There's a very implicit assumption in this text that the encyclopedia you buy in 2024 does NOT contain all sorts of AI-generated stuff.
Not sure I have that much faith in the world anymore to trust that to be true.
Not sure I have that much faith in the world anymore to trust that to be true.
I wouldn't want to make that bet for an encyclopaedia published in 2027 though.
Two opposite and equally nonsensical reasons to dismiss physical books. The more confused people get the higher the value of clarity of mind.
I also don't think you actually read the articles you linked to. The first one says that the publisher is planning on using AI, it doesn't say that they have already. As I said in the initial comment:
> I wouldn’t expect a publisher of encyclopedias to implement AI tools for years.
Meaning, an encyclopedia bought today won't be using these tools. Which is what the grandfather comment was about.
May 16th: Wiley shuts down journals over AI use: https://www.theregister.com/2024/05/16/wiley_journals_ai/
June 13th: Wiley book series is going to be using AI: https://qz.com/for-dummies-books-publisher-john-wiley-sons-a...
My interest in Wiley is that my late uncle helped them to "go digital" decades ago, and remained remotely as a proofreader, editor and fact-checker for physics titles. He would be spinning in his grave if he weren't cremated. Honestly depressing to see it happen.
Ten bucks says there's LLM-generated content in encyclopedias today because LLMs have been mainstream for a few years now, and even prominent scientists and experts have been caught out "writing" with LLMs.
https://arxiv.org/abs/2404.01268
Why wouldn't there be LLM-generated content in encyclopedias today if it's been (ab)used for several years now? Seems a little naive, especially when even Wiley (wow) is openly embracing AI now with a sudden heelturn.
As I said, I wouldn't be surprised if these companies start using LLMs in the near future, but as I said in another comment, I think the average tech person vastly overestimates how much the average person keeps up-to-date on the latest LLM developments. They also tend to overhype how much society will be transformed by tech, largely because getting VC dollars depends on that hype.
These things are still in their infancy and I'd expect them to be used soon, but the people running encyclopedias are not so up-to-date on the latest LLM developments to have already incorporated them into publications planned years ago.
In general tech workers on HN tend to overestimate how familiar with technology non-tech workers are.