January 27, 2019 Snippets: The Double Bind

As always, thanks for reading. Want Snippets delivered to your inbox, a whole day earlier? Subscribe here.

This week’s theme: Silicon Valley rituals around competition and failure; plus Swarm raises $25 million in their Series A.


In last week’s Snippets, we introduced the idea that what really drives Silicon Valley has to do with people: the unique social rules, norms and rituals we observe around desire, and modelling our desire off of other people. This week we’re going to dig deeper into what the tech community really thinks about modelled desire: about similarity versus difference; differentiation versus undifferentiation. So, humour us for a bit while we go exploring, we might find something interesting at the bottom of all of this.

The most popular book on this subject in the tech canon is undoubtedly Peter Thiel’s Zero to One, which came out several years ago with some strong claims about differentiation: competition and capitalism are opposites. Competition is bad, because it’s destructive. Competition is for suckers. You want to be differentiated, not undifferentiated. Be contrarian. (It struck a chord in the community, and kicked off a recurring chorus from the crowd: “I, too, am contrarian!”)

His point about competition being destructive is important, but reading Zero to One you get the feeling that Thiel is not telling the whole story of what he really thinks; like he’s hiding something big from us. It’s actually sitting right there at the outset, when he shifts seamlessly from one chapter saying “Competition is bad, because it’s destructive; wealth and societal value are created through differentiation.”, and then in the next chapter suggesting, “What if the 1999 bubble was actually good, because for once we actually had a collective moment of clarity that we need to innovate our way forward by building the future?” The 1999 bubble was many things, but one thing I would not associate it with was differentiation. It was a bubble! Bubbles are periods of extreme undifferentiation; where everyone temporarily becomes the exact same; pursuing the same things; copying each other more aggressively and outlandishly than usual, as we race to build the future. That’s good? Wait, I thought that was bad! What are we supposed to think?

If you look at what drives the creative energy and relentless work ethic that we see in the tech industry, and especially in early stage startups that are creating something out of nothing, we have to face something uncomfortable: in many of the ways that count, all startups are basically the same; and this is an important part of why the startup ecosystem works the way it does. They all compete for the same money, in the same cap tables, with the same SAFE notes. They all compete for the same engineering talent, for the same developer positions, to solve similar software challenges. For the most part, they’re all graded on the same rubric: monthly growth rate. At a deep level, they compete internally within Silicon Valley off of the same value proposition: We are a lottery ticket. Who wants to buy us? Y Combinator, who understands this perhaps better than anyone, makes this explicit: Everyone gets the same terms. As far as your cap tables are concerned, all startups are the same. (We’ll talk about this more next week, but I think there’s a case to be made that there’s only one thing we universally recognize as legitimately differentiated in Silicon Valley, and that’s founders.)

All of this similarity is a feature, not a bug. Yes, everybody resembling and imitating everybody else leads to some funny little absurdities, like moving from one startup with Kind Bars and exposed brick walls to another startup with Kind Bars and exposed brick walls. But all of this sameness, as we discussed last week, does something important: first, it validates what any one crazy little startup is doing, because their peers are doing it too. It’s a whole lot easier to raise money to build a smart toaster company if everybody else is also raising money to build smart toaster companies. And, perhaps more importantly, startups using the same methods, the same technology, and the same playbooks as their peers creates an important mutual validation of one another: we’re using this tech, or raising money in this way, or using this go-to-market strategy, and the fact that all our peers around us are using it too helps alleviate a lot of doubt and hesitation that would normally be overpowering. All of this cross-validation collectively removes a lot of friction from the ecosystem, and makes it easier to just get started now and ask questions later. Second, it creates rivalry: the more closely we resemble one another, the more inevitably we will become competitors. Bubbles are what you get when this duality of mutual validation combined with mutual rivalry unlocks enormous amounts of resources, and drive everyone to temporarily throw away everything they were doing and dive headfirst into whatever the new thing is.

What this all boils down to is a hidden jewel at the core of human behaviour called the ‘Mimetic Double Bind’, a derivative of Gregory Bateson’s Double Bind concept:

I love you for copying me, because it validates me. And I hate you for copying me, because it makes us rivals. No escape route is available.

The Double Bind has been deeply internalized by the Silicon Valley tech community. We don’t openly talk about it, but we revere it. We understand, at are core, that it is our energy source: that Silicon Valley works because we let it go to work for us. But we also acknowledge that we have to be careful: if we let it run too wild, it runs away from us. The bubble bursts and we all lose. If you want to see this acknowledgement and reverence in action, look at one important ritual that we unfailingly perform, month after month: the ritual of what happens when a company shuts down.

Primitive societies from early human history, which similarly to Silicon Valley did not have the layers of repressions and taboos against open copying or envious behaviour, pretty much all had one alarming and violent ritual in common: human sacrifice. The specifics of how these rituals are performed vary from culture to culture, but they all seem to follow one important general rule: they’re a ritualistic representation of undifferentiation and the envious violence that comes with it (which, back then, was real violence) being resolved, through a death of a victim, into renewed differentiation and the return of peace. Through the ritual, the community is effectively doing three ritualistic steps: 1) acknowledging undifferentiation as the source of violence and problems; 2) transforming the victim into a surrogate, or a kind of token representation, for that undifferentiation, and then killing them; then 3) everyone collectively celebrating the symbolic “death of undifferentiation” as a kind of salve; like opening a release that lets off steam and brings the community back to peace, even though in reality, undifferentiation and envy in the community has not been actually resolved. But, for a little while, peace is restored. We celebrate death as the source of new life.

Now, let’s look at what happens when a well-known and generally well-respected startup, which at one point was a high-flying growth story, announces that it is shutting down. (This week, it was Munchery. Next week, next month, next year, it’ll be someone else. It could be anyone.) We go through a very deliberate ritual as a community that, although there’s obviously no literal violence or human sacrifice involved, replicates the exact same steps as those primitive ceremonies.

First, we talk about how it’s so brutally competitive being a venture-backed startup; “another promising company bites the dust to the brutal reality of venture-backed competition”, acknowledging undifferentiation as the source of rivalry, problems, and ultimately, of the death of many promising companies. Then, we do something really interesting: over the course of the day, the dead company gets stripped away of many of its distinguishing factors, histories, and even products. They’re then commemorated specifically in terms of how much money they raised and spent: a perfect representation of undifferentiated competition if there ever was one. It becomes eulogized, almost without fail, as “Company, who raised X million dollars, has shut down.” There is almost no character assassination; very little personal criticism of the founder, not like you’d see in other industries in a notable bankruptcy; instead, we congratulate them and give them EIR positions. Blame is shifted away from company specifics and into the idea of the rat race itself.

The whole process is a ritualistic portrayal of undifferentiated competition as the source of our problems. Internally, we quietly reflect: “Well, they died so that the remaining companies can now live, as profitable businesses that have exited the crucible of competition.” The problem of competition itself has not actually been vanquished in any way whatsoever, but we pretend it has, at least for a little while. Publicly, in contrast, we go through a period of performatively condemning undifferentiated conflict: “Is raising so much money good for startups? Does VC kill good companies?” Then, finally, we perform the last rite of the ritual: we announce, to no one in particular, that Silicon Valley is ‘a place where we celebrate failure’.

Whenever companies die, we go through this ritual and make an important point of condemning all of the undifferentiation and the competition, but then do not actually do anything about it, because we need it so very much. The fact that we never actually take any kind of action as a community, while making a repeated point to talk about how bad this all is, is a kind of respect for the double bind: condemning it with our words, but embracing it with our actions, as a way of keeping the bubble in check, a little bit each time.

One final note before we break for next week: there’s an old psychological and sociological concept called “the founding murder”, which goes back to Freud and others, which says something like, “all periods of peace and prosperity must have begun with an acute act of violence.” One interpretation is that, in order for lasting stability be possible, there must be some collective memory in the community of an event, whether real or imaginary, that everyone fears and reveres as a founding moment where the undifferentiation and chaos of the pre-history was resolved, in a moment, into the newly differentiated stability. What might that moment be for the modern tech industry? What is the founding murder of modern Silicon Valley?

I think it was the dot com crash. The bubble in 1999 was indeed a moment of clarity, like Thiel suggests in Zero to One, and it was a moment that embraced the mimetic double bind as an energy source. We tell and retell hyperbolic stories about our ancestors, the dot com companies, who IPOd out of living rooms with nothing more than a business plan (as the legend goes, so was everyone!) and who were all truly identical to one another in the sense that 99.9% of their market cap was the same fumes and hype as everyone else was breathing. It was a moment where we realized, at some subconscious level, that the double bind could be liberated as an energy source to motivate founders, build startups, to build the future, and to change the world. But we liberated it too far, too fast: we lost our minds, the bubble burst, and most of those companies died. In the aftermath, the true, modern form of Silicon Valley was built from the ashes, in the restored differentiation: between those who survived and those who did not. We learned an important lesson: let the double bind work for you, but do not let it run completely free — keep it in check, like through rituals like the ones we’ve talked about. Our legends and our collective communal memory about 1999 and 2000 take this general form in one way or another. (For reference, I’m 30 years old; in the year 2000 I was eleven. I do not have any actual memories of the dot com crash. But I absolutely have memories and reverences and a great cultural fear that have been instilled on me by the tech community: that this was “the before-time”, when gods roamed the earth and brothers fought each other with sticks and spears.)

Next week, we’re going to look at the other side of the coin: if so much of what we’ve discussed this week is about undifferentiation, then what in Silicon Valley is recognized as unique and special and valuable? It’s not companies, it’s not money, it’s not technology, it’s not ideas, it’s not even growth. It’s founders.


In this week’s highlighted reading list, one really interesting read (h/t to Taylor Lorenz for the suggestion) is Kyle Wagner’s article on Gamergate from back in 2014, in a piece that reads as brilliant foresight. It also adds another brick to the steadily building case that the gaming community is increasingly the origin of mainstream culture, which is pretty interesting but also quite worrisome!

The future of the culture wars is here, and it’s Gamergate (2014) | Kyle Wagner, Deadpan

The Information also had a great profile on Andy Jassy, AWS’s CEO, one of the most influential but least known tech company CEOs out there. It also features this excellent line from right up front:

“I would not want to compete with Andy Jassy,”, said Mr. Larson, whose company [makes] the Taser electroshock weapons. “The guy is clearly a force.”

Amazon’s Cloud King: inside the world of Andy Jassy | Kevin McLaughlin, The Information

And finally, this is a very interesting article by some folks from Cornell CS, MIT and Dropbox who help explain why sometimes it’s a good idea for login pages to allow people to successfully sign in even if their password is slightly wrong. Tl;dr: denying access because caps lock is on, or if the first letter of the password is accidentally capitalized on a mobile device, for instance, may do more harm than good: it’s unlikely to thwart any attackers, but can cause real harm when forcing users to reset their passwords introduces new opportunities for bad security hygiene.

pASSWORD tYPOS and how to correct them securely | Rahul Chatterjee et al.

A pretty wild theory that has some compelling evidence behind it: Alzheimer’s Disease as a downstream consequence of gum disease?

We may finally know what causes Alzheimer’s, and how to stop it | Deborah MacKenzie, New Scientist

Chronic oral application of a periodontal pathogen results in brain inflammation, neurodegeneration and amyloid beta production in wild type mice | Ilievski et al., PLOS One

Porphyromonas gingivalis in Alzheimer’s Disease brains: evidence for disease causation and treatment with small-molecule inhibitors | Domniny et al., Science Advances

While other avenues of scientific progress keep making steady gains:

Advances in weather prediction: forecasts really are improving | Richard B Alley, Kerry Emanuel & Fuqing Zhang, Science

A gut punch fights cancer and infection: evidence is accumulating that microorganisms in the human gut give important support to the immune system in fighting infections and even cancer | Nathan Reticker-Flynn & Edgar Engleman, Nature

Good news in the ebola crisis: vaccines are working, and supplies are expected to last | Helen Branswell, STAT news

Posts from around the portfolio:

Making the cloud more secure and effective for developers | John Spurlock, Cryptomove in partnership with AWS

The ultimate guide to remote meetings 2019 | Deanna DeBara, Slack

Salesforce’s Mike Kreaden on how to build a platform to drive growth | John Collins, Intercom

Other reading from around the internet:

Eugene Wei — Tech, Media and Culture — Investor’s Field Guide Podcast with Patrick O’Shaaughnessy

Seth Klarman, in a rare interview, offers a warning | Evan Osnos, The New Yorker

The case against behavioural advertising is stacking up | Natasha Lomas, TechCrunch

On the ebb and flow of luck: a visualization of the variability of annual returns and risks | Nick Maggiulli, Of Dollars and Data

Befriending peacocks to being #Humbled: Indian prime minister Modi is winning Instagram to woo millennials | Nandita Singh, The Print

When Charlie Munger calls, listen and learn: a self-published author gets an unsolicited phone call | Jason Zweig, WSJ

Where Amazon returns to go get resold by hucksters | Alexis Madrigal, The Atlantic

And a throwback article from another time:

The Dirty Trickster: Jeff Toobin’s 2008 piece on Roger Stone that makes for some good reading this week | Jeffrey Toobin, The New Yorker


In this week’s news and notes from the Social Capital team, Swarm had a big announcement this week — they’ve officially raised a Series A investment of $25 million from David Sacks’ firm Craft Ventures along with friend of the Social Capital family Sky Dayton.

Swarm raises $25 million Series A to build world’s lowest-cost satellite networks | Sara Spangelo, Swarm

Satellite startup Swarm raises $25 million for space-based internet plan | Aaron Pressman, Fortune

It’s a great step for Swarm, who will be able to use the new cash to continue building out their world-class (really we should say Space-class) team, and to take the next step in deploying their global constellation of satellites into orbit. It’s also pretty stunning to think that a team only a couple years old and with only a dozen-odd members could have multiple satellites already in space, sending and receiving messages, and jumping over technical and operational hurdles that used to take massive teams and billions of dollars just to get to the starting line.

It’s worth taking a minute to appreciate what this really looks like: where are those satellites in space, right now? On Swarm’s website, you can track their satellites in real time through LEO Labs; at the time I grabbed this screenshot (Sunday morning around 9:30 AM east coast time) a couple satellites were over the South Pacific, a few more over Antarctica, one over Greenland, and the rest over the oceans — check back another time and you might find them somewhere new, even maybe overhead.

In the old days of satellite internet, not only were launches expensive and time horizons impossibly long, but it was also hard to say exactly where the urgent demand was coming from. The hope that if you built it, they would come was just that — a hope that needed to be tested over a very long time. As Sky Dayton, who successfully took the opposite path when building EarthLink and Boingo, observed: “I had always avoided the satellite space, because it’s traditionally required a billion dollar investment followed by 10 years to find out if you were right. Swarm’s approach reminds me of the early years at EarthLink — stay super scrappy, serve customers and generate revenue quickly.”

Today, not only have the economics of satellite launches changed significantly (and are poised to continue doing so, thanks to a renewed space race and the imminent arrival of new startups like Relativity Space entering the launch industry), there’s also a continuous need for more connectivity, everywhere. CES this year was all about connected devices, especially car companies who want to make sure their vehicles are in continuous communication through satellite links, including Swarm’s partnership with Ford through Autonomic.

With this Series A financing, Swarm will continue to build and deploy their constellation of 150 satellites over the next 18 months, and then the real fun begins: building and delivering connectivity solutions to a wide variety of problems and industries. The first batch of applications they’ll be targeting, with major agreements already in place, include diagnostics and emergency messaging from connected vehicles, agriculture sensors in farmlands outside of cell coverage, shipping containers and asset tracking across oceans, water monitoring devices in remote African countries, smart meter reporting in remote locations, and connecting people through text messaging in rural and remote areas.

On a personal note from the Discover Team at Social Capital, Jay, Nicole and I along with the whole team at Social Capital couldn’t be any more excited and proud of how far the Swarm team has come, and how high they’re going to be able to climb. As one of the first companies to be incubated through Social Capital’s Discover program, Swarm represented everything that we care about when we ask ourselves what companies to look for or to help start: are they helping to solve one or more of the world’s hardest problems? Absolutely: if just one out of Swarm’s many potential applications is a game-changing success, they’ll have done something very special to help the world. Are they tackling a problem in a brand new way that most people think is impossible? Do they know something others don’t? Absolutely: prior to Swarm, it was considered pretty settled knowledge that CubeSats had reached their smallest dimensions possible, because they’d hit some fundamental limitations around thrust and stabilization control. But Sara and Ben knew something others didn’t: Swarm’s special Secret Sauce that allows their satellites to be dramatically sized down to the size of a grilled cheese sandwich, which in turn unlocks the launch economics, deployment speed, and everything else that brings their vision within reach. Are they a special founding team? I think at this point it should be clear to everyone; from the very beginning, it was pretty clear to us. It’s been such a joy working with the Swarm team, and days like today are a good reminder of why we do what we do.

Keep in touch with Swarm’s incredible journey by subscribing to their newsletter; you can also find career openings — including in business development, hardware and software engineering, regulatory and operations, and more — here. If you’re interested or know someone who might be interested in joining the Swarm team, please send them Swarm’s way.

Have a great week,

Alex & the team from Social Capital