As Bitcoin hashrate crosses a Zetahash, reminder that security is not correlated with hashrate
According to my math, the difficulty adjustment that happened Sept 18, 2025 put the required number of hashes to expected to find one block to be 6.113568244e+23. Divide this by 600 (seconds per ten minutes) and we get a hashrate of 1.018928e+21 per second, that is about 1.02 Zetahash per second. (It's official when it's recorded in the difficulty, so it is now official.)
That's a lot. Up by a factor of 2500 over ten years ago and 8-9 from five years ago.
But is Bitcoin more secure? There's a widely circulated video from about ten years ago when hashrate was 3 orders of magnitude less, which unfortunately has poisoned the minds of many folks as they started to think bout Bitcoin security. If you haven't watched it; it's short and amusing. Andreas Antonopoulos responds to a question of whether a nation state could develop enough hashrate to attack Bitcoin.
I'm going to call the misconception present here the Antonopoulos Video Fallacy. Not because Andreas Antonopoulos isn't smart enough to know he was answering a question about an unrealistic attack vector for a few laughs, but to give a name to this idea that everyone associates with the video. People have misinterpreted this video to think he is answering the much more difficult question of whether Bitcoin could ever be attacked.
The Antonopoulos Video Fallacy is the idea that anyone who attacks Bitcoin must covertly generate the hashrate de novo.
The reason this is flatly wrong is the very simple tautology: 51% is less than 100%.
It's really that simple. As long as there is any positive amount of hashrate in the universe, 51% of that hashrate can be used to attack the network. Nakamoto consensus was meant to protect the network not from folks who don't have hashrate, but from miners who do have hashrate.
So once we reached a place (we're about ten years in) when the dominant portion of hashrate is generated from ASICS, the amount of raw hashrate and the security guarantees are almost completely uncorrelated. The entirety of the security model rests on the incentives of the folks who have the hashrate, and, more importantly, people who have any leverage or power over them.
In other news, Jimmy Kimmel's show has been suspended indefinitely by ABC. The suspension is purportedly due to comments he made last Monday (September 15) not about Charlie Kirk, but about the political response to the shooting. The chronology; the show aired Monday night, then he ran another show Tuesday night, and then Wednesday Brendan Carr made the following statement on a podcast
“We can do this the easy way or the hard way. These companies can find ways to take action on Kimmel, or there’s going to be additional work for the FCC ahead.
Obviously, this was taken as a threat. By the end of the day, ABC announced Kimmel was out. Not for nothing, Kimmel's Tuesday monologue was largely dedicated to mocking Trump, including his lawsuit against the New York Times (lawsuit has already since been dismissed.)
Just to be very explicit about what happened. Within a few hours, a threat from the FCC turned into a major programming change by a major network. This was most certainly a "business decision" in the sense that it was in the financial best interest of the corporation to not square off with the FCC, instead opting to drop Kimmel and accept whatever public blowback was generated from lowering the cancel hammer. The calculus implicit is that whatever hit they would take from losing Kimmel, the hit they would take from the FCC cancelling their entire business would be much worse. And it was that easy.
Now why am I doing a woke rant in the middle of a Bitcoin security post?
Let's talk about who owns much of the hashrate these day: It's increasingly publicly traded companies who are enmeshed with other businesses such as AI/HPC and the energy sector. Today hashprice is very near to $0.50/ EH, pushing more home miners out, and concentrating power among large miners who have access to the latest ASICs and obtain sweetheart deals with utilities to take electricity behind the meter. I'm guessing that within a year or two, US publicly traded miners will be generating near 50% of the hashrate.
This all means that these publicly traded companies are subject to easy leverage. Many of these companies now have their fingers in many pies, and are actively involving in mergers. Any financial interest can be used as leverage against them by a heavy-handed regulator who is seeking some objective.
The question of Bitcoin security (at least as far as the nation-state attack) is not a question of how much hashrate needs to be generated from zero, but how many arms need to be twisted, and how exposed those arms are to being twisted.
The pivot to AI/HPC represents an unmistakable threat vector. Now that the Bitcoin mining corporations have large revenue streams (outside of Bitcoin mining) not only can these streams be leveraged, but if the Bitcoin mining industry were to suffer consequence from some hard or soft government takeover, these corporations will all land on their feet. There is no more guarantee that "He ought to find it more profitable to play by the rules" when it's easy to picture a regulatory landscape in which the board determines that the corporation ought find more accretive to shareholder value to comply with the regulators.
It's also worth noting that the Market Fragility Hypothesis probably will not change the calculus. How many people, in 2025 (or 2027 or 2029 or whenever a heavy-handed administration decides to use leverage over mining pubcos) would actually sell their Bitcoin in response to some sort of nationalization of mining?
Just imagine if we were to discover today that someone had built an assassination market on Bitcoin, and this was being paid out on the blockchain. What would the response be? Would Greg Abbott not leverage all control over ERCOT to have miners try to reverse or censor the transaction? More importantly, what could anybody do if miners decided to comply?
