Why blockchains can be really bad. Or: How techno-futurists can ruin things.

Here’s what’s particularly disturbing about techno-futurists right now: Software is so widely distributed, so persistent and attainable, that any fantastical idea is just leading to more everyday destruction.

Take, for example, the DAO, or the Decentralized Autonomous Organization, a company governed entirely on a blockchain. That means its daily operations — from payments to investment plans to corporate governance — are handled via code. More specifically, they are handled by a series of contracts triggered by a piece of software sitting on the Ethereum blockchain, a competitor to bitcoin.

The conceit is that anyone can join, raise money or propose new investments, and then everyone gets to vote. It’s similar to an old-fashioned venture fund, except decisions don’t rest with an elite class like a partner group or a board of directors. It’s ... democratic.

The description on the website is typically breathless: “[I]t is a new breed of human organization never before attempted.”

Basically, the code rules, not a CEO or a board of directors or shareholders like a normal company.

Why? Well, because blockchain. To the techno-futurists running the world today, anything blockchain suggests a purity of purpose that a publicly traded company or a democratically elected government or even Mother Teresa herself could never achieve.

If you think this is just another wacky Reddit-sourced scheme, the DAO has raised more than $100 million so far, making it one of the best capitalized fin-tech startups around, according to the Wall Street Journal.

The problem is, the DAO was hacked on Friday. More problematic — and why it’s potentially even more destructive — is that even though you’d think such a breach merits a corrective patch, that’s not necessarily going to happen.

Here’s a great interpretation of the DAO’s rules and one explained by Bloomberg View’s Matt Levine, and both are worth reading in full, but the gist of the argument is the code wasn’t hacked. The “hack” wasn’t a bug so much as a feature that hadn’t been tapped into until now.

Remember, the code rules.

On its face, that argument makes zero sense. But if the conceit is that blockchain is inherently more democratic and the code can’t be corrupted in the way a human can, then you can’t blame the code. You also can’t blame the person who made the best possible use of that code, since the code, again, rules.

This reasoning is especially scary since it’s a belief so commonly held in Silicon Valley. Larry Page, the founder of Google and head of Alphabet, adheres faithfully to the notion that software, or “the algorithm,” is much more efficient and effective and pure than any human system could ever be. The sooner we digitize the entire world and let software run it all, the sooner everything will be made better.

He’s not alone; Facebook’s Mark Zuckerberg, Tesla’s Elon Musk and Amazon’s Jeff Bezos all submit to a similar logic, to varying degrees. While software in many cases is just better, mostly, and while it has fueled the rise of tech, mostly, it can ultimately be horrifying since the code enslaves common sense.

It may just be the “smart people can be dumb” aphorism, but when applied to people who can actually effect so much change, it’s scary. Or, maybe it won’t be. Maybe it’ll be, just as the website says, “a new breed of human organization.” Like a computer.

This article originally appeared on Recode.net.