Why would you think they make good targets for vandalism? A major data center would have very restricted physical access.
FaceDeer
The 2022 estimate for how much water was used in Texas in total was 15.2 million acre-feet, or approximately 5 trillion gallons. So these AI centers are accounting for 0.00926% of Texas' water use.
I would suspect they're thinking of robots with "brains" that are separate from the physical chassis, perhaps in a cloud server somewhere. If you've got a factory full of robots tromping around under wifi guidance you'd want that to be a reliable connection.
The lawsuit between NYT and OpenAI is still ongoing, this article is about a court order to "preserve evidence" that could be used in the trial. It doesn't indicate anything about how the case might ultimately be decided.
Last I dug into the NYT v. OpenAI case it looked pretty weak, NYT had heavily massaged their prompts in order to get ChatGPT to regurgitate snippets of their old articles and the judge had called them out on that.
I was specifically speaking in USD terms, take a look at the page I linked above. It has graphs with both USD and Bitcoin on them. In Bitcoin terms Lightning's capacity peaked in early 2023. In dollar terms it was December 2024. The line is squiggly and has a general long-term upward trend overall on the net dollar capacity, but it really doesn't look very impressive compared to Ethereum's layer-2 architecture. And that's the only line where I see a long-term rise, the rest have been stagnant or declining for years.
I stand by my overall view that Bitcoin's technology is simply obsolete. It doesn't do anything, it just sits there being valuable because it's valuable. I don't think that's going to endure forever.
There's already been a summary judgment in this case ruling that the AI training activity was not by itself copyright violation.
First-mover advantage is powerful.
To comply with copyright law, not to skirt it. That's what companies that scan large numbers of books do. See for example Authors Guild v. Google from back when Google was scanning books to add to their book search engine. Framing this like it's some kind of nefarious act is misleading.
No it hasn't. Again, according to that link I provided, the total capacity of Lightning peaked in December 2024. These are not the graphs of a growing layer 2, it's been stagnant for many years.
Bitcoin simply wasn't designed for this sort of application, and Bitcoin's foundation layer is absolutely frozen due to the ideology of its users and developers so I don't expect the situation will improve. If you want to do a layer 2 then why not use a blockchain that's specifically designed to support it? If you use Ethereum you can even use token-wrapped Bitcoin as your medium of exchange. There's $14.4 billion dollars worth of WBTC on Ethereum available for exchange, as opposed to the $440 million worth in Lightning channels.
The comment you're responding to linked to a page giving statistics about the Lightning network. The number of channels peaked in 2022 and has been going down ever since then.
This was a really interesting reply, thanks. I'd leave a longer response, but honestly I really need to be asleep right now.
No problem. It's past my bedtime too, but I'm really pleased that I'm able to discuss this stuff and I'm not getting downvotes or called a shill simply for providing information. It's always been a big area of fascination for me, the technology is really neat. :)
You can definitely get started for $1000.
Sure, you could set up something that can process blocks. But there's no way you'd be able to make a profit with something that small. One of the fundamental tenets of cryptocurrency is that it doesn't rely on anyone acting altruistically, it assumes that everyone involved is in it for the money. It leverages greed to ensure that everyone "follows the rules", by making it so that if you break those rules you make less money. So I wouldn't consider a blockchain to be secure if it depended on miners who mined at a loss out of the goodness of their hearts. When people worry about centralization they overlook that Bitcoin has economies of scale that massively favors the bigger mining operations, the dollars-per-hash are much lower for the warehouses full of ASICs next door to a power plant than for the guy with a graphics card in a closet at home.
I did also mention that you could get involved in staking on Ethereum for much less than $120,000, at the cost of depending on third parties to handle the actual validation. You can do that either through staking pools or liquid staking. Essentially, you own a "share" of a single validator's stake and get a proportionate portion of the validator's rewards, minus a fee that the validator charges for actually running the validator.
That is how water use works, yes. The water goes back into the environment and is later reused.
Also, there's a good chance the AIs are not being trained in the same facilities that they're later being run in. Different sorts of work is being done.