A few points:
1) maybe you didn't follow my argument: that's why it's farcical to believe that these companies are wasting money buying/destroying them when they are digitally available. Why would they waste $$ they don't have to?

2) FWIW, that's the 1901 edition, the 1915 3rd ed was (as the author says) largely rewritten with updated information not available for the 1901.

3) Do you realize your comment is literally justifying scanning and preserving books that aren't yet available?

Isn't there a clear bright line of liability here?
If the hugging face event cost hugging face (hypothetically) $250,000 to fix, how is the ai creator not instantly on the hook for a multiple of those damages?

It feels like we should quickly establish say a 10x multiplier as standard penalty for ai caused damages.

I know it's a shitty post facto tool and yes, it has failed plenty of times, but product liability is a well-precedented path (ie no new law approach has to be created) to compelling corporate responsibility. If this were the case, even a trillion dollar market value would feel shallow compared to potential liabilities.

Someone please name one rare, valuable book that has been destroyed. I'm still waiting.

Nobody has answered my criticisms of this particular Frankenstein story:
1) there's essentially nothing IN rare books that LLMs benefit from
2) rare old books are difficult to ocr in the first place, when reprints are both easier to scan AND cheaper.

This entire contrived hysteria is even more amusing as people basically don't read anymore, not generally. Oh heavens! They're destroying the books NOBODY READS? Really? That's what you're weeping about?

The only remaining meaningful contemporary book sales are women reading their own particular flavor of smut.
https://www.pensandpoison.org/...

Less than half of adults say they've read ONE BOOK in the last year.
https://www.theatlantic.com/ma...

I love old books. I'm currently reading Rogers 3rd ed (1915) of A History of Babylonia and Assyria and it's delightful. An absolute pleasure to read from the skill of the author, the comprehensiveness of the information (without a tedious agenda to flog, imagine!), to the very tactile experience of leather bound, deckle-edged paper. A joy every minute I'm reading it.

I would genuinely join you if there truly was a threat against old books.
But this? This is a manipulated hysteria contrived for some reason, maybe an experiment to see if they could spur nearly-illiterate sheeple to rise up to 'save' something they absolutely care nothing about.

The new system is simply another set of inputs. You will still get search results even if you don't click on anything. In fact, I would bet that actual user interaction with this new system is going to be very small. Most of the input that Google will get will be people trying to game the system.

I didn't even know about Kagi. In fact, it has been a long time since I have even considered using someone besides Google. I just resigned myself to the idea that things were forever broken. Thanks for the insight.

That's literally what I said, you dumb pillock.

The fact that Canada is insisting on their right to propagandize their own populace doesn't change that it's propaganda.
Of COURSE it's still 'soft power', even if the target is their own disaffected population.

Duh?

Yeah, you're not understanding what I'm saying: Azure doesn't just host Linux VMs, significant parts of the infrastructure are built on Linux. Microsoft has just as much (if not more) to lose from a resurgent SCO lawsuit than anyone, they aren't going to be the ones rocking this boat.

You mean the site whose water discharge processes were indeed certified shortly after that article in 2025?

https://www.faa.gov/media/9434...

Water discharge is covered in section 3.2.7, fyi.

"why (is) the U.S. government is so invested in the issue?"

Claiming it's 'perhaps' about the US wanting power to project US culture...as a way to deflect from Canada ACTUALLY OPENLY AND CLEARLY DOING THAT is propaganda-judo at Iraqi-information-minister level.

Even *asking* the question is hilariously obvious. So Canada is entitled to use its streaming channels to promulgate Brand Canada (and their obligatory Francophone appendix, always...) but the US is quasi imperialist - not for demanding the right to do so, just interfering with Canada's 'right' to do so.

And offering some sappy Ontario postdoc as the "quoted expert" instead of, say, an actual Canadian gov't official to defend their overt policy is even funnier.

And before the nabobs come chattering into the conversation, yes, OF COURSE the US promotes itself in media and indeed that is soft-power, but let's be clear here that the US isn't asking for the right to do so, just that Canada NOT do so. And oh, when's the last time you saw the majority of US media (incl Hollywood, to look at the bastion of American cultural power) saying /nice/ things about the US?

Except:
A) we are talking about sea temps 10m below the surface across a global data set. Some Enlightenment polymaths taking readings with their home made thermometers in Europe wouldn't impact that data set at all.
B) "guestimates" would be hilarious to include in a conclusion being reached based on HUNDREDTHS of a degree (0.10 vs 0.09)

My supposition would be this is modem data only, eg since the 1990s MAYBE.

That is a very good point. duckduckgo.com (to use an example) does not summarize the search results for you, and, once you know what to look for the ads are clearly marked. To be honest, purchasing ads on duckduckgo.com is likely a sign of good taste on the part of the advertiser as well. Honestly, I hadn't even considered using a non-Google search option for a long time. That is almost certainly part of the problem. Thanks for your response.

One could argue that no one is doing it in six days without a shit ton of automation support. Some cherubim in ops saved eons when he wrote createstar.pl and it's still running in production today (though it looks like creategalaxy.pl stopped executing some time ago and hasn't been restarted and/or patched yet).