The posters decide

Who picks Silicon Valley’s winners?

It’s a tough call to choose which ChatGPT wrapper to invest in. They’re so different.

Personally, 2020 is not the year I would choose to go back and relive, but earlier this week, the United States decided that is what it wanted, and nominated Joe Biden and Donald Trump to run for president again. 

They both won their respective elections because they were ostensibly the most preferred candidates by American voters. However, there is a theory among political scientists that voters have less say in who wins presidential primaries than it appears. Nominees are chosen by party insiders and Washington power brokers, who typically coalesce around a preferred candidate, and then court donors, secure key endorsements, and recruit talented staffers on that candidate’s behalf. These efforts tilt the playing field towards their candidate, by giving them material advantages like more money and a better team, and by creating media narratives—three Super Tuesday governors all endorsed Bill Pullman! Key donors are having second thoughts about supporting Billy Bob Thorton!—that define one candidate as a frontrunner and others as fringe. 

Importantly, this isn’t some conspiracy theory about stealing elections, or smoke-filled secret societies, or the Illuminati; voters still vote, and their votes still count. It’s a theory of influence: Parties have influencers; those influencers have money, followers, and connections; those resources can be used to alter public opinion. Sometimes, influencers will disagree; sometimes, their influence isn’t enough to sway voters; sometimes, voters sway influencers’ opinions. Still, the argument goes, we shouldn’t see elections as decided by an entirely free market—and the more the insiders agree, the less say the everyday voter has.¹ 

But this is all a Washington thing; this blog is about Silicon Valley; Silicon Valley is not like Washington. A few months ago, Bill Gurley, a prominent venture capitalist and the personification of Silicon Valley swagger, said in a thundering presentation that “Silicon Valley is 2,851 miles away from Washington, and as these people put their eyes towards us, I would state the following: The reason Silicon Valley has been so successful is that it’s so fucking far away from Washington, DC.” The line got a standing ovation. Here, we don’t have cabals; we have free-market capitalism. There is no “party” that “decides;” there are no monied superdelegates; there is no invisible primary. There is only the invisible hand—users who choose, customers who buy, ideas whose time has come.² This is a meritocracy, if Washington will let us keep it.

Or at least that’s a nice story? Because Silicon Valley sure seems to be driven by the exact same dynamic that people say drives presidential primaries: Elite insiders choosing who gets money and who doesn’t, influencing who can hire and who can’t, and tilting markets and (social) media narratives in favor of their preferred technologies, companies, and founders. In other words, in the way we chose our successes, Silicon Valley doesn’t seem very far from Washington at all.

Same medium, subtle differences

When people ask me about the very early days of working at Mode, I often say that I wrote articles for our corporate blog. This isn't entirely true. The initial posts that I wrote weren’t actually published on Mode’s website, but on a blogging platform called Svbtle. I wrote my first blog post as a Mode employee two days after we started the company,³ which was well before we had a corporate brand—and therefore, well before we had a website on which we could host a blog. So we had to choose a third-party platform instead.

Back then, the two up-and-coming options were Svbtle and Medium.⁴ As products, they were basically identical—both were very easy to set up; both embraced the crisp millennial minimalism aesthetic that was irresistible in 2013;⁵ both came with nifty charts that would show me the dozens (dozens!) of people who read what I wrote; both had nascent distribution and discovery features that would, in theory, push my posts out to new readers via some algorithmic recommendation engine. I chose Svbtle, and passed on Medium, for none of these reasons. I chose it because it was still invite-only and using it made the egomaniac in me feel special, and because Svtle’s “like” feature—called Kudos⁶—was fun.⁷ 

Most of Silicon Valley’s elite chose Medium. It raised money from some of the industry's most followed venture capitalists, including Greylock’s Josh Elman, a16z’s Ben Horowitz, and Google Ventures’ M.G. Siegler, all of whom began posting on Medium. Though Svbtle had fans and notable backers of its own, most prominent writers in tech—Hunter Walk, Julie Zhuo, David Heinemeier Hansson, Jason Fried, Jason Calcanis, Stewart Butterfield, Anil Dash, Chris Sacca, Dustin Moskovitz, Brad Feld, even Bill Gurley—coalesced around Medium. Within two years, people were declaring Svbtle dead, and had accepted Medium as Silicon Valley’s default blogging platform. 

Of course, products succeed and fail for lots of reasons. We’ll never know exactly why Medium won and Svbtle lost. Maybe Medium’s editor was a little better; maybe its viral distribution loops were a little tighter. Maybe Medium was right to hire their own writers. Maybe bombastic VC grandstanding is better paired with an elegant serif than austere sans serif. Maybe Svbtle was too exclusive for too long. Maybe product science really works. I don’t know. 

But as someone who was publishing on Svbtle at the time, it felt like the only thing that mattered was Medium's momentum—and specifically, its momentum among people who had big existing followings. When I shared my posts, people periodically asked me why I was using Svbtle. Wasn’t Medium the thing everyone was using now? Could I not get an invite to it yet? If you used Medium, you were a respectable writer with an independent voice. If you knew about it, you were in the club. IYKYK.

Very little of this perception came from people’s lived experiences of writing on Medium, and even less of it came from people who had tried both products and found the nuances of Medium to be more effective or more agreeable than Svbtle. Medium’s momentum came from important people in Silicon Valley posting on it and posting about it on Twitter, and from the rest of us following them. That doesn’t discount the quality of Medium’s product—it was good, and Silicon Valley’s influencers surely talked about it in part because it was good. But it feels incomplete to tell the story of Medium’s success, and of Svbtle’s failure, by just talking about the product. Surely, the party also played a role. 

The emperor occasionally experiments with editing

If I’m really honest about it, watching this unfold probably had an overly formative effect on me and my sense of how Silicon Valley works. But still, once you see it—once you see elite opinion become public perception, and public perception become at least some sort of reality—it’s hard not to see it everywhere. There was a moment in 2014 when Slack was suddenly the only product anyone could talk about. There was a moment in 2016 when Airtable, a fairly confusing product that wasn’t easily identifiable as anything, was broadly hailed as the new way to build everything. There was a moment in 2017 when VCs and startup founders were obsessed with—and obsessed with talking about—Superhuman. 

There was another moment like this earlier this week. A small startup called Cognition Labs launched Devin, an AI agent that can develop software applications on its own. The video blew up, with dozens of notable Silicon Valley influencers posting about how Devin was a remarkable leap forward for both AI and for software development. Within a few hours of the announcement, people were already suggesting that it might be the next $100 billion company, which is 25 percent more than even OpenAI is currently worth.

Which, maybe! A hundred billion seems high, but I don’t know! Cognition CEO Scott Wu can apparently solve impossible math questions in less time than it takes most of us to read impossible math questions, so if you have to put a bet down on something, I guess Cognition is as good as anything. 

What seems notable, though, is I’m not saying that because I’ve used Devin, which is still behind a closed beta, nor am I saying it that because the launch video was that unique or impressive. There are tons of companies building AI agents like Devin, and you can never fully trust a pre-recorded demo. Instead, my belief in the potential of Cognition comes entirely from other people’s belief in the potential of Cognition—which was, at least to some degree, astroturfed by Cognition’s investors and PR team.

The point here isn’t that the emperor has no clothes, or even that they don’t have $100 billion clothes. The point is that nobody has seen the emperor. Investors have, and we’ve seen curated demos, and the emperor is very good at math. But the clothes are still invite-only. And yet, all at once, the entire internet began telling the same stories of their wonder.

Imagine that you didn’t see the Devin demo because a Silicon Valley celebrity tweeted about it. Imagine that the video only had a few thousand views. Imagine that you found it on the second page of Product Hunt, on a slow Thursday afternoon. Imagine it had fifteen upvotes and a lonely comment from the “maker,” saying they were very excited to share their project with the world. Imagine that the thing itself was exactly the same, but there was no party and no posting around it. 

Would it be impressive? Would you think anything of it? Would you want to invest? Would you roll your eyes at what was probably the fourth AI bot that you saw on your bored scroll through Product Hunt? 

I really don’t know. I doubt I would’ve thought that it was any different than the other belts I’ve seen. Which may well be wrong—Devin really might be different, and in ten years, Devin may be building all of our websites, and our phone apps, and even future Devins. But without the hype, I don’t think Devin and its five minute sizzle reel would register to me as anything that was particularly noteworthy. 

But with the hype, I would invest in it? Partly, perhaps, because I’ve seen the light, but more because Silicon Valley’s influencers—the party—have decided that Devin is important. Cognition can now easily court investors, secure key partnerships, and recruit talented staffers. The playing field is tilted in their favor; they have material advantages like more money and a better team, and a media narrative that defines them as a frontrunner in a very noisy space.⁸ And once you’re on the ballot, you’ve got a shot.

From a pile of “stuff”

On one hand, all this is very silly to say. Of course this is how Silicon Valley works. It's how everything works. Party elites don’t influence presidential primaries because of some unique culture of cronyism in Washington DC; they influence primaries because social systems have hierarchies, hierarchies have people at the top, and the people beneath them pay a lot of attention to what the people above them do. If Taylor Swift likes Travis Kelce, we all like Travis Kelce. If Oscar de la Renta does cerulean, we all do cerulean. If the president says it’s a revolution, we storm the Capitol. If engineers who work for OpenAI say that Devin will transform software development, we tweet memes about how Devin is five years from taking our jobs. If important people post about something a lot, we start to post about it too—that is not a novel point to make.

On the other hand, it still seems useful to say? The conversations about these dynamics in Silicon Valley tend to exist on the poles. There are conspiracy theorists on one side, who say that Silicon Valley is controlled by mob bosses and hitmen, and there are heads-down hustle heroes on the other, who say that the only thing that matters is what you build. 

The reality is somewhere in the middle. Investors and tech twitter personalities aren’t all in one group chat, masterfully pulling our strings in unison, but they aren’t external observers either. In that, Silicon Valley isn’t exempt from the same dynamic that drives every industry: There are influencers, and there are the influenced. Though companies ultimately live and die on their own merits—in recent years, Medium’s road has been rocky—the party decides which companies win the early-stage primaries.⁹ And our predictions about breakthrough technologies and breakout successes aren’t the product of lived experiences and independent analysis. They are selected for us. 

1. Prior to 2016, some people believed that, if voters and insiders were at odds, the insiders’ candidate would win, which led to some ill-fated predictions about Trump’s chances in the primary (a primary which he has now won three times). A seemingly more accurate version of the theory makes adjustments for how much the insiders agree. If they all coalesce around one candidate—say, Hillary Clinton—they can have a lot of influence. But if they don’t agree, they tend to be less effective. Succeed together or fail alone, I guess. 

2. #Bitcoin 🚀

3. The blog’s inaugural post was about Miley Cyrus, as every blog post that introduces a technical B2B SaaS tool should be. That’s what it was like back then—I used to write about Miley Cyrus, and now I write about liquidation preferences. Time has done changed me; that's fine, I've had a good run. I know I used to be crazy, that's 'cause I used to be young.

4. In hindsight, it probably would’ve been better if people at Mode told me to choose a third platform: www.creedthoughts.gov.www/creedthoughts.

5. Apparently, Gen Z is over millennial minimalism, and is now into cluttercore. Which, ok, but will that eventually make its way into web design too? Are we going to go back to the MySpace aesthetic of websites looking like scrapbooks, with overlapping UI elements, spinning gifs, and background music? 

6. Am I the only person who will never be able to separate this word from this granola bar?

7. You can still check it out! Scroll to the bottom and hover over the circle! Very cool!

8. This tweet, for example. Before Devin came out, would you look at this resume—an engineer who spent some time in sales—and think “that is a great founder, I must invest” or “huh, that’s weird?” Both seem like reasonable reactions. But after Devin comes out, the second story is the one we tell ourselves.

9. Is that bad? No? It’s probably fine? You could argue that this is the point of venture capital, to have experts comb through haystacks of startups looking for the needles with potential. The social media blitz around Cognition may well be a good thing, if Cognition is the company that we need to make better software, and if the hype around Cognition is what it needs to realize its full potential.