TL;DR - skip down to Current Intermediate Implementations for what I’m looking at in crypto-based economics, the preceding passages are for historical context & defining firms, markets, & estates.

Earlier this year, I wrote some thoughts down around the current trends of artificial intelligence & geopolitics.

Last year, I wrote some thoughts around decentralized economics & the possible peer-to-peer approaches that could manifest between independent firms.

While these perspectives were broad and strictly theoretical, I’ve been thinking about how to compound these two topics into a comprehensive philosophy for trends that can be developed sooner for a general economic benefit. One could argue that many of these ideas are easier said than done because they are not idealogically impossible, but operationally infeasible. I would tend to agree, and the reason that I feel like writing this is more my own benefit in determining the next step to engineer for what is a wholly battletested, coherent North Star.

Walmart, Sysco, & McDonalds

In many conversations over many years, a common business model has permeated our memes. I’m confident many readers have seen the “capitulated McDonalds employee” meme. In the United States, Walmart is almost the consumer’s version of that.

Walmart is an interesting outcome for the “American consumer”, but when you look at the context, it’s hard to not see how it’s inevitable. Throughout the Western hemisphere (as well as some modernized areas globally), there is a heavy sense of desolation in wide areas. The town center is the chosen spot for socialization, the lack of infrastructure suppresses job availability & employment, and culturally there’s little reason for mass influx. However, Wal-Mart seems to have optimized placement of its locations within this desolation, and the reason is simple: it works on top of an interstate highway network that already exists. The same applies to McDonald’s, which many believe to be the widely available bottom rung of national employment. It’s ubiquitous because McDonald’s optimized the quickest roadside restaurant & the most effective franchise structure for an effectively global car-centric culture. It may be a meme but the business model has been copied many times over, as well as form the basis of the gig economy.

Why group Sysco with Walmart & McDonalds? They’re not B2C, and they depend on more than just highway trucking. There is a reason: all three corporations conform to Ronald Coase’s “Nature of the Firm”, which states that a business firm will grow to internalize many classes of cost in order to minimize the sum of those costs in production. All three firms have already grown to an almost monopolistic state in part because they could benefit from the Federal-Aid Highway Act of 1956 and progressively stable global trade routes. In both cases the nation-state assumes some infrastructural costs but the subordinate firms are free to capitalize the benefits. I might consider myself a libertarian, but that’s a pretty compelling reason to pay some taxes. Coase also explained overhead limits to the size of the firm, which also apply to taxes, but we’ll revisit that later.

There’s also an interesting sociological cost that decreases over time: the nuisance of commuting. From walking to horseback, from cars to bullet-trains, we tend to budget for a constant commute time. This otherwise described as Marchetti’s constant, typically one hour daily or half-hour for a one-way trip. The progressive implication is that lifestyles change to include more range over the same time expense. If Thanos snapped his fingers and manifested a high-speed rail system within the walking distance vicinity of every population center on the continent, Wal-Mart & McDonalds would accordingly downsize some of their locations. Also, a lot more jobs would appear & get filled. Which means a lot more downstream business cases would suddenly become viable. In a similar sense, the online user experience saves time, which increases the potential downstream business and decreases the burden of distribution. Which, of course, we’ll revisit.

White, Gray, & Black

Everyone’s heard the term “black market”, but only a small fraction will have heard of “gray markets”. They both follow the same principle: policing costs cannot exceed the total global production of capital. Of course, the more the good or service is more ethically castigated, the further policing costs will be allowed & spent. In Gray Markets: Prevention, Detection, & Litigation (which I highly recommend), several black markets are described: street drugs, human exploitation, wire fraud, you get the picture. However, the tolerated policing cost drops outside of the obvious examples. So when there’s arbitrage of a tariffed good, or unofficial reselling of products, intended for end users, to other resellers, there’s quite a bit of volume that can’t be policed. In fact, unregulated financial markets might trade instruments like commodity futures in gray markets.

This is not to say that black markets are incapable of operating despite the policing. The blackpill is that black markets are better at self-regulating despite liabilities like their ethical nature. The more voluminous the economy, the less effective a flat policing cost.

There’s the historical examples of organized crime thriving in more severely-policed states. Since the advent of legalized marijuana, one of the prevailing arguments kind of implies that overregulation, by driving centralizing growth into a black market & increasing the efflux into the prison-industrial complex, is its own form of racketeering. Coase might just disagree and claim that this is overhead cost of a firm that’s overgrown.

Perhaps criminals & police are just opposite nodes of a generative adversarial network. Traffickers learn how to smuggle in one way, the police soon find a way to detect that traffic. An area experiences more B&E, and potential victims buy better locks & surveillance. In that lens we might see one of the most important facets of the black market: what kind of self-government is most selected for by the market. And these opposing groups are overlapping; a police officer in many countries can be quite corrupt as a societal norm, the more organized the crime & the weaker the rule of law. Zooming back out, we can see that the economy contain certain motives & certain costs. As Charlie Munger would put it, “show me the incentives and I’ll show you the outcome”. It should come as no surprise that the economy expands to join supply & demand despite societal & cultural costs.

This is the focal balance of intermediation: an entity or element can internalize production & grow to meet the consumer base of any market. However, there will be an increasing limiter the larger the firm, and there will be adversarial firms that can grow to adapt to the circumstances. For both, if global society has already socialized some cost, both firms will autonomously grow as much as the cost, not the culture norm, justifies. If there’s some improved method of production or some reduced benefit for criminal activity, the market broadly lowers the societal cost. In a world that’s transitioning into a gig economy, this relationship determines the efficacy & network effect of that aggregate market.

Uber, AirBnB, & Amazon

How could I not mention the Internet by this point? One the biggest drug markets are so-called darknet marketplaces, and Walmart, Sysco, & McDonalds all use web apps to coordinate orders. In “Optimal Economics” I mention the ecological metaphor of a large ziggurat covered in more superficial “pyramids” & “obelisks”. Starting from the bottom, all the firms mentioned above rely on globally maintained infrastructure for the physical layer of internet traffic. However, they diverge at the higher digital layer (the network).

Darknets take advantage of onion routing (among other technologies), while the megacorporations might rely on cloud-based VPNs (among other technologies). There are additional trusted intermediaries the further a given firm operates within the scope of legality. For example, cloud service providers will be more hardened against DDoS, malicious or benevolent, than a random dedicated server hosting an illicit web business.

A low-effort criticism following the dotcom bubble: “Why spend money to develop a business online? It’s just speculative investment in a skeuomorphic business model. There’s no fundamental gain of function”. The answers to this take considerably more effort and more engineering, but there’s no question that the speculative bubble preceded a massive economic boom. More precisely, a lot of brick&mortar businesses are outcompeted by more nimble online firms with fewer total costs.

Amazon, for example, sold durable goods like books, and it’s no secret that brick&mortar equivalents like Borders and Barnes & Noble lost market share. Most durable goods can be outsold by online retailers, and so the trend has progressed. But what about services like transportation? It’s not quite as clear that Uber is the be-all, end-all, but something definitely happened to the taxi medallion scheme.

What about socioeconomic classism like land ownership? Do the AirBnBs and WeWorks respectively disrupt residential & commercial landlords? Probably not in their current forms, but there also might be a reason why they can’t, and that requires further critical analysis. Suppose we expand the definition of racketeering to non-explicit intermediation to solve some contrived problem. A homeowner has liabilities like bargaining costs, maintenance, and cultural minefields, and Airbnb pretty much confirms this in their feature page. There’s also the question of financing the firm, the sustainable losses, and perception of legitimacy & success. It’s a long story beyond the beginning of the Industrial Era corporation, yet we can see that a global public good like the Internet unlocks some economics, reveals some rackets, and advertises a competitive opportunity to more efficiently intermediate certain facets of a society with some technologically novel mechanism.

The Fifth Estate

There are three political estates in classic history: the clergy, the nobility, and the rest of society. The idea is simple: the clergy dictate a spiritual mandate, the nobility dictate prestige, wealth, and titles, and the rest of society cooperates to support a relatively balanced state. However, this tends to work best when the flow of information is restricted to certain titles. When the printing press forces the Reformation, there’s a fractious tapestry of slightly different spiritual mandates, the commoners can afford their own dictation of prestige, and the nobility has a bunch of war chests to resolve this paradigm shift and protect their wealth & titles.

The Industrial Revolution is effectively the end of the classic three political states. The “corporation” allows for further class subdivision, like the supervisor or union of the third estate, or the LLC, trust, or PAC of the second estate. Moreover, in both democratic republics and parliamentary monarchies, industrialization & corporations take control of the mass flow of information. By the early 20th century, a “fourth estate”, commonly seen as printed, newspaper-based journalism, becomes the much-needed check on regulatory capture, a form of special interest intermediation into the government of a formerly competitive market.

It should be emphasized at this point that the fourth estate is not necessarily incorruptible or effective. While Ida Tarbell revealed the anticompetitive practices of Standard Oil, there was also an endemic torrent of low-quality newspapers. It should also be noted that the fourth estate is not owned & operated by some non-corporate aspect of society; it would be more accurate to describe the fourth estate as the arena over which all corporations contend for the greatest flow of information.

The “fourth estate” includes other mass media like radio, television, and social media. Others have argued that these have become the modern equivalent of a religion, and so the concept of the “first estate” has been replaced somewhat by pop culture broadcasted through mass media. One can quote Nietzsche: “God is dead!” Coase distinguishes between the firm and the market as separate coordination mechanisms; the firm coordinates how to produce best for maximal capital gain, and the market coordinates optimal exchange of capital for maximal growth. With respect to the concept of religious mass media, one can assert that the market can replace firms (a given denomination) with diminishing societal upheaval (i.e. socialized capital loss) over time.

The Internet, and more broadly the Information Age, is a marketplace of non-scarce ideas. It’s uncannily cheap to consume information. There’s quite a bit to unpack: the “first estate” is subsumed by it, the “second estate” hosts it, the “third estate” can consume it, the “fourth estate” antagonizes it… what’s going on?

Starting with unlicensed pirate radio, the “marketplace of ideas” begins to populate a public domain without any form of corporate intermediation. Television is not completely disintermediated, but public-access television becomes more common over time. A casual observer of mass media can see an originally centralized mass media effectively transform into a self-regulating gray market. By the early 2010’s, this gray market becomes so integral to society that it turns into a deadman’s switch: if a given second estate turns off the market to suppress the fourth estate, the third estate riots & the world finds out. If any political estate attempts to subvert the political discourse, the second & third estates revoke their financial and/or ideological patronage.

The Internet hasn’t provided the third estate with a completely disintermediated market of ideas, but it has mandated some form of senescence. One counterargument to this is the “lindyness” of a trend/product, but look at how lindy brands like The New York Times actually are. It’s only existed for the last century of the printing press trend, yet only two decades of the Information Age has completely catabolized the revenue model that’s sustained it. One could state that the “fifth estate” began with the repurposing of the first & fourth estates into a self-cannibalizing, incorporate mass forum.

Now, I’ve heard many terms around crypto, web3, this or that, but all of those started a while ago. In 1995, Phil Zimmerman published the source code for PGP, in part because it had been overregulated as a no-export munition. In 1992, Cynthia Dwork & Moni Naor published a Proof of Work to combat spam. These both lead to opensourced forms of regulatory code with no trusted third party. While the fifth estate could be seen as just a replacement of the fourth estate, we also know it as a gray market of ideation with auxiliary, self-evident property rights. In this respect, while 90% of the internet can be derided as porn, spam, and trolls, there is a dense core of self-regulated digital state with senescent prestige & rights. In other words, the fifth estate not only inherited speech from the fourth estate, it has also inherited money, prestige, and rights from the second estate.

Since Bitcoin, the goal has been to refactor intermediate economics with a balance between five modern estates. In other words, buzzwords like “web3” point to a modern interpretation of Coase: the market can sustain a gray, supermodular economy of operational functions, and firms must internalize the most efficient executive stack of these in order to compete. The key design parameter is that all of these firms can share executive & operational circuitry as much as the current corporate structure shares socialized physical infrastructure. It’s unfortunate that the focus has remained on solving the pitfalls of an incorporated internet, just like it’s unfortunate that 90% of the fourth estate, the Internet, and crypto has been a deluge of grift. But this does not disqualify the paradigm before its full implementation nor does it qualify full disintermediation after the fact.

Current Intermediate Implementations

The key focus is the development of modules that support any third party in performing a given process of production. In the case of Walmart, Amazon, AirBnB, & WeWork, one would have to design:

• A socially recoverable reputation system for onboarding & offboarding workplace operators, long-distance freight transporters, vehicle ownership, and landownership within a logistical network. Gitcoin Passport can be a recursive, public VC, EIP-4337 & Gnosis Safe wallets can be socially recoverable (and recursive). Protocols like UniRep & Sismo are potentially anonymous methods of maintaining reputation without an unnecessarily invasive form of social credit. TaterDAO, a subDAO of LexDAO, might be a suitable intermediate owner of real-world assets. Perhaps DAO delegation in voting systems will also lead to reputable “trustees” as appropriate operators of real-world ownership & feedback. It is important to note that markets like real estate brokerage fail without area-specific domain expertise of how the market is priced.

• A generalized tamper-evident, sufficiently failsafed architecture for physical security. So far I’ve seen physical primitives for tokens & cash. I’ve also seen proof of concept for physical gating with EIP-712.

This is a promising start, but let’s consider the real use-case in demand: locker hubs for temporary storage. This is direly needed in low-security areas like homeless shelters, and it’s also needed in high-security areas like workplace equipment storage. This becomes more salient as real-world “places of employment” become more seamlessly permissioned. The focal precursor of “permissionless contribution” is the inherent reduction of supervisory costs in a firm. However, physical security needs a minimum degree of supervision as a failsafe for physical failure. Ultimately, this means that the reputation system needs to temporarily permission a “guardian” to bypass some security, and it also means that the locker needs to log various degrees of acceptable deposit & withdrawal. In other use cases, borrowed equipment like rented vehicles needs to log various forms of usage metadata. This also raises the issue of how much finality & surveillance is necessary (imho, there’s no easy answer).

• In the case of last mile businesses like Uber, Amazon, & Sysco delivery (among many others), there needs to be a sovereign, compartmentalized database of critical information. In the case of Uber, a reputable user still needs to hide their destination & other personal details or risk denial of service. Moreover, both the driver and the user need to rely on a censorship-resistant mapping & tracking in their respective mobile apps. A delivery of nominal goods, even on the last mile, can be excessively valuable cargo. There needs to be a layer of privacy even between the delivery operator and their cargo.

• All firms require a socialized bargaining resource that is some maximized combination of neutrality & nuance. One can argue that many public infrastructure projects have languished in part because the manner of execution is a point of contention between several self-interested factions: corporate, labor, and land ownership. Again, this may require intensive development of an anonymous reputation system, simply because neutral bargaining requires some check on retaliation, as well as subversion. Neither the second nor fourth political estates are a sufficient source of arbitration.

• Speaking of retaliation & subversion, new firms need protection from anti-competitive corporate practices. Legal intermediation is a necessary evil in this context, given the propensity of liability & litigation in today’s form of the second estate. However, it is not a mandate for broad implementation for all forms of corporations and all political estates. One aspect of NFTs is their potential for many object-oriented paradigms besides p2p collectibles; in the same respect DAOs have a potential for autonomous operations that fall outside of corporate law. While there may be many corporations LARPing as “limited-liability DAOs”, there are also superjudicial DAOs that need to compartmentalize separate operations to subordinate, limited-liability corporations.

• Firms that depend on labor costs often require continuity of nominal inputs, and they also might depend on continuity of nominal points of sale. In both respects there can be a standard of forming a corporation with allocable equity, as well forming a standard, continuous point of sale. We still haven’t found out whether crypto businesses can afford continuity of payment processing offchain. It is interesting that legal formation of for-profit & nonprofit firms can be so direct & seamless, and yet we are so early in testing what the limiters of those firms really are. This is where gig apps need to adapt as marketplaces of laborers & end consumers (which may not be sustainable w/o sweat equity).

• Firms also depend on continuity of infrastructure, but SMBs & consumers are more taxed than necessary for second estate fiscal policy to be less optimized than necessary. Though we’re too early to experience this tide shift, I suspect that one of the main “network states” will be a global DAO that strictly maintains public goods like infrastructure, and many firms around the globe will internalize that membership cost as well as the lobbying cost for reducing taxes globally. This also plays into the policing costs that maintain continuity of globalization. Additionally, this may be an entirely novel monetary policy similar to seigniorage, instead of an explicitly itemized cost.

These are just a few thoughts about where my focus is in 2023. It feels like we have a lot of real-world applications around the corner, but many of these are not just “X, but decentralized” business models. Intermediation is a nuanced, essential function in modern economics, but it is known to be an imperfect function in a society that requires checks & balances. To complement the ziggurat, the best forms of intermediation are specified as open-source standards for many firms to implement on their own. The major hurdle in 2023 & onwards is implementing more effective DAOs to internalize “the nature of the firm” & appropriately compartmentalize costs in already decentralized, autonomous markets.