@user2963623 without proof of work, a malicious actor can perform a sybil attack by spinning up millions of nodes to convince your node that their version of the blockchain is the canonical chain.
@user2963623 If nodes see a new chain which is longer/heavier, they will switch to it, even if it means abandoning a chain that they've been on for years.
This might seem stupid, but there's a reason for it. Specifically, if a new node connects to the network, it's not going go know that the old chain is older. We want the new nodes to do the same thing as the nodes that are already on the network.
@NickODell That makes a lot of sense.. so would it be correct to say that POW is a deterrent to this scenario because of the involved computation cost rather than a safeguard?
@user2963623 Yes, that's the point. The protocol doesn't know which one is malicious or not. The way it decides it by making it costly to produce a chain.
If an attacker is willing to pay the price to reorg the chain, they can, arbitrarily deep.
However, the system is designed so that that should be against their own financial interests.
The most upvoted answer here appears to be quite dangerous advice. Exposing your wallet seed to an online tool might lead to all sorts of trouble, no?
What to do in cases like this? Comment/downvote/flag/meta post?