If Crypto was a house


Buying a house

I can buy a house wherever I want as long as I have the money, and the seller is willing.

To do this buy/sell transaction, we need to trust the city to vouch for the property titles, and we may pay title insurance to protect against errors.

I trust the government to give me the land and not take it from me by force (and protect me if someone attempts to do so). Buying a house in most places is an open, trust-dependent system. Anyone can do it, but they need to work with, and trust, the government to get it done.

50 years ago, my grandparents could not have bought a house in my neighborhood. Not because the property rights system was not trustworthy, but because they would not have permission. At the time, the city charters prohibited jewish families from moving in, so that ruled them out. In this case, the system was trustworthy, but not permissionless, and hence less open.

The proof of ownership of the house is recorded and validated by the government. Because people can successfully lie to the governments, and because the government can lie and cheat to its citizens, different governments have different levels of trustworthiness.


Cutting out the middleman

It is really important when buying a house to know that property claims over it will be respected.

In some countries, you can’t trust the government to give you legitimate titles, or to keep the terms in which they give the titles somewhat constant, or to respect and enforce the rights of those titles. In some countries, people lose their houses to lawlessness, coming from fellow citizens or the government itself.

In the ideal scenario, buyers and sellers could transact without the need of a third party like the government. But without some form of enforcer, validator and witness to the transaction, how do we keep people from cheating? As a buyer, how can I know the house wasn’t sold to someone else 3 minutes ago? Or that the government will take it away by force if they choose to?

The grand contribution of the Bitcoin whitepaper was the invention of a mechanism where independent agents can find consensus and common truth, without the need to trust someone else. No government or any other form of trusted third party is needed. Bitcoin does this specifically for the transactions that take place in its network, commonly denominated in BTC.

Bitcoin and other legitimate Crypto protocols are permissionless, trustless systems. These systems are open to participation without asking for permission, and agents can participate without having to trust a third party to coordinate their interactions.


Keeping my house

For people living in mature democracies, sometimes it is hard to imagine why the government would want to take someone’s house, and even harder to think that the house the government could expropriate would be their house. Unfortunately this happens in many parts of the world, often for retaliation purposes.

One can imagine an example where the government dislikes my political opinion, and use that as an excuse to take the house from me.

There are two main protections I could have against a government attack.

One, is to make my ownership anonymous. It’s hard to imagine doing this with the house I live in, but easier to imagine with a house no one knows I own.

In this case, privacy is really important for me, since it protects me from abuse from others that can be stronger and more violent. Maintaining privacy with physical assets is super hard, but increasingly possible with digital assets.

Some Crypto projects focus on how to make transactions that can be not only trustless (no need for third parties) but also, private. There are different levels of privacy which can be achieved by different means, which go from bundling transactions so one transaction can not be separated from a larger group, to using advancements in math like zero-knowledge proof which allows people to transact without revealing identities.

The other protection against someone taking your house - and here I am stretching the housing metaphor quite a bit - is if no one but you has the one key to get in, and your house is designed in such a way that it is impossible to get in without the key.

Private keys are a cryptographic tool to keep information private (present in Bitcoin and onwards with other protocols). In Crypto protocols, private keys are used to make transactions; without that key no transaction can happen. If you have a ton of crypto and lose those private keys, you lost your crypto.

To manage these keys, some people use specialized hardware to keep them safe; others give private keys it to someone to look after them or have an account in an institution instead of keys(Coinbase business is to have the keys and create an internal account for their customers), but as you can see, if you give the keys to someone else, now you have introduced a third party again who you hope will take good care of those keys.


Getting help to buy more houses

If I had tons of money and loved houses very much, I could have an army of agents buying houses for me. I would give them certain criteria including total yearly budget and per house budget, and they would go buy them for me. This system would be distributed, but not really decentralized because I control it (I can fire agents, change compensation or buying criteria, etc).

In Crypto, a lot of people from many different places participating in a project is no indication of how descentralized is.


A co-op buying a house

A Cooperative (co-op) can also buy houses and in that system ostensibly any member can bring an idea for what houses to buy, everyone gets a vote, the votes get aggregated and the decision is made. Since the power is shared among everyone, the decision is decentralized.

Of course, a co-op with only one member wouldn’t really be a co-op.

A co-op with 3 members is more decentralized than a co-op with only one member, but still 2 out of 3 could collude against the remaining one. So, the more people there are, the more decentralized the decision is.

In this sense, decentralization means that every member of the co-op makes a decision, each decision has equal weight, and the final decision is the aggregate of all.

A good co-op has tons of members, but the more people it has, the harder it is to coordinate. Also, the more varied the objectives of the co-op, the harder to coordinate: It is easy to coordinate if we all want the biggest house possible. But if some want a big house, others want the most walkable and the rest want the prettiest, coordination becomes harder.

Different Crypto projects have different ways to facilitate trustless transactions, but all of those depend heavily on decentralization. No matter how good the system, if few participate in it, or if few participants represent most of the controlling agents, then the system can be perverted.

But as with the co-op, decentralization in crypto projects comes at a cost. Depending on the nature of what is being transacted, or the participants’ tolerance to censorship, some protocols can make design decisions that make them a bit more centralized in exchange for being much faster, or cheaper. Solana is a good example of a protocol where it is much more expensive to help keep the system trustless, which makes at the moment a bit more centralized than say Ethereum, but in exchange it is much faster and much cheaper.


A co-op buying many houses

If the co-op wants to buy a ton more houses they can get more members, or they can also have the members agree on a process that can be automated. They can build their own co-op owned mini OpenDoor type system, and write a program that ingests a ton of data and build a model that outputs the right house to buy, and the target price.
The more narrow the objective of this co-op buying these houses, the simpler the algorithm will be: it is simpler to aim for the biggest house than aiming to buy the biggest and somewhat walkable and cheaper than average house. Creating and maintaining consensus from the co-op members on a simple algorithm will be easier than on a more complex one.

We tend to talk about algorithms like they are all the same, but there is a big difference between simpler algorithms that are understandable, and highly complex ones where we only understand what goes in (the data) and what goes out (did we buy the right house?) but not how the decision was made.

When people need to know the consequences they can expect from these algorithms, and the ways in which they can improve them, complexity of algorithms has a great influence on the ability to decide which one to use.


Typical houses and weird houses

Houses have tons of parts, but given the zoning regulations, their use is very limited. I can not sell stuff from my house, nor rent a bathroom per “visit”, or a room per hour. That limits the reasons for buying a house to a few: I either live in it, hold it for appreciation, or rent it (I could also rent a room, etc, but you get the point). If, however, I could do anything I want with the house, there would be more revenue streams, potentially many I can’t even imagine now, we would have more and more varied houses, and the usages could change and adapt over time. This house would be said to have more composability, because I can use each component - bathroom, each room, etc - differently and mix and match in various ways.

But if the co-op wanted to buy such a multifaceted and adaptable house, the decision would be even harder to make: there would be more things to agree on, and different members could want different, conflicting things.

In Crypto, the work of what the protocol does is programmatic. Simple programs with simple goals are easier to manage but can’t do very much. More complex programs can do more, but tend to be harder to manage, because different use cases may affect different participants (owner of the underlying token, user of the network, developers building apps on the network, and protectors of the trustless system - miners, validators, etc -) in different ways.

Ethereum is the first and biggest player in this space of multiple possibilities. As more use cases become possible (or are invented) more developers participate and more people use it. But also, as it serves many use cases with many constituents, it becomes a bit harder to manage, and competitors can emerge by focusing on specific use cases. As an example, the Crypto Kitties project, which started as a Non Fungible token (NFT) on Ethereum, built their own protocol, Flow, to specialize on that use case.


The value of a house

All markets, including the housing markets, need certain conditions to be a place where buyers and sellers can do business. Different conditions will affect the price of the asset.

For the housing market to come together we need more than trust and security. For example, if we don’t know if the market fees are going to be 3% or 30% when I sell, or if selling a house will take a week or 5 years, it is hard to price the asset and hold it long term. certainty might not be needed, but we can easily see how too much uncertainty negatively affects the price.

Or take a more radical thought experiment where unique conditions affect prices and use cases –

There is a cool idea proposed by Glen Wyl to combat price distortion in the housing market: imagine if every number of years, your house price was assessed by you, so in effect you decide the amount to be taxed by the city. Here is the kicker: at that price, anyone can bid on the house to take it from you; if you price it too low you pay little tax but risk losing the house; if you price too high, the opposite happens.

What we have gained with this radical system is better prices. What we have lost is predictability. What would happen to the housing market? At least for use cases that require long term predictability (say, raising a family, retiring and dying in that house), buying a house will become less attractive.

One of the biggest differences between Bitcoin and other projects is that it aims to be predictable. This makes sense since the asset it secures is Money itself. Other projects trade off against predictability in order to be generative: more use cases are possible when more things can be built on these protocols. Ethereum, as we mention, is the main participant in this generative space.

There are many different types of projects happening in Crypto.

Like with the house buying examples, some Crypto projects are open but not trustless, trustless but not decentralized, decentralized but not predictable.

Predictability can be good for some things (specially scarce and elastic goods - where price matters), and bad for others (generative, creative use cases which need composability).

Privacy may be a human right to some, and worthless to others (we might also be in one position one day, and in another in the future) and privacy may matter more or less depending on the activity.

Many of the scams in cryptoland are predicated on claims to have all of these qualities when they have some, often just one and on occasion none of these.

The legitimate projects focus on different use cases requiring more or less of each of these dimensions (trustless, decentralized, composable, predictable, private), and in these tradeoffs different opportunities emerge.

· startups