But that doesn't consider the actual dynamic between the miners - who are also speculators, mining difficulty, and the price of Bitcoin. Those parts are constantly moving that lead to whether a miner adds more hashing power, or reduces hashing power. That's the current reality. Running some of their hashing power for A.I. systems "may" become part of the rotation, but saying that Hashing as a Sybil Protection Mechanism is a "thing of the past" is simply wrong. It's the future.
There is no mysterious dynamic here, lower income less hashrate, that's it!
Q2 has ended and you can see that despite newer models replacing older with better ths/joules we have still gone down 5%.
Have the coin drop to 50k and you will see how much miners care about dynamic as they shut down!
Also, miners don't protect against a Sybill type attacks, that's what nodes do!
So you don't believe that computers will be faster 20 years from now ?
Check the CPU frequency 10 years ago, and now. Compare it:
There is one problem with that approach: verification. Sending the whole chain is not a problem. But verifying still is. And what is the bottleneck of verification? For example CPU speed, which depends on frequency:
Can you please explain how doge manages to validate 10x the blocks as Bitcoin with all this CPU limitations mumbo jumbo?
Do they use some much wow shiba CPU that are different?
Rather than keep on talking about CPU limitations how about we do something else, see what the limit of the average CPU now in verifying a block is compared to the average CPU back in 2009, cause there might be a surprise! While you're at it you might also explain to me why the mother of all shitcoin cumwrightvision didn't break with 100MB blocks!