00:00hi everyone welcome to the a6 and Z
00:01podcast I'm sonal and I'm here today
00:03with Chris Dixon a general partner on a
00:066 + Z crypto and Steven B Johnson who is
00:10the author of many books including where
00:12good ideas come from the PBS series how
00:14we got to know a book on play called
00:17Wunderland and his latest book is
00:19farsighted which is how we make the
00:22decisions that matter the most so
00:24welcome thank you for having me
00:26could you start just telling us a little
00:27about the book yeah this is a book that
00:29has been a long time in the making which
00:31is appropriate for a book about
00:32long-term decision-making I had a long
00:35incubation period one of the things that
00:37occurred to me that that got me
00:39interested in this topic is that there
00:41had been a lot of material written both
00:45in terms of academic studies but also in
00:47terms of kind of popular books but a
00:49disproportionate amount of that was
00:51focused on people making gut decisions
00:53or instinctual decisions like thinking
00:55fellow also blank is like that it is
00:58amazing the amount of processing and and
01:01all the heuristics we have for making
01:03short-term instinctual decisions but the
01:06decisions that really matter the most
01:08are slow decisions are decisions that
01:11have a much longer both time span in
01:14terms of how much time you spend
01:16deliberating them and then also the time
01:18span of their consequences and I got
01:22interested in what the the kind of the
01:24science is and some of the art in a way
01:26behind those kinds of decisions actually
01:29the book partially starts with the great
01:31excerpt in in Charles Darwin's Diaries
01:34where he's trying to decide whether to
01:36get married Edie it's it's a beautiful
01:38it's a beautiful lizards like okay
01:39against getting married I'll give up the
01:42clever conversation with men in clubs my
01:45favorite of against marriage was less
01:48money for books etc yeah and you know I
01:54thought looking at it it's kind of
01:55comical and and sweet in some ways but
01:58that technique of creating a pros and
02:00cons list basically that was state of
02:03the art in 1837 1838 and it's still kind
02:07of state of the art for most people
02:08that's the one tool they have for making
02:11a complicated decision and actually we
02:13tools and we have a lot more insight
02:15about how to make these things seems
02:16like there's two questions right there's
02:17a descriptive and the normative question
02:19kind of like the script like how do
02:20people make these decisions
02:21yeah and how our societies are you know
02:23governments or whoever the actor might
02:24be and then there's a second question
02:26how one should make these decisions I've
02:29got more and more interested in in the
02:31second question right like how what are
02:34the tools that you can really use to do
02:35this in your life can you get better at
02:37it yeah and it's a tricky one I was
02:39really grappling trying to take very
02:40seriously the legitimate objection to a
02:45book like this which is that it is in
02:47the nature of complex life decisions
02:50career decisions should I get married
02:52decisions should I take this job
02:53decisions that each one is unique right
02:56that's what makes them hard is that and
02:58they're made up of all these multiple
02:59variables and competing value systems
03:01and stuff like that and it turns out
03:03really that a lot of the the the science
03:07of this and the kind of practice of
03:09making a deliberative decision is a set
03:12of tricks to get your mind to see the
03:15problem or the crossroads or whatever
03:17you want to call it in all of its
03:18complexity and to not just reduce it
03:21down to a series of predictable patterns
03:22or cliches or stairs and that's where
03:24actually the advice I think is useful
03:26and so that's like the scenario planning
03:28where where there's sort of a discipline
03:30around what's the upside case the middle
03:34case it frameworks to for forcing
03:35yourself to kind of mentally traverse
03:38different future paths yeah exactly well
03:41one of the big themes of the book that
03:43runs throughout it in lots of different
03:44ways is the importance of storytelling
03:47yeah and in all these different ways
03:50scenario planning is one example so and
03:52that's usually used in a kind of
03:54business context right so you're like
03:55okay we're trying to decide should we
03:57start this launch this new product let's
03:59generate some scenario plans for what
04:00the market is gonna do over the next
04:02five years but let's generate multiple
04:04ones yeah let's not just per day
04:05different ones who worked at a large oil
04:07company and their scenario planning
04:08group and I you know first doesn't sound
04:10like that interesting but it turns out
04:12these large oil companies like whether
04:14oil is thirty dollars or hundred dollars
04:15you know it's a lot of money's at stake
04:17and so they had this infrastructure like
04:19thousands of people it was like the
04:21State Department or something you know
04:22yeah it was quite fascinating to hear
04:24about and like what if there's a war in
04:27and oil drops this much and what do we
04:29do and like just the level of rigor I
04:31never imagined it was as complexes you
04:33know as sophisticated as it was well I
04:35had some great conversations over the
04:36years with with Peter Schwartz whose
04:38shear in the Bay Area and and he was one
04:41of the pioneers of scenario planning and
04:43one model that that he talks about is
04:45you do three different narratives one
04:48where things get better one where things
04:50get worse and one where things get weird
04:53yeah I love that because I think that
04:56all of us kind of intuitively build the
04:59like it gets better it gets worse kind
05:00of scenario plan in our head it's useful
05:02to actually walk through and and do it
05:04and tell that story but the weird one is
05:06what's cool because I'm you're like what
05:07would be the really surprising thing and
05:09and the funny thing at least if you look
05:11at history the weird is often the case
05:16and I think it in a key part of it is
05:19that if the predictions don't even have
05:22to be right on some level for it to be a
05:25useful exercise because a lot of this is
05:28about recognizing the uncertainty that's
05:32involved in any of these kinds of
05:33choices it's creating a mindset that's
05:36open to unpredictable events so going
05:39through narratives where you imagine
05:40alternatives even if they don't actually
05:43turn out to be the case they they get
05:45you in this state so that when you do
05:46encounter an unpredictable future
05:48whatever it happens gonna be you're
05:50you're more prepared for it you've
05:51thought about at least some of those
05:52variables but the other thing I was just
05:54gonna say on the storytelling front that
05:56one of the places where it kind of came
05:57together there's there's a lot in the
05:59book about collective decisions like
06:01what do we do about climate change or
06:03what do we do about the potential threat
06:05from super intelligence and AI write
06:08something that we think about a lot well
06:09here I'll see generational days yeah
06:11it's super long term decision right
06:13right and one of the points that I tried
06:16to make in the book is while we have
06:20this cliche about our society that we
06:22live in this short attention span world
06:24and we can't think beyond 140 characters
06:26and all that stuff the fact that we are
06:30actively making decisions that involve
06:33changes to the environment that might
06:35not happen for another 20 or 30 years
06:37and we're thinking about what the planet
06:41a hundred years is something that people
06:43have not really done before they built
06:45institutions designed to last for longer
06:48periods so they've built pyramids
06:49designed to last but they weren't very
06:52good at thinking about you know we're
06:54doing these things now what will be the
06:56consequences eighty years from now from
06:58these choices for me regardless of what
06:59you think about whether we're doing
07:01enough from climate change now the very
07:02fact that it's in a central political
07:05topic is is not that was not the case
07:08hundred years it's a sign of progress
07:10and superintelligence is even better
07:12example of it I think because the fact
07:14that we're having a debate about a
07:16problem that is not at all a problem for
07:18us now but that potentially might be a
07:19problem in 50 years now that is a skill
07:22that human beings didn't used to have
07:24that we have now when I was talking
07:26about this once with Kevin Kelly out
07:28here another Bay Area person he had this
07:30great point which is like this is why
07:32science fiction is such an important
07:35kind of cognitive tool because you run
07:38these alternate scenarios of the future
07:40and they help us kind of imagine what
07:42direction we should be steering in even
07:44if their made-up story
07:45don't people actually say that science
07:47fiction is the only way to quote predict
07:49the future in terms of what you can
07:51actually think of for very complex
07:52technologies or I feel like I've heard a
07:54statistic or observation to that effect
07:56I mean I certainly think that you would
07:58you would find more things that ended up
08:00happening in fictional accounts then you
08:03know official people making predictions
08:04about the future outside of a fictional
08:07context yeah my bias has always been
08:09towards history for example like the
08:11only way you're ever going to possibly
08:12get a lens on how to predict the future
08:14is to read a lot of history understand
08:15how things work because of such complex
08:17systems that you're not gonna you know
08:19have empirical data and polling and
08:21everything else to analyze this yeah I
08:23wonder if to what extent are ways of
08:25thinking about these things in academic
08:26literature and things like this have
08:27been shaped by the kind of the the you
08:29know look we require everything be
08:31testable right you also dramatically
08:33narrow the things that can be tested
08:34yeah right or the things that can be
08:36tested it's a is a subset of the things
08:40that are interesting and worth exploring
08:41in the world and you get steered towards
08:43those things I made this decision with
08:46my wife to move to Northern California
08:47having lived in Brooklyn and New York
08:50for a long time and you know when you
08:52think about a choice like that there are
08:55variables they're variables about the
08:57economics of it they're the the kids
08:59schools do you want to live in a city
09:01you do you want to live near nature all
09:02these different things is incredibly
09:04common all the second-order things you
09:05could never predict right well what will
09:07the concentrating the changes yeah you
09:12know you're you know particularly with
09:13children you know you are changing the
09:15overall arc of your kid's life by making
09:17a choice like that and that's scary and
09:18but to your point that kind of decision
09:21well certainly I would say it's one of
09:23the most important decisions that I ever
09:24really thought about it and kind of
09:26worked through with my wife
09:28how would you study that in the lab
09:30right yeah everybody we've got ten of
09:34you that are going to move you mentioned
09:42the book simulations we have like a
09:45actually some investments in this area
09:46but like the idea that computing is
09:49getting powerful enough that you could
09:52you could ask questions like we want to
09:54fix the New York subways and we want to
09:55shut down these subways how does that
09:56have what are all this you know
09:58consequences of that or we change
10:00interest you know there's always there's
10:01always been this interview for the Santa
10:03Fe kind of yeah I think it's still kind
10:08of this fringe I always think about I
10:09have friends who did machine learning in
10:12the 80s and back then it was this kind
10:14of rebel fringe group an AI right so
10:16mainstream AI back then was heuristics
10:18based just like okay we're gonna win all
10:20these things by you know literally
10:22putting in these rules and teach em
10:24computers common sense and there was
10:25this kind of rebel group that said that
10:27will never work you need to use
10:29statistical methods and have the machine
10:30learn now fast word today like machine
10:32learning and AI are synonymous right
10:33yeah it feels like simulations today are
10:35this kind of Fringe group over time like
10:37it just seems like a far better way to
10:39test these really complex things like
10:41what if you could run a simulation I
10:42don't know if you could run a simulation
10:43for moving to California but you could
10:45run a simulation for changing interest
10:46rates now or for closing down a bridge
10:48those things I think are fairly limited
10:50today you could imagine them getting
10:51orders of magnitude more common more
10:53sophisticated right there's so many so
10:55many things to say that so the first is
10:57it actually gets back to that classic
11:00book that David Guillermo wrote in
11:07theme post after that is one of my dear
11:10favorite people you know I wrote I read
11:12that book when I was I guess in just in
11:15grad school it was one of the first it
11:16was one of the first technics where I
11:18was like ooh this is really fascinating
11:19some some ways my first book was shaped
11:21by that Marc Andreessen Elsa says is a
11:22huge influence yeah yeah and so so we
11:25will I think that is that it's something
11:27that's coming we should explain Sameer
11:29worlds the idea is that there's I recall
11:31you kind of have the whole world
11:32instrumented with IOT devices and things
11:34and you have then the mirror world is
11:37the computer representation of that and
11:38the two can interact and really it's
11:41basically you have all that every single
11:42object in in let's say we're talking
11:45about a city yeah you know is somehow
11:47reporting data on all of its different
11:49states and and then and the computer is
11:52just some massive supercomputer although
11:54it was super greater than his day and
11:56now it might just he by the way today
12:00argues it's just streams of information
12:02right yeah yeah what was that thing he
12:03was like life streams now he's he thinks
12:07about in the context of streams as like
12:09browsers Twitter like the automation
12:10that we constantly live in so you
12:13basically have you know software that's
12:15looking at all that information and and
12:18then the idea would be that it would
12:20develop enough of a kind of an
12:22intelligence that you could say okay
12:23given the patterns you've seen over the
12:25last ten years with all these different
12:27data points if we close that bridge or
12:29if we you know switch this one
12:32neighborhood over to commercial
12:33development what would it look like
12:35press fast forward it becomes a kind of
12:37SimCity simulation but based on actual
12:40data that's coming from the real city
12:41and it was just it's just one of those
12:42ideas I think there's a whole
12:43generations game in the holding where he
12:48actually was playing a simulation then
12:49you realize in the end it's actually the
12:55real war that he's fighting in the final
12:57simulation so the other thing about
12:59simulations is a big theme of the book
13:01it's one of those kind of ways in which
13:04the book connects to storytelling as
13:05well because I think the personal
13:07version of this for the should I marry
13:10this person or should I moved to
13:11California this is actually what novels
13:13do right and that we don't have the
13:16luxury of simulating an alternate
13:17version of our lives because we can't do
13:21yet and we probably won't be able to
13:22prolong it I'm particular the kind of
13:23emotional complexity of choosing to
13:24marry someone or something like that but
13:26we do spend in an order amount of time
13:29reading fictional narratives of other
13:31people's lives and the idea is that
13:34that's part of the almost like
13:36evolutionary role of narrative is to run
13:39these parallel simulations of others
13:41live right and and by having that
13:44practice of seeing oh it played out this
13:46way with this person's life this book
13:47that this other person dies and in the
13:50novel's ability to take you into this
13:51psychological in emergency of what's
13:54going on in a person's mind a great
13:55biography will do that yeah yeah so
13:57reading history as you said is a part of
13:59that but it's in fact the first draft
14:02this book was had just like ridiculous
14:05amount of Middlemarch in it and in the
14:10first draft it I think my editor was
14:11like um this is great but like I don't
14:13know if this is what people need it's
14:14interesting how we spend so much time
14:17either kind of daydreaming about future
14:20events or reading fiction we soar
14:22watching fiction and on TV we spend so
14:25much time immersed in things that are
14:26not by definition not true yeah they
14:28haven't happened or they haven't
14:29happened yet and I think the reason we
14:32do that is because there is an
14:33incredible adaptive value in running
14:36those simulations in our heads is that
14:37then it prepares us for our we're
14:40building kind of the emotional like
14:42logic space or something and I don't
14:44know expanding I always think that cool
14:46like I always get this feeling when I
14:47read a good book it's I think someone
14:49said it makes the world feel larger
14:50right and I think that's another way of
14:51saying it kind of expands the you know
14:55the possible like trees of possibility
14:57yes like your mental sample space yeah
14:59the world is bigger history and you just
15:01feel like it's big anywhere you read a
15:03novel and you feel like the emotional
15:05world is bigger right and that's sort of
15:06more possibilities and and and I
15:10interesting so you're saying that's what
15:11almost like an evolutionary yeah I need
15:13to do that to sort of adapt to be more
15:16sufficiently sophisticated there's a
15:18there's a great essay by tooby and
15:21cosmides I believe names are pronounced
15:24about that kind of evolutionary function
15:26of storytelling and they the things that
15:29I talk about is the precisely this point
15:31that we spend an inordinate amount of
15:33time thinking about things that are not
15:34that would seem to be actually a waste
15:36of time but in fact there's a whole
15:39range of different ways in which things
15:42are not true there's though she said it
15:44was true but it's not true or like this
15:46might happen and thus might be true but
15:48it's not true now or you know I wish
15:50this were true and our brain is
15:52incredibly good at bouncing back and
15:54forth between all those kind of
15:55hypotheticals and half-truths and I
15:58don't mean this in a kind of fake news
16:00kind of way like this is actually a
16:01really good skill the ability to conjure
16:03up things that have not happened yet but
16:06that might is one of the things that
16:09human beings do better than any other
16:12future do it in a like Aristotle said
16:15the point of tragedy was that you could
16:17experience it with an emotional distance
16:18yeah right so you can go that's the
16:20other value of narrative right as you
16:21can go and you can experience and like
16:23look at the logic without so you you can
16:26go and and think about tragedy and how
16:28to deal with it without actually being
16:29overwhelmed by the emotion of it right
16:30and so you're involved but not so
16:32involved that you can't sort of parse it
16:34and understand it right and that's a
16:37great point and and the other thing I'd
16:39just last point on simulations we were
16:41talking about how it's hard to simulate
16:43these types of decisions in the lab but
16:45the one place in which we actually have
16:48seen a lot of good research into how to
16:51successfully make complex deliberative
16:54decisions is another kind of simulation
16:56which is mock trials and jury decisions
16:59right and that gets you into group
17:01decisions which of course is a really
17:03important thing particularly in the in
17:04the business world so like what are the
17:06key I guess components both to the group
17:09composition and also to the process to
17:12determine you know to get you the right
17:13answer so the biggest one which is
17:14something that's true of innovation as
17:16well not just decision making is you
17:19know diversity it's the classic slogan
17:22of like diversity Trump's ability which
17:24is you take groups of high IQ
17:27individuals who are all from the same
17:29say academic background or economic
17:32background and have them make a
17:35complicated group decision and then you
17:38take a group of actually lower IQ people
17:40but who come from diverse fields
17:43professions fields of expertise or
17:46economic fields whatever cultural back
17:48that group will outperform the allegedly
17:51smarter group is that because the more
17:53diverse group will Traverse more future
17:55paths of the tree of the pasta tree of
17:57possibility so the assumption was always
17:59the diverse group group just brings more
18:02perspectives to the table right so they
18:04have different you know it's a
18:05complicated that's a variant to earlier
18:07framework is that like good bad weird
18:09like they'll just simply bring up and
18:11and explore yeah more possibilities
18:14because of their more diverse experience
18:16there's no doubt that it's part of it
18:17right when you're dealing with what
18:18makes a complex decision complex is that
18:20it has multiple variables operating on
18:23kind of different scales are different
18:24you know and it also turns out that just
18:29the presence of difference in a group
18:32makes the the kind of initial kind of
18:36insiders more open to new ideas if you
18:39haven't kind of an insider group a
18:40homogeneous group and you bring in folks
18:42who bring some kind of difference even
18:44if they don't say anything the insider
18:46group gets more kind of right okay they
18:49challenge their assumptions internally
18:52more so there are exercises you can do
18:55to bring out the the kind of hidden
18:57knowledge that the diverse group has
19:00they took the technical term for it is
19:02hidden profiles and so when you put a
19:05bunch of people together and they're
19:06trying to solve a problem come up with
19:07the decision there's a body of kind of
19:09shared knowledge that the group has this
19:11is the pool of things that everybody
19:13knows about this decision that's obvious
19:14for the group to be effective you've got
19:17to get the hidden pieces of information
19:20that only one member knows but that add
19:22to the puzzle right and for some reason
19:23psychologically when you put groups
19:25together they tend to just talk about
19:26the shared stuff like there's a human
19:28that you know kind of desired I can be
19:29like well we all agree on this and so
19:31some of the exercises and practices that
19:34people talk about are trying to expose
19:36that hidden information and one of them
19:38is to just assign people roles and say
19:40you were the expert on this you're the
19:42expert on this you're the expert in this
19:43and just arbitrarily so they yeah okay
19:45my job is to go and be the expert on
19:48this and therefore all more likely
19:49surface hidden knowledge yeah it
19:51diversifies the actual information that
19:53shared not just like the profiles I have
19:55a question about this because I thought
19:56I found that fascinating that you can
19:58essentially define expertise as a way to
20:01of seeking common ground but then later
20:04you talk about this difference between
20:05the classic phrase of boxes and
20:08hedgehogs yeah and how actually it's not
20:11hedgehogs that are deep experts in a
20:13single thing that perform well in those
20:14scenarios but foxes that are more
20:17diverse in their expertise so I couldn't
20:20reconcile those two pieces of
20:21information it's a great question so so
20:23just to clarify it so comes all this
20:26famous study that Philip tetlock did
20:28super forecasting yeah and expert
20:32political judgment and he did one of the
20:35most amazing kind of long-term studies
20:37of people making predictions about
20:39things and it turned out kind of
20:41famously that all the experts are like
20:44worse than a dart throwing chimp at
20:46predicting the future and the more
20:47famous you got the worse you were at
20:49having it but he did find the subset of
20:51people who were pretty good you know
20:53significantly better than average at
20:54predicting kind of long-term events
20:56which of course is incredibly important
20:57for making decisions because you're
20:58thinking about what's going to happen
21:00you can't make the choice if you don't
21:01have a forecast or some kind and what do
21:04you found with those people he described
21:05them in the classic Fox versus Hedgehog
21:07which is you know that Hedgehog knows
21:09one big thing has one big ideology one
21:12big explanation for the world of Fox it
21:13was many little things and as a kind of
21:16monolithic thinker but has lots of kind
21:19of distributed knowledge and so the
21:21reason why that I think is in sync with
21:24what we were talking about before is in
21:26that situation you're talking about
21:27individuals so it's it's a fox and a
21:30hedgehog and what the what the Fox does
21:33is simulate a diverse group right he or
21:36she has a lot of different eclectic
21:38interests and so inside his or her
21:42that's one of the reasons why you know a
21:44lot of the people who really are able to
21:46have these big breakthrough ideas one of
21:50their defining characteristics is that
21:51they have a lot of hobbies that's so
21:53true I used to give the tours at Xerox
21:56for all the visitors and actually one of
21:58the big talking points was when we had
22:00like these big muckety-mucks coming
22:01through was how like there'd be a
22:03material science expert and he'd be an
22:05the world's expert in like goat raising
22:07right there be someone else who's the
22:09father of information theory for
22:10computers and he's like a world-class
22:12surfer yeah they all had one
22:14specific like music whatever yeah
22:16there's a funny connection actually to
22:18Wonderland my last book which is all
22:20about the importance of play and driving
22:22innovation and and so much of kind of
22:24Hobby work is his people as opposed I
22:27mean a classic post on this unlike the
22:29things that as far as people doing again
22:30is what the rest of the world be doing
22:31ten years later I remember the way I was
22:34thinking about it is there's so many
22:35things in life especially the workplace
22:38are governed over you you basically have
22:40a one to two year horizon right yeah
22:42like and that's particularly because
22:43business people almost by definition
22:45right if you work at a public companies
22:47that they're moving by quarter by year
22:48and so where are the places in the world
22:51where you actually people actually smart
22:52people have a ten year plus horizon
22:54behind and it's like probably academia
22:56and then my model would be sort of
22:58technical people on the weekends right
23:00nice and weekends right like this is I
23:01think there's more than it's more than a
23:03coincidence that so many of these you
23:05know Wozniak and jobs and just was a
23:07whole but you know the internet early
23:08internet and all these other things
23:09started off as he's like homebrew clubs
23:11and weekend clubs and things like that
23:12right because it's just simply time
23:13horizon right I mean I think it relates
23:15to your book but like yeah so much of
23:17what we've done what we do in the
23:18business world and just the whole kind
23:20of system right is structured around a
23:22relatively short time probably I think
23:24about it in terms of like what we do in
23:26our job like one of our big advantages
23:27right is the fact that we are able to
23:29take a longer-term perspective just
23:30facedown right where our capital come
23:32from and all the other kinds of things
23:34um and that just lets you do invest in a
23:36whole bunch of things that you just
23:37other people just simply can't because
23:39they're under a different set of
23:40incentives well that's I mean one of the
23:42one of the great things that I got out
23:45of actually deciding to move to
23:46California is spending a bunch of time
23:48with the folks that the long now
23:49foundation you know and you're really
23:51trying to encourage not it's not ten
23:53years it's you know the ten thousand you
23:56know basically to be as long as to last
23:59as long in the future as civilization as
24:01old I tell some people about that that's
24:06problems but so many of the problems we
24:09have now come from not having taken that
24:11kind of time right and in fact one of
24:14the other rifts in the book I started
24:16thinking about like okay if we are now
24:17capable of thinking on longer timescales
24:19if we're thinking about climate change
24:21on a hundred year scale if we're
24:22thinking about super intelligence on a
24:23fifty or hundred year scale like what's
24:25the what's the longest
24:27decision that one could contemplate and
24:29actually Xander Rose who who he hear me
24:36talking about this and he said haha
24:38we're working on this on this project
24:40with on this group called Metis which is
24:42a group that is debating whether to and
24:45what they should if they decide to send
24:48as a targeted message to planets that
24:51are likely to support life now we've
24:54identified their plants whatever and
24:56it's it's a it's similar to super
24:58intelligence and it's a surprisingly
25:00controversial project and there are a
25:02bunch of people including the late
25:03Stephen Hawking who think it's terrible
25:05idea and if you read the three about
25:06your Parliament yeah it's the worst idea
25:08everybody problem I'm sure a lot of your
25:10listeners provoked by definition they
25:14are going to be more advanced than we
25:15are which is a whole complicated reason
25:17why that is but they will be and every
25:19in in of course of human history every
25:21encounter between a more advanced
25:23civilization and a less advanced about a
25:27rooted in the Drake Equation and in the
25:29dark forest analogy and the dark force
25:31idea right is that then therefore the
25:33best strategy was the Fermi's paradox
25:40because right it brings all these
25:44concepts together what I just love about
25:45it is just just because of the speed of
25:48light and the distance you have to
25:50travel to these planets this is a
25:52decision that by definition can't have a
25:54consequence for at least you know 5,000
25:57to 50,000 years depending on the planet
25:59or targeting maybe a hundred thousand
26:01years and so the idea that humans are
26:02walking about be like alright I think
26:03we're going to decide to communicate
26:05with these aliens that we gonna saw
26:07their planet and we'll get the results
26:08back in a hundred thousand die it's
26:15often not about but something that I
26:16think is very confusing about making
26:18decisions in this framework is that you
26:20know we can't predict ten thousand years
26:22ahead but nor can we predict immediate
26:24second and third order effects of things
26:25we build today so my question is I mean
26:28it sounds like a terrible question ask
26:29if in the book is about making better
26:31decisions or but why bother making a
26:33good decision why don't we just sort of
26:35let it work itself out in a series of
26:37complex little because there's you can't
26:41you can't really do you can't predict
26:43the future I mean we don't know how
26:44things are gonna play out yeah well the
26:46question is can you and can you get
26:48better at it I think that was meaning
26:49that I think that's one of things that's
26:51important about Tet locks work which is
26:54that first book was about people being
26:56comically bad at it but he did carve out
26:58this zone and said some people actually
27:00have a strategy that works and seems to
27:03be better than just flipping a coin or
27:05or you know just making it up and so I
27:08think that the you know it's not there
27:10is definitely not a crystal ball for
27:11this and there's not an applied strategy
27:13that works in all situations but I do
27:14think you can kind of nudge it and
27:16because decisions are I mean that is
27:17kind of the definition of wisdom is how
27:19you make the right choice make life
27:20right so I have a question - so you
27:22talked a little about the Fox and the
27:24hedgehogs one of the things you
27:25mentioned your book is the role of
27:26extreme person vs. mainstream and I
27:29thought that'd be really interesting cuz
27:30we think about that a lot like where
27:31ideas come from on the fringes well it
27:33all kind of revolves in the story but
27:35the high line in in New York right the
27:37the now iconic Park there was an old
27:39abandoned rail line one of the great
27:41yeah one of the great urban parks
27:43created in the 21st century and for you
27:46know 20 years it was an abandoned rail
27:48line and I saw a public nuisance and so
27:50on and so one one thing that the book
27:54argues is there's a stage in
27:57decision-making in the early stage which
27:59one should consciously kind of seek out
28:01to do which is to diversify your options
28:05right and folks have looked at one of
28:07the key predictors of a failed decision
28:09is it was a whether or not decision
28:11there was just one alternative like
28:13should we do this or not in a company in
28:15a company yeah but I think it applies to
28:16a lot of things when you just have one
28:18option on the table those decisions are
28:20more likely to end up in a kind of
28:23failure of one form or another so part
28:26of the strategies is that you should
28:28when you're at that early stage and get
28:30you know let's do this versus this
28:31versus this multiply your options in the
28:34case of the High Line for 20 years to
28:37debate about the High Line was basically
28:39should we tear it down or not and it was
28:41really even agree that we should tear it
28:43down it was just who's gonna pay for it
28:44it was like it's a rail line that's
28:46nobody using industrial rail is not
28:48coming back to downtown Manhattan
28:50whatever and so it was just stuck in
28:52this kind of whether or not form
28:54and then this interesting bunch of folks
28:57who to your kind of point about extreme
29:02positions who were not part of the
29:05official decision-making process of what
29:06to do that was the city problem was a
29:08debate between the rail lines and you
29:09know I said but then you had you know an
29:11artist and a photographer and a writer
29:13who kind of gotten attached to this idea
29:15that maybe you could do something with
29:16this space and it was this kind of
29:20marginal set of folks who were not part
29:23of the official conversation about what
29:25to do with this who added a second
29:28option or you know it said listen what
29:29what if we kept it and turned it into a
29:31park that would be amazing because our
29:34politics are so contentious and
29:37polarized there's this kind of default
29:39you know anti-extremism now like we want
29:42to get out you know we get rid of this
29:43extremist but in in a society there is a
29:47certain level of extremism that's really
29:50important so sometimes ideas that are
29:52important and it need to happen come
29:54into the mainstream from the margins so
29:56it's trying to get what I call is the
29:57optimal extremism like how do you and
29:59it's a tricky one I don't have actually
30:00a clear recipe for this but it's I think
30:03when you're making a decision are you
30:05you know are you bringing in those
30:08fringe voices to at least have a seat at
30:11the table yeah one thing I was relating
30:14to the internet like one thing I think
30:15it's so contentiously great about the
30:18Internet is you have all of these niche
30:19communities you know subreddits and you
30:22know crowdfunding what we know we were
30:24investors in in oculus and I don't think
30:26oculus would have ever gotten initially
30:27funded had it not been for the
30:28crowdfunding I mean there's obviously
30:30been you know bad things in the internet
30:32as well but I think for the most part I
30:33believe has allowed some of these kinds
30:36of more interesting and potentially
30:38positive fringe groups to get together
30:41whether that will continue you know as
30:43the Internet has become more and more
30:45centralized and as a topic to we've both
30:47yeah I've talked about before you wrote
30:50a really interesting article for the New
30:52York Times last year about about
30:55something I spent a lot of time on it
30:56was that kind of adaptation of your work
30:57actually I think so you know I think
31:03this the issue was talking about of sort
31:05of the centralization the internet and
31:06how do we make sure that
31:08the internet stays interesting and
31:11diverse and yeah I think good for small
31:16businesses and creators and all sorts of
31:19other people right and and it is an
31:21issue that I think a bunch of people are
31:22talking about right I mean you see it
31:24discussed and when people talk about
31:26these issues like demonization D
31:27platforming you see people talk about in
31:29terms of regulation should should these
31:31platforms be more regulated are we
31:32headed to an Internet that's that's
31:35similar to TV we have like four channels
31:37or control everything you know Google
31:39Facebook Amazon etc and then you wrote
31:41about there's kind of you know fringe
31:43movement that is trying to kind of
31:45through technology principles and
31:47innovations create alternatives
31:50infrastructure yeah there was a direct
31:53connection actually between far-sighted
31:54this book and and that that piece for
31:58The Times Magazine and really the thing
32:00that began in all was Walter Isaacson
32:02wrote a op-ed I think in the Atlantic
32:05saying the Internet has broken you know
32:06and we need to fix it it has these
32:08problems and he kind of listed a bunch
32:10of problems which I thought were
32:11reasonable and so I sent him a note and
32:13I said you know I liked what you wrote
32:15how would we go about fixing it like
32:18what would be the decision-making body
32:21that would decide these are the fixes
32:24and we're gonna apply them and he wrote
32:25back and he said you're right it would
32:27be impossible in this polarized age you
32:29know we can't do it and I thought if
32:34we're just stuck with the infrastructure
32:35we have then really that's that's really
32:38depressing right so I slowly kind of dug
32:41through the writing about it and and but
32:44you know about halfway through it I
32:45began to think that some of the
32:47blockchain models and some of the token
32:50economy stuff that that you've written
32:52about as a way of creating sustainable
32:57business models for open protocols
32:59basically which is what what we really
33:01kind of need I think one of the reasons
33:02that piece worked is that there were
33:05million pieces written about the
33:06blockchain but I didn't actually set out
33:08to write a piece about the blockchain
33:09not to write a piece about how would we
33:11fix this problem and I got organically
33:13led towards the blockchain meanwhile as
33:16that was happening all the crazy ICO
33:19scams were happen again like it was like
33:22my culture exploding I think that mean
33:26so Walter I think I read the same thing
33:28Walter I think he articulates very well
33:30the the negative side of it I think the
33:32positive side is I would argue to things
33:33like one is just the nature like the
33:35architecture specifically the you know
33:37Internet Protocol yeah
33:38being very presciently designed as a
33:42dumb layer yeah that's in a good way
33:44right so that you can reinvent the
33:45Internet is reinvented if the nodes and
33:47the internet appeared themselves right
33:49yeah and so I think of Internet
33:51architecture as the intersection of
33:54incentives and Technology design right
33:57so you have to create better kind of
33:58software that runs in those nodes and
33:59then you have to provide the right
34:00incentives right yeah and it's one of
34:02the fascinating is about the Bitcoin
34:03white paper is it's essentially you know
34:04eight pages of incentives and if you do
34:06the incentives right the Internet is
34:08able to sort of heal itself yeah or
34:09upgrade itself I should say or change
34:11itself and then the question people are
34:13looking at is can you take those that
34:14interesting incentives design and can
34:16you apply it for things that are more
34:19useful than simply solving cryptographic
34:20puzzles like Bitcoin right yeah and and
34:22and incentivize new behavior so that
34:25that's like the other thing I would
34:26think about is is so many of the models
34:28we use and and our hardware based and
34:31including like I know I've read all your
34:33books and like the people you talk about
34:36writing them just by definition or
34:37building usually physical things right
34:39because that's what they were doing pria
34:4020 years ago right and you think about
34:42like you know once you build like the
34:44combustible engine you've basically
34:45built it I mean you can improve it you
34:46know you build a car you basically build
34:48it where software is fundamentally
34:50different this is a Marc Andreessen
34:51point software eats the world like he's
34:53always talking he just thinks people
34:54fundamentally misunderstand software and
34:56keep applying these old you know
34:58physical models of how innovate Carlota
35:00Perez and like all these you know which
35:02a great frameworks but they're all kind
35:03of based on on how Hardware cycles yeah
35:06I guess the one thing that I would kind
35:09of bring out is that I actually didn't
35:11get to in that crypto piece and the
35:13times was the importance of governance
35:17structures inside of these crypto
35:20protocols and platforms and you know
35:23there's always been some level of
35:25governance involved in software in the
35:26sense that you had a corporation or you
35:28had a standards body that was you know
35:29deciding what the actual software
35:32package should be or what features
35:33should be included but now really for
35:36time the governance is actually built
35:37into the code if you think about
35:39decision-making that is what you know
35:42you have governments like we have we
35:44have embedded in this code a set of
35:48rules governing like what we
35:49collectively are going to decide for the
35:51future of this platform and that the
35:55fact that that's now being built into
35:56the software the idea right is the point
35:58of this movement right is that
35:59decentralized to take the power away
36:00from an individual and therefore you
36:02have to think about well then how do you
36:03how do these systems upgrade themselves
36:05and govern themselves and who gets to
36:06decide who gets a voice and all these
36:08questions right yeah who's in the PAP in
36:10the old model you just said okay the CEO
36:12right now it's like well there's no CEO
36:14so how do you figure it out
36:15for masses of people to decide yeah and
36:18coordinated activity at unprecedented
36:19scale so there's been great we've been
36:21talking about decision making and how it
36:23plays out you know in crypto and create
36:25an innovation and also then even in
36:27personal lives like Darwin or even
36:30novels and literature like Middlemarch
36:32but what are some concrete takeaways or
36:35advice not just for how to think about
36:37decision making and being farsighted but
36:40for what both people and companies big
36:42or small could do so for instance one of
36:45my favorite kind of tricks in the in the
36:47book is this thing that I'm Gary Klein
36:50came up with which is a a technique also
36:53to deal with kind of the dangers of
36:55groupthink and making a you know let's
36:57say a work decision where you've got
36:58your team and you've decided we are
37:00going to launch this project where this
37:01product and we're all really excited
37:02about it and so he created this kind of
37:05technique which he goes a pre-mortem and
37:08I love this idea so both more time
37:11obviously the patient is dead you're
37:12trying to figure out what caused the
37:14patient staff a pre-mortem is this idea
37:17is going to die a spectacularly horrible
37:21death in the future tell the story of
37:23how that death happened right in five
37:26years this will turn out to have been a
37:28bad decision tell us why and that
37:31exercise even again it's like scenario
37:34planning it's a kind of negative
37:35scenario planning even if it ends up not
37:37being true the exercise of forcing your
37:39brain deal with the scenery right it as
37:42opposed to just saying hey guys any any
37:44you see any flaws with this plan do you
37:47guys do that when you talk through deals
37:49I think a good investor discipline is to
37:51do something similar to that yeah you
37:53know were you kind of and frankly an
37:54entrepreneur like I think that one of
37:55the myths around entrepreneurship is
37:57that there that there I mean the risk
38:00takers I that said mantra to take risks
38:02but good entrepreneurs are very good at
38:06ordering the risks and then
38:08systematically trying to mitigate them
38:10right I mean that's not to say that they
38:12don't take big risks but you certainly
38:14don't want to take on this see risks
38:15right so I think what you're good
38:16entrepreneurs doing is constantly
38:18thinking about all the different
38:19scenarios how they'll go wrong you know
38:22kind of rank ordering them taking a
38:24bunch of risks but saying hey and the
38:26big you know so my key risk and this is
38:27you know it's sort of like this type of
38:29business it's all gonna be about
38:30financing risk and this one will all be
38:32about talents and this one will be all
38:33about you know yeah well I go wrong and
38:35you see enough of it and of course it's
38:37a very rough and imperfect science but
38:40right but you better than you feels like
38:42you can get it seems like you get better
38:43over time yeah the original patent that
38:45Google filed for the self-driving car
38:47projects included in it is this thing
38:49they call the bad events table okay
38:52basically it's like at any moment in the
38:55car is the cars driving it's creating
38:57this bad events table and the bad events
38:59are range from I'm gonna you know dent
39:02the you know right side mirror by
39:06accident you know just scraping against
39:08this car - I'm gonna collide with these
39:10two pedestrians and we're you know
39:12they're gonna die and there's like 15
39:15bad events that can potentially having
39:16happen given the circumstance on the
39:18road and not only do they kind of list
39:21the bad events but then the software is
39:25calculating both likelihood of the event
39:27happening and then in the magnet the
39:29magnitude of the of the risk right so
39:31two pedestrians die very high magnitude
39:33but if it's very low probability you
39:36kind of measure it and I think of that
39:37as in a sense the car is doing that at
39:40the speed of instinct but in a way
39:43that's the kind of table that would be
39:45really nice to put next to a pros and
39:47cons to you know what are all the
39:50terrible things that could happen yeah
39:51let's rank them with probability and and
39:53with magnitude and just to see it I
39:55think about this all the time actually
39:57in terms of how people make pros and
39:59cons list and how they're so flat
40:00variable wise and if you've gone through
40:02any statistical training
40:03the first thing you learn in any linear
40:05model is how to wait your algorithm and
40:07you wait the variables and I always
40:08think about that later
40:09well I'm gonna give this like a move to
40:11California at 10x weight and my move
40:14back from New York and you'll give
40:15something else to X and you you multiply
40:17all those probabilities in those weights
40:19to come up with your decision I think
40:20that's a very good way of thinking about
40:22it you know pros and cons tables date
40:24back to this famous letter that Ben
40:26Franklin writes to Joseph Priestley who
40:27Quin cently was the hero of my book the
40:29invention there but he's like explaining
40:31this technique he has which is basically
40:32a pros and cons list and he calls it
40:34moral algebra what what gets lost in the
40:37conventional way that people do pros and
40:39cons lists is Franklin had a kind of
40:42waiting mechanism where he basically
40:43said okay create your list of pros and
40:45cons and then if you find ones that are
40:48comparable kind of magnitude on one side
40:51and the other cross them out we would do
40:53it differently now it was a way of
40:54assessing okay these two things are kind
40:56of minor and I got one on one side one
40:58of the other so I'm gonna cancel you
40:59know that's great I think some of those
41:01exercises are really important I think
41:03cultivating a wide range of interests
41:07and influences is a really important
41:08thing to do both in terms of innovation
41:10and creativity but also in terms of
41:11decision making yeah
41:12and I think it's very important to stop
41:16and say okay what would the alternate
41:18scenarios be what what if it gets better
41:20what if it gets worse what if it gets
41:21weird and the other thing about the kind
41:23of diversity point I think that's gonna
41:24become increasingly important the
41:25diversity is actually going to be also
41:27machine intelligence - right and
41:30increasingly part of that intellectual
41:32cognitive diversity is going to involve
41:34machine intelligence and so it's going
41:37to be you know not just you know making
41:39sure you have a physicist and a poet on
41:41you know in your kind of posse that's
41:43helping you make this decision but we're
41:45gonna see more and more people making
41:47decisions for instance you know there's
41:48a lot of interesting research into in
41:51the legal world bail decisions that
41:53normally a judge would make a decision
41:54okay this person should be let out on
41:55bail for this amount or not let out on
41:57bail whatever and there's there's some
41:59evidence now that machine learning can
42:00actually make those decisions more
42:02effectively it's not that we want to
42:03hand over the process to the machines
42:06entirely but the idea that you would be
42:08assisted in making a choice like that I
42:11think it's gonna be something we'll see
42:13more and more I mean I think we already
42:14seeing hybrids of that play out like
42:16quant strategies etc but you're saying
42:18something even more you're saying it's
42:20like a partner in decision-making yeah
42:22it's it's a collaborative model my
42:25friend Ken Goldberg it's at Berkeley in
42:26the robotics program there he talks
42:28about inclusive intelligence the idea
42:31that it's it's not just about you know
42:33just human intelligence versus
42:35artificial intelligence but actually
42:37these this kind of dialogue that you can
42:39have with with the machine you might say
42:40I think I should release this person on
42:42you know a very low bail and the machine
42:45comes back with while looking at all
42:46comparable case studies I think he
42:48actually you know shouldn't be released
42:50at all at that point you're like okay
42:52that's interesting I'm gonna question my
42:54assumptions here and think about what I
42:55might have missed and you don't you
42:57might not change your mind but having
42:59that extra voice in the long run will
43:02probably be better for us right it feels
43:03like crowd intelligence on a whole
43:04different scale yeah were there any
43:07qualities of people that you've seen one
43:09of the things that you put in the book
43:10was that one of the key factors is an
43:13openness to experience as a real great
43:16predictor of very good decision-making
43:19prediction etc and I thought that was
43:21fascinating cuz I thought of immigrants
43:23it's like a defining quality of
43:24immigration and what brings people to
43:25different places it's what you know it's
43:27one of the big five personality traits
43:29openness to experience
43:30it's another fruit and I love the word
43:33curiosity but openness to experiences
43:36it's slightly different way of thinking
43:37about it but you are walking through
43:38life looking for you know this I'm open
43:41to this thing that I've stumbled across
43:42and I want to learn more and tetlock
43:45predictors the super forecasters that
43:48we've talked about they were they had
43:50that personality trait in in spades in
43:52general so it's a it's a wonderful thing
43:55and and it's related I think to another
43:57quality which is empathy right which is
43:59also by the way one of the very things
44:01that fiction helps Oh exactly exactly so
44:04it's when you get into the world of kind
44:06of personal decision making novels in a
44:09sense train the kind of empathy systems
44:12in the brain because you're sitting
44:13there like projecting your own mind into
44:16the mind of another listening to their
44:18inner monologue their kind of
44:20consciousness in a way that almost no
44:21other art form can do as well as a novel
44:24can and so that exercise of just what
44:26would that other person think what would
44:29in so many decisions that we have to
44:30make you have to run those simulations
44:33in your head right because your
44:34decisions up consequences to other
44:35people's lives and if you aren't able to
44:37make those projections you're gonna be
44:41missing some of the key variables that's
44:42great and then finally what do you make
44:44of all those folks that have like these
44:46lists of tips and advice like when they
44:48think about like Jeff Bezos does this
44:50and Elon must do that I think you
44:52might've written mothers in your book
44:53about how Jeff Bezos believes that you
44:55should get to seventy percent certainty
44:57yeah I actually I like that technique
44:59which is to say don't wait for a hundred
45:01percent certainty there's a lot of the
45:03challenge with these complex decisions
45:04is they're just you cannot by definition
45:06be fully certain about so the question
45:08is where do you stop the deliberation
45:11process so you don't just freeze and by
45:14measuring your certainty levels over
45:16time taking a step out of the process
45:19say like okay how certain am I really
45:20about this yeah I think that's a really
45:22good actor so I think those little you
45:24know I definitely included them I tried
45:26it with this book to try and hit the
45:27sweet spot of like these are kind of
45:29interesting tools that have been useful
45:31and that have some science behind them
45:32but also then to just look at the kind
45:35of broad history and some of the science
45:37about the way that people make decisions
45:39and some where have it kind of be a mix
45:41of those two things I think it's great
45:42and especially because we as Homo
45:44sapiens are very unique in being able to
45:46actually have the luxury of doing this
45:48well thank you Stephen for joining a 6nz
45:51podcast he is the author of the new book
45:54just out farsighted how we make the
45:56decisions that matter the most thank you
45:59thank you very much it was great talking
46:00to you Stephen I love it thank you
