AFAIU, smart contracts on the ethereum VM can be arbitrarily complex, so you pay the network to execute them, or a random user with an infinite loop would bring down the network.

You are indeed renting a machine to run some code, and if you want many people to use your code you want to make it cheap. There's a trade off.

You can fuck up things on the BTC blockchain too, "burning" crypto by sending it to a dead address has been a thing for a long time.

It always seemed stupid to me that it was possible, compared to sending money to an invalid IBAN, but I'm not a crypto enthusiast so I may be biased.

On BTC, you could send things to the wrong address, yes. But you can't send the wrong type of currency or send it to a nonexistent address. In this case it seems like the contract has created the black hole (not the Ethereum blockchain itself), but that's even more absurd since someone ultimately should have control over everything that was put into the contract, regardless of the source.

> But you can't send the wrong type of currency or send it to a nonexistent address

Of course you can do both things, which is why catastrophic financial ruin is a daily fear when dealing with cryptocurrencies.

https://www.quora.com/How-can-I-find-my-Bitcoin-cash-that-I-...

That's impressive, the number of scammers on that thread.

Oh wow, the "answers" on that question are cancer.

Sending currency to non-existant addresses is how people encode arbitary data on the bitcoin blockchain, so definitely possible.

Giving a person control over the funds allocated to their smart contract probably opens holes where they can steal the smart contract's money, though obviously creating software that handles money and can't be updated is its own kettle of fish.

The thing is, the address exists - only (probably) nobody has its private key.

Why would one need to send to non-existent address? Any transaction can include data, one can donate for example.