Politically Opinionated Platforms
|Venkatesh Rao||Feb 15, 2019|
Two years ago (March 2017), I wrote a newsletter issue titled [Politically Opinionated Products](http://The difficult case is products where you are a prosumer, one among a bunch of amateurs in the multi-player community of users around a "democratized" capability that previously used to be the preserve of expert single-player producers and consumers. The trick to democratizing a capability well is to not just build opinionated technology, but politically opinionated technology. When you do this well, people come for the tool, but stay for the network, as Chris Dixon put it (I am borrowing the single/multi player terminology from the linked post). When you do it poorly, you get a toxic shitshow.), in which I argued that prosumer products featuring significant social experiences should be politically opinionated by design. To quote myself:
The difficult case is products where you are a prosumer, one among a bunch of amateurs in the multi-player community of users around a "democratized" capability that previously used to be the preserve of expert single-player producers and consumers. The trick to democratizing a capability well is to not just build opinionated technology, but politically opinionated technology. When you do this well, people come for the tool, but stay for the network, as Chris Dixon put it (I am borrowing the single/multi player terminology from the linked post).
This opinion has strengthened over the last couple of years. In the 2017 take, I meant something fairly mild by "politically opinionated" -- an extension of "opinionated" individual user experience design to the group experience. Now I believe in a stronger and simpler form of the thesis: a platform-scale social technology is a political ideology in the conventional sense. Its nature dictates a particular political "grain" and it is hard, perhaps futile, to try to go against directly.
If you want open, pluralist platforms, you're going to have to mix "pure" technologies in specific proportions, and arrange them in a conscious balance of power. If you can't achieve the balance you want with the pure technologies available, you need to invent new technologies that exhibit the ideological predispositions you think are needed to restore balance. Regulation (by institutions that themselves rest on balances of power among politically opinionated industrial-age technologies) is not a powerful enough force to do the job here.
Niels Bohr once said, the opposite of every great truth is also a great truth. Pluralist platforms are best understood as balance-of-power equilibria among competing great truths.
Let's start by plotting 4 major "pure" emerging multiplayer platform-scale technologies on the well-known political-compass meme 2x2. I originally tweeted this classification as a joke, and then realized I believed it seriously.
Technology on the poliitical compass
1/ Technologies are media. Media have messages. Large scale, multi-player social technologies have political and ideological messages. Platforms built out of these have emergent political messages that are a function of the component messages.
2/ Good social platforms recognize and accommodate the messages of the constituent pure technological media. They do not try to make the media work against their natural grain. That basically never works.
3/ To the extent that the ideological messages of various media, in purist form, are toxic, the trick to counter-programming them is to blend them. Technological homogeneity is a recipe for ideological cancer.
4/ If you want moderate, balanced, overall centrist, pluralist platforms, you're going to have to mix "pure" technologies in specific proportions, and arrange them in a conscious balance of power.
5/ If you can't achieve the balance you want with the pure technologies available, you need to invent new technologies that exhibit the ideological biases you think are needed to restore balance.
6/ Regulation (by institutions that themselves rest on balances of power among politically opinionated industrial age technologies) is not a powerful enough force to do the job here. The political character has to be a natural one. Regulation can only provide minor finishing touches.
7/ Based on their early histories, arguably, 4 of the most important emerging technology sets today -- deep learning, GOFAI 2.0, VR/AR/IoT, and blockchains -- have clear, fairly pure, ideological grains.
8/ Deep learning has an authoritarian-right bias. It feeds on vast data sets created by natural behavior, has a tendency to inherit and reproduce endemic biases, and codify them in favor of conservative authoritarians who see the incumbent balance of power as natural and just.
9/ Trying to "regulate" the functioning of deep-learning algorithms directly, through human political processes, or by demanding Justifiable AI that can "explain" itself, is a fool's errand destined to throw out the baby with the bathwater.
10/ GOFAI -- "Good Old-Fashioned AI" -- is not new, but is seeing a 2.0 revival due to the rise of deep learning. Technically, it is based on formalist methods -- symbolic logic, theorem-proving -- that have a natural authoritarian left (ie, bureaucratic, impersonal) bias.
11/ These will likely see a powerful revival soon, as methods from the last "AI winter" era of the 1980s are revisited, upgraded and strengthened. But these are also naturally an authoritarian rather than individualist technology.
12/ In fact, I predict a new AI winter soon, as disillusionment with purist deep-learning approaches drives a retreat. This will be a healthy period of learning to pragmatically blend GOFAI and deep learning. Check out David Chapman's deep dive for more insight on this.
13/ Next, we have the 3 technologies I've lumped together: AR/VR/IoT. What they have in common is that they are democratized production tools that take computing beyond the screen-and-keyboard interaction paradigms (which has been with us since 1961).
14/ Democratized production tools tend to encourage individualist divergence driven by creative effort and personal growth. There's a reason creative/artistic communities lean left-libertarian. Creative tool-use at once encourages individuation and social connection driven by openness to experience, novelty-seeking, and diversity-seeking.
15/ As someone who identifies as left-libertarian, one of my political opinions is that this quadrant is in deep recession right now, and we need a big boom to counter-program imbalances in the zeitgeist due to the deficit here.
16/ Finally, the darling of the last couple of years, the blockchain, is perhaps most obviously a naturally right-libertarian technology. It encourages deeper intimacy among homogeneous in-groups, and social disconnection with out-groups.
17/ It does this by introducing the mediating effect of market-like low/zero trust technologies that are preferred by self-segregating small homogenous groups for interactions among groups. It encourages an attitude of "don't try to influence/be influenced by people who aren't like you. Just trade with them using zero-trust market technology."
18/ None of these 4 technology sets, in pure form, can form a politically healthy, generative environment for humans. Only playpens for ideological robots. In fact, in pure form, they are so politically unstable, I suspect technological paradigm-purity is a really good predictor of failure.
19/ In other words, the purer the technology set from which a social platform is created, the sooner it will fail (if it gets off the ground at all). And for a platform to be "pure" it need not use only one technology set. It is sufficient for a "one technology to rule them all" architectural commitment.
20/ For example, a platform where "blockchain" is the only first class citizen, driving all architectural decisions without challenge or compromise, with the other 3 being second-class citizens that must conform to it, is as fragile as one that is pure blockchain. The "mixing" is shallow.
21/ The question then, becomes: how do you blend politically incompatible technologies architecturally? How do you avoid Frankensteinian inartistic lumpings on the one hand, and "one ring to rule them all" fatally compromised designs on the other?
22/ The first step is to invite and embrace "traffic jams" of technologies colliding. Technologies must be experimented with in ways that force mutual compromises and accommodations. For example, maybe you shove some deep learning into some blockchain by compromising both the "pure unsupervised" aspect of deep learning and the "zero trust" aspect of blockchains.
23/ I wrote about this a little in my April 2017 newsletter, The Anatomy of Future Fly. Note a slight difference: in that newsletter I defined the Big 4 as AI, VR/AR, IoT, and blockchain. Now I lump IoT with VR/AR, and break out AI into GOFAI 2.0 and deep learning.
24/ The trick is to apply Marshall McLuhan's idea of "medium is the message" in a balance-of-power way. A blend of all the Big 4 technologies is one in which "blockchain" is at once the message of the "deep learning" medium, and vice versa (and so on for all 4).
25/ This is not going to be easy. We're used to what you might call auteur-driven platform design. A single visionary seeing an elegant way to blend seemingly Fraknensteinian parts together into a beautiful whole. That's how we got, for instance, the iPhone.
26/ But to take that approach to the next-generation social platforms is to end up with purist Digital Marxism, Digital Fasicsm, Digital Anarchism, and Digital Libertarianism and a series of Platform World Wars. A dystopia of Platform Determinism. I think that's where China is headed.
27/ What we want is something that starts the process of platform evolution with a relatively chaotic collision of component technological parts with the right ideological proportions. The right "pH" and fragment-distribution in the primordial soup so to speak.
28/ Then, you take a somewhat laissez-faire approach to emergent, directed evolution that nevertheless produces harmonized, elegantly intertwingled organic realities with enough time. This should sound familiar: this is how modern cities emerged.
29/ City-flavored politics emerged from the older agrarian-flavored politics over the two centuries of political development in the industrial age through exactly such a process. Chaotic beginnings do not necessarily mean disorderly outcomes.
30/ The sociopolitical nature of New York, for example, is the result, most recently, of a collision of visions between Jane Jacobs and Robert Moses. This was not a purely sociopolitical outcome. Technologies with ideological grains, like skyscrapers and highways, played a huge role.
31/ San Francisco of course, is now famously a locus of colliding ideologies: techies, NIMBYs, YIMBYs, affordable-housing advocates, service workers, environmentalists. But because there's a "one ring to rule them all" condition (Proposition 13, a GOFAI-like sacred-cow rule in a rules engine), the system is unable to work itself into a new, healthier. equilibrium.
32/ With digital platforms, we are currently where we were with cities at around the end of the Napoleonic wars. Between 1812 and World War 2, a massive migration to cities, coupled with rapid architectural evolution from pre-modern cities, created the built environment where half the world lives today.
33/ Invariably, when you look at cities around the world, you can trace pathologies to insufficiently artistic political blending. Some cities are too pure-paradigm (industrial towns for example) to grow to metropolitan size. Others muddle their way through to metropolitan size, but once there, slight imbalances create impossible gridlocks, or "captures" by one group.
34/ The very best cities by some measure -- the ones that manage to keep evolving with the times as economic engines and great places to live -- are ones where no one politically opinionated technology (be it "highways" or "single-family housing") can prevail absolutely.
35/ Early digital technology platforms like MySpace were often so fragile they barely lasted a few years. Like supernovae, they grew big fast, but then collapsed onto themselves, and exploded, creating the raw material from which today's platforms were built.
36/ The current generation of platforms has figured out longevity. Google and Amazon are 20+ years old. Apple (as a platform induced by the iPhone), Facebook and Twitter are 12 years old. Long-lived, city-state-like platforms that keep changing with the times are here for good. These are going to evolve like rainforests, not fruit flies. We are entering a digital permaculture era.
37/ How should platform architects work to steward and evolve them? How should they create the right kind of balance of power between technological and human parts to maintain generative, pluralistic condition capable of recovering from big and small troubles, fires and floods?
38/ I don't have answers, but I do have calibration. Think not like architects, urban planners, or CEOs, but like mayors of city states. A good model who's been on my mind a lot lately is Havelock Vetinari, the Patrician of the city-state of Ankh-Morpork in the Discworld novels, who oversees the evolution of the city from medieval slum to industrializing world power. Here's a twitter thread I did about Vetinari.
39/ We're in this for the long haul. Like it or not, a massive human migration, comparable to urbanization, is underway. More humans are living out more of their lives on city-state-scale social platforms than every before. Not because we're being forced to, but because we like to, fake news and all. Waldenponding has always been a minority tendency, historically.
40/ Humans are not just a social species. We are a scalably social species. Our weird brains make us crave social experiences beyond just 150 companions in stunted-growth Dunbar communities like other apes. At some level, we are much Borg as we are Federation, Ferengi, or Romulan.
41/ We use social scaling technologies to explore connected existence at every scale from 150 to 7.5 billion, and inhabit social conditions at every possible social scale where stable configurations are achievable. If you think there's a "natural" right size of human sociability you're going to hate the next century.
42/ Some humans like small hunter-gatherer sized troops of 12. Others like tribal village realities of 150. Still others like dense, beehive like metropolitan environments. And still others like the idea of there being an emperor who can yell "off with his head!" on occasion. Like it or not. The answer to "what's the future going to to look like?" is "all of the above."
P.S. All 4 sessions of the Season 1 workshop are now online.
_Feel free to forward this newsletter on email and share it via the social media buttons below. You can check out the archives here. First-timers can subscribe to the newsletter here. More about me at venkateshrao.com. _You can follow me on Twitter @vgr
Copyright © 2019 Ribbonfarm Consulting, LLC, All rights reserved.