SCIOPS 01.23: Platform Cooperative
Hello, and happy Labor Day to my correspondents in the US. If you don’t know, Labor Day is a wonderful holiday where we celebrate the violent shutdown of the Pullman strike by the federal government. Thirty workers were killed, Eugene Debs was imprisoned, and the Pullman workers pledged to never again form a union. Hurray! We shall not be moved! Solidarity forever!
Well, the new industrial revolution is here, and in the spirit of Debs I’d like to offer some radical solutions to the current economic paradigm. The Fordist worker-consumer economy is long gone, the US is in a drastically unstable “growth period” fueled by monetary policy tricks, and the platform capitalism of the tech boom is steering us toward massive unemployment and unprecedented monopolies.
In SCIOPS 01.22 , I spoke about this platform model and asked: *”
What can we do about it? What other equilibria exist in the slippery new internet world?”*
This week I’ll share the key idea to making the next economy humane and equitable: the platform co-operative . If the current capitalist model is to build a Schelling point and then charge rent to the buyers and sellers who seek each other there, the only equitable solution is to make Schelling points that are collectively owned by the people who use them .
This conversation often starts something like this: “What if we, the users, owned Facebook? What if our posts and likes and shares actually paid us, rather than just creating value for Silicon Valley dorkwallets? We would get paid to use social media , dude!”
That’s the wrong approach. The (alleged) value of social media is not the people who read and thumbs-up your posts. They don’t pay a subscription to see pictures of your filthy children. The users are not the customers. The advertisers are the customers.
And yes, we could make a platform where the users get dividends from the advertising revenues, which are in turn based on how much the users buy from the advertisers. But that sounds like a vortex, a black hole for money where no real value is being created. Besides, the focal points have already been captured. People have been trying to make a decentralized Facebook for a while now; they don’t succeed, because Facebook has the network effect. To make a social network, you have to have the users, and to get the users, you have to have a social network. Catch-22.
So where are the openings? What domains have yet to be monopolized by a successful platform, and what efforts are being made to build co-operative ownership thereof? Here’s three such projects, handpicked for cogsec practitioners:
Journalist and activist Barrett Brown, in his new column at Motherboard, has set in motion a new project for investigative journalism and activism. It’s called Pursuance , and it’s designed as a “process democracy” (he’s not super clear on the definition of the phrase), where individuals can collaborate on collective investigations into fake news, abuse of power, and other dystopian features of the modern civic landscape.
If one were to start with a sufficiently resourceful group of initial participants with broad agreement on keystone issues—opposition to the drug war, police state, and mass surveillance, for instance, with these issues chosen in order to establish a reasoned polity sharing common values, if not ideological unanimity—one could expect such a system to quickly expand into a vast and formidable new force in world affairs, capable of advancing reform and confronting criminalized institutions across the globe.
– Barrett Brown, What Is to Be Done?
Crowdfunding has shaken the traditional practices of manufacture – instead of getting a loan, sinking that money into production, and finally selling products to end users, the Kickstarter and Patreon models involve getting committed buyers before production. That way, the money spent to make the product is collected directly from the consumers, and the creators don’t have to pay interest on a business loan.
That’s great as far as it goes, but the reason it hasn’t completely toppled the economic system is simple. In a successful campaign, the crowdfunders will get what they paid for – a product. But the companies that grow out of these products can keep generating revenue long after the initial crowdfunding seed. The early backers get no equity – it’s a donation, not an investment. Famously, the Oculus Rift VR headset was funded from over 10,000 Kickstarter donors, but paid them exactly nothing when it was acquired by Facebook for 2 BILLION dollars.
Swarm.fund proposes to remedy this situation, by combining the crowd-funding power of cryptocurrencies with legal structures that can own real property and receive dividends. Swarm (which holds its public ICO on Thursday, if you’re interested in joining) is a co-operatively owned, democratic platform for facilitating investments. It could be bigger than Kickstarter – hell, if it actually monopolized the market for investment it could be bigger than anything. Their target is the 3 trillion dollars currently sitting in hedge funds with a less than 3.5% return. If they can successfully use the hive-mind to get better returns, some of those trillions will be distributed among the democratic member-owners of the co-op.
Artificial intelligence is not exactly an unoccupied market. All the top tech companies are becoming AI companies, as it represents the fastest way to reduce marginal costs and disenfranchise labor. It’s a classic capitalist gambit, as automation has always been.
But AI has needs. It can’t grow in a vacuum: it needs a deluge of processing power and megatons of training data. Right now, these data are produced by peripheral means: the spyware companies, like Google and Facebook; siloed efforts of researchers at universities and tech giants; and automated training programs like reCAPTCHA . The data must be annotated by humans to be useful for training AIs. After all, how can you score a test unless you already have the answers?
John Robb, the notoriously accurate spook futurist, proposes the AI@home Consortium, which would combine decentralized processing (on mobile phones, laptops, etc) with a co-operatively owned pool of training data. If this model snowballed, the network effect would be give AI@home the monopoly on datasets and processing power. The revenue from selling those data would be distributed among the worker-owners, giving them an incentive to make the biggest, best datasets on the market.
After all, this is similar the model Facebook uses: you give them clicks and likes and shares, they concoct AI based on those data (its only purpose the targeting of ads) and they make billions by selling access to that AI. The only problem is that the people making the data are not the people getting the billions.
That’s what I’ve got for you this week. Really, check out those three projects and tell me what you think. Is this the labor movement of the future, or a dead duck like so many other tech utopias? There are lots of other platforms that could be cooperatized, like healthcare, education, and (of course) ride-sharing. What co-op would you join?
Thanks for reading.
is a cogsec newsletter. If you liked this issue, forward it to a friend. If you’re not signed up, you can sign up at
. Question, reflections and criticism can be levelled by responding to this email.