Internet Capital Markets are heating up. The dream of funding internet-native businesses is back!

But here’s the trillion dollar question:

Once the project is funded, what is the best way to apply the capital toward achieving its mission?

Well, if you want a typical startup risk-return profile, you'd probably plow the funds into the founders’ bank account or multisig and hope for the best.

But as we’ve seen over and over again, that doesn’t usually work out too well…

On the other hand, if you are one of the few decentralization purists left, you might try starting a DAO and giving real smart contract-enforced ownership of the project to tokenholders.

And inevitably everyone would look at you like you have three heads because we all know that "DAOs don’t work."

...but could they?

We’ve been grinding for years to figure this out. We think there’s one key insight that might help us all get to the root of all these failures over the last ten years.

It starts out like this:

“DAOs don’t work because voter turnout is too low.”

Well that’s a symptom, not the root problem. Voter turnout is low because most of the votes don’t really matter… Then you arrive at:

“DAOs don’t work because voting is the wrong decision-making mechanism for most decisions.”

Okay, that’s interesting. Well then what is the right mechanism for making each decision? When should you use DAO votes, versus the wide range of other possible options? Is there a principled way to approach this?

This question is where the Hats Protocol project began going down the rabbit hole 3.5 years ago.

We’ve written a lot about our learnings over the years. But to save you some time, here are the big ideas:

1. DAO voting gives us capture-resistance. The democracy at the top level of the DAO is an absolute necessity in order to maintain capture resistance, and align the platform creators, investors, and users so that nobody gets screwed over.

This is essentially what is different about Web3, and why there is any hope at all in this approach for slaying moloch, solving alignment, and creating a user-owned internet that is fundamentally different from the maximally extractive landscape we have today.

2. DAO votes have very high costs. While the DAO’s democracy is very useful, it is extremely expensive to use. For this reason, we need to minimize its usage while ensuring it is always available as the last line of defense in decision-making and resource allocation in the organization.

3. The DAO must delegate to practically achieve its goals. Any time you want to have a decision made by a body other than the full DAO, you are delegating. (Not the “delegate your airdropped tokens and never look at them again” delegation we’re all too familiar with. We mean the operational delegation of specific work and responsibilities by the whole DAO to other contributors.)

Delegation introduces the principal-agent problem, or the problem of ensuring the body acting on behalf of the full DAO actually acts in alignment with the DAO’s best interests.

A counterexample of this would be: dumping 25% of the org’s treasury into Day 1 post-TGE hype.

But what’s unique about blockchains is that we have smart contract-based hardness as a building block to consider in the design of our delegation systems.

So if you want to have the founding team do “onchain market-making,” you could transparently deploy a smart contract that had specific buy and sell constraints, preventing them from self-dealing without you needing to trust them.

4. Delegation in DAOs creates a trilemma. What you end up with, as you break this down, is a series of tradeoffs between these three dimensions. You want to maximize effectiveness of decisions, while minimizing costs (moving away from full DAO votes), but without giving up hardness (reliable commitments to take aligned actions). And any action you take to gain more of one dimension makes tradeoffs against the others.

That’s it! This is the real problem with DAOs.

Every time a DAO (or a set of tokenholders) wants to accomplish something, they must delegate. And by doing so, they must make a series of tradeoffs within this design space.

Of course, you will want to make the best possible tradeoffs for each decision you are making. One of these tradeoffs is how much effort to spend on optimizing the delegations. The act of delegation (designing the delegation, making tradeoffs, and actually implementing it with smart contracts), is itself costly.

As the technology for delegation improves, this cost will decrease, and the delegations (and therefore entire DAOs) will become more optimized over time.

Through this lens, everything we’ve seen unfold in crypto makes so much more sense:

1. Bitcoin. Bitcoin intentionally made it very hard to change how it works in order to be resilient to attacks, corruption, and misalignment. Low effectiveness, high hardness.

It is extremely expensive to operate Bitcoin. But it turned out the that extreme value of hard money made it so that Bitcoin as a distributed organization had enough resources to pay that high cost.

2. Real DeFi DAOs. Similarly, Aave, Maker, and Lido succeeded as DAOs precisely because they operate in markets where users are willing to pay a premium for hardness.

However, there is a lot more going on technically, so these organizations have to manage a lot more complexity in their operations. Because of the premium, these organizations had enough resources to make the tradeoff of paying very high costs in order to achieve both effectiveness and hardness and in their operations.

3. ICO + Founders in Labs Corp. This is the common alternative to DAOing, which roughly results in entrenched and powerful founders with stodgy, static community committees. "We can't actually be decentralized because we won't be able to get anything done." Medium to high effectiveness and low costs, but dangerously low hardness.

4. Facebook/Meta. If you broaden your aperture a bit, you can even look at traditional companies through this lens, considering the mechanisms they use for hardness (institutions, legal systems, and state monopoly on violence) in place of smart contracts.

Facebook/Meta exemplifies traditional company optimization where Zuck maintained near-total control from startup to public corporation, establishing high effectiveness culture ("move fast and break things") while screwing over employees, users, and platform developers. Despite the legal contracts in place, in practice, the hardness is relatively low.

The Delegation Trilemma is the real problem with DAOs.

We hope this new lens helps the industry makes better sense of the challenges at hand, and helps make the application of funds raised from internet capital markets as strong of an innovation as the upgrades to capital formation itself.

Internet Capital Markets are heating up. The dream of funding internet-native businesses is back!

But here’s the trillion dollar question:

Once the project is funded, what is the best way to apply the capital toward achieving its mission?

Well, if you want a typical startup risk-return profile, you'd probably plow the funds into the founders’ bank account or multisig and hope for the best.

But as we’ve seen over and over again, that doesn’t usually work out too well…

On the other hand, if you are one of the few decentralization purists left, you might try starting a DAO and giving real smart contract-enforced ownership of the project to tokenholders.

And inevitably everyone would look at you like you have three heads because we all know that "DAOs don’t work."

...but could they?

We’ve been grinding for years to figure this out. We think there’s one key insight that might help us all get to the root of all these failures over the last ten years.

It starts out like this:

“DAOs don’t work because voter turnout is too low.”

Well that’s a symptom, not the root problem. Voter turnout is low because most of the votes don’t really matter… Then you arrive at:

“DAOs don’t work because voting is the wrong decision-making mechanism for most decisions.”

Okay, that’s interesting. Well then what is the right mechanism for making each decision? When should you use DAO votes, versus the wide range of other possible options? Is there a principled way to approach this?

This question is where the Hats Protocol project began going down the rabbit hole 3.5 years ago.

We’ve written a lot about our learnings over the years. But to save you some time, here are the big ideas:

1. DAO voting gives us capture-resistance. The democracy at the top level of the DAO is an absolute necessity in order to maintain capture resistance, and align the platform creators, investors, and users so that nobody gets screwed over.

This is essentially what is different about Web3, and why there is any hope at all in this approach for slaying moloch, solving alignment, and creating a user-owned internet that is fundamentally different from the maximally extractive landscape we have today.

2. DAO votes have very high costs. While the DAO’s democracy is very useful, it is extremely expensive to use. For this reason, we need to minimize its usage while ensuring it is always available as the last line of defense in decision-making and resource allocation in the organization.

3. The DAO must delegate to practically achieve its goals. Any time you want to have a decision made by a body other than the full DAO, you are delegating. (Not the “delegate your airdropped tokens and never look at them again” delegation we’re all too familiar with. We mean the operational delegation of specific work and responsibilities by the whole DAO to other contributors.)

Delegation introduces the principal-agent problem, or the problem of ensuring the body acting on behalf of the full DAO actually acts in alignment with the DAO’s best interests.

A counterexample of this would be: dumping 25% of the org’s treasury into Day 1 post-TGE hype.

But what’s unique about blockchains is that we have smart contract-based hardness as a building block to consider in the design of our delegation systems.

So if you want to have the founding team do “onchain market-making,” you could transparently deploy a smart contract that had specific buy and sell constraints, preventing them from self-dealing without you needing to trust them.

4. Delegation in DAOs creates a trilemma. What you end up with, as you break this down, is a series of tradeoffs between these three dimensions. You want to maximize effectiveness of decisions, while minimizing costs (moving away from full DAO votes), but without giving up hardness (reliable commitments to take aligned actions). And any action you take to gain more of one dimension makes tradeoffs against the others.

That’s it! This is the real problem with DAOs.

Every time a DAO (or a set of tokenholders) wants to accomplish something, they must delegate. And by doing so, they must make a series of tradeoffs within this design space.

Of course, you will want to make the best possible tradeoffs for each decision you are making. One of these tradeoffs is how much effort to spend on optimizing the delegations. The act of delegation (designing the delegation, making tradeoffs, and actually implementing it with smart contracts), is itself costly.

As the technology for delegation improves, this cost will decrease, and the delegations (and therefore entire DAOs) will become more optimized over time.

Through this lens, everything we’ve seen unfold in crypto makes so much more sense:

1. Bitcoin. Bitcoin intentionally made it very hard to change how it works in order to be resilient to attacks, corruption, and misalignment. Low effectiveness, high hardness.

It is extremely expensive to operate Bitcoin. But it turned out the that extreme value of hard money made it so that Bitcoin as a distributed organization had enough resources to pay that high cost.

2. Real DeFi DAOs. Similarly, Aave, Maker, and Lido succeeded as DAOs precisely because they operate in markets where users are willing to pay a premium for hardness.

However, there is a lot more going on technically, so these organizations have to manage a lot more complexity in their operations. Because of the premium, these organizations had enough resources to make the tradeoff of paying very high costs in order to achieve both effectiveness and hardness and in their operations.

3. ICO + Founders in Labs Corp. This is the common alternative to DAOing, which roughly results in entrenched and powerful founders with stodgy, static community committees. "We can't actually be decentralized because we won't be able to get anything done." Medium to high effectiveness and low costs, but dangerously low hardness.

4. Facebook/Meta. If you broaden your aperture a bit, you can even look at traditional companies through this lens, considering the mechanisms they use for hardness (institutions, legal systems, and state monopoly on violence) in place of smart contracts.

Facebook/Meta exemplifies traditional company optimization where Zuck maintained near-total control from startup to public corporation, establishing high effectiveness culture ("move fast and break things") while screwing over employees, users, and platform developers. Despite the legal contracts in place, in practice, the hardness is relatively low.

The Delegation Trilemma is the real problem with DAOs.

We hope this new lens helps the industry makes better sense of the challenges at hand, and helps make the application of funds raised from internet capital markets as strong of an innovation as the upgrades to capital formation itself.