0:02
Hi, I'm Jim O'Shaughnessy and welcome to Infinite Loops.
Search this video
0:02
Sometimes we get caught up in what feel like infinite loops when trying to figure things out.
0:07
Markets go up and down, research is presented and then refuted, and we find ourselves right back where we started.
0:13
The goal of this podcast is to learn how we can reset our thinking on issues that hopefully leaves us with a better understanding as to why we think the way we think and how we might be able to change that to avoid going in infinite loops of thought.
0:36
We hope to offer our listeners a fresh perspective on a variety of issues and look at them through a multifaceted lens — including history, philosophy, art, science, linguistics, and yes, also through quantitative analysis.
0:49
And through these discussions help you not only become a better investor, but also become a more nuanced thinker.
0:56
With each episode we hope to bring you along with us as we learn together.
1:02
Thanks for joining us, now please enjoy this episode of Infinite Loops.
1:11
Disclaimer: Jim O'Shaughnessy is chairman and Co-Chief Investment Officer of O'Shaughnessy Asset Management, where Jamie Catherwood is an associate.
1:17
All opinions expressed by Jim, Jamie and podcast guests are solely their own opinions and do not reflect the opinions of O'Shaughnessy Asset Management.
1:26
This podcast is for informational purposes only, and should not be relied upon as a basis for investment decisions.
1:30
Clients of O'Shaughnessy Asset Management may maintain positions in the securities discussed in this podcast.
1:39
Jim O'Shaughnessy: Well, hello everybody.
1:39
It's Jim O'Shaughnessy with another episode of Infinite Loops, and today, I have a very good friend and mutual from Twitter, Johnny Nelson. Welcome Johnny.
1:50
Johnny Nelson: Thank you for having me.
1:53
I've always been excited to hear your voice and see your expression when you said that, because you can hear your smile when you say it, and now I get to see it.
2:02
Jim O'Shaughnessy: It's a really weird thing.
2:03
I'm looking into it at the whole idea about voice and whatnot.
2:08
But I want to start with you.
2:08
I want to start with your handle on Twitter.
2:08
Okay, so, wannabe breaker of bad loops. Can you explain?
2:14
Johnny Nelson: So it might be very, very perfect for the context here.
2:19
I know for me, and for a lot of other people, there seem to be these totalogies we get stuck in for what we believe.
2:24
Or, you might call them local minima, and we get trapped by them.
2:30
And I, for me, would like to break them, because I think they really hold you back in belief systems for other people, and obviously, for me, it's the same process.
2:42
These things tend to really produce behaviors at odds with what we want.
2:48
So when I say, I want to break the bad loops, those are what I'm talking about.
2:53
The things that really, exactly as the intro to this show talks about, those are the things I'm interested in breaking.
2:59
And I say wannabe, because I don't really know how to do it.
3:03
You can't just poke at someone's bad belief and say, "Rip this out."
3:07
It doesn't work that way. Jim O'Shaughnessy: No.
3:10
Johnny Nelson: Yeah, exactly.
3:10
That's what I want to do, and that's what I'm spending all my time trying to do.
3:12
And it's hard, because it's not something that really works that way.
3:17
Jim O'Shaughnessy: Not at all, as a matter of fact.
3:20
I think as we've chatted about, people, if you present arguments that are rational and filled with facts that violate their belief system, they'll double down on their belief.
3:31
They won't listen to your facts, they'll hate you, and then- Johnny Nelson: That's exactly it. Jim O'Shaughnessy: ...
3:37
you'll be put in the out-group and they'll double down on their beliefs.
3:40
Johnny Nelson: Yeah, you're going to train them to not trust you and somehow reinforce what you think was a mistaken belief in the first place.
3:48
Jim O'Shaughnessy: Exactly.
3:48
It's human OS, man, and it's something that has fascinated me forever.
3:52
It's going to dominate a large part of our conversation today.
3:57
So Johnny, you have a quote that you put up almost every day that says, "Attention is scarce and must be preserved," and that's very, very different than, "More information is better." I love the quote. Explain, please.
4:09
Johnny Nelson: So, I did a PhD in computational social science, and Herbert Simon is a titanic figure in that field.
4:19
So I read a lot of his stuff.
4:19
I don't have intellectual heroes, but he's the closest thing that comes to someone I'll go to first and assume he's probably correct.
4:25
So we had this talk, it was, I want to say 1970.
4:31
And if you read the talk, it's really incredible how it could have happened this week.
4:36
So it's aged really well.
4:36
And the whole thing spoke to me, especially the part about attention, because, I guess, the cobbler has the worst shoes.
4:48
I think I'm particularly bad at narrowing my attention.
4:54
My instinct is to rabbit-hole 42 layers deep until I've actually lost why I wanted to do this in the first place.
5:02
It stops being, "I want to solve this problem," and becomes, "I just need to know."
5:06
And it's really funny you're seeing that a lot in data science right now.
5:12
There are good dashboards that tell you exactly what you need to know to make a decision, and then there are these other dashboards that have 42 tabs and can tell you anything.
5:24
And there's this instinct that, well, you have more information there, that's better.
5:28
But what you actually have is a recipe for paralysis in decision-making.
5:32
The quote for me is actually, I've said this a few times, the quote actually isn't for other people on Twitter, it's to start my day and my notifications tab where it's people liking that, who appreciate it, too, and it's reminder for me, it's 9:00, there's stuff you want to do today.
5:47
It might not necessarily be spend the next hour on Twitter seeing what happened last night.
5:52
So it's mostly for me, but I really, really appreciate that other people appreciate it.
5:57
Because I think it's one of the better things I do on Twitter, is just that silly, automated tweet every morning. Jim O'Shaughnessy: Yeah.
6:04
I think it's great because I agree with you, as you know.
6:08
Many people are reaching their Shannon limit, and they don't understand that they're doing that.
6:15
And so they become confused, and rightly so, by the way.
6:15
It's no fault of their own, right?
6:20
So it's happening and people who know about this stuff need to do a better job.
6:26
That's why I love that quote by the way.
6:26
Because you know about this stuff, it's up there every day.
6:30
A certain number of people are going to see it and go, "Okay, I wonder what he means by that?" And then, look into it.
6:37
But the thing that really puzzles me and I've been trying to develop a variety of hypotheses and thesis on it, is this whole idea of human beings interacting with this thing called social media.
6:52
As you know, I'm wildly bullish on many aspects of it, but I'm not ignorant of the many, many downsides of it.
6:59
And it's just like the...
6:59
I was rereading Rules for Radicals.
7:05
I'm sure you're familiar with that book.
7:07
Johnny Nelson: Haven't read it in a long time, but yes.
7:09
Jim O'Shaughnessy: So again, like you said about your professor's mantra, you could read it as if it were written today, and you could also read it as if it were written specifically for social media.
7:19
In other words, make your opponent, label them, define them, make them live up to their rules, or you're not living up to yours, ridicule them, which goes in with memetic theory, right, and just make it personal, right?
7:34
It's like the old thing, Stalin probably didn't say it, but, "One death is a tragedy, a million are a statistic."
7:48
And it really takes advantage of all of those things that I really think are part of our human OS, right? Johnny Nelson: Yeah.
7:57
Jim O'Shaughnessy: When Danny Kahneman, who is well known in behavioral finance, started out, he actually started out looking at people's pupils, and they would show different pictures to them.
8:06
And essentially, what would happen is dilation or constriction- Johnny Nelson: Yeah, pupil arousal. Jim O'Shaughnessy: Yeah.
8:13
And then, he walked away saying, "There's no rationality in the base code of humans."
8:15
This is something that we put on ourselves.
8:22
So long-winded way of getting to the question that I have for you, which is, the current social media platforms, what are they getting right, and what are they getting wrong?
8:30
Because I know you've written extensively on this. Johnny Nelson: Okay.
8:37
I guess, I'll first start by saying, I actually am a, not unbridled optimist, but pretty close to that with social media.
8:44
I'm very, very bullish, too.
8:44
Which is frustrating online because, as you pointed out, I criticize a lot of what's happening.
8:45
Especially in tech circles, there is this perception that some people have about me that I'm just there to criticize it, which is not true.
8:51
I really think it's incredible.
8:51
The abolition of time and space from who we can interact with is incredible.
8:58
It's never happened before, and I think just the potential is still far from being realized.
9:02
So that's why I criticize.
9:08
I'm more interested in that future.
9:08
So I think the fundamental problem is that the benefits of social media largely do require amplification, whether it's social amplification over the graph or algorithmic amplification that's prodding you towards things.
9:23
I think they're both, on balance, beneficial for the discovery process.
9:30
In fact, I don't think you can remove them and still have those benefits.
9:35
That's where I get into arguments with people who identify algorithmic mediation as the problem.
9:40
I think it's a problem with how it's implemented.
9:46
I don't think it's the problem.
9:46
I think algorithms are absolutely necessary to navigate this gigantic space of possibilities that time and space, when you move them, you can talk to anyone.
9:57
So how do we filter that down?
10:00
That part, the social graph and algorithmic amplification, and amplification, in general, are necessary.
10:06
The problem is there is one set of incentives that is largely dictated by one institution, at least on Twitter.
10:16
There's one group, it's Twitter Corp.
10:16
They get to decide the evolution of the system.
10:22
And I think that's bad, because it's, essentially, monocropping incentives.
10:29
Different groups might have different things they care about.
10:33
So there are these little patches in space that have different rewards and benefits, but we're still playing exactly one game in a abstract sense.
10:38
And I think that's where we keep on having these problems, because the system can't evolve fast enough.
10:46
There isn't an ecology of different groups imposing different incentives.
10:59
And this lack of heterogeneity causes these things that...
10:59
I've said this before and some neurologist is probably going to correct me and say, "That's a really bad opinion," but it seems like we're really prone to these seizures, where it's the same incentive structures and lots and lots of different people collide, especially when you see trending topics.
11:15
And they collide, and they think they're having this discourse, having this debate on something, and all they're really doing is pounding their chest, but then learning who's pounding their chest against them. That's the whole thing.
11:28
So I think for the system to be able to evolve fast enough to mitigate these particular pathologies that have a tendency to manifest really quickly, you have to have a landscape that's a lot more varied.
11:40
And I don't know how to do that unless you are in control of the client and the algorithm.
11:46
I think you have to be able to impose your own, this is the best version of Twitter that I can have, or the best version of YouTube that I can have.
11:58
People have to be able to select that, so it creates a lot of variability. Jim O'Shaughnessy: Yeah. I agree.
12:04
And this was your piece that I thought was great, which you titled, Debate Me, and where you point out it's like that meme with the guy sitting there and it says, "I believe in some ridiculous thing, change my mind." Johnny Nelson: Yeah. Yeah, yeah, yeah.
12:18
Jim O'Shaughnessy: And so, one of the things that I've known for a long time, because I've been around for a long time, is that people don't want actual debate.
12:25
What they want, when they're asking for your advice, they're asking for your acquiescence to their belief. Johnny Nelson: Yes.
12:36
Jim O'Shaughnessy: And if you don't give them that support, they are going to react badly, for the most part.
12:45
And so the pretend debates are exactly that, they're pretend debates.
12:45
You and I had a correspondence a couple of months ago, because one of my thesis was that, we're domesticated primates, and so our brains are releasing the same chemicals, whether we're having a physical battle with blood in the air, or whether we're having one over cyberspace, right?
13:04
And so my position at the time, which you and I went back and forth on was, maybe, maybe, social media serves as a steam valve where people can get on, shout into the void, so to speak, and then feel good about that and move on.
13:22
What do you think about that?
13:26
Johnny Nelson: That is true, and it's also one of the things that I think you're not allowed to say, to some degree.
13:29
I think social media, this back and forth adversarial process, causes us to believe a lot of really bad things that are pathological.
13:40
The joke that Twitter isn't real life, which I think is a bad joke.
13:40
But the truth part of it is, the things on a daily basis that people are screaming about, tend to scream about and then go away immediately.
13:53
There is this pressure release that seems to happen.
13:58
And the part of the pressure that it release...
13:58
I mean, it could be good or bad.
14:03
It could be releasing pressure that should be applied productively to a real problem.
14:08
And that's the flip side of how people get distracted.
14:11
But there is some real, something is shitty, I need to say it just so I can share it with other people and they can appreciate my perspective.
14:17
That bleeds a real type of pressure, and more than that, it actually is just validating that, if it's a non-bullshit opinion...
14:23
Yeah, this is happening to me, too.
14:29
You're not crazy, this is a real thing.
14:29
And unfortunately, if it's a bullshit opinion, yeah, the earth really is flat. I believe that.
14:33
Let's yell about it some more.
14:38
So it absolutely does have this, the ability to be seen, which is related to Twitter's incredible power of discovery, is good.
14:43
And if that sometimes manifest as bleeding the pressure release valve, that's probably good, too, because people do need that.
14:49
It's just for the other side of that is, if it is a real problem that's not actually, in any way, solved by yelling, where's the motive force going?
15:03
Jim O'Shaughnessy: Right. Which I totally agree.
15:03
And my way of dealing with Twitter is as an introduction.
15:07
When I find somebody like you that I think is really interesting, I contact you in DMs, and then we talk.
15:15
And in a perfect world, we meet in real life.
15:22
And so from my perspective, it's Twitter used that way is fantastic.
15:22
I mean- Johnny Nelson: Yeah. Great way. Jim O'Shaughnessy: ...
15:30
I just hired a guy who lives in India, Vatsal.
15:32
He's already supercharged.
15:32
The podcast, and that's his whole- Johnny Nelson: He caught a bug.
15:35
He caught a bug in my writing. I was like, "All right."
15:39
Jim O'Shaughnessy: Yeah, yeah. He's good. He's good.
15:39
My bigger thesis is what I call the great reshuffle.
15:41
And we're in media res, I think, of the great reshuffle, and a lot of really great things are going to happen.
15:48
People like Vatsal now have a worldwide job market that they can compete in, and some bad things are going to happen.
15:59
Which is why, for example, I changed my attitude about guaranteed basic income.
16:05
Because a lot of people, through no fault of their own, I want to stress that, it's going to be a different world where we value different skillsets.
16:10
And it's already happening, obviously.
16:15
I mean- Johnny Nelson: Yeah. Last year. Jim O'Shaughnessy: ... at the Forbes... Yeah.
16:16
You look at the Forbes 400, the first one from the '80s, half of, more than half, was inherited money, and the other half tended to be in physical things, like oil, real estate, steel, etc.
16:30
If you look at it now and it's all symbol-manipulators all the time, right?
16:30
So the nerds won, and that means there's a lot...
16:38
we're not going to make this about my thesis, but it does lead to the question: take us through your ideal social media experience.
16:42
So if, let's say Jack hears this and he's like, "I hire that guy.
16:50
Hire that guy, I want him to build some stuff and a platform that maybe you wish existed, what would that look like? Johnny Nelson: Okay.
17:00
So, that's a big question, let me wind up here. So I wrote...
17:02
Here is something that basically all of the people I talk to in DMs about social media did not like.
17:08
I got a lot of shit for it, and I think they could be totally right, like how it was something that I was on the fence on to begin with.
17:13
This just happened to be what happened, four hours before Jack banned Donald Trump from Twitter, and then every social medium followed through, infrastructure followed through.
17:23
It was like this very broad thing.
17:28
Four hours before that I said, "don't ban Trump".
17:28
And I don't want to turn this into a political thing, but this kind of gets to the point I want.
17:33
I think it's really important to be able to have conversations between two consensual parties on any medium.
17:44
Because if you don't have that, you start splitting social space.
17:44
You start partitioning social place into groups who will not communicate.
17:48
And I think that's bad because it can become kind of a reactor for shitty beliefs and only partial...
17:54
just in terms of human experiences, if you are partitioned on some particular dimension, your different social experiences might not cross that divide.
18:06
So you're making decisions on partial information, which generally isn't good.
18:11
So, I do believe you have to somehow preserve the ability for people to communicate on your platform.
18:16
On the flip side of that, the ability to mediate your experience by something like 'mute inbox' is wildly insufficient.
18:22
It's like the most kludgy tool imaginable.
18:27
And you can, I forget what the limit is on Twitter right now, it's like 500 phrases.
18:32
Jim O'Shaughnessy: I'm approaching it.
18:32
Johnny Nelson: I hit it all the time and then I have to decide what thing do I think isn't going to become another bullshit week of discourse, and I'm usually wrong.
18:42
Jim O'Shaughnessy: Exactly.
18:42
It's key that you realize that. Johnny Nelson: Yeah.
18:45
You can't really predict what's going to happen, so if you have this constraint that's really tiny, it's not going to work.
18:53
So, I spent nine months trying to figure out a way to create a social immune system that modulates who you're muting, who you're blocking.
18:58
One of the reasons why it couldn't work was because of that 500 limit for muted phrases.
19:03
And that's partially an engineering problem for them.
19:08
I understand being able to do that will tax their infrastructure in a way that is bad, but that's what I want.
19:12
I really want the ability to completely mediate your experience, which also means bring your own algorithm.
19:17
If something is trending, and if it's in that side trending bar on the right, I want a button that just says, "ignore everything that's happening here", because it's rarely worth my time, but I still get sucked into it, and then I feel stupid for having done that.
19:31
So everything for me is just that Twitter as a medium...
19:34
let me define what I mean, as Twitter, as a medium.
19:34
These small bits of text that approach the chunk size of a thought, that is novel enough, but you can comprehend it quick enough. It's really perfect. At least for me.
19:43
And the ability to propagate these over a social graph of people I trust. That's incredible.
19:49
I mean that's the core of Twitter to me.
19:53
And I don't think I can improve on that.
19:53
Even Retweets, Quote Tweets and replies, and threads.
19:59
That whole structure is perfect to me.
19:59
Everything breaks down in that I can't control the algorithm, I can't control muting and blocking, I can't control what I see.
20:10
And by that, I don't mean...
20:10
I don't want to program all of that myself, even though I have, because I spend too much time on Twitter.
20:18
I want to see a marketplace for that and be like, "yeah, let me try this out.
20:18
See if this is a better experience for me.
20:22
Let me try this particular way of meeting and blocking".
20:26
The way people do it now is with block lists.
20:26
And I think they have really bad properties, because you're adopting, say, 100,000 blocks from someone who...
20:31
you don't know why all of those people are in there.
20:37
And again, you're starting to partition social space with, really not a lot of good evidence for why.
20:43
And I don't think that's good.
20:43
Even when I was trying to do muting and blocking, I think there should be some sort of built-in decay for it.
20:47
So if I block someone who I think is...
20:52
they could just be on tilt this week, they could just be really angry.
20:56
I want to block them because it's fucking up my day, it's not helpful.
20:59
But that doesn't mean I should never see them again, because it could have been a bad day.
21:04
Or if I start blocking people conditioned on one thing, I might be suddenly losing the ability to perceive stuff outside of my experience.
21:09
So, I think there should be this built-in statistical forgiveness.
21:15
If I block someone 10 times, okay, that's good go away.
21:19
But there should be this ability for the social graph to heal.
21:25
So, I guess that was a very long-winded way of saying, I think users have to have a lot more control, not just as some abstract go users, but because you need this really rich landscape, that's dynamic enough to keep up with all the bad events that happen on Twitter, these storms of anger and bullshit. Jim O'Shaughnessy: Yeah.
21:47
So, it resonates very strongly with me.
21:47
In my circumstance, I'm 61 years old.
21:54
I've always been a very early adopter of tech so...
21:58
Johnny Nelson: I can see that very easily.
22:00
Jim O'Shaughnessy: But one of the things that I've just trained myself to do is, I have trending blindness. I don't see it.
22:03
I literally don't see it. And people...
22:11
My son will complain to me about all the Instagram ads that are getting served up to him.
22:16
And I'm like, "what ads?"
22:16
I literally don't see them.
22:16
And that's through a lot of training over the years of what is important to me, what isn't.
22:23
And the other thing that I see a lot of people doing, especially smart people, you can't know about everything. Johnny Nelson: Yeah.
22:37
Jim O'Shaughnessy: Too big.
22:38
Johnny Nelson: The world is huge.
22:38
Jim O'Shaughnessy: It's just too big, it's too big.
22:39
And so, what happens is, you get infected, especially if you're smart.
22:44
You get infected with just enough knowledge to say really stupid things about something.
22:51
And so, one of the things that, to key into what you've said, the in-group, out-group thing is a real problem.
22:58
And so research has basically shown that even people who are randomly assigned to a group, right?
23:03
So, Johnny you're in group A, Jim you're in group B.
23:09
We will have an affinity for the strangers who have been assigned to our group, which is absolutely illogical and a bit insane, to the point where, let's keep it current, if I have given a hundred vaccines, and I'm randomly assigned to group A, my likelihood of giving almost all of those vaccines to people in group A, and none to group B, remember randomly assigned. Johnny Nelson: Yeah.
23:42
Jim O'Shaughnessy: And so, I worry about discovery a lot more, because I think only if you're open...
23:46
So, I intentionally read and follow people I know that I vehemently disagree with.
23:51
And I do that as a reality check because I'm usually wrong.
23:56
I mean, come on, let's be honest.
23:56
And again, I've just finished reading, rereading, David Deutsch's The Beginning of Infinity, and error correction is a must for progress to happen, right?
24:08
So I think, just listening to what you said, could this be a company, could this be an app that you could sell to people who use it as an add-on to Twitter or YouTube?
24:19
Johnny Nelson: That was my goal.
24:21
I was in academia for really weird reasons.
24:21
There was probably never going to be a future in academia for me, we can talk about that actually.
24:27
I don't want to come on that tangent, but the second I left academia, two weeks after, I started working on this.
24:35
The problem is there is insufficient control over Twitter's API to be able to do this in a competent way.
24:43
And the second part, and this might not be true, I've heard some rumors that it might not be true anymore, but the second part was...
24:48
A few years ago at one of their developer conferences, they were very clear in saying, "we don't want you developing a client".
24:58
That's Twitter, that's ours.
24:58
That's like our flag planting.
25:03
So, I was coming up with all of these creative, fun experiments for me, being...
25:03
You know, on the social science side it was fascinating, but I also thought it was a business.
25:08
I think there was this huge gap in how we should mediate our experience.
25:12
And then I, while also researching the business side and everything, I come across this talk, and I'm like, "okay, that, that might be a little bit of a problem in my plan".
25:21
But, the bigger problem was the limitations of...
25:28
If you can't control the client, you have to use their API to say, "mute this", "block this".
25:33
If it's your client, you can do it for anything.
25:33
You can have it really automated, you can use machine learning in the Twitter feed, even, you just say, " move this away".
25:40
If you can't control the client, you don't have a sufficient grip on the problem.
25:46
And, what the conclusion I eventually came was, even if I did, I'm not sure it's right, because it's essentially building a really elaborate sandcastle on someone else's private beach.
25:59
If this starts interfering with Twitter proper, I have no rights to their platform, they can just knock it down.
26:07
So that's, oddly enough, why I actually started getting into the crypto space, was because, I think there is going to be a lot more ownership over the medium and network layer.
26:17
So that enables a lot of things that I care about.
26:17
The problem with that is Twitter is a juggernaut in terms of network effects.
26:25
I'm probably not going to leave Twitter.
26:25
It's really great for me.
26:25
I've spent a lot of time here. It's really fun.
26:29
So I really do hope Jack is sincere with the blue sky effort to, kind of, decentralize it.
26:35
But I posted about that, and your expertise particular, how do you go to stockholders, who you have a fiduciary responsibility to, and say like, "yeah, we're going to blow up our big, powerful note and do something else"? I don't know how he...
26:51
If he can pull that off in some way, then he is this kind of genius that a lot of tech people think he is.
26:55
And I will, I will say that forever if he can, but I don't know how to do that.
26:59
Jim O'Shaughnessy: But on the other side of that argument is, you could make a very strong case that Jack was not serving his fiduciary responsibilities to his clients by, for example, not enabling the ability to trade via Twitter, and there's Robinhood.
27:15
Johnny Nelson: Twitter could have had that.
27:18
Jim O'Shaughnessy: They could... Stock symbols.
27:18
My friend Howard Lindzon told them, "do this, do this, do this", they wouldn't do it.
27:21
So, he makes a fortune with Stocktwits.
27:26
Johnny Nelson: And it's a great site and I love it.
27:28
Jim O'Shaughnessy: And so, the point being that Jack has had...
27:30
I want to stick with your metaphor.
27:30
So, Jack has had the best real estate on best private beach in the world, and hasn't put up any resorts.
27:46
And so, that leads to another question: if it were 20 years ago, would we be saying, "there's no way we're going to dislodge Myspace"?
27:53
I mean, everybody's on my space.
27:53
Why would we even think about doing that?
28:00
So, I do not think it is inconceivable that, at some point, and I'm like you, Twitter is where everybody is. It's a selling point.
28:07
And, it is until it isn't, right?
28:13
And so, let's think about that for a minute.
28:13
Is there a way that there could be an alternate site that was so compelling that you would get... Let me ask it this way.
28:20
Is there a way that some group of brilliant engineers working with psychologists and everyone else, could come up with a site that is so like catnip to human beings, that people like you and me would be like, "Hey, let's go over there". Johnny Nelson: Yeah.
28:34
So, I did have a theory about that, and I started working on it last year and I won't go into too much detail.
28:38
It's a classic mistake, I solved the wrong problems because I...
28:43
This is actually kind of a Jed thing, I've talked about it a little bit, there's some really terrible stuff happening in my family.
28:50
I started working for the sake of finding the hardest part of a problem, so it just went away.
28:54
But I think the original thesis was right.
28:54
And I don't want to get too long into it, 'cause it'll just take way more of the show.
28:59
But I think here, I'm going to answer your question with almost the opposite.
29:03
I don't think people care about privacy.
29:03
I don't think that's going to be the 10x that gets people to switch.
29:10
So, there's a lot of people I really respect who are working on that problem, and it's something I care about.
29:14
Anyone else I talked to, who's not a techie doesn't care. So, I don't...
29:20
decentralization, huge part of the crypto space.
29:20
No one's going to be sold by decentralization.
29:25
Twitter works by, like all of these providers that I go to or are centralized, they provide a service. That's what I want. It works great.
29:33
And related to that, when you talk about being censorship resistant, which again, I care about.
29:40
Any two people want to communicate who are consensual, I think they should be able to communicate.
29:43
Jim O'Shaughnessy: I agree.
29:44
Johnny Nelson: And, and when you extrapolate from that, that kind of means, groups who are consensually communicating should be able to on that platform too.
29:47
So you can kind of say, "well, decentralization sort of has that built in, privacy sort of has that built in", but I don't think that's the thing that gets users to switch.
29:57
My boring answer for the things that get users to switch is really just a better product for them.
30:04
And those things don't fit that.
30:04
So, a better product for them means it's more tailored to them. And I think that is...
30:11
It is really a boring answer.
30:16
I just want variability in the user interface.
30:16
I want to be able to try on this user interface.
30:19
I want to be able to try on this particular algorithm, cause maybe it works in my day better.
30:24
Even the case of context.
30:24
Sometimes, I go to Twitter and I'm in a professional mood.
30:27
Like I'm thinking about, say, a problem in recommender systems or data science, and I'm just stuck.
30:32
So, I want to scroll that side of Twitter for a while, to maybe get an idea, maybe see what people were working on.
30:37
And right now I manage that with lists and it's not particularly good.
30:42
And I just think we can't create the varied contexts we want right now with Twitter, or really any social product.
30:53
So, I guess it kind of is a boring answer, and it kind of sounds cliche to say we need more control over the experience, but that's what it is to me.
30:58
I want the network layer to be really, really plug-and-play amenable to experimentation.
31:03
And it's not right now, the only people who can really strongly experiment with Twitter itself, is Twitter.
31:14
Jim O'Shaughnessy: And so, I have a thesis that if you follow the money, you will be enlightened very quickly [crosstalk] as to why things really get done the way they get done.
31:26
And it's certainly accelerated in the social media age.
31:26
But this idea of falsification of one's preferences is rampant, especially among younger people.
31:39
I work with younger people, I'm kind of a mentor and it doesn't have to be in finance.
31:44
But one of the things that seems consistent to anyone I'm working with that's under age 35, one of the first questions I say is, do you falsify your preferences on a routine basis? And they all say, yes.
31:54
And so that's the intransigent minority, that...
31:54
I talked about rules for radicals, pick your target, label it, freeze it, personalize it, polarize it.
32:06
And so people would rather just say, "you know what? Yeah, you're right.
32:09
Fine", even though they don't believe that.
32:09
So, I think it bleeds into my question for you and, you're not really anonymous, but sort of, right?
32:22
I didn't know your name was Johnny until we got to know each other.
32:22
What do you think about anonymous accounts and, being a presence on Twitter?
32:26
I have kind of a take, I'd be much more interested to hear yours though, about being anonymous and what that means because, full disclosure, our first OSAM research associate is a person who is anonymous on Twitter, and he goes by the handle, Jesse Livermore, And we hired him on the strength of his work. We know who he is.
32:49
Johnny Nelson: Great pseudonym for you.
32:52
Jim O'Shaughnessy: Well, tell me, what do you think?
32:53
Johnny Nelson: So, I am public.
32:56
It doesn't take long to click one or two links from my profile to find my name.
33:01
I don't have my name just 'cause I like the breaker of loops thing.
33:04
Jim O'Shaughnessy: I think it's great.
33:04
I love the explanation too. Johnny Nelson: Yeah.
33:07
So, on the flip side, I should say, I do this once every two or three months.
33:12
I create a new account, completely anonymous, and I start playing with ideas on that, just because I can't do it on my main account.
33:17
There's no way for me to navigate that space because I like asking certain ideas, can come off as offensive to people who follow me, or kind of do the alternative thing, and I think this is even a bigger part, a lot of accounts, I guess I technically count as a big account.
33:36
You're a big account, in my opinion.
33:36
I'm on the cusp of a higher number account.
33:36
Me asking certain questions puts me in the category of, this person is starting to do culture war bait shit, which is annoying, and no one actually wants it.
33:47
So I regularly create accounts, not because I'm trying to bait people, but because I want to see the reaction that people give to particular questions.
33:59
People that, and it's kind of shitty to say, sometimes I'll be the person asking people I do follow as an anonymous person, what they think.
34:04
Because I actually do want to know, but I don't want them to perceive it in the wrong context.
34:16
So, anonymous accounts in this case, and at all, really have a weird privilege in that they can do that without feeling as if their reputation is at risk.
34:31
It lets them probe things that my main account, or your account, can't get at without having a really shitty day afterwards, because your replies and DMs are just going to be this awful experience that makes you want to log off. So, that's really good.
34:48
On the flip side, I think there's this sort of other movement that's happening right now, that seems to be your reputation as social capital should be this thing that's actually financially capitalized, using crypto or something.
34:58
And there's a lot of smart ideas around it, but I also do think that, when you're saying something, you are putting something at stake, and if it's purely detached from this long-term durable reputation that you're trying to establish, it actually makes everything harder to navigate.
35:19
So, there is this balance.
35:19
You need to be able to ask questions in a way that doesn't create a miserable experience that, you very quickly learn: stop asking questions.
35:30
That hides a lot of things from people, especially within your own particular groups that you identify with.
35:38
Once you learn that there are some things that, within yourself, even in a context of trust, you can't ask, you do stop asking them.
35:49
[inaudible] creates these blind spots that manifest in bad ways.
35:55
I think I went off on a little tangent, but yeah, that's my answer.
35:55
I think there's good and bad parts of it.
35:59
And I follow, at this point, I don't even know, I probably follow just as many anonymous accounts as I do named accounts.
36:03
They're everywhere and they're really useful.
36:08
It helps me focus more on what they're saying, than what identity X is saying .
36:15
Jim O'Shaughnessy: Yeah, but it also leads me to the idea that when I look at...
36:16
So, I left regular media, I used to do a lot of regular media.
36:16
So, I was on [crosstalk] Johnny Nelson: I've seen some of the [inaudible] clips, Jim. Jim O'Shaughnessy: Yeah. Right.
36:23
So, I used to do that a lot, and then I had been making a study of it, and kind of around 2017, 2018, I was like, "I'm switching to total social", because that's where the puck's...
36:35
To use a horrible cliche, that's where the puck is going and that's where I want to skate to.
36:40
But the idea that really resonates with me in terms of what you're saying is, people are, because of Overton windows, because of taboo subjects, they are getting incomplete information on really important things that we as a society really do need to have kind of conversations about and come to some form of resolution.
37:03
And I don't have the right answer there, but I think it's a question that we need to continually ask, because the world has literally, and by the way, it's one of the reasons why you're seeing these hysterical hissy fits that mainstream media are engaging in.
37:18
They know what's happening.
37:25
They know they're losing.
37:25
They know that Substack is going to at one point grab all of their best talent, unless they're willing to just pay ridiculous sums of money.
37:37
And so, going social definitely different way to behave, etc, etc. But it's so hackable.
37:45
And one of the things that really I would never do, but I see it happening all the time, and you and I have discussed it in emails and on Twitter, a lot of people don't realize that when they self-select into a group that, really, they're following some usually charismatic person who's the head of that group.
38:02
We won't name any names here, but I have said in the past, they don't really realize that they're non-player character canon fodder.
38:15
And if that group has a really bad day, for whatever reason, it's usually not the charismatic person at the top who's going to suffer.
38:23
It's some lower sacrificial lamb that's going to do it. So, let's talk.
38:29
I mean, you wrote an essay, Where Did the Future Go? Right?
38:29
And I thought that one was really interesting, because it was talking about Cloud computing with, the world is our oyster, but then it became ad-based. Is there any way...
38:36
Let me see if I can combine these questions.
38:42
Is there any way, do we have the technical stuff in place where we could prevent a lot of this ability to hack, human and OS, or does it always have to be ad-based?
38:57
Does there always have to be a revenue stream attached to it?
39:01
Johnny Nelson: So, I've actually moderated my opinion from that essay a little bit.
39:03
I'm blanking on, there's two people I'm thinking of.
39:03
I'm blanking on their names.
39:08
One problem with my strong criticism of ad-based does seem to be how many people it will exclude from the conversation if you transition out of ad-based. And I'm not that person.
39:22
I would pay, so it's not as strong as an issue for me, but when you actually start thinking about who's online, the entire world, and I live in Palo Alto, I'm basically saying without ads, yeah, all the people, unlike me, which is this very small segment of people, I don't want them here.
39:41
So, once someone played it back to me that way, it was like, "Okay."
39:41
Maybe my opinion is really bad.
39:46
So, I dislike the idea of ads, because I feel like ads are part of that hacking process.
39:52
There are some things I don't want to reach me.
39:52
And if you could pay to do it, it is kind of a nuisance.
39:56
So, if I could pay to opt out of ads, that's great.
39:56
And if ads are also part of it, that's great.
40:01
Maybe both is a better mixture.
40:05
I think the technology to create mediums where we don't need them and can still have communication are emerging.
40:13
That is a lot of the decentralized web tech.
40:13
The cost of being able to communicate with people over the Internet in a social way is lowering at the same time the technology is emerging.
40:25
And that was what I was trying to work on next year.
40:25
The only problem is it's not a one person effort.
40:31
It was a little very silly and arrogant for me to start doing it that way.
40:35
But I do think we're going to start seeing that.
40:35
A lot of people are working on it.
40:39
Paul [Froese], I'm mispronouncing his last name, I think, is working on something called Citizen, that's built on federated identity that gets to do something like that. Very Twitter-like. Mastodon tried to do it.
40:49
I think Mastodon was a great effort, and it was really cool.
40:54
It's just kind of what you said earlier.
40:54
You really do have to make something profoundly better than what we have to get people to overcome the network effects, and creating something that is Twitter but federated isn't that, because all the people aren't there.
41:09
And that's exactly the experience you have.
41:13
So, I don't know what it's going to be.
41:13
I think you need to have a very clever hack that gets people to join this new thing and Twitter at the same time, and slowly peels them off.
41:24
And lots of social mediums are starting to do that.
41:24
Take Clubhouse, for example.
41:24
Clubhouse, you had to join with your Twitter.
41:31
Well, you didn't have to, but it was part of the onboarding process.
41:36
And it went through your phone and started recreating your social network.
41:36
So, I think it's going to be that social experience, where your your social graph gets peeled off onto a new medium that allows people to communicate with each other without a central meeting influence.
41:51
Not because they care about decentralization, but because it's going to facilitate this richer experience that they can tailor exactly as they want.
42:00
Jim O'Shaughnessy: So again, incentives. Right? Johnny Nelson: Exactly.
42:02
Jim O'Shaughnessy: So, what I hear you saying is that this is mostly, we've got to figure out a way if you and I were in business together, right?
42:06
And the question I would ask is, "Okay, Johnny, how do we incentivize this group of people, this huge group of people over here on Twitter, to come to our app," whatever?
42:17
The challenge that I see there, and I don't have any formed thoughts on it that make any sense.
42:24
But Clubhouse is a good example, so I'll just use that.
42:31
So I went over, I got on Clubhouse.
42:31
I listened to a couple of them.
42:37
There was a really interesting one that Marc Andreessen and his partner did.
42:43
And they had open subjects that I really enjoyed it.
42:43
And I was like, many people close to me were like, "That sucks. Clubhouse sucks."
42:49
And then, someone said to me, basically, not joking.
42:56
He was like, "Listen, man, you will be the king of Clubhouse if it gets there, because you can speak continuously and extemporaneously about a lot of shit."
43:08
And so, I very flattered, but guess what?
43:08
Twitter spaces came, and I mean, I have a Infinite Loops Clubhouse at Clubhouse, but I was talking to the team, and they're like, "Jim, why would we go over to Clubhouse when you have 135,000-" Johnny Nelson: [crosstalk] capital. Jim O'Shaughnessy: Yeah. Right?
43:27
So, do you think that there will be an opening, so to speak?
43:27
You wrote an interesting piece about Ethereum, and you called it Ethereum's God, which I loved, and you can't kill it, right?
43:39
So, it's anti-fragile, but is there a way?
43:39
Is it a crack, or is it a doorway, I guess, is what I'm asking?
43:43
To incentivize people over to kind of this world that I certainly support, and you've thought a ton about.
43:50
Johnny Nelson: So, I would say it's like Amos's character in the Expanse, I think there is, it will be a crack, but more specifically, it will be weeds growing in a crack that you don't pay attention to at first.
44:01
And then all of a sudden, there's going to be this robust thing.
44:05
And Andreas Antonopoulos, the big Bitcoin guy, I'm mispronouncing his name, but it's going to be something that you don't notice that goes through a lot of rounds of staying a lot.
44:15
And the snowball effect, where it grabs more people, and all of a sudden, someone, say Twitter or Facebook, either has to buy it, but they won't be able to kill it.
44:21
It has to be something like that.
44:27
And the problem is you can't predict that. Jim O'Shaughnessy: No.
44:32
Johnny Nelson: I mean, try predicting Bitcoin in the year, I don't know, 2000.
44:34
Jim O'Shaughnessy: Impossible. Johnny Nelson: Yeah.
44:36
You might quote some Neal Stephenson stuff, but you would never anticipate this.
44:41
It's the creative effort of so many people going into it, so I think that's why all of the things are community-based that are people trying to explore things in a lot of different ways, and they find solutions, and it becomes part of this no Megatron Transformer process.
44:56
Jim O'Shaughnessy: I agree.
44:56
And that is really interesting to me, because I was joking, somebody was talking to me about, they wanted to do a creative treatment or something.
45:04
And I'm like, "Listen, you want the hack here? I'll give you the hack.
45:08
Go back and read history." And I gave two examples.
45:08
The first is read the actual accounts of people who were present at UFO sightings in the US in the 1940s and 1950s.
45:21
Guess who always showed up?
45:21
Johnny Nelson: The military? Jim O'Shaughnessy: No. The men in black.
45:27
That is how they were described by the people in the '40s, giving a contemporaneous account.
45:33
Even better, Dan Brown of DaVinci Code fame. Guess what?
45:40
It comes from him reading history.
45:40
There was a long dispute about this French family that said that actually they were the only direct bloodline to Jesus Christ himself, because he was married, had children. And our line is it.
46:01
Well, hello, DaVinci Code.
46:01
So, there's so much stuff in history.
46:01
And maybe, I don't know, maybe you could build a really cool algorithm that looks for very disparate sources, if you get my meaning. So, like men in black? You see what I'm saying?
46:15
And we could machine learn our way into innovation.
46:21
Johnny Nelson: I mean, I don't know if I should be saying this, but I'm going to say it anyway.
46:24
I downloaded, there's a very, very popular pirated source of journals and books that everyone goes to that gets in trouble a lot.
46:35
Before I was working on my dissertation, I downloaded the entire thing, which was very large.
46:40
And my point was that I would write a paragraph in my dissertation, would be a paragraph.
46:47
And then, I would have natural language processing algorithms search this whole thing, this whole corpus, and find the things that sort of similar in this embedded space to bring up.
46:56
And it was funny how often, I mean, funny now, painfully humbling in another way, how often my original insights were said 5,000 times by people going back a really long time ago.
47:10
So, I think history is great, because there's a saying, reality is stranger than fiction.
47:17
That's true, because there's a lot of people in reality doing stuff.
47:20
And for a long time, there's a lot of biographies you can sample over.
47:24
So, anything that can surface that better is great.
47:24
Unfortunately, copyright law seems to prevent most of it.
47:30
But the thing I made for myself, I can't give to anyone.
47:34
I'd probably get in trouble if they just looked at my hard drive.
47:37
Jim O'Shaughnessy: But man, I've got to tell you, and maybe we'll talk about this in a different conversation.
47:41
There is a pile of Bitcoin or gold there, because if you can mine successfully through machine learning the past, by the way, all of that stuff is out of copyright, for the most part.
47:57
Johnny Nelson: Yeah, that's true.
47:57
Jim O'Shaughnessy: Man, you want innovation central?
47:59
You've got it right there.
47:59
Because again, I often put up the gift from Fight Club, "Everything's a copy of a copy of a copy."
48:06
And rather than get annoyed about that, I think it's the recombining of ideas that lead to really true innovations, right? Johnny Nelson: Oh, yeah.
48:22
Jim O'Shaughnessy: So I mean, this little bad boy is rewiring our brains as we speak. I'm holding up a iPhone.
48:25
But if you're a Star Trek friend, they had the communicators.
48:31
If you're a Dick Tracy fan, he had the watch communicator, right?
48:36
So, I think that, man, we'll talk more about, actually, I might even take this out, because this could be such a huge profit maker, because it's this recombination, right?
48:42
That is where a lot of this stuff lives. Johnny Nelson: Yeah.
48:52
So, I don't know if we can expand a little bit.
48:52
There's this part of it that I want to talk about. Jim O'Shaughnessy: Go.
48:56
Johnny Nelson: So, I get angry at people a lot who look at a particular moment in history, and they see a particular bad thing happened, and they take that really particular event and ascribe it to the entire system in place that survives.
49:12
And most of the time, what you're actually describing is how human beings fail in groups. Jim O'Shaughnessy: Yes.
49:18
Johnny Nelson: Or you're describing a problem of social processes that is unconditional on any particular system, and you're ascribing the fault to the system.
49:23
So, my snotty answer a lot of the time when people talk about history as a guide to now is that history itself is N of less than one.
49:38
You can't actually port what happened and say, "This is what's going to happen."
49:42
And we're having a lot of that in the United States in the past four years, where everything is either an incident of obviously just around the corner Nazi-ism, or obviously just around the corner communism, because they're talking about events from history and saying it's the exact same symptoms, so it must be the same disease.
50:00
So, I get annoyed about that.
50:00
But on the flip side, history is a large corpus of things that actually did happen.
50:06
Being able to point to real things that did happen is worth a lot more than being able to talk about abstraction, because people get stuck in abstraction, which I mean, I do too, because that's the way my brain works.
50:18
I like to think about things at an abstract layer and figure out how they all work.
50:22
But until it actually meets reality, it's a fiction. And all models are bad.
50:28
And your model, if it never touched reality, is probably really bad.
50:33
Jim O'Shaughnessy: Right. I completely agree.
50:34
And I think that's one of the reasons we're so simpatico, because I process the same way. I prefer abstractions.
50:40
I prefer, but then I do a lot of alternate histories, so I learned this.
50:45
My grandfather called this premeditating, but just think about it as a kind of antiquated Monte Carlo simulator.
50:54
Johnny Nelson: Oh, that's, I listened to one.
50:54
So, probabilistic reasoning over multiple paths. Jim O'Shaughnessy: Yes. Right.
50:57
And so, I talked to a friend who's a true genius at machine learning.
50:57
And one of the reasons you know that immediately when you talk to him is he's like, "So, if you heard anything about machine learning from a marketer, it's bullshit. Ignore it. It's not true."
51:13
Johnny Nelson: Frustrating as a consultant too, because people have very different expectations than what I can actually do.
51:19
Jim O'Shaughnessy: Exactly. Right?
51:22
So, there's a great joke about that was when you were a prospect. Now you're a customer. Google it, anyone.
51:26
It's a good joke if you want a good, funny joke, but this idea that, yes...
51:31
To put it into perspective.
51:31
idea that, yes... To put it into perspective. So, I posited a theory that maybe we're wrong, and maybe the period between 1946 and, say, 1965 was the most incredible anomaly that could happen will probably not happen again, unless the world engages in a war where the
51:59
victorious state, in this case, the United States, so obliterates the industrial or knowledge capacity of their foes, Germany and Japan, such that in the '50s, America is making something like 80% of all the cars that are sold worldwide, because they didn't bomb GM. Right? And so, Right?
52:14
And so, it just always amazes me where people are like, "Well, look at how great the US used to be."
52:26
Well, yeah, because we wiped out all of our opponents. Johnny Nelson: Yeah. It's the same lesson.
52:30
History has a lot of lessons to tell you and a lot of lessons that aren't actually there that you want to draw.
52:39
So, my dissertation was on belief systems in American politics, so I talk about American politics a lot.
52:43
And every argument about what the founders intended annoys the shit out of me.
52:50
Jim O'Shaughnessy: I bet.
52:51
Johnny Nelson: Because, I forget, I'm going to mis-remember which city, but the population of the United States at the time could fit into Chicago now, the whole thing.
52:58
Jim O'Shaughnessy: Right. That's absolutely true.
52:59
Including- Johnny Nelson: So, you might have some scaling. Jim O'Shaughnessy: ...
53:03
Including the loyalists who hightailed it to Canada.
53:05
Johnny Nelson: Yeah, exactly.
53:05
So, we might be going through some scaling problems with the way our government works.
53:07
That's not to say they didn't really know what they were talking about, and were really incisive and brilliant at the time for what they were doing, but to port three centuries forward, people and their ideas, and say, "These are a gift handed down by God," is absolutely wild to me as a premise.
53:28
And I don't know how to relate to it.
53:28
Johnny Nelson: It's very hard. I really like history.
53:30
Right now I'm reading about Genghis Khan, because it's something I don't know about, and it's really interesting, but I'm not going to then say, "This is how I'm going to conquer the step."
53:40
Maybe it was a little different conditions back then.
53:46
Jim O'Shaughnessy: Maybe, slightly, slightly.
53:47
Well, the other thing, so one of the best things that I ever saw, and I can't remember whose idea it was, but it was, okay, so it was two ideas.
53:52
The first idea was every law sunsets.
53:57
So, the law is set, and it expires in however much time you want. Right?
54:05
And then the legislature can either re-up it, or say, "Yeah, that sucked.
54:11
Glad that's ending," right? So, sun setting.
54:11
And then, there was, I think it was Rory Sutherland who advanced this idea.
54:17
Why don't we just, and I actually had it in a fiction that I wrote in the early '80s that, that Washington had walled itself off wink, wink.
54:23
And that politicians each had their own journalist who followed them and issued blippo-second reports about the importance of that particular legislator.
54:35
The key thing was that the rest of America, outside the wall, had stopped paying any attention to them years ago.
54:40
And literally, the scene opens with a meeting in the fields of Iowa, where people are deciding who gets the nuclear football, but it's done by lot.
54:52
And you see that no one wants it. It's like, "Oh, fuck."
54:59
But this idea that Rory had is what about picking legislatures by lottery? Like [crosstalk]. Johnny Nelson: Yeah.
55:07
Sortition, it's a great, I mean, I like it, because I like random algorithms, and this will get me in trouble with the political scientists, sortition, term limits, all these things.
55:17
There's a lot of evidence against them, but this is where I go back to the thing, N of one, is there any way to experiment with this?
55:20
Even in some limited sense, and if the answer to that is no, I get suspicious that the danger of studying history is that you tend to reproduce it.
55:33
If you say this is going to happen, because here's the statistical evidence, so I'm going to act exactly on this statistical evidence, you're just reproducing the statistical process.
55:41
Jim O'Shaughnessy: Exactly.
55:43
Johnny Nelson: So, I don't know what to do with that, but to me, sortition, the idea that there should be some subset of people who get to be in a position of decision-making power as part of an assembly or whatever, it's basically saying...
55:55
There's ways to manipulate that, but it's also a defense against certain types of cabals or bad social learning processes, which I think is cool. Jim O'Shaughnessy: Yeah.
56:04
And so, Rory in defending his speculation, he's great, by the way, was like, "Hey, it's the way juries have been picked in Great Britain and the United States for 100s of years.
56:14
And it seems to work pretty well, because guess what?
56:21
You take a bunch of very different people, and they've got to work on a problem until they come up with a solution." Johnny Nelson: Yes.
56:27
Jim O'Shaughnessy: And so, you get cognitive diversity built right in.
56:30
Johnny Nelson: Cognitive and life experiences.
56:31
[crosstalk] very different.
56:32
Jim O'Shaughnessy: And life experiences.
56:32
I mean, what's not to like about that? Johnny Nelson: Yeah.
56:35
Jim O'Shaughnessy: But I think it's also right now, as part of my thesis what I call the great reshuffle, government can't keep up.
56:38
And they can't keep up, because they're designed not to be able to keep up.
56:43
And the problem when that happens is when they're the guys with all the guns, and they're the guys with the ability to tax, you want them to be somewhat competent.
56:54
And what's happening as we iterate, the flywheel is just going so fast right now.
57:01
And you're seeing power law distributions.
57:01
And you're seeing all of this sort of thing that, pretty easy to think about how that is going to continue to happen.
57:12
And if government can't keep up, we've got a problem.
57:16
And then, we also have the very social problem, in my opinion, I'm going to ask yours.
57:21
Is back in the heyday of the trust-busters, right?
57:21
They went after Standard Oil and the Rockefellers, and they went after AT&T and everything. Well, guess what?
57:28
Society was like, "Yeah, go for it. We hate those guys."
57:34
Now, imagine them trying to go after Amazon, where millions of people fucking love Amazon to the point that they don't go to the supermarket anymore.
57:50
You're going to get very different outcomes.
57:50
So, I just think a little randomness, intentional, might be helpful. Might be helpful.
57:57
I mean, have you got a solution for this?
58:00
Johnny Nelson: I'll talk about it at the very end of the show, but no.
58:01
So, to your point, fundamentally, governments are conservative organizations in the sense that they're supposed to provide a sort of ballast that's a continuous ballast, so that we're not sliding all over the landscape, because that has consequences too.
58:11
I think the problem has become, and it's different sides of the same problem, the government cannot react fast enough, because it is, socially too many problems are becoming intractable through government.
58:31
Too many groups are tied together and the attention is always focused on exactly one issue, that you can't...
58:36
the idea of representative government sort of breaks down when everyone's looking at everything.
58:40
If you can't trust the two parties in government to be able to identify a problem and come up with a solution, which necessarily requires compromise.
58:50
If you can't have that, you're essentially saying you are committed to being unable to learn at that level of institution and adapt.
58:55
So if the government can't actually adapt, it's supposed to be a conservative organization that's slow.
59:01
If it can't actually adapt and approach challenges, it's again, reproducing them and everything becomes this sort of locked into a state of paralysis.
59:10
And I don't know how to solve that.
59:10
You know, again, like the idea of injecting randomness to me is saying maybe that'll break up some of the tendrils that are binding groups apart and in this locked adversarial battle. But I don't know.
59:27
It's the same reason why people say you can't have term limits because there's a lot of institutional knowledge that gets lost in them.
59:31
For me, it's like maybe some of that needs to be pruned.
59:36
Jim O'Shaughnessy: That's a good thing.
59:39
Johnny Nelson: Yeah, shoot.
59:39
I forgot what I just went off on a tangent.
59:42
What was the beginning part of this?
59:43
Jim O'Shaughnessy: The beginning part of it was that I feel that as part of this great reshuffle, there is going to be some sort of exogenous events, and other people have already decided that that exogenous event was the worldwide government catastrophic failure to COVID. Johnny Nelson: Yeah. Yeah.
1:00:02
Jim O'Shaughnessy: I'm not so sure. I don't know about that.
1:00:03
But there's going to be something where...
1:00:09
and it's probably going to be a worldwide thing.
1:00:09
Like, and the only reason that I pay any attention to the people saying that COVID was...
1:00:14
and it was the first time, at least in my lifetime, that everyone on the planet was facing the same challenge. Johnny Nelson: Yeah.
1:00:22
Jim O'Shaughnessy: But where it becomes painfully obvious that the government simply isn't up to the task.
1:00:32
And I don't mean our government in the US.
1:00:32
I'm talking all governments- good, bad, indifferent.
1:00:39
And I just think that there's going to, at some point in the next 20 years, there's going to be some very, very bright people with very disparate backgrounds that need to discuss this and get social buy-in.
1:00:48
I have, by the way, I have no idea how that's going to happen. Johnny Nelson: Yeah.
1:00:53
Actually no, I do have an answer to this.
1:00:54
No, you know what, I do have an answer to this. Well, not an answer.
1:00:54
I do like a course heading, and it's part of why I'm so optimistic about social mediums.
1:00:58
I think what's actually happening right now with the internet is we are kind of growing this new nervous system for the social organization of the world or the human super organism or something like that.
1:01:14
I think we're still in the early phases of growth.
1:01:14
So predicting what happens when...
1:01:20
it's like predicting infant neural development, like there are stages.
1:01:20
We just don't know what the stages are yet.
1:01:24
And I think one of those later stages, there's going to be a new type of a way of communicating.
1:01:28
And there's going to be a lot of pruning going on and who communicates.
1:01:33
Somewhat down the line, we're going to recognize that a lot of the institutions that serve us now no longer fit, and it's going to be this punctuated dislocation.
1:01:46
And I think it's exactly what you said.
1:01:46
It's going to have to be a worldwide moment where with the ability to perceive things that new mediums allow, it's going to become obvious to the degree that we're not going to be able to undo it.
1:02:00
And I don't think COVID was that because in the historical perspective, a one-year event...
1:02:05
and it's not a one year event, we're still going to be dealing with this for a while, so maybe they're right.
1:02:08
But it just wasn't enough of a durable dislocation.
1:02:08
And now everyone I know, it's a little bit frustrating to me has just said, "Yep, it's over", at least in the United States.
1:02:21
"It's over, I don't care anymore, it doesn't matter.
1:02:21
Let's have this summer and let's create the new roaring twenties", or something like that.
1:02:29
I think more likely it's going to be a massive economic event. Not in the COVID sense.
1:02:29
I think there's going to be some cascading economic failure combined with a new ability to look at it socially where people are like, "Okay, wait, we don't have to do it like this anymore just because we always had". Jim O'Shaughnessy: Yeah. Yeah.
1:02:48
I love your metaphor because I use evolution as a metaphor in a lot of things I'm trying to explain to people and your idea that we are evolving a worldwide nervous system...
1:03:01
I actually have that written down somewhere.
1:03:01
So I obviously am very simpatico with you on that.
1:03:05
And, so again, to bring Deutsch back into it, the reason I think his book it's so important that people need to read it is he does a really great job of explaining, A) why pessimism is so seductive because it's always an immediate threat that's going to wipe us out, right. That's their pitch.
1:03:24
But if you kind of look back, it's always been their pitch.
1:03:28
"You just wait", you know?
1:03:28
But he also is very good at explaining we need to understand that we can't prophesize.
1:03:33
No matter how smart we are, we can't prophesize what this new learning is going to look like.
1:03:44
We can't prophesize what it will cover.
1:03:44
And like your idea of this emergent new, global nervous system is, I think, a good metaphor here.
1:03:52
I think within 20 years, there's going to be things that you and I right now can't even conceive of...
1:04:00
Johnny Nelson: Yeah, absolutely. Jim O'Shaughnessy: ...
1:04:02
that literally will be very commonplace and that we will both just be like, "Oh yeah, of course".
1:04:06
Because of the way that, the way the brain works is we'll change our memory.
1:04:11
That's one of the reasons why I keep journals because, listen man, our memories are unreliable narratives.
1:04:16
Johnny Nelson: You're wrong way more than you remember.
1:04:17
Jim O'Shaughnessy: Oh Yeah.
1:04:20
I mean, and that's the base case.
1:04:20
The base case is you're wrong, but that's okay.
1:04:25
As long as you say, "I'm probably wrong.
1:04:25
So I'm going to ask a lot of questions and people are going to get annoyed by me, but I'm going to try to figure this out". Okay. So final question.
1:04:35
Do you think that this ultimately, and this is such a loaded question, it's really not fair of me to ask, but I'm going to load it and launch it at you anyway. Do you think that...
1:04:46
and you provided me the perfect metaphor here, the global nervous system.
1:04:54
Do you think that what we're growing right now will ultimately, fundamentally change basic human nature?
1:05:04
And by this, I mean, I find humanOS to be a really good way to look at it because people get it. I'm not attacking them.
1:05:10
I'm saying we all, we all have humanOS, but is that code going to get overwritten? Johnny Nelson: Okay.
1:05:18
So I guess my answer would be almost if it's a fundamental change in override over the OS, I hope not because I wrote this before because I think human beings and all our weird buggy quirks are remarkable for overcoming any problem.
1:05:31
And that means a lot of our robustness is that we make a lot of stupid mistakes and find new landscapes to explore.
1:05:38
Chaosprime has a great quote on this.
1:05:45
I'm actually not going to say it because I'll butcher it.
1:05:45
I'll maybe I'll post it afterwards.
1:05:49
But I think like our humanOS is very error prone.
1:05:49
And as a benefit of that, it's almost like the same thing.
1:05:56
It means we can inject a lot of noise into our processes, a lot of randomness and we make a lot of leaps and a lot of those leaps are completely detached from any reasonable expectation.
1:06:13
And sometimes they pay off.
1:06:13
And a lot of other times they're there they're premised on a hunch that has the tiniest kernel of associated possibility, and someone or some group of people was willing to take that leap.
1:06:26
And I think that's great.
1:06:26
And they're never rational.
1:06:31
Any rational person would not do that and that's great.
1:06:31
And I think that's part of what's built in.
1:06:34
But I do hope whatever this thing we're growing or building towards...
1:06:34
and I personally think we should be engineering a little bit, but I'm also a little bit afraid of the technocratic approach.
1:06:44
So that's why I like variability.
1:06:44
I want a lot of people to be focusing on this.
1:06:48
I do think that this thing we are growing is going to fundamentally change the way we relate to each other.
1:06:55
And that's a really easy thing to say because it already has, but it's also exactly like you said.
1:06:59
Like 10 to 20 years from now, what I expect is wrong, almost...
1:07:04
I'm willing to say I'm almost definitely wrong because I can't predict it.
1:07:09
And I think anyone who believes they can predict 10 to 20 years in the future is full of shit.
1:07:12
But I do believe, and this is the strongest thing I'm willing to say is I still think we're really, really early in the developmental process.
1:07:20
So it is going to change us in fundamental ways.
1:07:20
I don't have kids yet, but when I do have kids, they're coming of age and adult experience is going to be far removed from anything I possibly could have thought.
1:07:32
And the way they relate to the world is going to be different.
1:07:36
And the optimist in me says, I hope it's going to be better.
1:07:36
And I trust that it can be and will be. Jim O'Shaughnessy: Yeah.
1:07:44
So we, on that, are incredibly simpatico. I am...
1:07:44
what a gift to be alive at this point in history, the way I look at it.
1:07:50
I mean, I am so incredibly optimistic.
1:07:50
And I think that if optimism sold, you'd see a lot of it, right? Humanprogress.
1:07:57
org is a good place to go.
1:08:03
Billions of people who were dying just two decades ago are no longer dying because of the progress that we've made.
1:08:10
And we're just at the start, I agree with you.
1:08:10
I think we're just at the start.
1:08:14
Will there be a ton of mistakes? Absolutely. That's the way it...
1:08:19
Johnny Nelson: Absolutely terrifying ones.
1:08:19
I'm sure we're going to hit too.
1:08:21
Jim O'Shaughnessy: Absolutely.
1:08:23
You got to get ready for that. That's absolutely true.
1:08:23
But I just think, obviously, it is an amazing time.
1:08:28
I think that the more that we can get people who, like yourself, are really, really gifted at certain aspects of how we could help here...
1:08:42
let's make sure that you're not it on Twitter.
1:08:42
That's Alex, my buddy, Alex Danco. He's like, "No, no.
1:08:49
Your one thing is just, don't be it on Twitter". Johnny Nelson: Yeah. I actually...
1:08:52
I've listened to all your podcasts with him. I really like him.
1:08:55
Jim O'Shaughnessy: He's a really great guy. All right.
1:08:57
So we've got a question that we ask everyone at the end and what we're going to do, Johnny, is we're going to make you emperor of the world for one day. You can't kill anybody.
1:09:09
Can't send anybody to camps to reeducate them.
1:09:09
Can't do any of that stuff, but you can incept two ideas that after they wake up in the morning after your day as emperor, the inception they think is their own idea and they just begin to adopt it. What two things...
1:09:30
I'm going to give you two things.
1:09:30
What two things you got for me? Johnny Nelson: Okay.
1:09:35
Since I'm bad at following directions, I'm going to twist it in the say inception is going to go to grok.
1:09:40
So you're going to grok two things. I think that counts.
1:09:45
Jim O'Shaughnessy: It's does.
1:09:45
Johnny Nelson: It's the same spirit.
1:09:48
Jim O'Shaughnessy: It is, absolutely.
1:09:48
Johnny Nelson: Number one: ensemble models are good.
1:09:50
And just to explain, I'm not sure how...
1:09:50
you have a quantitative audience, but what's incredible in machine learning and data science is that there's this instinct that you can find the best model that's always going to perform the best.
1:10:05
And it's like the strongest or the most powerful model.
1:10:05
And what happens a painful amount of time and an almost humbling amount of time is that the best model is a collection of models that have different capabilities.
1:10:18
They kind of focus on different parts of the problem, solve different parts of the problem.
1:10:24
And whether it's an average or something more sophisticated of what they say, that kind of group of...
1:10:29
that kind of collection of different constraints and different evidence and different experiences outperforms any one of them in isolation.
1:10:38
And it's an incredible thing because most of the time when you're doing data science or machine learning work, if you skip that step of trying to do an ensembel model, you're bad at what you do. To the general...
1:10:47
the person taking this generally is that look for people who have different life experiences from you because you're missing part of the puzzle that they focus on.
1:10:55
Look for people who disagree with you for the same reason.
1:11:01
they're operating under different constraints that you are and these different constraints together, there's this weird synergistic effect.
1:11:05
And that's hard because a lot of the people who have different experiences from you and different ways of looking at a problem, if you disagree with them, it's very easy to hate them in some way, or to almost bi-weekly distrust them.
1:11:22
And a lot of times you should.
1:11:22
There are a lot of bad people, but you really have to recognize that if you're only getting information for people who have similar experiences from you and different, similar constraints, you can't learn anything.
1:11:32
You're not getting new information.
1:11:35
You're just getting the confirmation.
1:11:35
So ensembel models are good.
1:11:40
Jim O'Shaughnessy: Yeah, they are.
1:11:40
And that has been our approach on the quant side, building algorithms from the get-go.
1:11:43
That you have to, and if you look at the wisdom of crowds, not the madness of crowds, but the wisdom of crowds...
1:11:50
crowds work well when you have heterogeneous expectations and you have a lot of cognitive diversity, then they're really great.
1:12:02
And I simply, when I'm talking to people about it, I simply say, "No, no, no, no, just don't don't think about it as people.
1:12:07
Think about it as models". Same thing with models. Same thing with models.
1:12:11
And so, the world is going to run a lot better with both of those.
1:12:15
If everyone on the planet crack those two.
1:12:18
Johnny Nelson: My second, my second, that was only... Jim O'Shaughnessy: Oh.
1:12:20
Johnny Nelson: My second is the one I'm very happy to give to you because I feel like you're one of the few people who appreciate it. Hail Eris. All hail Discordia.
1:12:26
So for the viewers who are not Discordians and Jim isn't because he's many things...
1:12:33
I will call myself a Discordian.
1:12:39
It's kind of a mock religion, but it's not.
1:12:39
It's kind of a way of looking at the world, but it's not. It's like...
1:12:43
it's hard to pin down, but it's for people like me and I suspect people like Jim.
1:12:47
I have an almost compulsive need to characterize uncertainty.
1:12:47
And the thing I liked about Discordian so much is it was the first kind of set of ideas and beliefs that said like, and almost recognized, a lot of uncertainty is irreducible and you can try forever and you're not going to get anywhere.
1:13:05
And not only is that not bad, but the inherent randomness is what life is.
1:13:11
Sometimes you need to be bumped along paths. If everything was...
1:13:16
if you can get probability of one for something, it's degenerate and it's not life and you're going to get stuck.
1:13:22
So Discordians embrace the randomness, and I think everyone should.
1:13:27
Jim O'Shaughnessy: I love it.
1:13:28
I have Discordian sympathies, as you know.
1:13:28
What I love about that religion in quotes is that it started as a prank, and it was done to demonstrate how silly other religions were.
1:13:41
And then all of a sudden it became a real thing. Johnny Nelson: Yeah.
1:13:45
There are on Twitter who I've become mutuals with almost because of this weird inside joke. Jim O'Shaughnessy: Sure. I love it. I love it. I love it. I love it. Well, this is great.
1:13:55
I will have you back on because I didn't even get to some of the other things that I want to start you about.
1:13:59
This has been a true pleasure.
1:13:59
Give us your Twitter handle and your blog address please.
1:14:04
We'll have them in the text too, but I want people to hear it. Johnny Nelson: Yeah.
1:14:11
Find me @Generativist on Twitter.
1:14:11
And my current blog is generativist. falsifiable. com.
1:14:17
The latter hopefully I'll start updating a little more. I've been lazy.
1:14:17
But on Twitter, if you follow me, just brace yourself for a lot of inane things.
1:14:24
I'd say one in every 500 times I'll say something clever.