Bitcoin mining has been too centralized for years, with just a handful of pool operators have controlled well more than 50% of hashing power. Recently, mining power has become even more centralized, with one mining pool (GHash.IO) likely controlling somewhere between 40% and 60% of hashing power. That isn’t good, and if you are mining with GHash.IO I would strongly urge you to try one of the smaller pools, or, even better, take the time to run bitcoind and p2pool. But it isn’t disastrous, either. Even if GHash.IO is evil and intends to destroy Bitcoin they would be able to do only two things:
The first thing they could do would be to double-spend already confirmed transactions. For example, they could send some bitcoins to an exchange, trade them for dollars, wire the dollars to their bank account, and then announce a longer blockchain where the transfer to the exchange never happened. Now they have dollars and bitcoins.
There are some practical problems with carrying out that attack, though. They are likely to get caught, because it is impossible to wire money to a bank account anonymously. It seems very likely they would find themselves in legal trouble for defrauding the exchange.
The second thing they could do would be to prevent transactions or new blocks from other people getting accepted, effectively stopping all payments and shutting down the network. I wrote about neutralizing that attack a couple of years ago.
I think either attack is extremely unlikely from an economically rational mining pool– blockchain history would make it obvious that they were mis-using their power, and I’m certain either technical or social solutions would be found to punish the bad behavior. However, this is a good time to re-iterate my standard disclaimers: Bitcoin is still a work in progress, and you should only risk time or money on it that you can afford to lose. Mining centralization is one of several potential risks; read Jim Harper’s excellentRisk Management Study for a clear-headed assessment of risks and consequences