Mayors with Boxing Gloves

Startup Cities will need basics like CSAT scores, vertical integration, and UX Research – not just sexy new tech

You’re reading Startup Cities, a newsletter about startups that build neighborhoods and cities.

This week:

• Markets vs. Liquid Democracy

• Mayor as Court Jester

• Do Customers Care About Transparency?

• DAOs: Cyberkluge of the Future

---

Startup Cities Need More than Sexy Tech

Urbanist Joni Baboci returns with another home-run essay: “Governing Futures”.

Baboci looks at how trends in technology will (and have) changed how people connect: the rise of maker-spaces, new social media like Clubhouse, and, of course, crypto.

Baboci’s vision is of cities with many layers:

An individual could belong to hundreds of community organizations that could interlock, connect and be dependent on one-another without the fuss and bureaucracy of the current system. [ZC: Big Bruno Frey vibes here!]

Baboci’s vision channels an older urbanism, when “little platoons” defined community. Maybe 21st century technology takes us back to the patchwork social order of 19th century cities? DAOs become today’s mutual-aid societies?

Source: https://triennale.org/en/magazine/governing-futures

I love the spirit of this. One of the reasons why Startups Should Build Cities is so we can see how successful companies weave digital innovation into the fabric of community. Startups are the ideal vehicles to carry tech innovation to market.

But, despite my enthusiasm for technology, I think we’ll discover that digital innovations do not solve the fundamental problems of today’s cities.

I don’t think Joni believes that technology will solve all ills, so this is a bit of a strawman— sorry Joni! But there definitely are technologists out there who believe that we’re one vote-tallying app away from Utopia. I think this is wrong.

Today’s Cities Lack Fundamentals

We live in the IBM Mainframe era of the city-building industry. Honestly, it’s worse than IBM Mainframes — at least IBM feared competition and tried to deliver value for their customers. Many modern cities seem to lack anything even resembling a customer mindset: some city leaders actually make fun of struggling customers.

To fix this industry we need startups. These startups must pursue industry-disrupting ideas like unified land ownership (rather than subdivision), for-profit management, radically different urban designs, new business models, vertical integration (or vertical disintegration like Sandy Springs), and product strategies like proposition or price simplification.

Startup Cities need meat and potatoes startup stuff: mainstays like UX research, A/B testing, unbundled pricing, customer lifecycle analysis, and the drive to grow at venture-backed speeds. These aren’t sexy. All the sexy tech in the world won’t change the need for fundamentals. As with Yoga class, the core must be strong.

Choosing Experts with Liquid Democracy

Baboci points to liquid democracy. Liquid democracy allows a voter to “delegate her voting stake to an expert of the domain on which a decision needs to be taken”.

Such tech might make city democracy “less about multi-year representation” and more about “individual participation” and “temporary short-term delegation”.

Vote delegation looks like this:

If the city needs to decide on the morphology of a new art gallery, a cultural operator might be better placed to make that decision...

So the voter delegates. Cool! Digital tech can make the process easy and transparent.

... But wait a minute! We’re voting on the “morphology of a new art gallery”? This sounds like a digital version of the same localism that breeds NIMBYs and blocks building in today’s cities.

It’s not obvious that “more voting” equals better decisions or even better democracy. One version of the Great Stagnation argues that the rise of “tiny democracies” like environmental review boards in the 1970’s caused NIMBYism in the U.S.

Voting is not a panacea: voting systems have a wide and disturbing set of failure modes. We shouldn’t expand voting onto more areas of life without a good reason.

Liquid democracy thrives on thoughtful delegation of votes. Voters delegate to experts. But the kind of person who’s thoughtful about vote delegation is probably thoughtful about their vote already. In Google’s internal experiments with liquid democracy, organizers found only 3.6% of people delegated votes.

The voting populace of this experiment probably respects expertise more than average. These voters passed Google’s notoriously difficult screening for intelligence and conscientiousness. They’re “above average”. And they still preferred to cast their own vote.

Mayor as Court Jester

Let’s assume people do delegate. Why would we assume that the median voter cares at all about expertise?

We don’t see expertise rise to the top in other voting systems. What protects liquid democracy from the demagoguery and other madness that plague today’s voting systems? One can imagine the Maduro of “cultural operators” amassing delegated votes through inflammatory bullsh*t.

Cities are not immune to gamesmanship. My current “favorite” mayor is Neto Bran from Mixco, Guatemala. He’s adored by voters. And not for his expertise.

Bran’s influence with voters grew because he flew into campaign events in an Iron Man costume, handed out Christmas events while strapped with a handgun, and fought a political rival in a boxing ring.

Mayor Neto Bran demonstrating his expertise to voters

(I admit the boxing thing has an appeal. Trump/Biden Royal Rumble 2024?)

Does liquid democracy delegate to expertise or to people with the strongest opinions or the most charisma? The Google paper takes pains to highlight 14 voters who delegated to a Vegan employee. The employee promoted themselves on Google+. Was this an “expert”? Or just someone who feels strongly about Veganism and made a lot of noise?

Neto Bran is a walking embodiment of why I’m so hot under the collar about startups that build cities. Entrepreneurial management is a different animal than traditional political management. Mayor-as-Court-Jester is clearly a valid strategy for hacking voters. But CEO-as-Court-Jester is not a competitive business practice. It can happen. But that company won’t last long.

Skin in the Game

Finally, why would we assume that the median voter is good at evaluating expertise? It’s hard to know who’s smart — especially in arcane or technical domains. It’s especially hard to know who has both good knowledge and good intentions.

In traditional voting, voters often solve this problem through group affiliation. Defenders of traditional voting argue that voters “rationally” use party affiliation as a way to economize information. This heuristic — voting with your party — does not honor expertise at all. It’s a blind groupishness. Why would this be different in liquid democracy?

The use of public funds for a rivalrous, excludable good like an art gallery invites conflict. Everyone will disagree on its value and how it should look. Voting just magnifies the disagreement. It gives veto power and therefore invites conflict.

Let’s say we don’t use public funds. If a non-profit foundation or a company wants to build an art gallery, why are we voting on it? Aren’t the people in these organizations the best ones to make the decision? They risk their own capital and reputation.

Baboci cites Skin in the Game by Nassim Taleb. Why isn’t this city-scale liquid democracy just a fancy way to encourage people with no (or superficial) skin-in-the-game to meddle in the construction of an art gallery by people who do have skin-in-the-game?

Choosing Experts Without Knowing It

There is a kernel of truth in liquid democracy. Expertise matters. Delegating to expertise is often a Good Thing. But a lack of liquid democracy doesn’t exclude delegation.

Liquid democracy operates via conscious delegation. We consciously choose an “expert”. But delegation to expertise already exists. When I buy from a company I unconsciously delegate to an army of experts.

Think of all the expertise embodied in a car, a MacBook, or even (as the libertarians like to say) the humble pencil. Sit in a Toyota and rejoice: you just delegated to 10,000 Japanese engineers.

Markets are the OG way we delegate to expertise. Why aren’t vibrant markets superior to liquid democracy?

Is this vision so unreasonable? In the pre-1971-stagnate-everything world, we just didn’t vote on a lot of things. This wasn’t a rejection of democracy. We just let things happen.

Cities like Tokyo or Taipei don’t have such constant voting on changes to their city. Were pre-70’s American cities really that much worse? Are Tokyo or Taipei failed cities? Are they undemocratic because they vote on fewer things?

Customers Don’t Want Transparency

Baboci also argues that blockchain technologies like DAOs create new frontiers for transparency:

The history of decisions, policies, votes, and results would be embedded on-chain forever and could be consulted by all citizens without fear of manipulation.

Transparency seems like a good thing. Back in 2013, I proposed a DAO-like idea called Munibit, meant to bring radical transparency to municipal finance. But I eventually gave up on it. Why?

Well, honestly, because I didn’t have enough money to build it! But also because I discovered that most people — and specifically users/customers — don’t care.

Data sharing and public sector transparency are hobby horses for tech nerds. Open source programmers make herculean efforts to digitize and publish data. There are even well-funded startups like OpenGov. These people are great. And the data helps researchers, which is good in itself.

But is availability of data for citizens really the constraint on city performance?

Once more, a product perspective helps. In general, mature products show less data to users, not more. There’s a whole concept in UI design — progressive disclosure — to describe how and when information should be shared. The default path for most users is to show as little information as possible.

What if a mature governing structure for cities looks more like a mature information product, where complexity and decisions are deliberately hidden from the user? This is edgy stuff. It seems to contradict proper thinking about democracy.

But why would the information processing system called city governance break with the pattern we see in all other information processing systems? Everywhere else in life, users want good defaults and simple choices. Users want to think less, not more. Under banners like “political ignorance”, academics love to ridicule how little the electorate wants to think.

I don’t get this. In every other industry we know that making users think less is a sign of good product design. Are we sure this is a failure of the populace and not a failure of city management? Well-designed products hide the gnarly details so the user can focus on what’s important. Are city services an exception?

One version of a Startup Cities future has people think about the data and decisions of their city less. Not due to some rejection of democracy, but because competent professionals solve problems for them. People are free to focus on their lives because cities treat them as valued customers.

Automation with DAOs: Glorious Cyberkluge of the Future

Finally, Baboci makes a bull case for DAOs:

Smart contracts are able to store, verify and self-execute rules agreed upon by parties independently and without outside interference.

As I wrote previously, I’m enthusiastic about how DAOs can aggregate capital around niche interests. But I’m an open-minded skeptic (please change my mind) of fully autonomous coordination with DAOs — at least for any non-trivial organization or problem space.

As with liquid democracy, DAO voting is not a universal solution to decision-making. Constant voting sounds like a failure state, not a feature, of a DAO.

I’m reminded of hanging out at Occupy Wall Street in 2011. Voting was so lionized that Occupiers burned hours each day arguing and voting. The voting became an end in itself, almost a religious ritual. Anything that resembled progress ground to a halt.

There’s a precedent for DAOs: publicly traded corporations. Shareholders vote on some issues, but mostly they don’t. Management makes decisions. People are cool with this, as evidenced by the ubiquity and value of publicly traded companies.

My current view is that successful DAOs will look a lot like public traded companies. They’ll be associated with something that looks like a traditional company, with traditional management. They’ll delegate to said management. They’ll vote on a subset of issues or a subset of budget. Or their votes will be more advisory than binding (a sort of internal futarchy). They’ll enjoy lower transaction costs enabled by the DAO when they need to vote.

Beyond voting, I suspect making DAOs truly autonomous will be hard for the same reason that self-driving is hard, AI is hard, and programming all the relevant edge cases in a software system is hard. In short: reality has a lot of edge cases.

My mentor in college, the irreverent economist Mario Rizzo, once told me: “Given enough people and enough time, everything stupid happens.”

Automation lives this maxim: I learned it the hard way as an early engineer at a fancy no-code/low-code automation startup. It’s really hard to make a tool that’s both general enough to capture a diversity of workflows and reliable enough to actually work. People do every kind of stupid thing. You can’t anticipate all the failure modes.

My favorite story of “everything stupid happens” is about an engineer mystified that his self-driving car prototype kept slamming on its brakes. The culprit: butterflies. The car treated butterflies as an obstacle that the car should not kill.

Within the hermetically sealed logic-world of Smart Contracts, things are “easy”. But once you introduce reality with external Oracles, you open Pandora’s Box. The state space explodes. (Seriously, could someone who knows Solidity well pair program with me one of these days?).

Markets tell us that automation is hard: Zapier — the digital equivalent of a cabinet full of little power adaptors — is a billion dollar company.

Tech-enabled delegation, transparency, and automation are great. But cities of the future will need more than digital bits to build the future.

---

Joni Baboci’s essay mentions Rose Island on Netflix, a hilarious (and tragic) movie about an eccentric Italian engineer that built a floating sea colony/micronation/swinger commune in the 60’s.

I recommend the movie. It’s fun. But it also shows the deep resistance by states to anything that deviates from their institutional monoculture.

Am I too bearish on the potential for these digital innovations? Tell me in the comments👇

... And don’t forget: Startup Should Build Cities!

---

P.S. Next week I’ll start to share some spicy conversations with innovators in the Startup Cities space. I focus on innovation enabled by Startup Cities and innovations that can enable startups to build cities. The first interview is with Sean Pawley, founder of Seshat Bank. Stay tuned!