Futurist Logo

Why Technology Always Wins


Brad Templeton

In this weeks mashup Breaking Banks and The Futurists podcasts, we interview pioneering technologist Brad Templeton. In a discussion ranging from the creation of the commercial internet,  the battle for our future, to ethical AI and the likely roll out of autonomous vehicles, we cover a lot of ground. But at the heart of this discussion is how technology is reframing society and why those looking to the past are likely to be left behind.

Analysis complete. No addtional information is required for context. Proceed with transcript display ...

View Transcript

document button

this week on the futurists you don't want to depend on others foryour information you can't depend on it if you're expecting other cars to communicate with yousome of them won't not all of them will have that equipment some of the equipment will be broken so what that means is you have to workextremely well without communicating with anyone it's not especially the first car on theroad can't talk to anybody so the first car on the road has to meet this goal of safe enough and that's a very very highgoal safer than human beings drive you have to meet that goal without any communication so the most thecommunication could do is take you from safe enough to safer enough[Music] welcome back to the futurists and uhBrad Templetonthis week we are partnering with our friends over at breaking banks and we have a special guest he's been onbreaking banks a few times actually brad templeton uh founded theworldsfirst.comback in 1989 he's chairman emeritus and futurist for the electronic frontierfoundation founding faculty and computing chair at singularity university he's consulted onself-driving cars transportation smart cities um worked on the google uhautonomous vehicle project or waymo as it's now known in the past brad talbotton welcome to the futurists goodto be here i mean usually we have you wrong when you're a long resume it just means you're old soi i guess so well that includes all three of us i guess um but uh it's greatperspective on the future yeah it does it does yeah you know i'm i i have a good friend who who's a a a futurist andhe always says um as jp rangaswamy out of the uk jp he was the chief scientist salesforce for a while um i guess you guys know him i know under british telecom as well right yeah exactly uh so jp uh hehas this great uh saying that to be a good futurist you have to be a good pastist because youhave to understand the past you have to understand human behavior um you know and so forth so brad you'vemade a lot of bets um you know in the past about the future you've gota a pretty good track record in in a range of areas but um let me start by this you know as afuturist um what is something that you predicted in the past that you're most proud ofwell i actually think my prediction record on self-driving cars uh has been fairly good even though a lot of peopleare going back and forth about is it going to happen soon is it going to happen in decades i actually uhdiscovered this interview with me from about eight or nine years ago where i laid out a timeline and we're actually on that timeline and that's odd becausei usually never want to do timelines and dates any futurist who does is an idiot because it's certainly going to comeback to burn you very rarely uh are you ever that correct uh or as i like to say all the predictions i'vemade that i remember have come true uh which is true of so many other people uh but uh uhi've been pretty good about that um i've been horribly wrong about a few thingslike the value of bitcoin and uh and stuff like that um and umi i you know i i think that um i also was pretty good at predictingsort of the a little bit about the course of the internet i mean although i certainly wasn't the only one and a lot of people like to think the internetjust in the early 90s just sort of sprung upon the world fully formed from the head of zeusand it didn't it had been going on for quite some time since the late 60s and i got involved in the late 70sand back then uh the equivalent of social media was the mailing list and the very not the very first main listbut one of the first and really interesting main lists was called human nets it was about what computer networksmeant for humans uh and back in those days on that uh that particular messageboard we had um a term we call world net and it was our dream of what in the futurewould be this big network that everyone was connected to and you would bank on it and shop on it and communicate withyour friends on it and do all the things that we do and and so we all had that rough picture in our minds now weobviously didn't predict facebook and we didn't predict uh all these other companies or we'd all be billionaires right we'd invested in them or builtthem uh so only a few of the people became billionaires but nonetheless it wasn't uh something that was a bigsurprise and uh i certainly laid that out and uh and got to talk about it inthe early days of it and many of the things i particularly came through obviously a lot of things i didn't even imagine happened towow you know it occurred to me reading up on you a little bit on your background on your website umUsenetthat you did a lot of work on usenet and now usenet seems like archaeology you knowlike i'm imagining half the people who are using smartphones today have no idea or have never heard of usednettell us a little bit about those dates tell us a bit about what you were doing in using it what was hard and what was what was it good for yeah it's umcertainly true if you're below a certain age you would never have encountered it uh it's interesting though because ilook at that history and try and map it into the future and trying to understand what you know today we call social mediabut back then we just called online discussion groups and bulletin boards and mailing lists and so on uh so socialuh interaction on computers began in fact with mailing lists when the uh the very first way that people could to havea social environment there was with the mailing list uh in about 1978 1979 bothbulletin board systems bbs's a term that people probably still do remember and uhusenet were born and a little bit not much longer after that commercial online uhservices like compuserve um and aol and soyal came later they were born um butfor the internet usenet was its community it was its place where you went to discuss things meet other peopleand unlike today there was only one um and unlike today nobody owned it it was thiscooperative this collective and basically because communications were more expensive in those days everybodywho had a it was always done on um you know what we might have called a mainframe but or many computers largercomputers that were shared rather than on personal computers uh i mean the cloud i guess that would be ourmagic term for today but anyway each was independent in a different place and you connected to your local computer and youput in messages and your computer shared its messages with all the people it knew and then they shared with everyone theyknew and they told two friends and they told two friends and so messages flooded around the world and you could read them locally on yourcomputer with the speed that involved so that was pretty good and from the late70s until really yahoo and a few other things arose in the early to mid 90sthat was the place of community for the online world and so it was very interesting it wasa more um you know not to be too elitist but a more erudite uh audience um there were there weretrolls of course uh spamming originated there so all these things that we know of that are bad originated therebut to be getting on it in the beginning you you generally had to be at a lab or universityand then later you could be in a public online service and that just altered the character of it a bit and so uh this isour error of course that we didn't um predict some of the things that are happening today whenlet's put it bluntly dark forces are on it deliberately trying to do things one of my great regrets and i shouldn't saymy regret because i was just on one of many contributors building these systems but that we builtthem uh maybe we should now say a bit naively about that we umyou know we we generally felt and and i i still mostly feel that giving peoplemore access to more information more communication the ability anyone to publish anyone to you know get access toall information that this is generally a good thing but what we didn't anticipate was ofcourse that if evil people decided to deliberately weaponize it use it for propaganda useit to bend people's wills uh or bend the sphere of information that people havethat this could have well obviously rather horrible consequences and one of the great problems i ponder today is howto find a solution for that a solution that still preserves what we value uh there are many solutions beingproposed to say well you know let's just censor the network let's just ask facebook to you know block everythinglet's get presidents of countries kicked off of twitter uh rather common topic ofdiscussion these days and uh those are i'm actually ieven would side a little bit with elon on that in the sense that he he has said that you know banning uh the presidentfrom twitter uh was going too far and a lot of people disagree with him on that uh that hewhat he would do if he owned twitter is he would you know ban specific tweets but he wouldn't uh take a person off umso uh i'm not entirely against his statement of that philosophy but how it's done is obviously going to be veryinteresting anyway so we've come to this time today where we're now pondering what to do about thisand unfortunately not as many people are pondering how to do it and still retain the values that we haveand that's a that's a big question i've been trying to even invent an entirely new moral theory in order to come upwith solutions to this one of the problems is bad news travels faster and when you have a retweetBad news travels fasterbutton that makes it as simple as a click um bad news gets amplified really fast yeahand fake news apparently it's not just that though uh i mean to be fair uh there's been some you know strongresearch showing that the way that the algorithms on youtube and facebook andmany other things not just retweeting uh have been amplifying the wrong thingsand not giving people what they want and of course addicting people yeah in ways that they don't want to beaddicted which at first i thought was all because all the media are sponsored by advertising basically the only way wereally came up with to make money on the internet 95 of the money came from advertising and of course that meansyou're working for the advertisers you're not working for the users but then i i met one of the executivesof netflix and i learned about their thinking inside netflix and they don't take any advertising at least not yetbut they are just as much about addicting you as they can as they can be regardless of where their money comesfrom so sadly uh it's not advertising which pushed us in this way at least notentirely i think advertising played a role um which is making it particularly hard to figure out the rightphilosophical principles uh to guide our online life and you know stop the badthings from happening and still preserve all the great stuff that i still believe in so when we we talk about that sort ofethical position you know you work with the electronic frontier foundation and so forth as we move into the metaverseas we move into artificial intelligence it's going to be making decisions for usback in those early days did you guys foresee these issues you know you've said you you're trying to work on sortof a new moral code for for online activity we know tim berners-lee is is doing something similar but um you knowwere some of these issues telescoped back in those early days as you were sort of trying to putpolicy and process together well so we had um spam of course originate in facti traced the history of spam and wrote some articles about it and found the first spam that we could think of wasdone in 1978 on on uh arpanet email so it goes back a long way but it actuallypretty much vanished from then until about 10 years later when it became a bigger thing againso these things were telescope we knew that there would be people who would abuse things in a way originally it waslike a small town where nobody locks their doors or leaves the keys in the car because you know you can trust yourneighbors and obviously as it grew it had to move away from being that small town there's always some regret whenthat happens um so some of these things were telescope the uh the idea though that uh we would see uhyou know deliberate propagandization but we actually of course propaganda is not a new invention and it goes back andit's had many masters in the past who have done horrible things obviouslybut um we are again this was a bit naive we thought we maybe had an answer for thatright that oh you wouldn't block the information right now in russia you can still watch youtube all of youtube canthey block facebook and a few other things but you can still get a lot of um of outside media in russia but they'remanaging to still have the majority of the population uh listen to only what the kremlin says and believing that ohwe're in the middle of this glorious war and we're just doing a small little military operation there and obviously there are real horrible consequences tothat happening we hoped that we had a solution it's funny in the end of the soviet unionin around 1990 many people pointed to the fax machine as one of the technologies which helpedend the soviet union because finally there was this means of communication within the soviet union that could notbe blocked as long as you could make a phone call you could send newsletters and publications so people startedfaxing the truth around uh that the kremlin didn't want people to hear and that was able to spread but everyoneknew back then don't trust the kremlin like that was not a secret that the kremlin lied to youfor some reason in this world where we now have all these technologies we managed to get a place where peoplebelieve the kremlin and that baffles us all they want to believe it's it's quiteSocial construction of realityinteresting right now you have it sort of choose your own reality and when you don't know which direction to lean or ifyou're unable to make a decision on your own you're going to be heavily influenced by the people the group that you want to belong to so it's like thesocial construction of reality is is what we're seeing happen and unfortunately the realities that are being constructedthey don't really mesh so well together so we have a lot of conflict in the world that stems from this misunderstandingi underestimated how tribal people are i mean it's not again a new discovery that tribalism isan important driver of our of our mental mind of our mindsets but um it's astonishing strong and we'reactually reinforcing the tribalism and tribalism goes together with everyone's favorite friend confirmation biasand so once you have your tribe picked then it's turned out to be much easierthan expected to get people to only listen to the information which confirms what they already know or want tobelieve i've actually grouped the world into two big tribes there obviously are many tribes and i'vegiven them names i call one tribe the kings and the other tribe the stewards now the kings areprobably the people listening to this show these are people who are keen on the future they think it's going to bebetter they like the idea of progress they like the idea of technology they tend to be secular they tend to bea little more on the left of political spectrum although not universally and then the stewardsas you can guess are people who value the past uh they want to protect the past they're stewards of it ummake america great again is a uh as a steward slogan as you would expect of course isis is also a stewardorganization that's trying to you know return to um a sort of a traditional caliphate umso the kings and the stewards are in battle and the dark secret is the kings aregoing to win uh and they have been winning and um the stewards know this sort ofbut that doesn't mean they accept it it doesn't mean they don't want to go gently into that goodnight they want to fight it if you used to be on top of the world and youknow the old order when i was young was white christian males were in charge andthat's that's just how it goes uh and even if you have come to believe thatthat was not morally right uh that uh you know it's not there's something inherently that says that white christian males are superior thereare some people who think that but there are many people who have come to accept that that's no longer what they would want to believe but even if you believeit doesn't mean you don't resist losing that uh that privilege that you got um and umso because of that we see the battle going on we see people forming into their tribes and doing things just because itdefeats the other charge so this is my explanation one of thousands of explanations of course it's not just one factor for things like the election oftrump and johnson and putin's power in russia and so on is this battle between the kings and the stewards and the mainconclusion i come out of it is that the keens who are going to win because they're richer and have bettertechnology because they love technology uh they must win more gracefullybut they don't want to um they must tolerate you know pullingthe others gently into the future that is the one we are going to buildand uh you know this is this idea is also expressed um slightly differently by virginia poster in her book thefuture and its enemies where he she identifies uh dynamists and statists forstasis not status but stasis i'm not actually i was thinking of that when you when you're talking about thekings yeah so virginia also has identified this before i did actually umand uh i think her thesis has also got a lot of validityuh and so this is this could you could call this the great futurist questions is this is the battle for the future umbetween these different attitudes now obviously there are sub-battles of all sorts but the weapons are gettingstronger and of course we keens are providing some of those weapons and uh what i'm afraid of iswhere we're going which is where the ai technologies that are beingrevolutionized right now uh are used to do things like you know build a model of every single individualvoter know as much as you can about that voter everything they've written on social media their political opinionsand then right for them i mean you've already people have been impressed with what gpt-3 can sort of pull out of a hatbut we're not too far some of it's garbage though but we're not too far from being able to build an ai that says here's everything about a personnow write things that will push their buttons exactly right okay andand and send only them that thing uh as opposed to old politics which has had to saylet's identify some common themes that we can rally our base around we have this concept of our base right therethese this group of people who are want to support us will believe what we tell them now you start expanding thebase now it's an arms race though so we have two different sides playing this game and the result is chaotic and so idon't uh i don't even know how to predict what's going to happen when this goes on and sothat's why we need to find a solution for it you know we were talking a little bit about business models a moment ago and how advertising shaped the webFacebooks patentsin february the financial times did an analysis of a number of patents that facebook had recently filedand in those patents they discovered that what facebook was trying to patent you know there's no guarantee they're going to implement this it's just thepatent it's just an idea um but what they're clearly trying to consider for the metaverse for their theirimmersive 3d world is to do exactly what you just described to build uh high fidelity replicas of the people who areusing the metabolism they're going to track your eye emotions your involuntary reactions your skin temperature uh youryour how your skin responds and so forth it's quite obvious they're trying to build like basically a high fidelity 3dreplica of their users so that they can test subliminal advertising on them and then saturate us with that inside of themetaverse um the the the article in the time is quite chilling the conclusion it came to because uh i don't think that'swhat people think the metaverse is about well no they definitely don't think it'sabout that well that's interesting that's an interesting segue into my explorations of a new moral theoryuh as a possible solution to this and the new moral theory basically says that it is wrongto exploit bugs in the human brain umand we already have a law that makes it wrong to exploit bugs and computer systems and we already have laws thatmake it wrong to exploit certain bugs in the human brain the most obvious of which is gambling addiction so there area variety we psychologists have clearly identified scientifically studied the phenomena of gambling addiction andother addictive behaviors uh and so we actually have laws that say it is illegalto exploit this and to try and you know get someone to gamble with all the money they have well what if we expand that theory butnow we have to be careful because taken too far this could go um into being too strongif it's a legal regime but you could start imagining saying that you know this idea that you aredeliberately trying to exploit things that people are unconscious of to make them do thingsthan in my test in particular is you make them do something which if you explain to them what you were doing theywould say i don't want that right so they have to be unaware you're doing it like i don't want to stop things that people voluntarily want todo you know and if people who aren't don't have a gambling addiction problem want to gamble want to take drugs i meani'm cool with that as long as they're fully aware of what they're doing i don't think it's my business to tell them to stop but it might be my businessto tell them to stop if the people are unaware of it uh and it's deliberate and it's a known scientifically studied flawin the human psyche and it's being used uh and so a brain actor yeah i am uh umi'm pretty much a free speech absolutist and so i am not on the side of those who say you know we should just ban thistype of speech or um you know force online services to delete this or you know remove peoplefrom twitter and so on um so i'm saying instead if if you're trying to hack someone's brain that's that's not aspeech that's an action and uh for people who are fans of free speech as i amwe tend to say we don't restrict speech and what we do is we restrict actionsand it's okay if we're restricting an action um the other thing that people don'tunderstand in the free speech debate is that it is not a belief that all speech is pure and good there is lots ofharmful speech there are many books that have caused great harm in the world that we don't ban and i wouldn't ban um thereality is that uh well you either they've either caused an action but the other thing is that wehave found throughout history that there is no way to give someone the power to pick good and bad speech that doesn'tbackfire right and so even though there is bad speech in the world if you say okay fine we will thenappoint mark zuckerberg or uh you know or a government agencyto be the arbiter of what speech is good and bad it always goes sour yeah and so it's not a power we hand over andbecause that power is not in our quiver we look for other things such as action regulation let me bring it to this pointBuilding consensusthough is obviously with the pandemic most recently we've had a lot of divisionaround that mask wearing vaccines and so forth we have the issue of inequality rightnow that that's a significant issue really accentuated by umby the inflation rates and so forth but for things like artificial intelligenceand climate change response in particular we are going to need to build consensus so just before we go to breakwhat are your thoughts on consensus building using the technology we have i mean this is the problem that i wastalking about with the tribalism the tribalism is making our consensus building be worse i mean the only reasonpeople got up in arms about masks is totally tribal so my tribe doesn't like masksi don't like wearing the mask either i mean i i get pretty uncomfortable in them but uh and you know i have apersonal connection to this but if it saves someone's life right well not just someone's life i lost a close member ofmy family to this disease it's not abstract for me it's not abstract for the relatives of amillion americans and five million people around the world it's very not abstract for usuh and so yeah it's a little uncomfortable but you know we put on pants just to stop people from uh seeingour you know well you don't want to see mine but uh you know that the point isuh we've been pretty easily come to consensus to do things like you know some pretty minor inconveniences inorder to uh make society work better but in this case to save lives uh but it became a whole thing and you know thegreat thing was that whichever you believed you could find a study that backed up what you believed you could find a study that said mass didn't seemto be effective you could find other studies that showed the masks were effective you believed the one thatmatched what your tribe wanted you to believe or or yourself it's not everyone is totally tribal but so many people areso the same thing has happened on global warming and you can tell this because there's a really strong correlationbetween one's political stripe and one's views on vaccines masking global warmingare probably three of the big issues and abortion is actually of course the new hot one which is flowing through ourveins the story behind abortion by the way is fascinating i don't think we have time for it but it was all very much adeliberate campaign back in the 1970s the evangelical community did not care about abortion they were not evenanti-abortion that was manufactured and it became here you're in our tribe areyou with jesus or are you with the baby murderers and you had to pick uh if you wanted to stay in the church and theyfound this way to create this divisive issue the way i sort of phrase it is that weare all passionate people most of us are passionate people but when you're a passionate person it means you're a gunthat someone else can aim and fire your beliefs are your own in many cases you believe what you believe for personalreasons in many cases but you can now be pointed and fired at some for someone else's purposes andthis is what vladimir putin does uh there are documented examples howthere would be two protests organized across the street from each other a left-wing protest or a right-wingprotest and a left-wing counter-protest in fact this idea of a counter-protest is a relatively modern invention andfacebook discovered that both of them both of the protests have been created by russia's internet research agencyputin's little propaganda arm not so little and in another case when this wasdiscovered and they talked to the to the people who were going to go to the protest and said we've discovered that it was putin whoactually instigated your protest they said that doesn't matter we still believe in this we still want to do it we're not going to we're not going tonot fight the alt-right just because putin made us fight the all right well that's rational in one sense but it'sdeeply wrong in another sense and uh it's very hard to come up with solutions for that so i wish i had a happy answerfor you on how we'll build consensus right now i'm mostly looking for how can i at least get out of being driven intonegative concern right right um before we can actually get to that consensuswell listen brad this is super interesting stuff what we'd really like to do though is we'd like to dive intohow to be a better futurist and uh um we'd love to hear about your experiencewith robocars and you know where you see uh autonomous vehicles taking uh humansociety after the break so stay with us everyone you're listening to the futurists we have brad templeton as ourguest we'll be right back after this break [Music]welcome to breaking banks the number one global fintech radio show and podcasti'm brett king and i'm jason henricks every week since 2013 we explored the personalitiesstartups innovators and industry players driving disruption in financial servicesfrom incumbents to unicorns and from cutting edge technology to the people using it to help create a moreinnovative inclusive and healthy financial future i'm jp nichols and thisis breaking banks [Music]Interview with Brad Templetonhey there you're listening to the futurists and this week we're in partnership with our friends at breaking banks i'm rob terczyk and my co-host isbrett king and this week we're interviewing brad templeton now brad at the beginning of the showyou talked a little bit about your timeline um and you were a little um cautious about futurists makingpredictions that are that are time based uh but you had said that you you had a umaccurate timeline for robot vehicles and you know if you don't mind tell me alittle bit about first of all where are we on that timeline and when am i going to have my robot car but secondly howdid you arrive at that we're really interested in methodology like how do you do your forecasting well there iwish i could tell you i have like you know a methodology or a formula for how you do it as i said naming dates isusually a mistake uh my colleague ray kurzweil uh he likes to say that we'reactually usually over optimistic about the near term and the pessimist and we underestimate the far term he's youmight have some merit to that thing again it's not a universal thing and and rey foolishly does name dates thesefurry figures been you know identifying uh when a.i will match human beings and when some sort of singularity willhappen so i'm not sure i would do that in fact when people ask me when can i get he's been reasonably accurate on anumber of things though he would have no ra ray is ray is definitely above average uh well above average even iwould not say that he he has that same phenomena i had about uh you know you always remember oror modify your memory of your predictions to make your predictions more accurate he ray will claim he'slike 97 accurate and and he isn't but he's still quite good so i'm not tryingto condemn him in any way there um but when people ask me when will i get my self-driving car my answer i usuallygive them is uh june 23rd of 2024 at 4 pm pacific timejust to tell them there isn't but a a real prediction actually the the more serious answer i give of course is thatit isn't in the same time in every different place uh never more have we seen the future arriving but not beingevenly distributed than this in the sense that if you live in suburbs of phoenix arizona you've been able to ridein waymo's cars since uh 2019 uh with nobody in them uh uh in more recenttimes that's now possible in san francisco from both waymo and cruz a unit ofgeneral motors it's possible in shenzhen it's possible in beijing it's possible in shanghai possible in guangzhou um andit will be possible uh and uh what spread brett has a yes i have a tesla too or you've tried you're trying to buyone yeah good luck i am no no i've got it it's coming on july 13 reportedlywith my with fsd so uh well so uh we could do an hour-longprogram about tesla's supposed fsd which is not f or s or d umbut it is uh it's a prototype it's interesting i have it on my car and i've uh i have reviewed it somewhatnegatively i'm afraid um but that is again the source of you know easily a few hours of discussion about how thatthing works but the real self-driving cars the ones that are operating there's nobody in the vehicle it pulls up picksyou up takes you somewhere else those are here if you are in these particular placesnow this is an effort to build what's called a robo taxi something it's basically an uber with no driver in ituh it's much easier to explain it to people by the way now that you can you know just point to uber in the past it was a little hard to imagine this cloudof cars that served you what tesla wants to build is um a privately sold personal car which hassome self-driving ability which is a different thing and and useful but actually harderbecause when you're building a taxi service your job is just you know make a commercially viable taxi service areawhich san francisco is that's where uber began just with san francisco new yorkis a variety of other places are as opposed to trying to build a car that can drive everywhere which is what isdemanded by car-buying customers if you bought a a car and it had this great feature you spenta lot of money from but it only worked in you know east los angeles well it's no you're not going to buy that or you could only sell them in east los angeleswhich is not of interest to any car company so tesla's also taking a much harderpath to doing it and trying to do it cheaply which is again a very odd philosophy so when they'll succeed ismuch harder to predict but the others have already succeeded sort of because even though uh waymo has nowdriven uh probably about 12 million miles around phoenix maybe maybe eight to ten they haven'tbroken down the numbers recently without ever being at fault in an accident that's uh the equivalent ofabout 10 to 12 human lifetimes of driving without being at fault in an accidentwhich is better than humans can drive so one could say they have achieved the safety goal we've all talked about sincethe beginning is how do we make this safe enough how do we show that it's safe enough i think they've achievedthat uh they still aren't deploying widely though uh one reason is that theydid that in an area that's easier to drive phoenix is one of the easiest places in the world to drivebut the other reason is i'm not sure they've fully conquered being a good road citizenuh which is to say that people honk at them sometimes because they're like grandmothers rather than uh thanordinary drivers and so you have to be both safe and a good citizen of the road somethingthat there's actually a trade-off between which makes it a little more challenging but just in the last few weeks we saw themexpand their service areas in san francisco we saw a whole bunch of new areas appear in china several othercompanies are doing taxi services robot taxi services with a human being still behind the wheel totake over in case of a problem uh in a number of cities around the world so that obviously is more experimental butthe ones that are running with nobody in them while still not profitable because their expensive research projects aremuch closer to you know achieving this goal and the question of when you'll get it so my answer to when you'll get it isthat i predict for this decade a land rush as companies do solve these safetyand road citizenship problems and then they go out to stake territory and unfortunately sticking territory isactually pretty expensive it needs a lot of capital now the company's doing this like general motorsand google and apple and others and amazon want to get in the game well they do have a lot of capital so that's notgoing to be a problem and even the startups are getting access to some of the biggestpools of capital that startups have ever received so it will be done uh even though it will be so capital intensivebut that means is you can't just deploy the whole united states at once even google doesn't have enough money todeploy the whole world or the whole us at once yeah tell me what's difficult here because my understanding is thatThe hard problems of autonomous vehiclesthey've got 80 solved and there's 20 remaining that is challenging maybe it's 10 remaining i don't know um talk alittle bit about the underlying technology and the hard problems of autonomous vehicles well i wouldn't usenumbers like 80 or 20 percent um in fact this is one of the mistakes that people make with the the vehicle that brett'sgoing to receive in a few months is they say you tell me yeah they see itdrive um and and work 99 of the time maybe even 99.9of the time and so um if you want a vehicle that only runs over one of a thousand pedestrians that it comesacross that that's great um but the real goal is99.9999 maybe even another nine on top of that six or seven nines uh and unfortunately the journey from ofthree nines to uh to to six nines is well it's a thousand fold so they're 0.1percent of the way there is how you might express it uh but on the other hand waymo and theothers are not 0.1 of the way they're they're they're much closer in fact i think in the safety number waymo hasachieved that role but what's hard well so actually surprisingly one of the hardthings is proving that you have done what you think you've done it's a bit like drugs right so we had the modernavaccine in february of 2020 but we didn't know it was safebut it was built like a month before there was ever a lockdown the modern vaccine was ready you could have gottenit injected in your arm and some people did who were doing the trials and it wasn't until november that we finallysaid okay now we know it's safe enough we'll put it in arms um if we had somehow magically known how to use it idon't know how we could have magically known this but if we had magically known it millions of people uh would haveavoided dying so this is no minor thing to to worry about this is very concrete sort of and tragic numbersbehind this but anyway with self-driving cars you have to prove that you've actually madeit safe enough just like the drug so you first you have to make it safe because you can't prove it by you knowdriving for um well human beings have a fatal accident about every 80 million miles in the united states so you can'tjust say let's just drive for 80 million miles and see if we kill someone uh or even drive for 10 billion miles and seeif we kill the average number of people because that's very time consuming you're always changing your softwarerelease every few days so you can't even drive that much except in simulator on arelease so proving it is problem number one the other problem that i think needs a little more work still todayis predicting the future being a good futurist and and in fact i often say that the task of a self-driving car it'soften expressed in terms of sensing and planning and all sorts of other technical robotics termsbut the real job is to predict the future and just not have any ones in which things get hituh so what you have to do is your sensors tell you where things are hopefully i mean that again is whatpeople work on is to get the census to be very good at letting me know where pedestrians and cars and so on are butreally what i want to know is where are they going to be um and human beings today are stillbetter at predicting what other human beings are going to do in the future than robots are and so that's one thing that we shouldimprove in our robots and if people are working very hard on exactly this you know get to know where things are goingto be we could still improve our sensing though i think getting faster responseFaster response timetime in certain highway situations high speed situations um is good it still human beings takeabout three quarters of a second to react to things they see on the road the robots are faster than that but i'd liketo make them even faster do you think the compute power will reside in the vehicle itself or will the compute powerbe like you know in the cloud or edge computing absolutely in the vehicle itself um andthere was a round trip the time the lag time no it's more it's more a question of of um reliability you want to testyou want to build everything if you're going to bet your company on the safety of a system you're deploying you don'twant to bet it along with betting on the reliability of verizon umyou know or or any other player who's not you um you you mostly want to bet it on andcontrol and test and certify and verify every component of it um you don't wantany kind of radio outage to suddenly mean you could have a safety incident so there'll be processor on the car youthink there'll be sufficient processing of power in the cars oh yeah yeah i know there's there's some there's more than enough processing on the power in thecar and how about mean the only thing sorry we'll say again what about karate car communication like what it makeswhat about for like 3d point mapping and stuff like that yeah yeah terrible or horrible mistakeum so that won't there are in fact i'm just doing having writing an article about uh um the the the the dumbness ofsmart roads uh to to put it in an ironic way um so there are a whole bunch of reasonsagain this is something that i've talked for an hour about uh butComputer securityyou again you you don't want to depend on others for your information for you don't you can't you can't depend on itif you were expecting other cars to communicate with you uh some of them won't um not all of themwill have that equipment some within the equipment will be broken so what that means is you have to workextremely well without communicating with anyone you it's not especially the first car on the road can't talk toanybody so the first car on the road has to meet this goal of safe enough and that's a very very high goal safer thanhuman beings drive you have to meet that goal without any communication so the most the communication could do is takeyou from safe enough to safer enough just a little tiny increment and so it's not really that valuablebut it creates a computer intrusion risk when you want to do good computer security the first thing you do withyour computers is teach them the same thing you teach small children don't talk to strangersuh and uh because that's where computer intrusion comes when you have a web server that will talk to anybodyuh when you have a something that will accept a communication or open a communication with random parties that's where most computer security violationstake place so the right thing is your car talks only to its headquarters it can't avoid talking to its headquarters but franklyit should be very afraid of even its own headquarters if it can be uh and uh there's really no need to gobeyond that people have dreamed up uh things they might do if they could talk to a traffic light or talk toanother car but almost none of them actually need that and so there's no reason to create a computer securityvulnerability that gains you almost nothing and also you don't buy into the conceptof like convoy you know vehicle to vehicle transmission so that you can have that you know sortof cloud computing power of the combined vehicles particularly in a highwaysetting well so for specific convoys that want to follow you know just a few feet from each other um there can bevalue in that although and there is a company that attempted to do that unfortunately they did go out of business there's a new company that'sjust started up doing that and they may do better um but no no generally not infact one of the problems with convoy is the there was a big experiment in sweden run by volvo and a european consortiumabout convoys and they found two problems with doing it one was if you were the rear core car in a convoyum your the front of your car got destroyed by the little pieces of stone on the roadwhich are thrown up if you follow too closely behind someone else because you know roads have got little tiny chips ofrock and stuff on them all over they're thrown up especially by the trucks and they fly you know 10 feet orsomething like that if you're following by the correct distance it's not a problem if you try and follow closely they had one car its radiator startedleaking just because hundreds and hundreds of stones had been embedded in it sofollowing too closely is not a good thing secondly even if convoying is useful for saving fuel it's notsomething you do on day one you sort of do that after everything else has been perfected and then you try getting alittle closer and a little more dangerous so it's uh even if it is a good idea it's some distance awaybut that's about it yes a radio might be useful for that although i think a laser would probably be better in that circumstance than radioum but that's uh that's a different matter there are advantages to both but no no absolutely you've got now forexample two million of the accidents in the united states out of the 12 million total every year car accidents are withdeer and uh i have not figured out how yeah i had not figured out how to get the deer to wear transpondersum so unless you can you're going to face the fact that you just have to make this work withoutcommunicating with anyone else as human beings do but you communicate with your headquarters you get everything that waze gives you you know you know ifCommunication between carsthere's something going on ahead of you on the road because a car up there told its headquarters that there's a problemhere there's ice there's an accident whatever it is stalled car um that already happens with ways withjust human beings they talk to the cloud and then your headquarters talks the cloud and learnsthat and informs you of things you want to know um you actually can do that now with latency of less than 100milliseconds so you can actually do real-time stuff that way and oddly enough this is this is veryweird so normally i'm a big fan of decentralization and distribution and you would say oh then you should lovedirect radio communication between the cars but it turns out that if you want to directly communicate between the carsyou have to standardize how you do it you have to come up with a way no one's ever built a a radiocommunications technology any kind of communications technology actually that did not provide high value to the veryfirst customer right if because if you have to have a to make radio communications between cars work you'vegot to have hundreds of millions of cars doing it before it's valuable if only 10 of the cars have it and by the way thatwould take many years to happen if only ten percent of the cars have it then only one percent of the interactions between cars can make use of itso the benefit is quite small and it only happens if both sides are speaking the same protocol and have the sameradio hardware and can talk to each other and have a line of sight between each other because this is all done athigh frequencies when you're talking to your headquarters you don't need a line of sight between the vehicles you don'tneed the vehicles to be compatible with anything but their headquarters it's only the headquarters who now have touh be compatible with each other which is a much easier place and then they can act as a gatekeeper though right yes soyeah no no this is one of the reasons why analysts say that tesla has a huge advantage becauseall of their cars are communicating with the headquarters and as a result tesla has a proprietary information advantageabout road conditions and how their cars are performing and where they're performing and so forth that other carmanufacturers don't have and might not catch up to you know in the near term well yes and no uh there's there's anisraeli company mobileye which is now belongs to intel although it will shortly go public again be spun out umthey are in 50 million cars now they don't get the same level of data access to the older cars that they're in butthey've now been working deals with the car companies who buy their chips and put them in their cars to increase thatamount of data access so they actually have a much larger fleet that they can gather data from than tesla can althoughthey can't get as much control and data as tesla can do and waymo has a much smaller fleet than tesla has but waymogets orders of magnitude more data from each vehicle so there's actually no uhone winning strategy among these three hey so brad let's change tack a little bit hereFuture goalsbefore we we close out the show um you know i want to go full futurist here right um soi told you 2024 june 25th what what is what is the stuff about thefuture that really gets you excited what do you think is going to be the most meaningful changes we're going to seeover the next 30 to 50 years that really is going to positively change the world for us uh you know those of us who arethe kings well so um because as you can see i bleach my hairwith peroxide every day so i can look more distinguished i will admit an age bias in this particularanswer which is i want to stop aging oh yeah and that unfortunately now trumps almostall i can't use that word anymore can you but um almost all other uh goals imay have for the future is that i have a future um and that it's it's a it's a youthful onerather than a decrepit one um and uh for that actuallyuh while there's lots of interesting research going on and we track it and i'm sure you've talked about it in otherepisodes of this show or will um but medicine is horribly archaic in its time and weModern medicinetalked a bit about how long it took to figure out the moderna vaccine could begiven to people um we were able to solve the pandemic in a week but our uhstyles of doing medicine do not permit that to happen there are other more direct stories there's a drug calledflavoxamine which is an uh an ocd drug it's been on the market for 40 years it'sas safe as any of those drugs are i wouldn't say perfectly safe but very well understood how safe it is and in june of 2020or maybe it was may some researchers discovered that it seemed to seriously reduce covet deaths and overtime much more research was done and found that yes i would cut covet deaths in half if you administered this drugand because it's well-known and safe and already approved it could be prescribed off label by anydoctor but none had the guts to do it i talked to many doctors during this period about it there were some doctorswho would do it i wanted it for my mother when she was diet coded um but you could not get it because it wasn'twell established about this now later in later 2021 more research came outconfirmed these numbers and now it is on the recommendation list for treatment uh it turns out we have a new antiviraldrug semi-new antiviral drug called paxlovid which is even better at preventing cover deaths and so fluvoxywon't except you've got to get it in the first five days and many people don't know they've got symptoms right wellalso it's by the way it's uh it interacts with half the drugs in the world so you also have to be very careful about how you take it um but itwell my point is though that this drug exists and it also took time to get it approved and so on um and so it's nolonger this abstract thing to discuss in journals about how many million people died because we weren't ready to usethat drug or to as easily use these drugs i mean we're talking death tolls that exceed any of the wars um you knowwell sorry not a death toll that exceeds the total death toll in the second world war which is about 60 million peoplebut the death of the united states certainly in that war is far exceeded and any all the wars in fact in thehistory of the united states so um we really have to rethinkhow we do now one thing i have proposed is that the united states should create a board certification in experimentalmedicine which any doctor could uh study and get board certified in and they would then be allowed todo to prescribe experimental medications in a modern world whereevery patient you prescribe an experimental medication to has a smartphone not a not a big leap anymoreum has with the smartphone uh these medical sensors you can get now that you can hook up to your phone withbluetooth it can do your eeg they can read all your vitals they can be tracking all things even in some casesdoing blood tests which we're going to get better at so we start getting this idea of making it much easier to bothwork with experimental medicines and also to immediately detect when they're going wrongand if we can do that we can much more safely uh test and deploy and figure outwhat medical therapies work and also possibly ways that treat aging because if even if there's a thing that's beendiscovered which will give you eternal life today it might not be ready by the time i'mold um and uh and and i'm not that old yet um sothis is actually we're on the clock where all of us are on the clock we're all on the clock um in fact uh what i what i've actuallystudied uh or read some research on um there's a very simple piece of math you can do that our current regulatoryregime the fda it can take uh six to ten years to get a drug approvedand so they were able to simply calculate let's take all the drugs that didn't get approved because they were bad they did something wrong they heardsome patients in the trials the safety trials so if you took those drugs you imagine they were used 10 times as much or50 times as much in experimental medicine trials managed by these board certified doctorsand you could calculate just how many people would have been harmed because of that but remember everyone's got the smartphone so as soon as something goeswrong they're reporting it any pattern in things they're reporting are being detected by ai within days we're sayingoh look this drug is causing some people to have heart murmurs let's immediately look into that let's find out what's going on so you can work that out theother number you can calculate is what if all the drugs that did get approved and ended up saving lives and curingdisease 10 years later what if they'd been out and ready you know six or sevenyears before that and how many people would have been saved and it's not even in the same ballpark uh you the numberof people harmed by that regime is like orders of magnitude smaller than the number of people saved by that regimebut it's all a data problem in the end you know it's all about getting getting the right data and being able to makethese contextualized decisions right data will maybe guess around the real problem thereal problem is are now archaic principles of medical ethicsright do not hurt the patient wait till you hear a symptom before you start a treatment rather than trackingthe likelihood of a particular condition genomics gut bio blood work you know internalsenses all of those things hey brad brad you know i know um you know we could continue this this conversation butwe've we've run out of time unfortunately yeah no worries um so episode two of the bad templates a veryironic a very ironic statement to make when we're talking about aging absolutely we have run out of time we'rerunning out of time um but thank you uh for your uh enormous generosity in giving us your time wherecan people find more about the the stuff that you're writing and and the stuff that you talk about well i have a blogideas.for brad and a website templetons.com i also write about transportation on the forbes.com websiteso you can find all those things if if you could somehow get a giant network of connected computers where you could lookup things you'll probably be able to learn where those things are or there'll be links in the podcast description probably therewill be there will be thank you very much brad great to see you again and uh you know i hope youhope you get to uh to wait out the cure of longevity like uh like all of us butuh thanks for joining us you've been listening to the futurists so this week partnering with our friends over atbreaking banks talking to brad templeton a uh a thought leader and futurist inthe technology space over many years it's been been great if you like the episode don't forget togive us a five star review on your podcast platform of choice make sure you tweet us out or share it on linkedin orwherever it is that you share such things with with your friends and your your tribeum but thanks for listening and we will return next week with more of the futurists until thenwe'll see you in the future [Music]well that's it for the futurists this week if you like the show we sure hope you did please subscribe and share itwith people in your community and don't forget to leave us a five star review that really helps other people find theshow and you can ping us anytime on instagram and twitter atfuturist podcast for the folks that you'd like to see on the show or the questions you'd like us to askthanks for joining and as always we'll see you in the future

Related Episodes