The Future of Work

Stephen McKeon is a partner at Collab+Currency

“You will probably work for a protocol someday.”

These are among the last words I say to students at the end of my course on the crypto economy. When I began teaching the course in 2017, it sounded absurd to many of them. Today, less so.

Crypto economies have already begun to shape the future of work. They blend how we play, learn, organize, socialize, and create, with ownership and income generation. In this sense, what is currently transpiring is much broader than work.

I’m simultaneously posting an article detailing the components that facilitate this new form of coordination, but in this piece I focus on forward looking applications that convert human labor to income within crypto economies.

45 years ago, economists Jensen and Meckling famously wrote:

It is important to recognize that most organizations are simply legal fictions which serve as a nexus for a set of contracting relationships among individuals.

The “nexus,” whether it be a corporation, LLC, non-profit, or government, has been required in order to coordinate all of these contracts, and in particular, operationalize the movement of resources between individuals, according to the contractual terms.

In the crypto economy, we have nexuses of smart contracts, where the movement of resources is enforced by code rather than an intermediary relying on a legal layer. What this means is that individuals are now able to collectively form complex contracting relationships in the absence of a legal fiction serving as the coordinator. That’s a game changer for human organization.

One mechanism to coalesce a nexus of smart contracts is a Decentralized Autonomous Organization (DAO). You can think of it like a shared treasury among the members, with hard coded rules that govern how funds can be disbursed. DAOs can be used for any number of collective pursuits, much like an LLC can be formed to pool resources for any number of objectives. Let’s compare and contrast these two organizational forms using the example of an investment club.

Say that 50 friends decide they want to build a shared collection of baseball cards. They form an LLC, open a bank account, each deposit $100, and buy baseball cards off eBay. Here are the logistics for each of those actions:

Form an LLC: Draft an operating agreement, distribute and collect signatures from all 50 members. Pay filing fees to the state. Ongoing annual reporting requirements.

Open a bank account: Additional paperwork, but more importantly someone needs to be designated as the signer on the account. This requires that all of the other members trust the signers.

Deposit funds: If members are located in multiple countries, there is some friction here. Someone needs to reconcile the membership list and deposits.

Buy assets: Requires participants to trust that the bank account signer will follow consensus. Internal controls are required for monitoring and enforcement of custody.

These issues become even more complicated when the group is 1,000 people. We rely on the legal layer to adjudicate disputes and produce trust in LLCs.

Compare the scenario above with Flamingo DAO, a group of 64 collectors, of which I am one, who pooled resources in a DAO to build a shared collection of NFTs. Here are the logistics:

Form a DAO: Mostly a function of replicating software. There are tools being built (e.g. Tribute DAO) that will make this easy for anyone to do with a few clicks. No filing fees since it exists only in the cloud. No annual reports since all activity is visible on-chain in real time.

Open a bank account: Unneeded. The DAO smart contract acts as a digital wallet.

Deposit funds: We all sent ETH to the DAO smart contract. It took 30 seconds for me to complete. Since the smart contract has hard coded rules, I have no concerns that it will be disbursed without consensus.

Buy assets: We vote on-chain using signatures from our personal digital wallets to reach hard consensus about what to buy. Once purchased, the assets are held by the DAO smart contract, so there are no concerns about one of the members running off with the assets. Most Flamingo members did not know each other when the DAO was formed, and this was not a problem.

Investment DAOs are just one example. DAOs can be used to coordinate almost any activity within digital environments and in time, many physical environments too. Today there are grants DAOs, doing the work of nonprofits and community development funds, protocol DAOs operating numerous decentralized finance applications, media DAOs driving the future of journalism and information dissemination, social DAOs turning social networks into economies, and service DAOs doing work of so many varieties that I can’t even adequately cover them here. DAOs are, or will be, the organizational form for many of the applications I describe in the remainder of the article.

DAOs have at least three distinct advantages over traditional organizations: (i) speed of capital formation (ii) voting is transparent and easily executed, and most importantly, (ii) they mitigate the trust requirement around jointly-owned assets.

I would also hasten to note that DAOs are not without challenges. Decentralized governance is hard. Free rider and blockholder problems continue to exist. I touch on these issues in the participate-to-earn section below.

It’s also worth noting that although the smart contracts relieve us from trusting each other, we now have to trust the code. Using battle tested code is one solution, but we also have an army of shadowy supercoders on our side. These individuals are a feature, not a bug. White hats like Samczsun, Immunofi, IntoTheBlock, and many others are serving as sentries, catching code vulnerabilities on our behalf before they are exploited. Compensation comes in both non-pecuniary reputational capital, but also explicit rewards like bug bounties. Any atomistic coder can join this army, no ID or resume required. They will be part of the future of work story.

The best deep dives on DAOs are Aaron Wright’s journal article and Coopahtroopa’s Mirror post, which points to many additional articles if you want to go further down the DAO rabbithole.

[ ]-to-earn models

Capital and labor are two well-known factors of production. Each can be deployed to earn income within crypto economies.

Capital-to-earn models have been around since the day Bitcoin was launched.

In proof-of-work protocols like Bitcoin and Ethereum, anyone can deploy capital to hardware and electricity to mine blocks, and begin earning.

In proof-of-stake protocols like Solana and Avalanche, anyone can deploy capital to the native asset, stake it, and begin earning.

More recently, there have been numerous options to engage in liquidity mining and yield farming, which involves deploying capital to, say, a lending protocol or decentralized trading venue, in order to begin earning.

These have historically been the dominant sources for earnings within the crypto economy and they will continue to grow. For example, staking rewards are positioned to become one of the largest categories of fixed-income instruments in the world. But that’s a story for another day.

Today, I’m focusing on labor-to-earn models, which are more fascinating because they leverage these new coordination mechanisms to convert human input to capital. This means that earning opportunities are available to a much wider swath of people. It is the democratization of economic empowerment on a scale we have not previously witnessed.

I will disclose up front that Collab+Currency has invested in many of the projects I mention below, so I am not an unbiased observer. Rather, we are severely biased. Our view is that these economies will drive the future of human interaction and we are investing in that future.


Axie Infinity is a play-to-earn game that has proven to be the breakout of the year, not just in gaming but in crypto generally.

The key feature is that the players own the game. They own the board (Lunacia), they own the characters (Axies), they own the ability to create new characters (breeding), and they own the in-game currencies (SPL and AXS). It is as much an economy as a game.


Players can earn the in-game currencies by playing, which is a model that has been around for years. However, unlike in-game currencies of the past, Axie’s in-game currencies can be exchanged for real world fiat currencies. Earning rates vary by skill and how much time you put in, but can reach hundreds of dollars per day for top players. In many parts of the world that dominates other earning opportunities.

But, you may ask, where do the earnings come from? And I may respond, where do any earnings come from?

The answer is applying capital and labor to generate economic activity.

Here’s a non-crypto example: Say I buy a cow and a bull. I apply labor to milk the cow and sell the milk to earn income. Later, I breed them. I’ve created more value (the calf). Perhaps someone wants to buy the calf for milking or breeding, so I sell it and earn a return on my investment of capital and labor. That’s an overly simplistic example, but it’s roughly the same thing with Axies. You buy them (capital), you play them to earn SLP from the game economy (labor) and then you breed them. SLP has value because it’s an input to breeding, and breeding has value because it generates a saleable good.

The other aspect to understand about play-to-earn models is that the assets are independent from the game. The players own the assets, not the game developer, which means that other game developers can build games utilizing the same assets. One day I’ll be able to play my Axies in all kinds of different games, further driving demand for Axies beyond the original gameplay.

One thing to understand about user growth for play-to-earn games like Axie is that, while people enjoy playing them, they don’t have to be more fun than games like Fortnite to attract new users. They only have to be more fun than work. That’s a pretty low bar for adoption. As a result, user growth is exploding.

Axie recently crossed 1.7M daily active users. That’s a lot for a crypto-based game, but it’s microscopic compared to the addressable market. If we focus on the countries with greater than 50% internet penetration, there are 1.9 billion internet users living in countries with an average income below $20/day.

The play-to-earn revolution is in its infancy and it will play a large part in the future of work for emerging and frontier markets. There are also avenues for expansion around the financial services being built on top.

For those interested in learning more, I’ll point you to Packy McCormick’s comprehensive deep dive.


The model for education has predominantly been pay-to-learn. Students, taxpayers, and/or donors bear the cost of education. This is even true of technical skills, for example, Microsoft charges you to get certified on their products.

RabbitHole is turning this model on it’s head, enabling learn-to-earn. Protocols sponsor quests, which new users can complete to earn a reward, paid by the protocol.

It would be somewhat analogous to Adobe giving you a few shares of stock for learning how to use Photoshop. However, these aren’t centralized software developers, they’re networks.

Networks become more valuable as more people use them, but potential users first need to learn how to use them. By rewarding users to go through the process of becoming informed, the networks get user acquisition, which enhances value, while the users learn a new skill and earn ownership in the protocol for doing so. It’s a win-win for both parties.

Aside from the direct payment for learning how to use the protocol, you also build your on-chain resume. Since all activity on-chain is visible, you can reveal evidence of what you’ve done, and when. Rabbithole translates this activity into experience points (XP), which determine your RabbitHole “level.”


When we hired our last intern, I asked him to send me his wallet address in advance of the interview. By reviewing what he had done on-chain, I knew that we were probably going to hire him before the interview even started. Next time, maybe I’ll just ask for the candidate’s RabbitHole level.


The impact of endowing property rights on digital assets has been historic for creators. We’ve entered a renaissance period in terms of the ability to earn income through creative labor.

Here’s an anecdote to illustrate how fast this vertical is scaling. When Derek and I wrote *You’re Sleeping on Crypto Art *in Sept. 2020, not a single cryptoartist had crossed $100k in sales. Yet, we hypothesized: “The advent of millionaire digital artists is within view.”

Six months later, Beeple’s piece sold for $69M at Christie’s.

Now, million-dollar sales are commonplace — every recent drop on the Art Blocks curated section has made the artist a multi-millionaire. We may be talking about billionaire digital artists someday.

This same theme is playing out in music, photography, essays and many other types of creative digital works. There is no dispute that if you can create desirable digital content, then you can earn a living in the cryptoeconomy.

Further, many of these platforms are permissionless, meaning you can start today. On Audius you can join and upload music tracks right now. You can curate playlists right now. If your work ends up getting traction, you’ll start earning AUDIO tokens.

Derek and I have written several deep dives on these topics. In addition to the cryptoart article linked above, you can check out our posts on Audius and blockchain-based generative media.


This category is driven by the difference between hard and soft information. Hard information is quantifiable and objective, like: what was the final score of the Knicks game. Oracles pass hard information to smart contracts for execution. Obviously, we need to trust the data source, but integrating hard information into smart contracts is relatively straightforward.

Soft information requires judgement. It is qualitative rather than quantitative. Consider someone’s participation in a Discord channel. The number of times they posted is hard information. Whether those posts added any value to the conversation is soft information.

As organizations migrate towards becoming a nexus of smart contracts, we will need to inject soft information into the contracting environment. Terms like “best effort” are both common in contracts and soft in nature. Who shall determine best effort?

We will see the rise of “subjective oracles’’ to address this issue. People will be able to serve as judges via decentralized protocols, and will be paid for their input. They will build a reputation, which will impact their earnings and future opportunities. Wikipedia has illustrated that people are willing to contribute, judge and curate for reputation alone. Adding economic incentives to this model will supercharge it.

Some applications of judge-to-earn:

(i) Decentralized justice: Early efforts can be found within UMA, Aragon Court, and Kleros.

(ii) Subjective appraisal: Upshot is working on peer-prediction models for valuing NFTs.

(iii) Curation: SuperRare and JPG are working on decentralized curation for cryptoart. I wouldn’t be surprised to see decentralized curation expand to other areas that utilize a subjective mechanism to screen on quality, such as journalism.

(iv) Credit: Historically, banks have been the arbiters of credit decisions. Loan officers synthesize hard and soft information about the borrower to arrive at a lending decision. DeFi has already solved the collateralized lending market, but undercollateralized lending is a tougher nut to crack. We need on-chain measures to assess probability of repayment, and act as deterrents to default. Some of this will be automated, but I suspect that we’ll also see the rise of decentralized loan officers. Individuals will be empowered by holders of capital to synthesize soft information about potential borrowers. Smart contracts will retain dominion over the assets so that the individuals acting as DeFi loan officers (i) do not take custody of the assets, and (ii) retain skin in the game while the loan is active, which exposes their cut to loan default.


Decentralized organizations are going to need many types of human participation and we should expect that participants will be compensated. They need community managers, governance participants, regulatory advocates, operational support, and of course developers. Instead of negotiating a salary with a superior, you’ll more likely put a token grant request through the governance process with the community.

The governance process itself is going to require a lot of human input and there’s still plenty to figure out here. It bears repeating: decentralized governance is hard. Most governance structures across all types of organizations use representative governance. Corporate boards, school boards, and Congress are all examples. Years ago, when these structures were developed, coordinating votes among all constituents on every issue was logistically infeasible. Crypto solves these logistical challenges — voting via a digital wallet is quick and easy.

However, it’s not solely a coordination problem, it’s also about the cost of informed participation. Most citizens don’t want to vote on every bill before Congress and most shareholders don’t want to research and weigh in on every board decision. So we elect people as our representatives with the expectation that they will bear the cost of becoming informed, and we compensate them for doing so. But electing representatives as our agents gives rise to agency problems, which lead to all sorts of suboptimal outcomes.

It’s possible that crypto governance will dispense with the concept of elections and fixed terms in favor of dynamic proxy. In other words, I can assign my voting rights to a delegate who I believe represents my interests, but I’m free to reassign those rights to another delegate (or myself) at any time. Logistically, this is pretty easy in crypto networks. It remains to be seen whether this feature mitigates agency problems.

These delegates, or protocol politicians, will eventually be a meaningful segment of the participate-to-earn opportunity set. How they will be compensated is a topic of ongoing debate.

Today, they are generally not compensated explicitly, for example, a16z states that the only payment for their delegates is expense reimbursement. The accumulation of reputational capital by serving as a delegate may be compensation enough, if it leads to additional opportunities to earn ex post. This is not unlike other types of political appointments. Direct pay is relatively small compared to earning opportunities after serving.

However, I personally believe that delegates should be compensated above and beyond reputation. One avenue is for the protocols to compensate delegates directly (i.e. governance mining), similar to the compensation paid to others who participate in the network such as validators or liquidity providers. It’s early days yet for these models and the incentives will need to be fine tuned, but once they’re deployed at scale it’s potentially a huge unlock for participate-to-earn.

Crypto networks facilitate new forms of human coordination and set the stage for the future of work. It’s all being developed right now, in real time, and you’re invited.

Your digital wallet is the passport to this land of opportunity.

Join us.

A special thanks to Morgan Housel, Dan McKeon, Michelle McKeon, and Derek Edws for their feedback and insights during the construction of this piece.

Disclosure: Collab+Currency is an investor in many of the projects mentioned above, including Sky Mavis (Axie Infinity), Art Blocks, Audius, SuperRare, Ethereum, UMA, and FWB.

You can find more thoughts from our team on Twitter: @Collab_Currency