Telling the tech story
I'm joined this week by Erik Torenberg, who runs Turpentine, the network which produces Complex System. The tech industry has a somewhat contentious relationship with the journalists who cover it; we discuss some of how that came to pass, including the history of it and some institutional factors which are perhaps less frequently mentioned publicly.
We also chatted about why some public intellectuals end up hanging out their own shingles.
[Patrick notes: As always, the below transcript includes my commentary, set aside in this fashion.]
Sponsors
Building an enterprise-ready SaaS app? WorkOS has got you covered with easy-to-integrate APIs for SAML, SCIM, and more. Start now at https://bit.ly/WorkOS-Turpentine-Network
Check is the leading payroll infrastructure provider and pioneer of embedded payroll. Check makes it easy for any SaaS platform to build a payroll business, and already powers 60+ popular platforms. Head to checkhq.com/complex and tell them patio11 sent you.
Timestamps
(00:00) Intro
(00:27) Fiction and Finance: The power of narrative
(01:41) The Social Network's impact on career choices
(03:34) Cultural perceptions and entrepreneurship
(06:04) Media influence and tech industry perception
(11:01) The role of tech journalism
(14:15) Social media's impact on journalism
(19:39) Sponsors: WorkOS | Check
(21:54) The intersection of media and tech
(39:22) Public intellectualism in tech
(57:40) Wrap
Transcript
(00:00) Intro
Erik Torenberg:
Patrick, it's great to be back and turn the tables again and interview you again. People loved the first time.
Patrick McKenzie:
Thanks very much for having me or thanks very much me for having you. I don't know exactly which.
(00:27) Fiction and Finance: The power of narrative
Erik:
Yeah, exactly. Well, let's start by getting into the piece that you most recently published, Fiction and Finance, and maybe zoom out. Is your mental model of fiction, the quote, all models are wrong, some are useful? How do you think about it?
Patrick:
I love that quote.
[Patrick notes: This is part of the extended lore of the tech industry, which (as I have noted before) is far more an oral culture than anyone would guess. You’ll pick up bits like this over the years through repetition, particularly with talking with your resident loremasters, and perhaps eventually graduate into being one yourself.
And therefore it falls to me, young ones, to tell you that this observation was from the great sage statistician George Box. We remain intensely attached to it in the industry because it bookends two fundamental truths of engineering: that you can actually understand the world we live in (and need to, in most cases, to successfully impact it in a planned fashion) and that no understanding of anything is ever complete, comprehensive, or entirely accurate. The physicists say differently, which is wonderful for them. Much like economists, they are accountable for no physical results in the world.
Oral cultures can include winking humor with a point underneath it.]
Patrick:
I definitely don't think that all fiction embeds a useful model, and specifically not in the way that I intended that essay. But there exists fiction, there exist games, there exist other things that strike us as trivial or “unreal” that are in fact good pedagogical devices. They're also good advocacy devices, sometimes intentionally and sometimes arriving at that outcome by accident. And the accidents happen.
Michael Lewis at least professes that Liar's Poker was not a sales pitch to get people to join the financial industry as traders. But it nonetheless was a very effective sales pitch over the years. People have told him, have told me, and have told all and sundry that being high testosterone swaggering bros with a high risk tolerance sounds like a wonderful life to lead, particularly when it leads to debauchery, and thus caused them to join the financial industry.
(01:41) The Social Network's impact on career choices
Patrick: That definitely isn't the only narrative that causes people to join the financial industry! But it crops up. The parallel I draw on the piece is to the film The Social Network. Byrne has an excellent essay on how the Social Network might be descriptively the most important movie of all time simply because of how much human capital it helped to reallocate from places like the financial industry to places like the software startup founding track.
[Patrick notes: I sort of wince on reading that, because I take many people’s argument that the proto-class distinction between founders and people who merely work in tech startups is disadvantageous for the industry and all concerned. That said, you do have to choose at some point to step over the founder threshold. The Social Network coaxed many people over it by giving them permission, as discussed below, both with regards to other people in their lives but also with regards to themselves. Many people who would have defaulted to e.g. becoming a product manager or engineer at AppAmaGooFaceSoft, or the many other places in the economy that soak up the time and attention of our broad tribe, started companies as a direct consequence.]
Patrick:
There objectively verifiable indicia [Patrick notes: a thing one notices about oneself over decades is that one has particular phrases that one likes far beyond all reason, and this is one of mine, it just comes out] that the Social Network happened and then there was a step change in YC applications that hasn't gone down after that. Another one, thatthat is more difficult to find publicl,y is founders will straight up tell you, “Yeah, I thought I was going to go get a job at Google or whatever and then I watched the Social Network and that sounded much more fun than configuring XML files for the next 10 years so I decided to do that instead and now I'm a billionaire.”
(One can imagine how many chats and who the speakers of those are but the price of learning some stories is you don't get to repeat them.)
[Patrick notes: As a metacommentary on how many layers of filters there are for certain observations about the world to bubble into the public understanding, note that this didn’t merely have to get through from a relatively small of potential speakers to someone who is notably indiscrete. No, this is really important to understand: this had to be both anodyne enough and common enough to not be personally identifying for me to be indiscrete about it in a way which doesn’t violate internal and external discretion norms.
Which is a long way of saying “I am skeptical of conspiracy theories but also really skeptical of people who don’t think secrets can be kept in the real world. Are you kidding? How much of the economy is based on successfully keeping secrets indefinitely. For that matter, how much active energy do governments and companies have to do to discover secrets they have invested no effort into keeping from themselves. The default number of minds for every novel human thought to ever encounter is exactly one, and getting that to two is already a monumental accomplishment. And so, yeah, there exist many consequential things known by hundreds of people that will never be known by the typical reader of the New York Times, including some that the NYT would aspirationally prefer to report.]
The broader point there is that the Social Network is not a true story. Facebook is a true story. The Social Network is a heavily fictionalized account that borrows some details from history. (I can only imagine like how frustrating it must be for the main dramatis personae, like Mark Zuckerberg. Some parts of it are well executed elements of pastiche or impersonation or dramatic convenience. And some are entirely made up. And when you are the screenwriter for you own life, as all of us must be, you’d probably spare a few words for e.g. your wife [Patrick notes: read this as deadpan humor, please], but the screenwriters for the Social Network did not consider her existence material to their narrative and so erased her from it.
There are many issues with the film, made more serious by how many people understand it to be a historical text. There are characters that come out of left field. The procedural history doesn't track the actual procedural history. We know that because history is not impossible to know, because some people were there, and because we have access to a written historical record like e.g. lawsuits and similar.
The narrative is sexed up massively. [Patrick notes: Literally and figuratively.]
But these points aside, many people watched the Social Network and said “You know, I want to live that life and have that adventure.”
(03:34) Cultural perceptions and entrepreneurship
For a bit of personal history here, and a bit of relevant context first, my wife Ruriko is Japanese. She, like many women, has a friend group of long standing where each is quite important to the others. I call them the “circle of girlfriends.” The circle of girlfriends grew up in central Japan and are, for various cultural reasons, not maximally positively disposed to one of the members of the circle dating an entrepreneur, because they think that is one step above homeless vagabond. [Patrick notes: The hypothesis “probably a drug runner” may also have been advanced. Oh he’s flying to Europe to do “consulting” for an unnamed software company—a likely story.]
And the circle of girlfriends attempted to stage an intervention on the topic of marrying me. Not simply because I was a foreigner, though that was already a check in the wrong column.
But because I was running a little software business over the internet and shouldn't Ruriko be able to land a salaryman somewhere? (I was an ex-salaryman at that point and pretty firmly “ex-”, although I kept options open when attempting to convince her mother.)
But then the Social Network came out.
This changed the circle’s consensus. He might not just be running a software business. We thought he was running a software business, in some inchoate way of understanding that claim. But he might be running one of those Mark Zuckerberg-y kind of software businesses.That's the kind that gets a movie made about you. And that would be cool if one of our friends' husbands had a movie made of him. So we approve.
This sounds like a trivial thing. It is not a trivial thing. It is extremely important to me that the people who matter most to my wife are happy with her major life decisions. When I talk to entrepreneurs and proto-entrepreneurs and people who are considering quitting their jobs in Japan, this comes up all the time. They are concerned about what their parents will think. They are concerned what significant others will think. And so they feel constrained in their life choices by what others want for them.
[Patrick notes: This is not a uniquely Japanese experience, of course, but I’m narrating the version of it that I’ve heard poured out over the most drinks, having a relatively weird life story which results in primarily writing in English but having a perhaps majority of all my conversations for 20 years in Japanese. And, before anyone needs to point it out, yes, Japanese men were a lot more likely to talk to me about their dating woes than Japanese women were, to the great relief of all concerned.]
Patrick:
I’m meandering a bit off topic, but this is enough of a factor here that Japanese startups, and other employees trying to hire from this population, have in some cases instituted a business process to manage this. At some of the savviest, it resembles a periodic orientation not for new employees, not for prospective candidates, but for the wifes and girlfriends of prospective candidates. That is largely designed to sell the most important person in the conversation on the opportunity, where the default would never have them meet a company representative.
[Patrick notes: There exists a cultural affectation in some parts of the United States to counsel people to rewrite the previous sentence to avoid assuming gender roles or heterosexual relationships. Note that I am describing reality, which has been known to include gender relations in it from time to time. If there is a company in Tokyo that runs a Sell Your Boyfriend workshop bully for them but I’ve yet to make their acquaintance.]
Patrick:
And so these workshops function much like the movie does: they are depictions of reality, constructions of reality, which show the reality of working in our industry through some distorted fun house mirror.
And that activity is instrumentally useful. The tech industry needs to successfully convince the right people to work in it. Spinning stories, ideally stories with lots of truth in them, is one way to accomplish this.
(06:04) Media influence and tech industry perception
You and I have noticed, as have many, that over the past few years, the construction of reality in the tech industry, how our story is told by the people who are in it and people who are observers of it, is… suboptimal. It sometimes sounds like we are looking through a glass darkly at the actual reality of it, in ways which cause me to lose even more hair than I already have.
And so let’s talk about tech-flavored media, which has come back before on this podcast in conversations with Kelsey Piper among others. And it has informed the rise of an emerging category of what might be called Internet-native media with a specifically tech or tech-adjacent focus.
Erik:
Yes. And one thing I just also want to point out even about the example you picked is that it's very illustrative of something broader, is those movies, you know, Social Network, but even I think the Michael Lewis movie about Wall Street were meant as cautionary tales. What was that again? *crosstalk*
Patrick:
The Big Short? [Patrick notes: He probably meant Liar’s Poker, given context.]
Erik:
Zuckerberg famously didn't like a lot of the characterizations, because they depicted him as a bad boy or a sexually frustrated social outcast. And in fact, he had found his college sweetheart early on. That fact didn’t match the narrative they were trying to tell. And so he didn’t love the depiction or film.
But this has a reverse, electrifying effect for some viewers of the movie. And so Zuckerberg mentioned, in a profile in the New Yorker, that he was at first frustrated but eventually organized a team outing to see it. It’s funny how the death of the author manifests in these artifacts: you can write with intent to create a hit piece, and instead create a glorification, a paean to, the intended antagonist.
Patrick:
I think it is almost literally impossible to make a movie in the way that we make movies and not have the protagonist of that movie come off as someone who a large portion of the audience wants to emulate. Cultural commentators sometimes complain about media literacy. They say that someone who watches Braeking Bad and ends up wanting to be like Heisenberg has missed a major point of the work. I don’t think this is entirely a media literacy problem, or if it is, the problem exists within the media itself.
[Patrick notes: Breaking Bad’s moral judgment of Walter White is extremely clearly communicated by the text! But most people who end up with sympathy for the devil, and some level of wanting to be the anti-hero, understood that message from the text. They just… on some level, in some sort of liminal space where one is allowed to play games without consequences, appreciate the power fantasy of being a hypercompetent man who rises to wealth, influence, and power while driving his enemies before him by being smarter and more ruthless than them. And so I think the media literacy critique probably needs to wrestle with some facets (which are certainly not covered here for the first time in cultural criticism), like escapist fiction being a space which does not necessarily 1:1 predict people’s behavior in the real world or their moral decisionmaking.
Where were we? Oh yeah, returning you to the construction of reality in the tech industry via depictions of it in the popular and less-popular media.]
Patrick:
We have this vast industrial complex which is very literally designed to make people look attractive to the audience. [Patrick notes: The most vain, most instrumentally effective, and most attractive person you know cannot possibly have devoted more hours in their life to makeup, wardrobe, and flattering lighting than the cast of a major production has devoted on their behalf, particularly normalized per minute of “screen time.” And this is all done very intentionally. The professionals who do this took classes on theory and then had multi-year apprenticeships with the goal of making the characters depicted more attractive to the audience than any human, inclusive of the people who are reading the lines, can be in real life!]
Huge amounts of human effort are dumped into that in every shot, in every frame, in every interview about the production, and in the media commentary around it. And then we are shocked, shocked that people might relate to the characters in ways which are unintended. [Patrick notes: Who are we kidding ourselves around which intent matters, the one which we justify to ourselves as we write the scripts or do criticism, or the one which we have built up an industrial complex to stamp out at scale?]
This is not just confined to industries where there is controversy in real life on whether the depicted deserve the audience’s sympathy. War Dogs, for example, sensationalizes the prospect of being low- to medium-competence gun runners. And if you come away from that movie thinking, I absolutely would not want to be a low-to-medium competence arms merchant, maybe you're in 5 to 10 % of the audience. [Patrick notes: Again, we (the societal we) talk out of both sides on our mouth here. War Dogs is very against war profiteering, a message which it has Ana de Armas read to the camera while wearing a bikini. No part of that sentence is an accident.]
Erik:
Yeah, it's very funny. There's this meme on Twitter that is going around that I think speaks to this phenomenon of like, the impact of the reader is opposite of what the writer intended. And the meme goes, “I already support him; stop trying to sell me.” (This came up in the context of an allegation that Elon Musk was secretly backing the Trump campaign to be the de facto president, using him as a puppet post-election.)
Patrick:
Without taking any sides here: I think that any serious charge that there is a conspiracy in action, which is moving multiple levers in concert to achieve a preferred policy outcome, always makes that conspiracy sound better than the real world. Wait, you can simply vote and then we’d get a competent government which had policy preferences and took reasonable steps to achieve them then successfully did achieve them? I'd like more of that, please. Which lever do I pull for that?
Erik:
Yeah, exactly.
Patrick:
If I may generally offer some messaging advice: Do not attribute power and likelihood of success to things you do not want to see in the world.
Erik:
Yeah, that makes sense.
(11:01) The role of tech journalism
Let’s go back to a broader question: why is tech journalism not great? Was it ever great? At the very least 15 years ago, it was more positive. And I would say that maybe there's some sort of tragedy of the commons here where if you are really interested in the technology and in business, you usually want to work in it in some capacity or influence it in some capacity.
Maybe that means you want to build it. Maybe that means you want to invest in it. Maybe that means you want to research in it. For all of these, one can find more interesting and lucrative positions inside the tech industry than outside of it in media.
Patrick:
And I think you can literally see that in people's LinkedIn and Twitter updates as evaporative cooling takes effect. Let’s see, what is an example I can use which is unambiguously not critical of an individual for making a career change.
Oh! Kim Mai Cutler! She was an excellent journalist and wrote the only piece in TechCrunch (an outlet I do not have a great deal of regard for) that was ever worth reading, on the Bay Area housing crisis.
Erik:
Yeah.
Patrick:
It turns out that if you are capable of writing like 600 page intensely researched cases for investment and malinvestment, you are inefficiently allocated as a journalist. Perhaps you would enjoy a career in venture capitalism or another capital allocation job instead. [Patrick notes: Many management functions in companies will involve this sort of decisionmaking process, and depending on the company Strategy/Finance/etc teams might also be good seats.]
Those jobs will pay better. You still get invited to very nice parties.
And so she made the leap (to being a VC), and then TechCrunch lost the one person who ever defaced its website with journalism. [Patrick notes: Did I mention I am not a fan of that outlet?]
Erik:
Totally. And the amount of venture capital jobs exploded over the past couple of decades, which means that more of these writers can become venture capitalists. But then also there are some people who love to do it for the love of the game, the Ben Thompsons and the Benedict Evans of the world and the Dwarkeshes, the podcast space, et cetera. More lucrative opportunities emerged for them too. Ben Thompson has built Stratechery into an eight-figure business, and doesn’t need the institutional backing of a publication.
The biggest factor, I think, is that social media disintermediated the conversation between experts and the public.
Patrick:
So, before I go into more considered thoughts, I should give the obligatory disclaimer. I worked at Stripe for a few years, including in their Communications department, and while I no longer work there I am still an advisor. Nothing I say in my personal spaces is necessarily endorsed by them.
And, oh boy, detailed commentary on how the media works is the sort of subject that Comms departments everywhere would probably prefer their employees never make. PR departments and the media exist in a… complicated, iterated game, in which it is very unwise to bite the hand that feeds you.
Alright disclaimer out of the way, let’s hit some factors one at a time.
(14:15) Social media's impact on journalism
Patrick:
The rise of social media gave companies an interesting incentive with regards to their labor forces. They started to realize that they were no longer in control of which pieces got huge hits and impact, where their primary control historically had been what they featured on the front page.
There was this phenomenon called “going viral.” And as Facebook and then Twitter especially became much more important for the distribution, they started to notice that the things that quote unquote “went viral”, which was difficult to track in the early internet, were very attributable to a certain seed set of users on places like Twitter and similar.
And they heavily encouraged their journalists to become dangerous at the act of making their own stuff viral on Twitter.
They said essentially, not that we pay you great amounts of money to work at legacy media institutions, but a de facto job expectation here is you are going to maintain a presence on Twitter where the rest of the industry, your sources, et cetera, are, and you're going to do your darndest to maintain reach for, maintain and improve the reach of your work in the future.
And we will show you a graph and we will award plums internally based on who does well on this graph. [Patrick notes: Many journalists have said that, fairly convincingly to me, that the allocation of status and material rewards internally is more complicated in some places than simply SORT BY hits DESC. But I think it is trivially observable that the graphs and leaderboards, formal and informal, bent behavior around them, as they almost always do. To the extent that some journalists sniff and say “My shop was different”, I will simply register my polite disagreement that their island alone was sheltered from the cultural tsunami.]
Patrick: I think employers got what was coming to them on that decision. [Patrick notes: I’m a capitalist at heart, but I come from an intellectual tradition of labor organizers, and here is how I square that circle: if you successfully incentivize someone to raise their human capital by performing unpaid work for you and then that bites you in the hindquarters at your next salary negotiation, well, there is no crying in capitalism.]
80% of the writers in legacy media have no following worth speaking of, largely because they don't write anything worth reading. I think this is sort of a robust criticism, but find me anyone professionally involved with media who disagrees with it.
For many reasons, including guild reasons (fun fact: the NYT union self-identifies as a guild), there is a structural wariness to compensate people in a fashion which tracks measurable outputs. Sure, there is a stark division internally in journalism between star writers and those who merely continue to allow us to publish the publication every day. But the salary bands and similar were heavily compressed.
[Patrick notes: Considerations of fairness drive some tech compensation decisions as well. It is an awkward fact of life that some people are 10X more productive than other people. When they both have a similar title and similar number of years of seniority, acknowledging this fact by paying one 10X as much as the other and making this fact publicly known is… considered socially non-desirable in many places. And so there are various mechanisms which allow it to happen without acknowledging that it, in fact, actually happens.
A classic one in the tech industry is that you’re allowed to acquire talent via a corporate transition with much more elasticity in compensation than HR would ever approve for people who actually currently worked for you, which leads to some “boomerang” behavior as certain goal-seeking technologists leave the company as, in effect, a drawn-out salary negotiation which causes as a side effect a filing to the state of Delaware.]
Patrick: Once it became possible for individuals to have a following, which was unambiguous because you can just look up somebody's follower count, you can just see which of their last five articles hit the top and trending sections or similar. You can see the hit counts in some cases. You can verify the following. You can see that they're doing very well.
The center cannot hold where the person who is the most effective at their job is getting paid within a 20% band of a similarly tenured individual who, for whatever reason, is not hitting those same numbers. After the center cannot hold, there starts to become some very difficult negotiating terrain for the traditional media providers with regards to some of their superstars. And unfortunately, media is a very hits-driven business in most niches of it.
It is institutionally incredibly aware of this because they have the graphs. And yet institutionally has to pretend that no, the New York Times is a brand and it has consistent brand standards. One standardized unit of New York Times content is exactly equivalent in truth and institutional accuracy and stamp of approvalness as every other standardized unit of New York Times content, which is why we're able to charge substantially similar advertising rates across all of it.
(People who work in media advertising departments are like, "Well, actually, we all know that world news and style are two different fiefdoms, and they get different advertising rates. But yeah, within style, sure. All style articles are the same to us. Or we profess that to be true with regards to negotiating advertising rates.”)
When you go out to superstar journalists, Ben Thompson, I don't know if he had a career in professional journalism prior to becoming Ben Thompson, the folks who have a differentiated opinion, who write really well, who are capable of doing their own PR, managing their social media following, surfing the zeitgeist, whatever you want to call that skill set, they should obviously capture superstar returns compared to the median entrant to the industry. That is difficult for the large media players to structurally offer them, and so the large media players mostly don't. [Patrick notes: There have been times where the industry flirted with exceptions. Nate Silver, for example, individually upended the entire field of political reporting on elections, and (compressing some corporate shenanigans here) was actually compensated and treated as if he had done that, versus having been an above-replacement-rate election reporter.]
And thus, a lot of the superstars have chosen exit over the years, and some of the proto-superstars have not chosen entry.
I was asked a time or two if I wanted to become a tech journalist. [Patrick notes: Both in the sense of “you should consider this professionally” and “We, an institution which employs tech journalists, have from time-to-time read your work and think there is potentially a future for you here.”]
“You enjoy tech, you write, why don't you do journalism?” My career planning is kind of boring, but the high order bit on that question is I'd have to be insane. “How about I take a 90% pay cut and also, by the way, work for a publication that has fewer readers than I do?”
(That might be a bit of an exaggeration, but not much of one with regards to the marginal artifacts.)
That is how some of the internal dynamics around compensation and status have played out. There are also more public discussions, and specifically something of a Red Queen’s Race on Twitter specifically, among journalists and tech journalists specifically.
(21:54) The intersection of media and tech
Patrick: Now, all industries are very small places. The collection of people who work in the Tokyo startup ecosystem, it's not literally the case that they can all fit around a dinner table. But they've all had dinner with each other and they all intend to have dinner with each other in the future. This is a community, rumors carry, reputations matter, your status within the community matters an incredible amount to you as a participant, et cetera. Journalism is no different.
The community that is journalism largely descriptively exists in a few neighborhoods in New York. And there is the normal amount of social pressure to vote with your compatriots, both in the literal and figurative sense of vote—to be interested in the things that they are interested in, default believe hypothesis that they all default believe, default be skeptical of hypothesis which they are all default to skepticism on.
There is a common phrase in Japan, often unjustly described as uniquely Japanese: "[Don’t be the nail that sticks up, because] the nail that sticks up gets hammered down." And that I think is a constant of almost all human cultures anywhere.
That culture plus the existence of Twitter and the existence of the Twitter/Slack bridge within journalistic institutions caused a hyperdrive into groupthink over the last couple of years. There were external influences here, which we can get into in a minute.
Many people spend too much time on Twitter. I certainly confess this about myself. Journalists felt professionally obligated to maintain an unhealthy relationship with the service. You outsource a tremendous portion of your thinking to the glass rectangle in the palm of your hand, in particular to the first couple of replies that hit you when you post something to the glass rectangle in the palm of your hand. And if those replies are skeptical of the thing that you've just spent the last two weeks of your professional life on, you feel terrible.
Would you, as a considered decision, really want to outsource how you feel next Monday to the opinions of Shitposter9000? Probably not. But you do anyhow, because that is the nature of these products. They are designed to achieve that outcome. (“Designed” here is pulling some weight, and might be contested by professionals who design social networks for a living. Consider me saying this in the purpose-of-a-system-is-what-it-does way.)
If Twitter did not succeed in greatly emotionally moving people, Twitter would not be Twitter. [Patrick notes: It would not hold the place in the hearts of users that it does, it would not have achieved the impact in the world it has achieved, it would not be the sub-rosa communication bus for the U.S. government, it would not feature the Secretary of Transportation and the richest man in the world hammering out infelicities in flightplans. All of these are downstream of having a powerful hold on the hand that thinks it is holding the device, and on the brain attached to it.]
Patrick:
And the first couple responders to someone who posts a journalistic article are often going to be journalism people on the same beat, partly because they all want to hitch their wagons to each other for professional reasons, partly because it's how you show your bosses, "Yeah, I'm still engaged in The Conversation," partly because in some cases you genuinely want to support your friend and colleagues' work.
And partly because of the brute reason of, if you're writing trade press, (the tech industry used to be covered as trade press and then became covered as a culture war issue, but that's neither here nor there), then the only people interested in trade press either produce it or consume it. That is a very select group.
You get positive feedback immediately if you say the thing that the marginal reporter would say about your situation, and you get visceral negative feedback immediately if you say the thing that the marginal reporter would not say, if you were perceived as carrying water for the industry. This social equilibrium caused a great deal of groupthink.
And then there's the Twitter/Slack crossover, which I think has been heavily under-discussed.
The Twitter/Slack crossover happens when there is something that is happening on the open Internet on Twitter. And coincidentally, there is something that is happening on the closed Internet, on Slack, typically within a particular large employer.
The vast majority of these incidents are like the vast majority of things that happened to the world yesterday. They are below the water line. The public will never see them, learn they happen, or understand them as having shaped the institutions which shape our world. But they happened nonetheless.
We have evidence of some Twitter/Slack crossover events, simply because the news media loves talking about itself. And thus we see a tiny bit of this societal iceberg, above the waterline.
For example, the New York Times specifically has had some crossover events which were both culture war flashpoints and irregularly, not-traditionally-conducted labor organizing campaigns. The specific format of that was essentially creating a coalition within the New York Times' newsroom that decided that it was going to post quite repetitive messages on Twitter in the support of a goal which was internally advocated for as a labor-organizing goal.
The specific flavor text for that doesn't really matter. It happened to be something around the safety of our colleagues who are African American. Don't interrogate that goal too much. It will break your brain. And it ultimately doesn’t matter: the thing that matters is a small group of employees used two parallel software systems which are not formally linked, with an assist from professional culture and U.S. labor protection laws, to attempt to root one of the most influential institutions in the world, against its own wishes.
Importantly: this is not a conspiracy theory, not in the sense that it alleges a shadowy group in smoky backrooms that probably doesn’t exist. No. This was conducted partially in the open by people under their own names; that was key to the mechanism working. And then there was a partially overlapping group of people who had, effectively, commandeered an office (or a private Slack channel; same thing) at their employer.
And they've said, as a considered strategy, we are going to take a break from reporting the news and engage in coordinated messaging via Twitter. We will then point to the coordinated messaging happening on Twitter as evidence to win an internal argument with regards to what this publication's policy should be.
That happened. Things that are similar to that have happened in many companies over the years. Stories which are not mine to tell are not mine to tell.
I would like to point to people, if you are in the position of having to design, I don't know, Large Company's social media standards, look very carefully at what happened to the New York Times that time. Consider what a small and very motivated group of employees could do if they decided to cause a combined Twitter/Slack event at your company. I don't know any company that has ever suffered a combined Twitter/Slack event and thought, "Thank goodness, we were gonna just step on a landmine managerially. But when several hundred of our employees did a coordinated labor action in response to a core organizing group that could fit around a ping pong table. They brought us into disrepute on Twitter. Thank goodness they did. That obviously maximizes for our true values."
And anyhow, this evaporative cooling of people who capitalism decided would be better used if they were in better jobs, and the sort of hothouse of the people remaining in journalism caused this sort of bifurcation in the industry, I think.
I also think that there was, you know, nobody's community is completely immune to things happening in the external world. The media and security state mutually talked each other into believing in the wake of a presidential election that the presidential election had been stolen.
Now, these days, it's much more popular to say that presidential elections never get stolen in the United States, but those of us who are older than the age of five or so might remember a time where that was a very live conversation and taken seriously by serious people. The theory for this was that Russian intelligence engaging with a British consultancy dropped six figures or so on Facebook ad spend for the election, and thus it was stolen. [Patrick notes: Cambridge Analytica is the search term, for people who want to read elaborate Facebook marketing fanfic. I will not rehash the reasons for my contempt for this theory here.]
Patrick: This was very passionately believed by a lot of people who were feeling pretty bad when their favorite candidate lost a presidential election. It happens to half the population every four years, but these folks, for whatever reason, felt like it should never have happened to them, at least not at the hands of This Guy.
And so they went looking for reasons, and they created a reason. And that reason had very little entanglement with the actual world we live in. But having discovered who their bad guy was, some combination of Russia and Facebook, and being unable to do much from their platform to make Russia's life miserable, they decided very overtly to make Facebook's life miserable for a couple of years. Then they used that reporting template to color a lot of their coverage on the tech industry, which previously had been some mix of like, "Here's the cool new thing you can do on your phone" and trade rag press.
And I say trade rag because trade rag is the term of art for industry-focused press. This is traditionally said in a sneering fashion by high-status publications, with regards to their relatively low-status brethren who are covering what is new and exciting with regards in publications like Food Sanitation Equipment Quarterly. [Patrick notes: Intended as a demonstration not an identification. I don’t work in the field or I’d presumably be more familiar with Food Safety Magazine.]
Trade presses, they need to exist in the world. All industries need some coordination mechanism which is larger than an individual firm but not as directly aligned as an industrial association is. They provide some value. And they also, bluntly, provide an obvious place for industry to create a Schelling point to do things like have trade events, conduct hiring of people, and have a messaging platform. Food sanitization device manufacturers in America who want to get their message out to the wider world, and the trade rag is a way for them to test that message against each other before they send it out to their lobbyists and try to get in the halls of power in Washington. [Patrick notes: Again, there is an actual need in the world for firms to discuss their preparation for the 2026 compliance date for the FDA’s Food Traceability Final Rule. And thus there is a podcast episode on that topic. The economy is a fractal, and anywhere you zoom into includes infinite detail within itself.]
Patrick:
Tech used to be covered like any other industry is covered. It isn’t anymore. This isn’t merely a conspiracy theory or sour grapes by someone who, admittedly, worked in a tech PR department. As reported by Matt Yglesias and then confirmed by a number of people with knowledge of the situation, the New York Times had a narrative setting from top down which said essentially “When we're reporting on the tech industry, we're going to find the maximally skeptical take on any given story, to hold the tech industry to account for their exercise of power.”
I think media's take that it necessarily holds powerful actors to account is very important to their self-conception. This might or might not match facts on the ground always. That is certainly not the way that people who are in the Washington sphere report on politics. Sometimes they report on politics as, this is self-evidently an important story. There was a hurricane happening right now, the government is responding to the hurricane. We would be derelict in our duties if we did not report what the government was saying right now.
Sometimes it's a human interest story. They become close to their sources, like people become close to anyone they spend a lot of time with. They are naturally sympathetic to the aims and worldview of these people, and have an uneasy relationship with needing to maintain their favor to preserve access.
Some portion of it is holding the industry to account, but holding the industry to account is not the only reason to observe a company acting within capitalism and report on their movements. Some of those movements are kind of important.
Like if you're reporting on automobile manufacturers and the only thing you ever report on is car accidents, you're doing a pretty shitty job of reporting on the impact of automobiles on human civilization. There is a wider context that you need to reflect. We have retreated from that wider context, intentionally, which is causing a bit of a reaction within the tech industry, partially gnashing teeth for a few years and then after it reached a certain crescendo, I think there have been more directed action than teeth gnashing.
Erik: Yeah, I spoke to one of the journalists I admire the most at The Information, one of the things they said that they care most about is holding tech companies accountable. And The Information is supposed to be the trade media publication. But who is served by keeping tech companies accountable? Not the tech companies themselves. So it's not trade media if you don't serve the people in the industry, you don't help them get better at their jobs. But to your point, there's this kind of adverse selection because people are getting 90% less compensation to be a journalist. You’re doing this because either you can't get a better job or you're ideologically motivated, i.e. you want to hold people or companies accountable.
Patrick: I'm minorly frustrated in that if you strictly think your job is holding someone accountable to some higher standard, you should probably be a regulator and you should probably have the democratic process backstopping the decisions you're making on a day-to-day basis. Otherwise, who picked you, bub? And a bit of the "who picked you, bub" is leaking into the decisions by the tech industry.
[Patrick notes: What do I mean here? I mean that, as a matter of legitimate exercise of power and how status should be allocated by those of us who don’t work for media companies, if you have a democratic mandate on the laws you are empowered to enforce, then those of us who build tech products for a living should give your point of view some respect. And if you don’t, best of luck on your blogging career, but we don’t owe you any special consideration, deference, access, or esteem.]
Patrick: Broader society may very well have a Who Picked You, Bub moment with respect to tech companies! And there’s an answer there!
Tech companies have largely accidentally found themselves in a position of power because they created, among other things, habits that allowed them to sell large amounts of advertising. [Patrick notes: The tech companies’ power and influence on the world is, in a broader sense, ceded voluntarily by everyone every day who decides that phones, social media platforms, spreadsheets, and web applications are superior ways to run their own lives and achieve their own values than alternatives. And the industry earns that power and influence every day, including from its critics directly.]
Patrick: Historically, the "Who Picked You" for the media, there's a bit of state involvement in there, but the biggest part is that they convinced the entirety of the middle class to accept a physical artifact at their door every day that was wrapped in advertising.
The trade, between the people who were the subject of news stories and the newspapers, was “We are a way for you to get your message out to the entirety of the middle class at once. In return, you keep giving us things that the middle class will find interesting. We will convert that interest into advertising revenue.” And when you say it that way, it sounds a little bit sordid. [Patrick notes: And if it doesn’t sound sordid enough, try to negotiate a PR embargo sometime.]
Patrick: It's not intrinsically more sordid than any other ad-supported industry or any other industry for that matter. The news media is a business. But as that business model broke down, it became less culturally accepted in the culture that was journalism to offer a quid pro quo to the companies that were continuing to bring them the feedstock that they would turn into advertising inventory.
And then the tech industry began to say, quietly and sometimes loudly "Wait, we had a deal. The deal isn't contractualized anywhere, but this is an iterated game with a handshake to it. And if you're not going to keep up your end of the deal… why am I talking to you again?"
This creates a lot of enthusiasm for building and owning our own platforms, both at the level of large tech companies and in this space of experts who write for the Internet. The entire middle class was habituated into accepting a printed artifact that was produced at enormous expense and logistical effort and dropped on their doorstep every day. That is not the primary habit of the entire middle class anymore. The primary habit is taking out the device in their pocket and accessing the internet, which, whoops, we didn't expect to do that, but we did.
And so tech companies have decide, "Hey, if the goal is getting to as many people as possible, through an internet platform that we control, and the media's core advantage there is simply a legacy user base which is dwindling every year and some amount of skill at what it takes to produce their printed/video/audio artifacts, let's just replicate the skill bits. We should control our own platform. Then we will renegotiate with ourselves how much of the coverage of our industry should be punching ourselves in the face."
And go figure, when the tech industry has that negotiation with itself, it does not decide to devote 100% of pixels to own-face-punching.
Erik: It's very funny. And I also want to get into going back to the tragedy of commons because most people who now work in the industry, they don't share their knowledge or they don't share their research to the extent that they could. If they were analyzing a company previously on behalf of a media outlet, now they're doing it for a venture firm. That venture firm wants them to keep their analysis internally so they can trade on it. And so as a result, there's just not a ton of amazing public analysis, although there's starting to be more because people are understanding the benefits of sharing your knowledge so that you could have demonstrated expertise in some capacity and recruit and all the sorts of benefits of having a media brand like you have.
And that's what we're trying to capitalize on at Turpentine, which is that experts who work in the industry have the knowledge, and maybe as an exhaust of them doing their jobs and creating this knowledge, they can become creators or educators or share their knowledge publicly. We're trying to make it easy for them to do so first in podcast form, but then later on hopefully in written form as well because there's a dearth of great content.
[Patrick notes: Never call it content! Taylor Swift doesn’t produce content! Leonardo da Vinci didn’t produce content! Calling what you do “content” invites the commoditization of it by your audience, by competitors, and by yourselves!
Sorry, I have strong opinions on this question. (And had “Content” on my business cards.)]
(39:22) Public intellectualism in tech
Patrick: I like to think of it as something like public intellectualism. The production function that is universities has any number of issues associated with it. But one of the relatively nice parts of that production function is as a necessary prerequisite for you climbing the ladder in universities and getting yourself on the frontier of human knowledge, you will contribute the things you learn at the frontier of human knowledge to the commons in a way which is very visible and legible to the other people in your field directly. And for some subset of university professors, the Tyler Cowens and similar of the world, they make an extra special effort to not merely keep that conversation within the walls of the ivory tower, but rather to let the rest of polite society in on it.
That's not as simple hanging out a sign saying “Interesting ideas within; all are welcome.” Writing is a lot of work. Writing well is a skill, and few people will devote 20+ years of effort to getting very, very good at it.
And so one of the things to be pretty happy about in the tech industry is that there are a number of people who for whatever combination of professional incentives, or psychological damage from when they were children [Patrick notes: I should use less self-deprecating jokes], decide to spend many, many thousands of hours writing about things and then publishing them to the internet.
Since the default way for publishing things is less gatekept and less institutionally controlled than it is in a lot of fashions in more traditional industries in the journalism industry and similar, I think you have higher variance with respect to the artifacts that actually make it to the internet, versus artifacts that go through a more controlled channel. And higher variance means more of the good stuff, and perhaps more of the bad stuff too. But due to the wonder and terror that is distribution on the internet, you are more likely to encounter the good stuff than the bad stuff.
(As long as you have not completely outsourced your media diet to whatever is trending on your social platform of choice. Which, by the way, there are some true knocks against the tech industry. And one of them is that if you, as someone designing a social product, know nothing about a user and you are optimizing for engagement, there are relatively simple ways to create Top Stories lists that are most likely to engage the user today. Many users of yours will adopt these features if you make them, because curating their own experience is difficult and only possible for power-users. And the sort of things that people will get when they consume that feature are going to be pretty terrible.)
Consider Twitter’s recommended feed (called For You in English) versus the followed-only sort-by-time-priority feed. The composition of For You has changed a little bit under new Twitter management, but it hasn't changed radically. It was a pretty awful place before, and it is a pretty awful place now, for following people that you actually want to invite into your life on a regular basis.
Anyhow, more choices is, generally speaking, good in the world. There are existing public intellectuals in the tech industry who have track records and some level of like, you know, the artifact is the artifact. You can just look at it and say, “Is that good or not? Do I want to invite more like this in my life or not?” And then very cheaply subscribe to the old internet phrases like, "I like your ideas and would like to subscribe to your newsletter." And now that it's actually a popular business model these days—props to the Substack team.
[Patrick notes: Every startup which is successful needs to motivate one major change in human behavior. One of them that Substack successfully convinced was convincing a very influential sliver of the population to move from a “gift economy” on the Internet to a system which blends patronage and paying-for-business-inputs. They didn’t invent the business model of newsletters, not by a long-shot, but they made it socially normative for many well-paid professionals to spend hundreds of dollars a month on niche publications in much the same way that it was previously normative to spend money on cable television, and simultaneously raised the status of writers who could sustainably produce in this economic model.]
Patrick:
If you were to sort by smartest people in tech industry or people who have the most interesting ideas and then intersect with who has an actively maintained say blog as of this week, those are not completely overlapping sets. Just like any other industry, the vast majority of the intellectual effort will be forever below the waterline. It's going to be in investment memos and in presentations to the board and in day-to-day work of running the world because the tech industry kind of runs a lot of the world these days. But when you look at public intellectuals in the tech industry, those leading lights who have repeatedly chosen to write for the Internet, I think our industry offers a pretty good showing.
When I look at public intellectuals across industries broadly, in regards to things that specifically implicate the tech industry, I think the industry tends to run the table on all commentary. There is not an easily accessible source of journalism or public investing commentary that is a better way to learn about the tech industry than the usual suspects. [Patrick notes: Markets in status are weird, aren’t they? To describe someone as high-status is to attack their status.]
Not that it's a contest or anything, but if you look at intellectual heavyweights in the sphere of creating public goods for humanity. If you were to come up with a list of like 100 people in the world or 100 Americans, and write down that list of like who are those best writers, and it included zero people from the tech industry, you kind of doubt the way that your ranking function has worked. (Not to talk up our industry too much.)
Erik: Yeah. Well, speaking of lists for a second, it's just funny. You know, Time had a list of like top hundred people in AI and Elon Musk didn't make the list, but Scott Weiner did or Scarlett Johansson did. And it just showed they’re off the ball.
Patrick: Yeah, people do not appreciate the media's production function in some respects. One, any list of topics is “Someone wanted something that was relatively cheap to produce but sustains a lot of attention on the internet and will sell a lot of ads.” Two, a frequent form of industrial organization: Time, Newsweek, etc, have been bought out by effectively private equity firms that wear those brands as skin suits, but they are not the journalism operations that contributed the prestige that we associate with Time and Newsweek.
I weigh these “Top hundred people in Field X” articles by Forbes or Time about as highly as I weigh the top hundred people on bobsblog.wordpress.com. This is internet ephemera, not really worth much time and attention, but because humans are social creatures and everyone likes to be ranked against people, these continue to be great ways to gain time and attention.
I'd also say, if one wanted to be maximally cynical and you were writing a list of the top 100 people and you wanted that to spread the most, you should obviously knock out some of the names that should be on the list just so that people say, "But hey, wait, what about Elon Musk, for example?" Because that will get people talking about your piece. And not saying that journalism is maximally cynical, but they're not stupid either. And this is like any product, this product is evolved by competitive pressures and also by the explicit theory of the mind that the producers have of the product of what the job is to be done that their users want.
So that is an important thing to remember when consuming journalism, regardless of whether it is produced by a truly well-credentialed institution such as, I don't know, the New York Times or by a PE firm which is wearing a legacy media institution as its skin suit.
Erik: Yeah, 100%. The type of media I'm interested in creating or producing for others to create is basically information that helps people get better at their jobs. I think Lenny Rachitsky and you and many others have done a really great job at helping people have a better sense for what it looks like on the inside to run or work at these companies, and how to solve tactical problems. I'm also interested in just the company analysis side.
If you take a company like Replit, and if you said, I want to have a sense for how Replit makes money or how the latest product that they shipped is doing, or what people think about it, or what is the state of the business, how does it compare to competitors? I wouldn't necessarily know where to go for that. You know, sometimes people have hot takes on Twitter. The memos, as you alluded to, are in private where people do that work. But I think we should have a public repository of companies that matter, that are interesting. We should track them. Sometimes the information will have details like Scale has increased revenue or something, more of the details of how they actually operate. So I hope to do a little bit of that research as well.
Patrick: I like to pull back the curtain a little bit on the operations of the tech industry. I think that is both instrumentally useful and kind of morally important for us to do. The vast majority of the people in the world will not get the opportunity to work at a tech company in their lives. And of the people we want to reach, most, unless they are extremely socially advantaged, will never get the inside view of a tech company. They will never have someone sit them down and say, "Here's what sitting around the table at a pre-launch meeting looks like."
Even people who are extremely professionally invested in holding tech accountable, whatever that means, don't understand how tech companies operate on a nuts and bolts level. [Patrick notes: Let he who would throw the first stone first describe what a pull request is.] And that is an extremely important thing to understand if you want to appreciate tech's impact on the world and how to influence that impact.
Big tech companies make decisions on a nuts and bolts basis. Who is in the room? What positions in the company really matter? When you say "Google," you're failing at Kremlinology if you think that Google is one institution that has unitary command and control. (Much like you are failing at Kremlinology if you say the Kremlin is a single institution that has unified command and control.)
We should have an appropriate amount of understanding of how that sausage is made, just like we try to have an appropriate amount of understanding of how the sausage is made in, for example, Washington. I think that democratic access of the constituents to the government that serves them is enormously morally important. And to the extent that one is worried about the amount of power the tech has accrued in the world, making that power more legible to the people that are impacted by it seems like a morally important point.
We have an ongoing discussion about how power is exercised in places like Washington. We have an enormous reporting apparatus that doesn't merely consider throwing stones, but rather actually reporting how the world works to be part of the job. And we have some amount of fictional content, like "The West Wing" and similar, that shows, "Okay, you'll probably never get invited to a cabinet meeting, but here are the players. And by the way, they have staff around them. And it's extremely important to understand, in the beast that is Washington, that staff are making a lot of the decisions." And let's follow some of those staff around.
"The West Wing," all models are wrong, some models are useful. I assume that people who have worked in the White House would not say that "The West Wing" is a fully accurate representation of how policymaking is done in the United States of America. I would also assume that they think it's a lot better of a representation than what you will get if you simply graduate high school with a typical American civics education. And I'll also think that many of them will say, "It's fictionalized. It has to make good television out of it. But I'm totally in Washington because of 'The West Wing.' That was amazing. I wanted to be like that when I grew up."
And I think producing that for tech is important. This industry isn't going anywhere. We need to recruit hundreds of thousands and millions of talented young people over the next 10 years. And that requires telling them true things. Like, this is an industry largely of talented, ethical people who understand these issues that get thrown in our face about governance and similar, who understand that we are doing extremely important work. We want you to help on the really important work. This is not like the caricature. It isn't a bunch of billionaires in a smoke-filled back room who are monopolistically determining the fate of the planet. It isn't the drug-fueled oversexed bonanza depicted in The Social Network.
Erik: Well, Emily Chang works for Protopia. It's sort of making it seem like Silicon Valley is having these insane crazy parties like it's Hollywood or something. Believe me, it'd be cool if they were. They're not.
Patrick: I was once asked in apparent seriousness by somebody, given that I am a somewhat insider after being an outsider in the tech industry, what the average rate per week was of drug- and sex-fueled orgies. I don't know what kind of meetings you have been hanging out with, but a good discussion around the campfire of the virtues and deficits of the book Seeing Like a State is much more likely than the other stuff. And honestly, I've never been around the other stuff. I don't know that the rate is any higher than it would be for the culture slash location that is the San Francisco Bay area, but not my scene.
Erik: I doubt it.
Patrick: Yeah, be that as it may, we're relatively regular people.
We were not bitten by radioactive spiders. That's another thing too. The tech industry has many very smart people in it, obviously. All places in American life where things get done on massive world-altering scale have smart, high-agency people in them. (I focus on America because I’m American, with a twenty year period of my life to the contrary, and also because when “tech industry” is used as a metonym that is a distinctly American institution. Not the semiconductor industry, nope, to the enduring dissatisfaction of many.)
There are many smart people here. We are not a blessed, anointed elite. Most of us do not inherit this job from our forefathers. Nobody ends up being a tech billionaire because your granddad was a tech billionaire. [Patrick notes: The New York Times, on the other hand, is a hereditary monarchy, as Byrne has pointed out many times.]
And that is one of the other things to tell people, as someone who had a bit of the imposter syndrome going on, certainly during my university and early job-seeking years and probably times much more recent than that. Like telling people, you might not think that you're exactly Google material. One, if you're capable of appreciating the question, are you Google material or not? You're probably already in the ballpark. Two, Google and tech companies broadly, they hire smart, accomplished, high-energy folks. Like many places in capitalism, hire smart, accomplished, high-energy folks. There is very likely a place for you here. Certainly don't take yourself out of the running for it.
This frustrates me enormously that the people who have self-appointed themselves to tell the story of the industry want to tell everyone that no individual of good moral conscience would consider working for these folks. Or there are more sophisticated ways to lodge that critique. Like a more sophisticated way is "I saw the best brains in my generation do nothing but optimize serving ads to users." That's a slightly more sophisticated critique than "no value has been created by the tech industry," which is obviously idiotic and we should say so.
But that slightly more sophisticated criticism is not true. You could level a criticism at the institution that is the US government to say, "I saw the brightest brains of my generation go into spending their entire career trying to kill the brightest brains of someone else's generation.” That's descriptively accurate for the job of the military-industrial complex. That's a relatively small portion of the operations of the government. It's also like a relatively small portion of the ethical calculus of having a military-industrial complex. And if you were to grapple with the ethical dimensions of that, and there are people who professionally grapple with the ethical dimensions of that, you would hope not to dismiss it with that one-sentence summary.
And for the people who purport to grapple with the ethical dimensions of the tech industry, simply dismissing it with the one-sentence summary is overly reductive and doesn't lead the audience to truth. Which is a plausible end goal.
The journalism industry used to be quite attached to publishing the truth. Perhaps we could return to our awesome traditions in that quarter.
(57:40) Wrap
Erik:
Right. Yes.
Patrick:
So, Erik, thanks for coming out today and discussing a little bit of this funny world where the tech industry is telling its own story and where there are interesting publications that's bringing up to fill this void for more informed commentary on the companies, people, and ethical considerations that are inherent to it.
Erik:
It was a great conversation. Thanks for having me.
Patrick:
I'll see the rest of you next week. Thanks very much.
