Nick Szabo  🔑

I've seen only one possibly good argument for reduced security under a tx-fee-only regime, and that's the volatility of fees, which seem to behave nonlinearly as blocks become full. Might lead to corresponding big swings in hashrate. But they don't even mention that.

I would think by then Exchanges might become miners as well.

It's prudent for businesses & users to take part in mining defensively to participate in security. Even post subsidy, I am not so worried because even less secure money has worked because people need it to, and there is market incentive to find ways to fund #Bitcoin security.

This smacks of asking for volunteers to mine despite having unprofitable rigs. How to make it incentive compatible?

If your businesses doing millions of dollars a day you have plenty of incentive to fire up rigs.

Free rider problem. More profitable to let somebody else pay for security, so nobody wants to.

If no one does they all go out of business. Of course no one knows for sure what will happen, we'll be going back to the experimental stage at that point.

Recipients of large value txs can: 1. try to mine the tx themselves by quickly leasing extra hashpower just for this block. may be lots of subtle problems with this. 2. wait more blocks

Why is lower hash rate long term not lower security?

Only big short-term hashrate changes or hashrates differences in the same algorithm used by more than one coin pose a potential security problem. Bitcoin hashrate wasn't less secure a few years ago than it is today, even though it was only 1/1,000th of what it is today.

Your statement is correct, but that doesn’t mean that less hashrate in the future doesn’t equate to less security. Hashrate means nothing without a correction for ASICs efficiency improvement applied to it. All that matters is the miner revenue chart.

Neither Sztorc nor you nor anybody I have read have made a reasonable argument that it does, excepting possibly the argument about fee volatility which Sztorc didn't make.

Let’s say the network hashrate was 10 exahash 3 years ago and let’s say today it’s also 10 exahash. Considering the improvements in mining technology, would you agree that in that case the network was more secure 3 years ago?

Let's say we stick to non-fantastic scenarios. Improved mining tech improves the hashrate. And the debate over whether transaction fees will be sufficient to secure the network involves decades not a mere 3 years.

But it’s hard to discuss non fantastic scenario’s if it’s a decades issue and bitcoin has only existed for one decade? What’s wrong with the assumption that mining cost ~ the miner revenue and an attacker needs to deploy >~51% of that cost to attack the network?

Lower hash rate isn’t less secure? Huh?

Was the hashrate 1/1,000th less secure a few years ago when the hashrate was only 1,000th that of today? No, the hashrate contributed about as much to security then as it does now.

Nick, I’ve reread this tweet 10x and thought about for 10 minutes and I’m genuinely perplexed. Do you mind expanding on this? Seems to me that if the hashrate dropped to 1/1000th of what it is today the chain would in fact have *less* than 1/1000th the security.