This is a clip from The Breakdown newsletter. To read the full editions, Subscribe.
“You have to enjoy failure.”
-James Dyson
In a 2003 research paper, Google engineers subscriber Their new approach is to build “trusted computing infrastructure from clusters of unreliable commodity personal computers.”
This was a radical change. Instead of paying for Sun’s state-of-the-art servers designed for near-perfect reliability, Google was adopting cheap commodity hardware that ran free Linux software. They then developed additional software to overcome frequent hardware failures.
They found that repeatability through size is cheaper than accuracy through geometry.
This is the model of modern AI data centers too: you can’t have a three-month training failure just because one GPU in 20,000 dies.
However, it is not a manufacturing model for GPUs themselves.
Instead, semiconductor manufacturing plants are set up to produce GPUs with a very low level of defects, requiring near-perfection at every step of an incredibly complex process — the opposite of Google’s recommendation for data centers.
But this is not engineering perfection. It is an economic necessity.
“Why do we need 99.9999% reliability at every step of the process?” Ben Thompson he asks. “It’s all capital equipment cost.”
That’s mostly because the $200 million ASML machines that every fab uses to make cutting-edge semiconductors “are the most over-engineered things on the planet,” Thompson explains.
This creates a feedback loop of escalating cost and perfection: manufacturing plants have to be nearly perfect to be profitable because ASML machines are expensive — and ASML machines are expensive because they have to be perfect.
As a result, the cost of building a single semiconductor factory now reaches $20 billion.
Thompson compares the process of making semiconductors with ASML’s highly complex machines to the process of making rockets at SpaceX, where Elon Musk chose to “blow up a bunch of spaceships because you don’t want to waste anything on complexity that’s not actually necessary.”
Detonating rockets frequently helps SpaceX find “the fine line where it wants to be,” Thompson puts it — the line between complexity and usefulness that allows it to avoid the astronomical upfront costs that perfection requires.
In other words, failure can be a feature, not a bug, of complex systems.
It’s about the journey…
Failure is a fundamental feature of cryptosystems as well: every blockchain is designed based on the assumption that a certain percentage of its nodes will be malicious or faulty.
This makes the blockchain resilient – like Google’s data centers, it continues to produce valid blocks despite individual failures. But this fault tolerance does not extend to the running code on blockchain.
Smart contract code, written by fallible humans, will inevitably have errors or design flaws. This can be expensive, because unlike traditional finance, there is no reversal of a cryptocurrency transaction once a flawed contract or protocol is exploited.
The code eliminates intermediaries, but at the cost of zero fault tolerance.
This makes code encryption the law of the system Philosophically Embraces decentralized fault tolerance however Architecturally It requires perfection.
There’s no way to design your way around that.
In traditional finance, where regulations are permissible and transactions are reversible, banks can aim for near-perfection by implementing stricter know-your-customer and anti-money laundering rules and hiring more compliance officers.
In cryptofinance, where systems are permissionless and transactions are irreversible, we can only aim to learn from our mistakes.
There have been plenty of them: from the DAO exploit that nearly sank Ethereum in 2016 to the Bybit hack that set a record $1.5 billion in 2025.
However, with luck, in hindsight these can look like Elon Musk’s exploding rockets. Each exploit exposes an attack vector — reentry, oracle manipulation, and flaws in the protocol design — that can be fixed in future iterations.
Like spacecraft explosions, the failures show what actually needed tightening versus what was unnecessary caution — each representing an opportunity to find the right balance between complexity and utility in a new kind of financial system.
Cryptocurrencies are often judged by their failures, which is fair enough, given their frequency and scale.
But failures are not mistakes in The process of building a new financial system. they We are This process.
To enjoy cryptocurrencies, you have to enjoy failure too.
a happy new year.
Get news in your inbox. Explore Blockworks newsletters:
