The difficulty will increase in response to the hash rate of the network, it doesn't directly correlate to the amount of coins in existence. So when the last BT

The difficulty will increase in response to the hash rate of the network, it doesn't directly correlate to the amount of coins in existence. So when the last BTC is mined, if people stop mining the difficulty will go down (after about 2 weeks I believe) and if more mine, it will still go up. The difficulty is designed to be at a rate that will see a block mined every 10 minutes. So the ideal max verification time should be about 60 minutes (6 blocks) on average, although you should see your transaction included in the next block mined (in 10 minutes-ish, assuming the miner included your transaction, which they have an incentive to do if you included a fee) but to stop a double spending attack, it would be unwise to consider a transaction verified after 1 block, 6 blocks makes sure your transaction is not lost in a blockchain fork or part of a double spending attack. Some people consider 3 blocks to be enough verification time, it depends on how much money you are dealing with but the recommended amount of 6 blocks is in the original paper. The last block to generate BTC that will be mined should happen in around 2140 (with the reward halving every 4 years) and then there will be a maximum of 20999999.9769 BTC in existence. Interestingly enough, if you flooded the network right now with your own transactions and included a higher than normal transaction fee on those transactions, you could conceivably block other transactions from being included in the blockchain, at least for a while. Last time I calculated the cost of such a scheme it was about 72BTC per hour, so even without fixing the scaling issues of transaction volume in BitCoin, the rise in price makes this sort of attack less likely (still theoretically possible).