Thread by @joekelly100

Formal science has entered the chat.

“BDoS: Blockchain Denial of Service”
arxiv.org/abs/1912.07497

Now we’re getting somewhere. You kill blockchains by breaking the incentives. Not sure the attack in this paper would be effective, but it’s where the vulnerability is.

1/

The goal of a sabotage attack is to upset the game theory, system-wide (not just miners), so that the whole thing falls apart of its own accord, in a decentralised way, via the spontaneous action of the individuals involved.

2/

The Bitcoin network is not a harmonious love-fuelled collective; it’s a disorganised body of ruthlessly self-interested and mutually distrustful individual actors. If the incentives are broken, Bitcoin is broken. Break the incentives, and you break Bitcoin.

3/

That’s why a government attack on #Bitcoin can move at the speed of information: you can break the incentives without actually going through the full process of the attack; you just have to *signal* what’s coming, and the rest takes care of itself.

4/

https://twitter.com/joekelly100/status/1294210039968862213

This paper assumes governments would want to be covert. I disagree: far better to announce your intent and exact plans out in the open. If the threat is credible, and the outcome predictable, it makes the final cost of any actual 51% attack low.

https://twitter.com/joekelly100/status/1294209666604503041

5/5