It might be one of the first cases of bribery where the briber doesn't have to talk to the bribee to discuss the object of the conspiracy, the price, or how the bribe would be delivered.
Kinda reminds me of Eliezer's concept of acausal blackmail
Comedy. Gold!
(Of course, this would likely drive the value of 1 BTC to exactly $0)
(Edit: This is all a demonstration of how the security of the blockchain absolutely depends on mining hashpower being widely decentralized and distributed)
> This is supposed to be impossible because it is incentivize-incompatible
Enter the wonderful world of coco values and side payments in game theory. Incentive compatibility and strategyproofness arguments break down when parties are compensated (implicitly or explicitly) out-of-band. That's not always a bad thing - it can lead to more efficient, utility maximizing equilibria. That said, under the status quo, we're trading one flavor of centralization (monetary policy) for another (core committers and mining pools).
So if you're playing along at home: the consensus Bitcoin chain is the longest one consistent with the rules of Bitcoin Core, the lineal descendant of the Satoshi client.
This consensus chain currently includes 1+ transactions T, which effected this theft.
The proposal is for a large number of miners to abandon the consensus chain and start mining a new chain at one transaction prior to T, replaying all real transactions minus T plus a new transaction which would moot T.
With enough hashpower, this chain eventually surpasses the current longest chain and at that point all points on network, not just mining cartel, will agree that T never happened.
This is supposed to be impossible because it is incentivize-incompatible, as the mining cartel would lose an incredible amount of segniorage to make it happen and it wouldn't be a sure thing. Point #2 is mooted by mining centralization, a ship which has sailed. Point #1 could get mooted by a payment outside the standard understanding of the Bitcoin protocol.
Comedy gold!