Today, I want to talk to you about
which blockchain is the “best” blockchain. Are you ready? Let’s ask the audience. Altogether, at the count of three, you shout
out which blockchain you think is the best. Three, two, one…
[Audience shouts] Fantastic. That is what subjective experience is all about.
We all have opinions, but opinions really don’t matter. It doesn’t matter if it is the “best”
blockchain in your opinion or not. Maybe it does matter if you are a trader,
and you need to guess what everybody else thinks it is. This is all about perception. I am not going
to tell you which blockchain is “the best.” I think that is a ridiculous question to even ask. It is a bit like asking, “What is the best car?”
“What is the best pair of shoes?” It depends on what you want to
achieve with this car or pair of shoes. If you ask someone what the best pair of shoes is,
they may tell you: strong hiking boots. If you ask someone else, they may tell you:
Manolo Blahniks. Have you ever tried hiking in Manolo Blahniks?
Have you ever walked a fashion show in hiking boots? They wouldn’t work so well, right? Purpose matters.
The purpose you want to build the tool for, matters. The purpose determines
what the best blockchain will be. The general question
“what is the best blockchain” makes no sense. You need to state the purpose. You can ask,
“What is the best blockchain for my purpose?” “What is the best cryptocurrency for my purpose?”
But you can’t just ask, “what is the best?” That makes no sense. If you start reading the marketing brochures,
I’m sure they will try to contradict what I just said. They will tell you, “This blockchain can do everything.
It is scalable, secure, and fast.” “It does smart contracts and sound money.” “It is quantum resistant, super private,
and uses military-grade encryption.” That last part just means 128-bit encryption… it’s bullshit, which is the purpose
of a marketing brochure, of course. People will [make a lot of promises] about their favourite
project, but you should really ask what the purpose is. What are you trying to achieve? In architecture, there is a great phrase from
the 1980s, which is: form follows function. The idea is, with any building, the way it looks
should communicate what it is used for. The form should follow the function; the architecture
of the building should reflect the purpose. The same thing applies to blockchains. The architecture of the system should
reflect what it is intended to be used for. But that is a very difficult thing to do, for architects,
software engineers, and product managers. It is even more difficult in open-source,
public blockchains. They are the only ones I care about
and think are interesting. It is difficult [to wade through] all of the opinions
about what a blockchain might do or will do. Pick any one of the open public blockchains
that exists today, and ask about twenty developers… working on it about what the primary purpose is.
“What is this thing supposed to do?” They will give you twenty different answers.
“Sound money.” “Digital cash.” “It will need big blocks.” “It will need small blocks.”
“It can handle smart contracts.” “It can do everything [you need].”
But what does it do for you? That is a more interesting question. You must decide
what fits the application you are trying to build. Unfortunately, that question is not asked a lot. You can tell, because people are still trying
to apply blockchain technology… to a whole bunch of things that don’t need a blockchain. I have considered wearing a t-shirt at all conferences
which says, “You don’t need a blockchain for that.” [Laughter] When I am attending a hackathon
and they present applications, someone claims, “We have put digital music on the blockchain.”
You don’t need a blockchain for that. Why would you need a blockchain for that? I should just write that question on
a sign and pick it up every time. What does a blockchain do?
It is not a content database. It is not just somewhere you record digital signatures.
You certainly don’t need a blockchain for just that. It is not scalable or efficient. It is decentralized, secure, and (for the ones
I care about) very robust against censorship. That means it should keep running even
when very powerful people or organizations… want to stop it from running. For that use case, you need a blockchain.
But what kind? How would you design it? Let’s talk about the engineering challenges when
building these systems. You need to make choices. Some of these choices come up at the very early stage. For example, do I want to build a system that
is very specific, or with a broad generic purpose? Why don’t we just do both? You can’t do both. Systems that are specific need
to be optimised for that purpose. In doing so, they are no longer flexible
enough to handle the general purpose. Systems that are general-purpose don’t have the unique
characteristics you need for specific applications. That should be a conscious decision at the beginning. Some blockchains are designed for rather specific
applications; some blockchains are designed… to handle generic applications. Both often claim that they can be both. Once you have made a decision on that first
question, you need to make further decisions, all of which involve more intractable trade-offs.
“Intractable” means, you can’t get both sides. I must choose [one or the other]. That choice then
determines the path that the blockchain will take. As a developer, designer, or architect,
you don’t [have the power] to tell the market… how they will use your product. If you try to be too specific, the market may
decide that it is not suitable for what they need. If you try to be too generic, but you only
have one type of application in mind, the market may decide to use it for
a completely different type of application. Hopefully, they will. As a developer, designer, or architect, you rarely have
the exclusive knowledge of what other people need. There is too much variety. You can barely imagine what people might need
in other places in the world that you don’t understand, under different circumstances than your own. You certainly can’t do that across a broad timefame. We are building systems that might last,
should last, will last, for decades. What do people need today?
What will people need in 30 years? Very different answers. The developers who are making these very careful
trade-offs, need to consider that they are… making those decisions with incomplete information. Even if they try to design someting with very specific
applications in mind, the market may decide… to do something completely different. Back to architecture history. In the 1960s and 1970s, in California, they
made swimming pools in organic shapes. No longer square, but with curves.
Have you seen those types of swimming pools? Then there was a big drought; for a couple years,
it was forbidden to fill your swimming pool… because there wasn’t enough water.
Guess what happened next? Skateboarding exploded in popularity across California. Skateboarders saw those curves and thought, ‘If I stand on the edge and kick down,
we could have some fun with this!’ All of the swimming pools which had been designed
with curves, became the first skate parks. Skateboarding wasn’t a big deal before then
and mostly involved [riding] in straight, flat lines. That sport was transformed. Do you think the swimmming pool designers
could imagine that, ten years later, some snotty nosed kid would go
wizzing around them on wheels? No. Form follows function… sometimes.
Sometimes, it doesn’t. Sometimes the form follows the function that
the designer thought it would have, and then the market says, “I have another function
that fits perfectly with this curve.” You can never know how that will play out.
The same thing will happen with blockchains. Ultimately, the users and the market decide. You can have all the ideas you want,
and yet here we are, arguing about it. Maybe we will have a panel of six very serious experts.
Some of them will say, “It is a store-of-value.” Somebody else will say, “It is a medium-of-exchange.” Another one will say, “It can’t be a unit-of-account,
because the volatility is too high.” And then someone will say,
“Maybe we should focus on smart contracts.” But they don’t really know what they’re talking about,
because they don’t get to decide [for everyone]. These are opinions, but they [don’t
represent every] user. Users will decide. You can make something that seems to fit better with
a store-of-value use case, or a smart contracts use case. The developers and designers of Bitcoin
had some ideas about it being digital cash. For some period of time, it plays as digital cash. For another period of time,
it plays as speculative gambling money. And in other times, it plays as a store-of-value, especially
in countries where their currency is distressed. Which of those will it be? We don’t know.
It depends on a lot of factors that we don’t know yet. It depends on what happens to national currencies
and inflation in the U.S., Europe, and Japan. It depends on how the world transforms. It depends on whether cash, as it exists today,
will still exist in fifteen years. All of these factors have nothing to do with Bitcoin. The designers of Ethereum had certain applications
in mind, mostly engineering applications. Did they think it would become a platform for launching
ten thousand scams and pump-and-dumps? [Laughter] No, but it was very good at that! A generic, flexible platform to build whatever
smart contract you can imagine, will attract… the kind of person who wants to build
a beautifully engineered pump-and-dump scheme. Does that change what Ethereum does
[as a whole]? No. It just means, it was a niche that
the market decided was really hot. Why? Because a whole bunch of naive investors,
and even more naive venture capitalists, were throwing as much money
as possible into this new space. Two years ago, you could attach “blockchain”
to any other word, and a venture capitalist… would throw a couple million dollars at you. Music? Blockchain. Movies? Blockchain!
Real-estate? Blockchain. Blockchain real-estate? Asparagus? Blockchain asparagus.
“Two million dollars for ou, sir. That sounds fascinating.” You could even combine it with the other ‘cool’ words. “We will cultivate asparagus with blockchain-based,
artificial intelligence directed autonomous drones.” [Laughter] That way, you could check all of the boxes,
and they would throw tens of millions of dollars at you. None of it made any sense! The market not only decides [rationally], but sometimes
the market is stupid, crazy, and irrational. It is driven by sentiment and emotion.
Over the next decade, things will calm down. People will figure out that
you don’t need a blockchain for that. How will most of them figure that out?
They will invest their money and then lose their money. They will invest more of their money
and then lose it again. By the third loss, most smart investors
would start picking up on that pattern. For the less smart investors, it may take them
ten or fifteen rounds of losses. The bottom line is, there is no “best” for every purpose. The people involved in making difficult trade-offs every
day, cannot design for a purpose that is too specific. That can be limiting and miss what the market
may want to do. They can miss the timing. They also cant design something that is too
generic, because it won’t have enough… powerful capabilities to solve
real problems that we have. They will need to make difficult trade-offs. You can’t simply be scalable, decentralized,
secure, and fast [at the same time]. In engineering terms,
there are some fundamental trade-offs. We categorize them as dilemmas or trilemmas. A classic trilemma is: security,
decentralization, and scalability. In a trilemma, you can only
pick two of the three [options]. If you make something that is maximally scalable
and secure, it probalby won’t be very decentralized. If you make something that is maximally scalable
and decentralized, it probably won’t be secure. If you make something that is maximally decentralized
and secure, it probably won’t be very scalable. Of course, there will be blockchain projects
that tell you, “We can do all three!” That means they don’t understand the trade-off,
which is even more dangerous. There are two possibilities: lying or ignorance.
Ignorance of the trilemma is far worse. This is what I used to tell my consultation clients:
“I can deliver the solution fast, cheap, or great. Pick two.” “I can do it fast and cheap, but it won’t be the best.
I can do it cheap and it will be great, but it won’t be fast.” “I can do it fast and it will be great,
but trust me, it won’t be cheap!” That is the essence of a trilemma we face every day
in life. Life involves choices. Think of it as a journey. When you go through the door on the left,
you may lose whatever is through the right door. You may never be able to go back and take
that road, and the roads that branch off from it. You have made your choice. When dealing with building the best blockchain
for a specific purpose, every choice you make, whether you know it or not, can close
as many possibilities as it opens. The next time someone asks you,
“What is the best blockchain,” instead of shouting out the name of your favourite,
ask “For what?” Thank you very much! [Applause]