboratory to use a computer and the pc kind of put one on everybody's desk it was kind of mind-blowing

SPEAKER_63: so did your school in um you know support computer programming at that time

SPEAKER_65: not initially so when i was in third grade i remember we had a teletype machine like you're talking about

SPEAKER_04: and like none of the kids used it it was often like the principal's office area where the staff and admins were but you could type things in it was like it was like a like a typewriter right and i think that's where paul allen and gates initially did their initial programming um so i had a briefest exposure to that but the overhead i can tell you was so high that it didn't really capture me in the same way right the delay the latency the lack of sort of a personal experience and it was in school and it was loud it just it just didn't i mean i played with it but it didn't really catch on for me then my high school which i was in at some great did there was a really great guy i remember

SPEAKER_29: his name was stutzman the teacher that supported this had both z80s which were the you know the trash 80s we called them which was the trs80 from radio shaft uh based on the z80 processor and the apple

SPEAKER_51: two based on the 6502 there it is this is the trs80 boom my school had this as well this was my first

SPEAKER_13: introduction to computing at severian high school in brooklyn which is like a dumb terminal uh now

SPEAKER_11: you must have gotten in trouble with some of this stuff like everybody who was on these things did

SPEAKER_04: something stupid or really did trouble with your story well funny i've never shared this nor have i even thought of it but the closest i got to because i didn't get much in the bulletin board systems where then you could network i mean i played them but i didn't use them enough to get in trouble there which probably is good but i did create a graphic representation of my most despised teacher

SPEAKER_29: i won't say who but it was an english teacher uh and uh some violent end coming to them uh and i think i got in trouble for that because the kids loved it was like look at the teacher opening the door and

SPEAKER_34: bet something bad happens like the rock falls in the head or something and uh but it was just an

SPEAKER_31: animation and the crude high-res graphics as they were called at the time um so that was about

SPEAKER_04: other than like just spending way too much time on it i didn't get into trouble there in fact if anything maybe it kept me out of trouble and i was more likely to get into trouble in the neighborhood

SPEAKER_56: um it's just idle time than i was you know what i did with the trs80 we had a dot matrix printer and as you remember at that time you had the you'd feed the paper in they were all connected and you

SPEAKER_11: have this big box right tall box wide paper it's a giant huge box you put underneath the printer so being the idiot i am i was like oh 10 uh say hello world 20 go to 10 30 print so i do this uh with a bunch of spaces and i start this program and i kid you not because i was just saying hello world one time per page it goes and the teacher had left the room he's in the back of the lab and i'm like oh my god so then everybody else starts sending print jobs to it the teacher lost his mind because somebody then just did one with blank pages it's shooting out the back but this was really like i think the

SPEAKER_87: start of elan's career as well is just getting onto one of these computers and then just going down the

SPEAKER_88: rabbit hole bulletin board systems as well hey listen we meet a lot of early stage founders here at

SPEAKER_24: launch my investment company and some they don't have a lot of traction yet they just have an idea maybe they haven't even finished their product they've just got an mvp but they still need investors and accelerators like ours to take them seriously and you know what we can't just wire money to your gmail or your paypal that's not how it works folks we need to know that you're a legit and official business we need to know your company is incorporated that's why you need northwest registered agent it's the service that will help you run your business the right way from day one in 10 clicks and in under 10 minutes you're going to file for your llc or a c corp if you're a startup get a domain name launch your official website claim your business email and even fast track your trademark application which some people forget to do we're talking about more than just company formation this is your entire identity as a business go to northwestregisteredagent.com twist and show the

SPEAKER_89: world you're in business and make sure you use that url slash twist so they know that we sent you

SPEAKER_91: tell me about the moment you heard the term venture capital ah i believe i was all the way at business

SPEAKER_04: school i think it was around 1993 or four probably where were you at school at stanford business locally

SPEAKER_31: yeah so i had followed a geek's path right i apple 2 programming electrical engineering studies did a bunch of stuff eventually even started a phd in double e focused on neural nets and ai of all things and how they map to parallel processing machines didn't want to go back to heal packard right and chip design so i did some engineering work over a number of summers there i did some

SPEAKER_04: summers at apple and next because i wanted to see steve jobs in action but that was in product marketing so i bounced around oh and then three and a half years in management consulting at bain for tech companies so i did consulting product marketing engineering i couldn't imagine 20 years of any of the ones that i had some exposure to um they were interesting learning experiences but even after a summer i might feel like you know how much more i'm going to learn in five years than i haven't already learned over a summer i mean pretty cheeky perhaps um so i was kind of a little lost but assuming i was just going to go back to bain the management consulting firm that was my assumption going into business school then out of nowhere in the beginning beginning of my second year i think it was probably uh october um i got a call from this guy chip hazard i remember specifically he was at graylock venture capital firm that i sort of maybe had heard of but they'd have no website there were no websites in fact there were only two venture firms that had websites a year later um so 94 there were none and um and i was like what is this all about but he i'd worked with him at bain uh he was

SPEAKER_31: here my senior he had you know already graduated gotten the graylock on the east coast and they wanted to create a west coast presence and grow the office out here and so lo and behold i'm like wow this

SPEAKER_04: from his description sounds interesting let me let me look at this um i had never met a venture capitalist um knew nothing about it and there was no internet so um i mean there was but it wasn't

SPEAKER_99: really like we know it today and it was bitnet it was arpanet it was oh yeah there wasn't like oh

SPEAKER_04: let me just look up what is venture capital and have the ai explain it to me right and so i did happen to find a gold mine for me was a fellow named chris alden who was a co-founder of a

David Friedberg: magazine called red herring yes i know chris upside was the other magazine right now and they were

SPEAKER_104: chronicling venture capital that's right in the early 90s exactly print format you get like a

SPEAKER_105: subscription and chris in particular did a uh monthly i think it was monthly uh column called vc whispers

SPEAKER_04: where he would talk to venture capitalists and ask them what do you actually do what what is your firm actually like and how is it different from others information that was impossible to come by otherwise and so he was really great at like pointing me to well you know graylops at that time

SPEAKER_29: uh they morphed a bit but at that time a very conservative white shoe uh buttoned down kind of firm like like the opposite of me same for sutter hill at the time uh and again they've they've

SPEAKER_31: changed a bit but at the time they were very different from the firm i joined which was called draper associates um but folks like chris helped point me out point me to the sort of wide spectrum of diversity there actually was within the culture and uh sort of strategy of venture firms some were entrepreneurial and risk takers others were very strangely very conservative and analytical most more like investment banker types um which would not have probably been a good match for me so i really credit chris and a couple other folks i spoke with to help steer me to well you know they're not all the same there's some firms that you might like more than others and um luckily ended up at the firm i did because it was very entrepreneurial and you know helped me pursue what was it ever changing not ever changing but sort of a drifting uh focus area over time from internet

SPEAKER_108: to deep tech um tell me about the first investment you were ever involved in because you must have diligence and met with hundreds of founders but there's always that first check uh that you are

SPEAKER_113: responsible for champion and putting in so i'm pretty sure the first three and i'll try to get the order right were fast parts interwoven and hotmail and two of those three were well no that's um well okay so

SPEAKER_31: i think fast parts might have been first and it's long gone no one's ever heard of it but it was a very

SPEAKER_04: unusual sort of b2b if you will uh trading exchange that dealt in the gray market of semiconductors and i had the only reason this one was really interesting to me is i had done a um i don't know if you call it a case study or an entrepreneurial exercise at business school almost with the same idea as the entrepreneur i was like whoa like it clicked on so many levels i'm like wait wait you're doing that um so to make almost very short there was a you know perhaps a a lesson there for me about not relying too heavily on something i think i know a lot about um instead of is this actually a good business opportunity independent of what i might know about so like the fact that it clicked was like oh my god everything lit up for me like that's something i was thinking about starting

SPEAKER_31: um but it ended up just going out of business um the interwoven went public uh it was a sort of web

SPEAKER_04: internet tools company for content management went you know as his websites and web properties got more complex how do you manage all these assets um almost like an infrastructure layer for the internet and then there was hotmail the hotmail was a really yeah that was like the one people have heard of boom like yeah that was the first one that got visibility bought by microsoft bingo for about 400 million dollars uh you know less than two years in and the thing that it really epitomized was first you know rapid internet growth for us second this thing we basically coined the term viral

SPEAKER_31: marketing because of hotmail trying to describe for a blog post what is it that hotmail did that was

SPEAKER_29: so different from other internet companies this idea that the message itself was the vector of uh of spread of marketing in a sense no like a virus which came from the signature when you sent a hot

SPEAKER_01: mail it said sent by hotmail or powered by hotmail yeah it's actually funny you say this it was the

SPEAKER_00: idea to do that was tim draper's idea 100 and i i not me and and i just want to give credit to him

SPEAKER_04: because it was incredibly cheeky at the time to embed a commercial message involuntarily to everything sent by your customers so you signed up for hotmail it feels like a normal email account but now every

SPEAKER_00: message you sent has this get your free email at hotmail.com right like like a call to action controversial exactly tim wanted it more controversial he was pounding the table that

SPEAKER_72: he wanted to say as if it was written by the sender p.s i love hotmail i love you

SPEAKER_125: guy the founders wanted like nothing to do that they're like spamming you know accusation people forget

SPEAKER_15: and you could explain to people that the internet was non-commercial that's right at its start and the idea of doing even a zima malt ice liquor ad uh banner ad on wired was was faced with fierce

SPEAKER_24: resistance from the venture entrepreneurial and early internet community that's right that's right so you

SPEAKER_04: know it was definitely an aggressive move and it made all the difference on their rate of growth because they spent nothing on marketing they spread globally uh the the actual number of users was scaling at such a rate that hadn't been seen before that that's what i wanted to describe what is this

SPEAKER_31: process by which they tell two people and then they tell the people you know this geometric explosion which by the way skype then used later for voice and other companies use for video and it sort of you know was a playbook if you will for how best to grow a consumer facing um business and there were a bunch of precepts and things that related to that like you know what is your viral coefficient can it can anyone who receives the message actually take action you have to be multi-platform there's a bunch of things that relate to it but it's but going back to your original question it was my first success that had visibility we had a whole bunch of internet companies by the way in the 90s so i joined in 1995 and that was perhaps in retrospect the best time to join the venture industry by just by pure dumb luck the internet was exploding it was like shooting fish in a barrel make money frankly in these companies for a period of you know four to five years then and we were the most active venture

SPEAKER_04: capital firm if you can believe it in internet investing in 95 96 we did about a third of all internet investments like amazing dfj of all yeah yeah it was called draper associates but it eventually became dfj a few a couple years later um but we did a lot and we learned a lot quickly and then something really big changed by the way around 99 i was still super gung-ho about the internet and its potential i know that because i wrote articles saying i'm super gung-ho but i pivoted completely away and stopped investing in internet companies and the reason wasn't because i saw a crash coming i was not able to foretell that with any consciousness it was that everything was looking the same it was really starting to get boring which is like derivative products yet another yeah way to divvy up a market and sell to consumers yet another way to do a bb training exchange those b2c b2c they open b2b and b2c and they all look the same just variants on a theme uh we had not yet had social media you know have its boom we not yet have things like uber and all those that were yet to come at that point it's like every business plan was like just another variation on a theme like this it's not interesting let me and in the end of gut sense it's not diversifying to portfolio it's probably not wise investment to just be paying higher and higher prices for more of the same even though a lot of

SPEAKER_31: other adventure rooms explicitly had that as a strategy like benchmark there's a book about let's just do nothing but x y or z because we're making so much money in x y and z like why look at anything else so you could have made the argument rightly so that it's foolish to invest in biotech it's foolish to invest in semiconductors what why nothing could beat the internet for rate of value

SPEAKER_141: creation for a while but i pivoted hard to something that actually turned out to be not interesting nanotech

SPEAKER_143: we all understand the importance of a crisp memorable easy to spell domain name one of those names you can say over the phone and people know how to type it in without asking you the spelling but let's get real the good ones are either taken or there's some poacher who's holding it and waiting for some huge payday and they don't reply to you if you even if you want to pay for a premium domain you don't want to use up all your runway on a domain name that's just the truth for a startup you want to put that valuable cash back into your startup's operations so you should consider this dot tech domain you can get a clean crisp super memorable name for your website and company and signal out loud to your customers and investors we're a tech company that's instant branding for you that's why over 500 000 founders have collectively raised over 5 billion in investment building their companies on dot tech so skip the hassle head to www dot get dot tech slash twist or go to your favorite registrar and

SPEAKER_145: grab your dot tech domain today you know if you will um but it was such an interesting space explain to

SPEAKER_15: people you know mit and and you know material science and what the promise was at that time because william gibson was writing these crazy stories of tiny little nanobots building a skyline in real time

SPEAKER_150: we thought that this was the micro robotics future that would change everything and it didn't why what

SPEAKER_31: was it yeah yeah that's a great point there were a lot of sci-fi influences you're right all those the

SPEAKER_04: sort of the neural lace that uh ian banks writes about and was the you know inspiration for neural link there was um eric drexler's book in like 85 i remember reading it maybe 86 i think it was 85 um engines of creation which was just a deep dive into if here's the big thought experiment if you could put an atom wherever you wanted and you weren't burdened by where we are and the tools we have today but if somehow we could imagine that future where we have atomic precision and you could imagine if you could do that you could also build machines that could build more machines of their same ilk you could have

SPEAKER_31: self-replicating machines uh there were a lot of just analyses being done by foresight institute and by eric himself that just pointed out like oh my gosh this is insane it's so mind-bendingly different from what we're used to uh the absence of friction on some of these rotary bearings the rate at which these things would mechanically move you could actually imagine a mechanical computer if you will uh that outperforms anything we know of with a better energy footprint and such and so there were books like diamond age by neil stevenson as well that talks about this future so a lot of sci-fi precursors or people were thinking about this and maybe i fell prey to some of that to say well that's motivating to look at how do we get there how do we get from where we are today to there and so i wrote some blog posts um called transcending moore's law was one of them it was actually of all things in a law journal

SPEAKER_04: focused in nanotech the first competition of all places but it was it seemed clear to me that there were there was a problem which is you can't manipulate atoms today with that precision there was like atomic force microscopes and stuff but nothing that's scaled so i described two ways to get to this future we might imagine one is the bottom-up kind of um organic bio-inspired uh path which is let's use the tools of biology like uh you know the the ribosome that that uh the conductings or uh you know now what we call crisper another molecular tool we didn't have those words back then but you know are there molecular machinery uh that we can harness perhaps using dna as a structural material perhaps self-assembling uh molecular films which i did invest in to create a better memory chip a variety of things where you engage processes that work already at the nano scale if you will and build up from there over time so that felt like powerful and immediate the other path with most people were thinking about was a top-down approach to say let's start with actual businesses that exist today

SPEAKER_31: that sound like this like the semiconductor industry hey they want to make things smaller they're

SPEAKER_04: trying to scale down to nanometer scales like why not just work our way down where you inherit the interfaces to the real world from above in other words if you had a chip of a certain scale and interconnect from broadcom whoever that works let's just figure out how we can make smaller smaller

SPEAKER_31: things but harness to what we have from above and that just was going to take like 20 years estimated maybe 30 but i was it was like a long time like there aren't nanotech opportunities there in the near term uh and that's probably still true today it's it's slow but sure we're

SPEAKER_04: invested in some things like lace lithography that'll hopefully get us there but it's it's taking a while the bottom up so this was in a long-winded way a gateway for me to get more and more fascinated about the information systems biology what we can learn from biology that applies to it the inner cross-pollination of ideas between what formerly were completely different investment domains like the biotech investors were different people right different people doing different investments that had no interface whatsoever to it you gave an incredible talk one of the highest

SPEAKER_13: rated talks we've ever had at any event i do this liquidity event used to be called angel summit

SPEAKER_150: we should pull up the deck here and just go to morse directly to the morse law slide and you know this is something that i don't think people talk about all that much but when we were coming up more's law kind of ruled everybody and then as you and i uh you 30 years me just over 10 in venture the power law these are two laws that i think rule and dominate our lives here's 128 years of moore's law walk people through this um and why it's so important and why you think about it

SPEAKER_31: yeah it's the thing i think about the most that also seems the most descriptive for understanding the

SPEAKER_04: world we live in today and where we're heading over the next five to ten years it has incredible predictive power so let me set it up and describe it so uh the years on the bottom pretty easy to understand we're covering 100 and almost 130 years here of time the dots are the best price performance computer ballpark you know like there aren't any dots above the line that we know of but there are plenty below the line sort of the frontier of the best price performance computer of the day and the

SPEAKER_31: axis most importantly on the y is the choice of what it is and that it's logarithmic meaning every tick mark there is 100 x 100 x 100 x right so a straight line on a graph like this is an exponential and if

SPEAKER_04: you eyeball it it almost looks like it's a slightly upticking curve on another double exponential but what the axis is showing is how much computation can you buy for a dollar constant dollar meaning

SPEAKER_31: inflation adjusted so what's fascinating about this is that first it's looks like it's on rails for 130 years this is kind of mind-blowing and it's covering entirely different technology substrates so on the far left you have mechanical uh devices you know that that did the census in 1890 i mean like literally machines going back to 1890 you have the relay-based computer that cracked the nazi enigma code if you watch the movie um imitation game you have vacuum tube-based computers that predicted eisenhower's win in 1956 you have discrete transistors which were all the rage uh in the 50s and 60s and then you have the integrated circuit era that started interestingly with the lunar module guidance computer which i have

SPEAKER_04: around the corner here in the office and one of the early ibm's uh machines of 360 were some of the earliest to bet on integrated circuits yeah the the amount of compute here

SPEAKER_24: and by year it is every 18 months it doubles is technically moore's law's definition yes that's

SPEAKER_04: a fun uh good that you mentioned that if you ask anybody on the street what is moore's law you will get different answers but probably most will say what you just said doubling every 18 months turns out i learned it in computer science school exactly it turns out gordon moore never said that in 1965 he predicted actually what he predicted it was very peculiar to the integrated circuit industry in fab yield optimization which was what will be the number of transistors on the ideal die size because you can choose they're on big chips with a certain error rate or small ones that have less errors but what's that economic trade-off point of the ideal sweet spot and he had like five data points and he just cheekily you know predicted a line but never said any text about what it was and that initial line was doubling every year then in 1975 you modified it to say doubling every two years and today people kind of wave their hand and say every 18 months but that's largely being dictated by intel in its personal trajectory it doesn't actually relate to this curve because on this curve we're not counting transistors no one buys transistors like that's a weird count like how many transistors on a chip who cares how about how much memory do i have and if it takes too many trans if it takes more transistors as it does today than ever before to store a bit of memory why are we counting transitions let's count actual things that matter memory storage or computation this is computation now this one has been doubling every year for 130 years and then that adds up by the way so this graph is covering a thousand billion billion fold improvement in price performance of computing and it is almost cosmologically bizarre that isn't that that we're on a curve like this but most of this time no one knew they were on the curve they weren't like building to the curve they didn't know no one had graphed this until 1999 time frame which was ray kurzwell and then i've kept it up with the colored dots since then as we've moved from one substrate to another and it's been exogenous to the economy world war one world war two the great depression had no impact on this compounding capacity to compute it's really wild right

SPEAKER_36: it's it's very strange nature has the universe dare i say uh has some rules to it that emerge and we

SPEAKER_13: become aware of them as you're pointing out when ray sort of made this chart it's like wait a second

SPEAKER_87: all right at my founder university my number one rule is to listen to your customers why well delighting people who use your product it's like job number one for founders but how do you know what your customers really think about you well we all know surveys kind of useless people just tell you what you want to hear or they just click the number eight of ten over and over and over again what you really need is perspective ai you just give perspective ai a simple prompt telling it what you want to know am i connecting with the right customers is this new feature working is my ui clear and easy to navigate whatever question you have as a founder of your company and they with their incredible ai interviewer will get to work talking to real people about your product they'll do interviews using ai we've been using perspective ai for just a few weeks now at this week in startups and we've learned a ton about you our loyal listeners for example joe wants to hear more about building ai startups and fine-tuning llms so we're putting that into our docket into the show and tony is a solo founder working in edtech and he thinks the show features too much political news so we're dialing that back and we were able to set this whole thing up and start generating these reports in minutes so here's your call to action sign up today at get perspective dot ai twist to get two

SPEAKER_182: months free you got to try this product it's incredible get perspective dot ai twist we're following

SPEAKER_24: a pattern that we didn't know we were on which makes one wonder today what patterns are we on

SPEAKER_13: that we can't see because we're too close to the proverbial elephant you know we see a gray wall but if we step back in 20 years you know with ai and we're going to get to that what is actually going on here and what your chart shows as you continue it if we fast forward past intel and that's really

SPEAKER_15: interesting about intel like in some ways you have nvidia taking over but intel was a bit of in their

SPEAKER_24: minds a marketing kind of channel management process to try to double every 18 months correct

SPEAKER_04: well they had a variety of tricks they they tried to build to the curve they did definitely with intel inside to try to corner the market but what they missed was a transition to a fine-grained architecture that you can just or another way of phrasing it is intel in their development of the cpu the you know the pentium onward um was using human ingenuity to try to build a better and better single processor and then a few multi-core but ultimately nothing like nvidia or the asic or the custom chips to follow that fundamentally ran out of steam there wasn't that much advance you can make while being backward compatible and the the later sort of late cycle intel chips were mostly memory by the way they were like 99.5 percent memory so much of cache memory basically local cache memory that improves performance and that's how they ate up tons of transistors but ultimately weren't delivering that much more value uh compared to a completely different huge array of computational elements in a gpu or an nvidia chip uh that is inherently better suited to ai workloads um which right going back to my old phd in double e's like that's what you want a bunch of local memory

SPEAKER_31: a bunch of local computation that's more akin to how the brain works frankly almost almost like recapitulating uh our own evolutionary um advance with the cortex over simpler sort of limbic brain region so there's like an analogy to how our brain evolved to how computation for ai is evolving but basically over 10 years about 15 years ago intel was no longer the frontier of moore's law really

SPEAKER_04: we shouldn't listen to anything they have to say about moore's law even today oh wow oh yeah that's cool

SPEAKER_72: that's the reactions on zoom exactly um in any case nvidia has taken that baton and has for the last decade

SPEAKER_04: and lately there are a lot of custom silicon solutions like google has their tpu and attention to the process amazon's developed their own chips open ai of course is working on their own chip

SPEAKER_31: all the major ai you know companies that you know of have their own semiconductor efforts underway because that is inherently better way to do ai workloads where you're doing the same matrix multiplying ad over and over and over again it's so interesting too if we go back to business because

SPEAKER_195: we're kind of looking at the spiritual here like how does this thing exist in the universe

SPEAKER_13: then you have the business of like well how do we capitalize on it and i should point it out like no matter what's going on in the world the vietnam war great recession you know whatever it is the depression great financial crisis it just keeps going okay this is huge mystery and super fascinating

SPEAKER_11: it makes you think about god and what's powering all this we can get to that too that's why i said

SPEAKER_04: it's almost cosmological it's like what is the point of intelligence what is the point of life and it might be an ever expanding um understanding of the universe that you can another way to frame it abstract is you can think that every idea is a recombination of prior ideas they're always building on the shoulders of giants before you like recombining two things or combining two ideas across academic disciplines that hadn't been combined before and therefore the number of possible ideas in the pool of ideas that humanity has is growing combinatorially the number of subsets you can draw around two three or four ideas that are interesting to recombine the new product um is is growing as reads a lot goes to the nth power so n ideas to the nth possible pairings or or subgroups um that could be the fundamental dynamo of this perpetual sense of accelerating change that we're that like every year feel like oh my god we've been so much more than the prior year lately we notice it year to year

SPEAKER_29: but throughout most of human history this would have been like you know a century passes and a little

SPEAKER_36: but sure like here's the steam engine oh wow the wright brothers were able to get off the planet earth for a couple of minutes like right super interesting how our brains can figure that out and if you look at a company like intel and there's been other ones ibm microsoft that missed paradigm shifts

SPEAKER_13: it's so unbelievable how predictable it is that they can't make the jump and we were watching the cpus

SPEAKER_87: from in the pentiums and the little ding ding ding you know sound that intel would make for their commercials intel inside yeah but they couldn't figure out that nobody was buying new laptops because the cpus did not change the experience now you said earlier in the conversation with the apple 2 you you took the chips you bent them you put them into the you know motherboard and you had a totally different experience but again that that wasn't going to change things and uh they eventually the determining factor when you bought a computer somewhere in the 90s when you were playing quake or you know call of duty or anything in between became the gpu and the processing power of that and that was something that consumers human beings organisms would respond to so there's like another layer here of uh mystery which is you're trying to innovate but then the consumer the human being stops responding to one idea and starts responding to another maybe incorporate that chaotic concept well yeah so i think

SPEAKER_31: it's fascinating to think about the the rise of gpus so first as you mentioned it was a way to do polygon rendering you know in high speed so at its core it's somewhat akin to the sensory cortex in a

SPEAKER_04: way that you have this massive um representation of computation in parallel right across a visual field and you're trying to in distributing computation across all that now it was all initially developed for gaming right okay we're trying to rep you could almost to represent the world in a simulation if you

SPEAKER_31: really want to get abstract about it yeah but the visual side of that it is almost beautiful and uh

SPEAKER_04: poetic that that exact substrate is so useful for various forms of scientific computing and there were

SPEAKER_31: early experiments about 20 years ago um there were about five of them that nvidia supported it was almost like a side project like a crazy little side thing where they're like could these gpus be used

SPEAKER_04: for something else and a friend of mine this guy paul rhodes actually started coming called evolve machines that was doing um neuronal modeling basically can we model how a neuron works and then a cluster of neurons then an entire maybe cortical column using gpus as a substrate i remember this was the first eye-opening moment for me um that he said i went to fry's electronics which is a local store that used to be here to sell stuff and i've bought the equivalent of like one of the most powerful supercomputers on earth for just a few thousand dollars like i literally in what i'm doing right here at home in his living room can out compete like the national labs in this molecular modeling i'm sorry this yeah it was sort of was ion channels molecular modeling and cellular modeling so this scientific modeling task with this handful of things i just bought at fries like on a weekend crazy and so i was like i did blog posts about that i'm like what is this thing how can this possibly be that was before the ai application came right so in 2012 with image net competition you know there's this thing called alex net that is famous this um this contest that fefe lee at stanford has been holding and the neural net approach just dominated and this was a guy that used like one or two gpus like that was it like you guys have all this computation i'm using a couple gpus and i'm blowing the doors off like finding you know is that a cat or a dog or a tractor you know what time period was this that this 2012 2012 2012 yep i remember vividly because as a you know someone who started his phd in this exact field on this exact idea which is how can you map neural nets to parallel processing machines i was like what and so we made our we started looking at ai investing back then there were no no venture firm had ai on their website as a sector of

SPEAKER_217: interest yeah by the way it's like there was machine learning on the margins right bingo yep

SPEAKER_04: and and and even that vernacular started to come in later when deep learning as a term came up and around 2012 2014 there was this just this renaissance we invested in our first ai chip company in 2014 this guy giving rao it started a company called nirvana and the idea that he and many that followed had was wow let's build a custom silicon chip that's even better than nvidia because if the idea was wow we can shoehorn the nvidia chip built for graphics and gaming and we you know at the time thought they're going to be focused primarily on gaming because that's where they're you know making most of their money and this ai stuff will be an afterthought that was in

SPEAKER_31: 2014 let's make a dedicated chip that does nothing but ai acceleration you know and optimize it even further still more units more local memory switch fabric like you'd have in any networking chip in the 90s and like boom you just throw it together and it makes perfect sense it's like in some ways

SPEAKER_04: so obvious that's why you have like 40 companies that tried you know about a five-year period to do something similar um and uh that's really carrying more's law forward now like the majority of compute deep in the walls of google and elsewhere on these custom chips and and it's kind of interesting if you think about this migration single processor you know massively parallel or you know fairly parallel gpu now massively parable uh parallel uh custom chips and then potentially we think the next step would be analog chips so going even closer still to how the brain works do things in

SPEAKER_31: the analog domain instead of digital just like our brain does you know this this is like 40 watts and

SPEAKER_220: nothing we're building in silicon today competes with it on power for calculation right but but it is

SPEAKER_150: possible it is interesting that you bring up the wattage this year actually just in the last 90 days

SPEAKER_13: people stopped talking about how many h100s they were building you know remember when colossus got built in under three months a hundred thousand h1s because this is crazy and now all the announcements are being done in how many measured in watts not yeah and so now you start to get to the the meaning of the universe or the drivers of the universe energy and the ratio of how much energy it takes to process the

SPEAKER_87: world and then you start thinking about our own biology to your point this giant brain being powered by some number of calories we've consumed some animal proteins some oranges whatever it is now and you mentioned stim simulation theory and the great breakthrough comes not from how we think but from how we see and process the world or visual creatures the world is visual you know now we start

SPEAKER_150: getting into consciousness you know and the nature of what's going on in our brains and what's going

SPEAKER_234: on in these giant clusters where does that lead you well it's an interesting line of thinking which is what is you know in what ways are there commonalities in the information processing dynamics of the world so

SPEAKER_04: you could for example make the argument that what we're doing with this incredible influx of information to our retina uh into our optics system is information reduction what are the heuristics simplifications ways that i can down grade what is a overwhelming amount of data if you just did the brute force you know how many you know how many pixels times what per second you know what is the input of information right now every day as we're awake with our eyes open into our brain it's overwhelming and the way we do simplification representation of the worlds in 3d um constructs and model building within cortical columns upstream of the visual system there's this sort of pattern throughout our cortex and in computation in our neural nets so many areas where they're similar in this regard where you're reducing uh you're basically finding you know that the hidden if you have latent space that represents what we're seeing and understanding in a more compressed form and the way that that computation takes place in our brain is similar to the way it does in our neural nets that we train the

SPEAKER_31: way that we in a sense grow these things the processing of of information to train these nets is is very similar and so some ways it's it's beautiful that the world allows itself to be to have this compressed

SPEAKER_04: representation that for example the laws of physics are all very low order polynomials that the the

SPEAKER_29: the hidden formula isn't all possible formulas to be like you know what is the trajectory of something if you're trying to just deduce it from data points like what what is this trajectory of an object

SPEAKER_04: it's probably going to be similar order polynomial in almost all physics that describe things so there's this almost natural way in which the world around us lends itself to neural nets and by analogy

SPEAKER_31: that neural nets really came from mimicking the brain but in an abstract way you know neurons weights are like you know like like an axon there's synapse the similarities of the weighting between these

SPEAKER_04: nodes that our brain had to do this like it was an evolutionary requirement to do data reduction model building on the fly and the way the whole brain works is it's predicting in a sense of the next token it's predicting the next thing you're going to experience across all your senses visual tactile auditory and only when something differs from what you just predicted are you even conscious

SPEAKER_36: that it just happened when you look at the landscape it's kind of like a security camera that's like nothing's changed nothing's changed and then mountain lion okay i'm going to send an

SPEAKER_242: alert exactly your security camera sends you alert somebody's hopping the fence at your house or you're looking across the serengeti and oh yeah there's a lion that's not good and things get triggered and we call it intuition we call it you know we have this crazy reaction system dopamine uh cortisol that dump these you know compounds into our body to generate a reaction and i'm trying to think here what what is the equivalent of computing when you know what what chemical do we dump into the computer to have it pay more attention i don't know if that's an algorithm or you know what the analogy is here and

SPEAKER_13: sometimes these analogies break down but i guess when we get off this chart uh and we you know we drop

SPEAKER_150: nvidia off what do you think's next is it something like cortical labs i don't know if you've seen this company we had them on the program uh actually growing neurons on silicon right yeah well yeah it's biological compute platform so they're they're literally combining lab-grown neurons with silicon chips and then make it available that seems like a moonshot and then of course there's quantum what

SPEAKER_247: what what takes the baton from you know these gpus right 20 years right so that asics are the obvious

SPEAKER_29: like we're in the middle of that transition in various places so one of the reasons i think the data centers talk about wattage is they don't necessarily specify oh is it h100 is like it's specifically nvidia or is it going to be my internal team that's been trying to build a chip that's compliant so they're trying to do apples to apples say we're not sure what the chips are going to be but they'll be this wattage level um i think analog is the next step and um and quantum is really a left turn we can get to that in a moment because there are things in quantum machine learning but it gets pretty complicated pretty quickly and so let's go back to what i think is in the immediate

SPEAKER_04: term there are a handful of companies that are not using little neurons because the little neurons thing i might put more in the nanotake bucket of um it's a difficult interface um to manage but if you go to just analog silicon you have this capacity to do some pretty crazy things so a company on mythic that we invested in recently uh for example can store eight bits of information

SPEAKER_31: so you think eight bits is you know like in the old apple 2 would have been eight of those chips in a

SPEAKER_04: row for you know a byte of information uh they can store eight eight bits of memory in a single transistor now that is my kind of mind-blowing because the way that normally happens on a digital chip is you have these sram banks you know static ram banks each of which have you know eight or so transistors per bit and then there's error correction code because they're so you know as they get smaller and smaller they get more erroneous you have all these extra ones you have readout stuff around them it's it's it's a lot of transistors but in the analog domain in a single flash memory transistor you can actually do the matrix multiply and add where the add is just the wire of common current and you have a row of these transistors make a long story short they've shown this you can make basically a neural network chip that works in the analog domain that's you know a thousand x better on power for example back to power per calculation so more akin to the brain so you know massively parallel slow and low power uh is where that vector would be taking us and there are others brand new our most recent investment just last week and conventional labs they aren't really saying exactly what they're doing but they're in this domain also of analog and biomimicry it's the same guy by the way he started nirvana in 2014 that i mentioned that when i first invested in semiconductors for ai then he started another company the databricks bought now this is his third startup anyway um so analog has a lot of potential it in some ways feels like the uh natural trajectory of this recapitulation of our biology if you will in the substrate and uh and it has applications not only in traditional ai as you might think of it but also really small neural networks that you might stick in everything so jensen ceo of nvidia a few couple years back said they were about to enter a area of explosive growth in ai like nothing we've ever seen before and he wasn't talking about any product and videos about the launch it was edge ai basically putting little neural networks trillions of little neural networks little brains into everything every security camera should have one obviously every car every car is going to have them every moving object every autonomous vehicle wearable every sensor could it's like wouldn't you want a little more intelligence than any consumer product you can think of yeah right um and by the way this so imagine like a voice interface for example that's speaker independent large vocabulary so it actually works uh could be in anything for less than the cost of the plastic buttons it would replace so if you have a roomba scooting around on your floor instead of bending over to push a button for any particular use case you just call alto and say hey like don't go in the corner hey can you get the bedroom or you know don't stop bugging the cat whatever it might be

SPEAKER_31: um it could be that intelligent and it would be local intelligence that doesn't require the latency

SPEAKER_108: or any of the overhead no cloud no security issues privacy issues yeah local local data retention yeah

SPEAKER_36: such an interesting concept i want to take a pause for the cause here and show a clip of you 10 years ago or so uh talking about robotics you and i haven't seen the clip my crack research team are going to play

SPEAKER_270: for us so we get our reaction to it in real time in real time so uh producer oliver or as i call him master oliver the young master oliver here we go let me just get my screen ready to go here here he is baby steve go okay try one more time with sound let's reshare and do sound then i'm going to

SPEAKER_150: do a pickup for you i found the uh i think i found clinton talking about nano which will splice into the other

SPEAKER_273: one here we go something near term that is kind of interesting led by rodney brooks um uh and there's

SPEAKER_275: more than one company doing this uh now which is humanoid robots for the workplace so the reason i hesitate about san francisco would be i'm not sure if it'll percolate into i mean even though i bought one of these just on a large um the average person doesn't really nor do i have a real use for it yet um but in a work context uh these are two-handed robots right now on a pedestal so they don't walk around but they you program just by moving the arms so anybody can program uh so yeah it was hard

SPEAKER_278: for me to even hear it you do you have a response to it steve sure so rethink robotics was a company

SPEAKER_128: rodney brooks started a famous mit roboticist who had a documentary once about him called fast cheap and

SPEAKER_04: out of control and he had used biological metaphors in a way to think about how we're going to build robotics and control systems and the insight that rethink had was wow if we could use the really cheap now available motion sensors that we all have in our phone that allow us to know exactly how we're tilting right this multi-axis accelerometer put several of them over an arm we use the feedback

SPEAKER_31: loop and control layers to use very cheap springs and actuators and create fluid motion in a humanoid

SPEAKER_04: robot form uh the second inst or thought was what would these robots be good for well as long as they have the same lifting capacity accuracy precision whatever as a human then you don't have to ask the question you say wherever you have in that in their particular case a sedentary human doing some repetitive task um or even you know less repetitive tasks we're basically sitting at a desk um you could just swap out the robot right at a much lower cost challenge uh for that company ultimately failed it came really close to having a great exit and an acquisition strangely by a chinese company that very

SPEAKER_31: much wanted this technology but the u.s sisyphus laws that basically prevented technology transfer shot that down and instead become another business uh brutal it's so hard when you have the right idea and the

SPEAKER_287: timing's a bit off or the go-to market give me the post-mortem here what didn't go right so what didn't

SPEAKER_04: go right was at the end of the day they called it sort of like the shaky bot there was this idea of am i drifting for what i'm doing let me quickly correct it there's some history system that so they realized towards the end as they were running out of capital that they needed to design their own

SPEAKER_31: motors and fast forward to today that's the same conclusion that xai and optimus and all the not xa excuse me optimus and let me just say other ai companies developing humanoid robots have all

SPEAKER_04: realized oh my god we got to build the entire stack the supply chain yes elon speciality is building the

SPEAKER_292: entire stack so that's his power alley yep yeah but that is today i didn't know it right i mean if we

SPEAKER_36: look back on the history of it one of the seminal moments at tesla i remember sitting with elon when these uh yeah dipshits who were making the first roadster came to him and were like the parts

SPEAKER_242: combined equal more than the cost of the car and you and i paid 150k for that roadster it was like wait a second you're spending more on the parts and he had to deal with part suppliers and he learned this brutally hard lesson which is your production is as fast as your weakest supplier and your your

SPEAKER_87: products as good as the worst component in some ways you know or it could be and man it's just amazing how over yeah close to 20 years now he's just decided it i'm making an hvac uh you know and

SPEAKER_150: i don't need the steering column i guess the model s the original steering column and the yeah they got

SPEAKER_128: from taylor yeah they got from mercedes yeah yep so that by the way uh you know when you're starting

SPEAKER_04: you don't have much capital so the roadster era they had to use as best they could off the shelf parts

SPEAKER_31: so famously the battery cell was the big leap forward every other electric car company that followed one of these prismatic pouch cells that are custom made for automotive and they're like let's just use the same same laptop cells dell's using and shipping and volume right and let's

SPEAKER_04: you know use lotus uh frame which had its own trade-offs but basically whatever possible use off the shelf but you're exactly right they hit all these enormous headaches like the transmission so that the roadster almost didn't ship ship because they needed a two-speed transmission they tried three different vendors like borg wagner and all these different ones and no one could make a transmission they could take the torque i mean it was so much torque just the two-speed transmission we just rip the transmission in half you just destroy it and it wasn't until the special bipolar transistor came out where jb struggle was like wow if we switch to this transistor the latest and greatest transistor we don't need a transmission we just have a single fixed gear ratio and you just basically have to go up to really high rpms they're really low and everything that you've experienced in every tesla from that point onward doesn't have a gear shifter right when you want to go backwards the engine just runs the motor it just runs backwards so there's no there's no shifting there's nothing like you have in every gas car that exists on the earth today so um that was just the beginning of a whole string of like everything you could imagine uh there was one time when elon personally went to fries back to fries to get you know ethernet cable that they needed to be able to

SPEAKER_31: keep shipping vehicles because the ethernet cable didn't exist from the some supplier same thing under the covers was happening at spacex basically vertically integrating if you think forward to like why the model s was such a breakthrough vehicle right the roadster had its challenges it wasn't for everybody um in terms of comfort handling i mean it was good you still have yours or did you give just miss

SPEAKER_257: something i gave it no i gave it to a different museum the um the peterson yes peterson automotive

SPEAKER_313: museum i think it's the largest in the us you know yeah significant i still have mine it's right over

SPEAKER_315: here at number 16 in the garage i got i redid the battery pack i have 17 of you do the roadster yeah

SPEAKER_13: i had you know how i got 16 there was a venture capital so i'm not going to say their name

SPEAKER_150: um who ordered probably right before you um and they uh i subsequently ordered like you know after that i might have been like 100 or something i was in the signature 100 was way up the list

SPEAKER_322: and um then he invested in fisker oh and he stabbed elon in the back yep so in retrospect by the way this is

SPEAKER_04: even memorialized in some of the books some of these people have written some great venture firms the best you've ever heard of were really confused about the difference between electric car and a hybrid car they would call fisker an electric car company back yes back then it was hybrid and it's a completely different design space it is such an albatross of a product and not good at anything because it was a hybrid car yeah and it was garbage they didn't get that the electric vehicle transition is very different and it doesn't leverage anything that the whole idea is to not go to the

SPEAKER_287: gas station and not have those parts and not have all the overhead of a gas tank like you basically

SPEAKER_04: double the complexity with a hybrid versus a simple electric car so it was bizarre and so um the person

SPEAKER_13: i think if i have my story correct uh proactively canceled their deposit happened to be sitting next to

SPEAKER_242: elon having you know a meal and he's like oh this person canceled the car and i said oh where are they

SPEAKER_119: in the the the order i said 16. i said oh i'll take it can you move me up yeah and just took us blackberry and moved me up the list and i got the first orange one i still think it's the best color

SPEAKER_326: what was your color you went red i went no i went for uh racing green which was their signature color i think i even have the uh i think the race in green was tight i think you might have been the

SPEAKER_328: first racing green for sure i'm not sure i'm not sure i don't know but i loved it really beautiful

SPEAKER_04: i got a custom uh roofs you know the the roof option i got it in clear unpainted carbon fiber

SPEAKER_335: because my bike my mountain bike at the time was clear unpainted carbon fiber my favorite rocket that i built was clear unpainted carbon fiber so i'd like there's something about that aesthetic that

SPEAKER_150: people really like i've been watching um corvette is having a renaissance they used to own a corvette before i got my roadster and i traded my corvette in for a roadster um they have this new one the zrx one which is the greatest hypercar ever built at this point 1 250 horsepower but anyway uh what corvette

SPEAKER_87: drivers like is just raw carbon fiber that aesthetic looks so pretty cool beautiful it just yeah and they're and they're me and it's like i guess it's still super expensive so when you put like the accents in the car it's like they're like oh four thousand dollars to do these three parts on the cockpit

SPEAKER_234: like really crazy why is it so expensive carbon fiber hand done i believe in the case of the roadster was in italy if i recall and uh and they wouldn't let me get the body in clear because there were areas

SPEAKER_04: where they knew it wouldn't look good that it was physically smooth but underneath you'd see that the cloth was overlapping in certain ways and so they didn't want to reveal that yeah i didn't want to

SPEAKER_36: reveal those uh secrets here's a clip of elon talking about 2025 and 2015. well that clip 10 years ago

SPEAKER_350: oh i was interviewing him your question so in in terms of what i think 20 yeah please um so for for sure ubiquitous computing um ai that's beyond anything uh like the public appreciates today um i think we'll have um most of the new vehicles being produced uh being electric um and we'll be probably have the super majority of energy being produced being uh sustainable so i think i think we're on headed solar primarily in your mind primarily solar yeah wow interesting you're right and that was by the

SPEAKER_72: way that was me interviewing him yes yes that's where i put it in sorry about that we should have

SPEAKER_354: included it no no it was just like whoa so i haven't revisited that clip so the first one nailed it like like your ai boom right you're soaking in it it's just dripping out of our ears it's so much

SPEAKER_357: literally we're like literally in the matrix right now in a bath of ai okay let's go to energy and solar

SPEAKER_04: and right so and ev so um so the part where so making a forecast of five to ten years is really difficult because of the inertia of where we are if you would instead ask them what are the three most important trends that will be let's say in ten years everyone will agree are inevitable but we may not have made the transition we may not have majority solar on the planet we may not have majority of vehicles electric vehicles simply because people hold on to cars for 12 years on average you know some parts of the world even longer no matter how good the new product is although he said new product sales so that was at least starting to hedge the you know it's not the swap out but the the new sale so there's inertia there's political shenanigans there's uh weird ways that people you know forestall the future and prevent the inevitable future from manifesting so look where we're incumbents right so i think one thing you can say is today in 2025 uh elon believes all three of those very strongly right even though we haven't fully realized solar's potential or the ev potential so i'll go further and say in 2015 it was obvious to me and it's even more obvious today that it is inevitable that all vehicles will be electric every train plane tank heavy machinery you name it and they'll all be autonomous so electric i would add autonomous that that's sort of the ai plus electric um solar he said sustainable and then i had him double click mostly solar that's his point of view and solar and storage solves the problem and as a company that makes a lot of solar storage or specifically storage you've got to understand why that's important i would just stick with this first answer which is

SPEAKER_31: clean energy will be the never future and that includes nuclear right so it's obvious that fusion fission solar are our energy sources of primary resort geothermal may make a major comeback if we can go really deep and hot rock that's that's a sleeper potential um but those that category of products really dominates all of it we shouldn't be burning coal we shouldn't be burning gas uh it's just but it will take time right because it's it's the drug much of the economy is addicted

SPEAKER_366: to right currently yeah and if you look at that time uh solar was like one percent of electoral uh

SPEAKER_150: in the us 2015. 2025 solar's 10 renewables you put wind in there you get to 17 according to

SPEAKER_125: claude ai and then i would put nuclear in there so there was a weird by the way marketing thing where at

SPEAKER_04: some point i think this was emery levins and some others uh put in uh this idea of what was it renewable energy there was some term for it as opposed to just clean energy right they they were trying to distinguish um uh nuclear from all the others and they're really if you think about it how about zero

SPEAKER_31: carbon energy or zero pollutant energy um that would be a better term for what we want and that would be solar wind geothermal and nuclear all types of nuclear right it's it's so that's the category

SPEAKER_29: that's inherently distinct from fossil fuels yeah and if you want to know why uh all of this happened

SPEAKER_87: this hatred of nuclear nuclear uh 1979 ah jackson brown crosby stills nash and young that's right

SPEAKER_377: john hall doobie brothers yep so interesting notice the bruce springsteen and this is no nukes this is

SPEAKER_04: the exact confusion understandable perhaps between nuclear weapons and nuclear electricity notice the

SPEAKER_00: image in the background that's not a nuclear power plant in the background right that's not three

SPEAKER_04: miles mushroom cloud that's not chernobyl that's nothing that doesn't happen in those plants right um and greenpeace was actually founded specifically a focus on anti-nuclear weapons and then it bled into nuclear energy as if they were the same and i gotta say the government did a bad job it was a militarized technology it of course had uranium going around and fears were easily stoked so it was easy for the public to be confused and to be distrustful and we've been living with that ever since

SPEAKER_31: which they are completely different applications of course it's just infuriating when you think about

SPEAKER_36: it that like uh the germans have shut down i think all six of their nuclear power plants at this point

SPEAKER_386: and decided buying oil from putin i know it's a better idea because of fukushima so get a little

SPEAKER_04: of this so they shut it down there germany has this perpetual propensity to be on the wrong side of history which is actually almost a direct quote from this book uh rad future where did i put it oh it's it's underneath my oh yeah by isabel yeah it's great i just just finished it a couple days ago so when uh putin invaded ukraine at that moment germany was sending 220 million dollars per day to russia because of the oil the gas dependency then they shut down the nuclear power plants as well around fukushima time and the estimates are that currently they're dealing with a thousand one hundred excess deaths per year already because of the extra pollutants from burning fossil fuels instead of nuclear um it's mind-blowing like at fukushima no one died from radiation uh about 2 000 people died because they did a botched evacuation of the region and the way in which they relocated people they died um hundreds are dying every winter now because of higher electricity costs it's just mind-blowing about 28 people died in chernobyl in total more have died from fear of

SPEAKER_335: nuclear than in the discussion about nuclear than from nuclear itself the panic attacks have done more

SPEAKER_36: and if you look at coal taking four million lives a year beautiful four million lives a year from coal particulates in the air it's unbelievable you know and and we got the president out here saying it's clean beautiful call and i interviewed chris wright twice in the past year on this issue and i'm just

SPEAKER_13: trying to you know get him to say that solar's great and he's like there's a place for solar but

SPEAKER_242: he keeps calling it unreliable and i'm like put batteries with it he's like oh there's no batteries i'm like what ask australia yeah like think it through uh think it through my guy you can you know they're

SPEAKER_87: here in the great state of texas uh there we have more solar here than any other state i can tell you it's not because they love solar it's not because they hate fossil fuels this place loves a good oil right it's an oil patch you know yeah they they do it because it's the most cost efficient it is the best bang for your buck long term so here we are and the other thing about fukushima which is so crazy is when you look into uh the design and placement of the reactor they told the province don't put it there and like we'll build a sea wall and they're like yeah but every 100 years there's going to be a storm and literally 50 whatever number of years later the storm arrives like don't put the

SPEAKER_04: digital gently on time below sea level so that if you had a wave it's gonna be like there so the same was with chernobyl by the way chernobyl uh was the worst nuclear accident we've ever had it's the only one that actually had some deaths associated with it and um and small small number by the much smaller than people think uh it's not a reactor design anyone builds today it's like you do learn from mistakes so to look and you know reference that to say we should do nuclear is as illogical as saying we shouldn't have cruise ships because titanic there was this one titanic and it sank and we learned how better to navigate and avoid icebergs and you know et cetera et cetera et cetera like we

SPEAKER_404: don't even think about that today because the lesson wasn't develop don't don't build another ship ever

SPEAKER_13: again build a better ship people exactly and and and here we go all right let's go i don't want to run out of time with you this has been incredible um but i want to get into ai a bit um in your mind the pace at which we're moving and getting to i'll just use two terms we'll define them here for

SPEAKER_150: ourselves people will debate them artificial general intelligence being the smartest human on the planet in any discipline lawyer accountant chess player whatever it is and then we have super intelligence being an intelligence that we can't comprehend you know a level of intelligence that you take all the humans together and then times it by a hundred and it's it's something else you can find questions to answer we don't even know to ask where are we in this timeline well

SPEAKER_04: there's a big architectural gap still to fill on what some people might hear the question or hear the term and think oh it's like a human meaning it's going through the world with agency and purpose it is making decisions and doing its own thing if you will that is i don't think they have a good term for that like uh like sentience or something like the art of it so if you just say will an ai um system you know like the gpt's and the xai's of the world of today as a framework you know outperform a human on any question or task that you present it um absolutely i mean that's that's that happens quickly right like is it a year or two you know by the way when that happens it's just give it another year and it'll be much greater than a human so like the difference between einstein and you know one of the less capable humans on planet earth is not that big a gap versus humans and you know pigs and and and you know lesser animals it's like on the scheme of things like those two humans on an intelligence spectrum of ai let's say just ai's trajectory is like you know going from one human to a collection of humans applying their intelligence collectively in a group kind of setting because otherwise it doesn't really matter how many humans we have on earth like you know if they're not acting in coherence as a team if it was twice as many or half as many it doesn't matter right what matters is are you a thousand times as smart as any human has been that comes quickly after even simply moore's law right and in ai actually we've had moore's law doubling every year and algorithmic improvements of a doubling every year for like 15 years now which is kind of astounding that makes a big difference so i think those come quickly and so some of the evidence of that you know these these models today already outperform humans at almost any task you apply them towards a little bit of specialized training you know there hasn't been a domain where you're like oh we can't do that i'll give an example medicine i was just in stanford last week getting the update on you know state of the art of large language models for health care and so the the sort of humiliating and and humbling takeaway in short is that the ai alone is much better than a human of course and it is much better than a human using the ai in other words take your best doctor in a field of medicine reading an x-ray doing whatever they're on a certain talent level if they start using ai they get a little bit better but if they just let the ai run without the human in the loop it does better still by far like off the charts and so the point is humans are just holding it back right then the next one is not just diagnosis but the course of therapy what should we do given what we just learned they also outperform there and then best of all blow the doors off the human on empathy as reported by the patient so if you have a chat interface where the doctor is talking to the patient through that same interface or an ai the ai blows the door off on truly understanding me conveying the situation and understanding on some really tough issues like let's say end of life care for a parent who you know do you pull the plug or do you you know like yeah schedule really tough conversations they blow the doors off humans so we're already there um and elon would talk about a number of you know you know phd level equivalents for all the phds um the challenge then is going to be you alluded to this earlier on the emotional response you know we're talking about the big systems and what have you earlier in our conversation there's still something missing about like obviously these systems aren't just going off and and doing interesting work now the agentic chain of reasoning is starting down this path and say i have a task i want you to do can you find me

SPEAKER_31: the best vacation plan and you know reaching out to all these different websites and like figure out where the flights and the hotel and the things i might do for kids of this age and you know pull it all together with a series of steps same thing agentic flows in programming as well right you can do encoding but there's something different still from that sort of spark of you know consciousness or

SPEAKER_04: sentience which is perhaps going to require some other it could be it could be an emergent property by the way of forecasting the future so let me share a bizarre thought i alluded to this earlier that what our brain naturally is doing is predicting the future and only when what we sense is different

SPEAKER_31: do we perceive it in any sense like if i grab this this this thing and it's much hotter or colder than i could possibly imagine i'll notice that otherwise i won't even notice temperature it won't register right because i'm working off predictions in fact they've done free will studies if you will that we

SPEAKER_04: we retrospectively rationalize what we just did like yes right and confirmation bias why all kinds of biases but it is very low like in the microsecond level that like i just did move this finger and

SPEAKER_113: then i was like oh i intended to move this finger okay yeah um perhaps that's what's going to happen with our artificial systems as well with next token prediction they'll be like how am i you know

SPEAKER_04: retrospectively making sense making of what i've done and there could be some vote taking circuits that we'd we might have to build some circuitry for this i guess the vote taking circuitry that's

SPEAKER_31: in the feedback loop of what we retain is novel that will then sort of bootstrap this sort of consciousness or intelligence the perception of free will the perception of consciousness might be a phenomenon of that there are others who think we need to take a neural symbolic approach that we need to literally recapitulate if you will these lower level primitive systems that are um emotive and and what have you to actually have an emotion as opposed to faking it well um i don't know the

SPEAKER_04: answer but i i uh i do think we can do a lot more experiments now than ever before we can run these evolutionary uh sort of feedback loops in in ways that will potentially bootstrap intelligence and it might come from some heterogeneity it might come from just the sheer approach that we're taking but it's

SPEAKER_31: not obvious it's gonna that we're there now but that like what we've built isn't like a baby version of oh it'll just naturally be on you know self-directed um that that takes a different bootstrap from the current vectors that are there so i think what we have is a hyper intelligent um

SPEAKER_04: adjunct or or or or colleague kind of like c3po if you will you know yeah way too smart for its own good chattering on when you don't want it to chatter on um but it can do a lot of things for you um

SPEAKER_287: can do a lot of things but it can't beat darth raider exactly you wouldn't just not gonna have

SPEAKER_429: smart hands solo what not happening there we go let's take an analogy never tell me the odds

SPEAKER_430: yeah it's not getting through an asteroid field that's right not a chance it's an interesting

SPEAKER_36: thing you bring up uh which is it's already beating the doctors on empathy and so then you have to think about um this next generation uh this concept of being one-shotted if you've heard it where your

SPEAKER_13: relationship with ai becomes the dominant one in your relationship sam waltman just announced he's gonna make it spicy you know make a little spicy in chat gpt for adults there are other companies that

SPEAKER_38: are i get pitched all the time the last couple years on like an ai therapist and i'm like you know

SPEAKER_287: that's obviously that's inevitable but i'm not investing in that right now because what if it goes wrong you know this seems like it could you know have some pretty bad outcomes on the

SPEAKER_13: margin so what are your thoughts on i don't want to get into ai regulation because it's kind of dumb but it's it's gonna happen in different countries but young people even older people using this as a proxy for a better companion and what that will do to humanity because yeah hey the doctor this is a better doctor i can ask it about private questions okay yeah fair enough doctor loses their job or doctor has a different job does something different you know but then your companion is always empathetic never angry at you and they never have a bad day everything is sycophantic this seems like a road

SPEAKER_427: to purgatory to me i don't think no it's it's a really interesting question and there there are

SPEAKER_31: small experiments that have gone on in a number of places even a few years back in in very rudimentary forms when some financial services firms put you know ai chat bots in place to try to do customer service you know try to lower the customer service workload they found certain subset of people became romantically attached to these people and these entities and it just couldn't accept the fact that they weren't real like like like even when told so they they had this so there's this natural human tendency i think to project um agency and intelligence onto things that aren't we do that with our pets we

SPEAKER_04: do that with a lot of things um and it's risky so i think when you mentioned sam altman that you know he also said you know we were cautious at first because there's some mentally ill people out there and we don't want the ai to you know if it's sycophantic or just repeating back to you what you want to hear

SPEAKER_31: amplifying some of our baser natures like oh yes you should do that terrible thing or yes you should be

SPEAKER_04: you know bad violent what have you whatever it might be and so um here's the challenge and you alluded to regulation you alluded to you know how do you direct these things is this whole mode of ai development is not like traditional engineering and this is something that i have yet to meet a regulator or politician who understands this at all you can't say take that let's say frontier intelligence of any kind so something that's like state-of-the-art something interesting like anything anything you've heard of um you can't say make it safe you can't say prove to me it's safe you can't say control it you can't say align it you can't say explain it it's not interpretable and there are definitely groups anthropic being one of them who think oh yeah we're gonna work on that we're gonna try to make that happen they have yet to make meaningful progress in this meaningful meaning a result that would give people comfort that this is doable and i would assert it may never work that we may never reverse engineer an intelligence uh in the time frame of relevance

SPEAKER_31: meaning we'll just have built a better one before we reverse engineered the prior one just like we haven't reverse engineered our brain these complex information systems both brain and and neural nets are inherently inscrutable we understand the interfaces the ins and outs but not the inner

SPEAKER_04: workings you can't cut and paste functionality you can't you know figure out the subsystems there you can't like draw a box around where there's the english-speaking partners we're just doing math like none of that really lends itself to uh reverse engineering uh it's the same for evolved artifacts like our brain by the way um and the implications are that you can't go in there and manage it like you would an engineered product so if i was to wrap all that up into a simple sort of euphemism it would be it's more like parenting than programming so we're talking about programming you and me early on and the little hacks that we were doing we knew what we were doing and if something went crazy it's like your paper clip of printing yes uh it's gonna run out of paper you can quickly understand what you just did and fix it right that does not apply to the complex interactions of ais internally nor are they back and forth with a human which is like two complex systems working with each other so whenever you hear someone say let's regulate ai substitute teenager for the word ai and then ask yourself what would that regulation look like so if you say let's make sure that the teenager is aligned with us how right let's make sure the teenager does no crime how the best you could do is be a good parent and have police to say i could have an after the fact that's either looking at what inputs this ai is getting let's like not let certain questions be asked or look at the outputs to say oh wait that just went down dangerous territory let's not share that result

SPEAKER_31: right and have a police police layer that you could do what you can't do is have any of these companies do what california bill initially proposed was like prove uh you know safety before you

SPEAKER_34: even start training i mean that shows you how ridiculous they work right yeah to your point

SPEAKER_36: you can have a speed trap or a dui you know uh trap where you you check people aren't drinking but you know people have a car that's right it has wheels and has an engine like they're going to drive it and some are going to go fast some are going to slow some might drink and drive and so there's regulations

SPEAKER_04: there but yeah by the way let me let me throw one premature absolutely not only premature but it never worked so i think it's a fool's errand to think about safety and alignment as if they were achievable in the sense that they're being described today so let me take alignment imagine you and i or whoever someone at xai or someone at um one of the other like closed ai uh says i want

SPEAKER_452: to align with my interests my woke whatever or my you know western liberal ideals or whatever i think

SPEAKER_00: conservative humanity is they might imagine you could you imagine you could do that which i'm saying you can't imagine you could um the authoritarian regimes will be able to do this too

SPEAKER_04: and you're gonna have a really dystopian world where some really bad ais will be all over the place most people live in authoritarian regimes today and under unfortunately authoritarian rule you do not want them having aligned to their cultures and norms ais right what you want you're going to have a lot more uyghurs in concentration camps yeah being tortured no idea and so i i really believe in elon which is well in general but in the case of the way to get through this quagmire is to have it be a truth-seeking out of not say i know what you need to do i so in other words instead of mind control right to say i'm going to make you think a certain way which doesn't work with teenagers right it cripples their ability to reason and the same is true of neural nets the more rlhf or the more in the sense of mind control you try to layer on late in the cycle the more you compromise the reasoning capabilities of the ai itself it's profound both humans and ai so many different ways so do not think mind control is the answer do not think containment's the answer it's more like parenting and policing and then i think you might finally get a regulatory regime that makes some sense

SPEAKER_108: so interesting you bring up the number of people living under authoritarian regimes you know even a steven pinker's book which is probably where we both got attuned to this everything's going great except for the spread of democracy and authoritarianism uh democracy going down

SPEAKER_150: authoritarianism going up 54 of people live in authoritarian countries that's why democracy is

SPEAKER_13: worth fighting for and it's an interesting experiment but it it might be the unnatural condition of human existence might be to have a democracy and the natural one might be to be dominated by other people that's what we've seen from the bulk of humanity which you don't have to make any of this political but if you see things being less democratic be concerned absolutely you see things being less fair you see

SPEAKER_458: things being more cruel to certain groups of people get curious as to what's happening here

SPEAKER_04: it is the path of doom um and it's it's frightening um the the way the way in which authoritarianism and theocracies both of different variants um can take over and are so difficult to uproot it it is a precious and delicate thing when a group of people in their founding like in the founding of the united states or you know the long arc of western civilization uh take a somewhat selfless act in the leadership

SPEAKER_31: ranks to say we're not going to try to just covet power i mean think about like kleptocracy like russia is the exact opposite like you know this this cult of an individual and has this madman who can launch wars around the country um is a symptom of oh wow that's over time it's just look at that democracy just boop boop boop yeah it is a precious it is it is the thing that allows progress by the way if you think

SPEAKER_04: about why i think authoritarian regimes ultimately will will fail it's because of the technology they don't embrace change either politically at company levels they they coddle the you know the power breakers that are and that in a sense is a architectural resistance to change and innovation and new entrance in in america at least in the west we have you know a system that allows and encourages entrepreneurship it allows disruption that allows overthrowing of the past and that is so precious that is really the vector of change because all meaningful change comes from new entrance it comes from entrepreneurs in the broadest sense of the term doing something new it doesn't come from big companies incrementally improving their core business it doesn't dictators or dictators saying you know

SPEAKER_36: planned economy let's do x let's do y that's never worked the top down works incredibly efficiently in very narrow fields in very short arcs of history you want to build a bridge fast you want to build a high-speed train and you want to run it through five neighborhoods and take away their rights and yeah sure great you can even run the people over and make them dig the ditches and then kill them and bury them in the ditches we saw this in china they basically said you know what entrepreneurship not for us we took it as far as we could jack ma disappears a bunch of companies disappear and uh now they regret it and now they've got you know population decline they got to worry

SPEAKER_150: about where everybody's gonna get their jobs they got 20 25 unemployment with young men you know and democracy finds a way and all these companies this is a really beautiful you know sort of interesting wrap-up point that we just kind of stumbled into which is what we're talking about what you and i do

SPEAKER_87: for a living what elon does what entrepreneurs do try to create something new that makes individuals a bit more free a bit more happy a bit more productive is the only operating system that seems to be pro-humanity and it really does start with entrepreneurship with people who are change agents and not the people who are seeking power not the people who are seeking control talk about star wars like the empire tries to control things the more they try to control it the more brittle it becomes it's hard it's hard to be an authoritarian because all the change is

SPEAKER_150: happening everywhere around you and you have to try to squash it and quell it and extinguish it it

SPEAKER_234: it's exhausting yeah to be an authoritarian putin lives at the end of a two mile long tunnel in intense paranoia it's kind of like hitler in his bunker like yeah but he's stuck and it's all stuck

SPEAKER_471: so his whole country is stuck under him can't you can't come you know this is you know when people

SPEAKER_36: talk about like uh trying to the peace dividend trying to work with these people i said to somebody

SPEAKER_242: at some point like you know what kim jong-un would really like like to come to the nba finals

SPEAKER_15: say what you will about trump when he went there and he crossed in the dmz and he was like you want me to come over like do you want me to step over into north korea okay i'm gonna i'm gonna do it i'm gonna step over here i got kim jong-un's face lit up like oh my god sorry that was good he's like i'm coming in somebody loves me i gotta tell you you you get those guys to come to vegas for a weekend get them to go to carbone and get that rigatoni you get them to come to f1 and hang out there's so much great stuff in freedom and democracy we just have to allow those people to experience it in their humor here it is and this is one of the great moments in history

SPEAKER_483: look here it is you just look at kim jong-un's face waiting i mean if they see a smile he's he hasn't smiled like this in his life look at him he is so happy it's like there's a peak

SPEAKER_480: experience for him i think this is the way to get these these dictators to flip just butter them up

SPEAKER_242: let them come to a music house tell us a burning man yeah go to all of it all of it would be amazing

SPEAKER_15: listen steve you gave me more than an hour and a half of your time you're one of the great humans on the planet great thinkers i love talking to you uh and thanks for sharing so openly with the group here i'm going to have you come back in six months please and uh we're going to do this every six months can be a little check-in and next time i just want to go through all of your portfolio all this amazing stuff if people want you as an investor and they're doing something absolutely crazy and it's a hell mary and it's going to take 20 years but we'll change the world and shadow the last paradigm how do they reach you sure how do they reach out with their business

SPEAKER_36: plan or their ideas and i don't mean to flood your inbox no yeah no i do this for a living

SPEAKER_00: absolutely our website is future.ventures i'm also most active on x socially but my email and everyone here at our firm is just our first name at future.ventures there you go all right everybody

SPEAKER_489: there's your 90 minutes with steve jervison we'll see you next time thank you thank you jason