I’m curious if any of these books are even claiming copyright. If they have registered for copyrights, that could be one way to combat this, because that would arguably be fraud. Copyright is only available to human creators. If the work is substantially generated by AI, it’s not going to be copyrightable. Which points to another possible way to combat these. Just copy them and re-upload them under your name. That’s probably a less attractive method, but it might at least reduce the incentive to make these.
I’m not sure if it’s just the emergent effect of a large number of self-interested scammers, a conscious effort by those looking to wrap up the class wars in favor of capital good and hard, or some combination of the two; but it seems like the creation of a lemon market for intellectual and creative products will have the effect of advancing basically all the downsides of ‘AI’ even faster than its advances in capability would otherwise support.
As we’ve seen with things like IVR systems in customer ‘support’ people are eager to adopt cheaper solutions despite their being years to decades from being remotely fit for purpose; and more or less the whole point of review stuffing and SEO bullshittery is to short-circuit any feedback mechanisms between quality and visibility; and both could hardly be more ideally suited to have just plain bad bot slurry push humans out of the market well ahead of when a contest, in isolation, between their respective costs and outputs would.
I doubt that they bother. This is a very low effort scam by some truly lazy bottom feeders. I doubt many of them have read a full book in their entire lives.
Luckily text ‘summarization’ tools are an entire genre of shifty AI startups.
Add another chapter to the “Monkey Selfie” copyright drama.
I wonder if the flood of AI-generated dreck overwhelming modern day search results could bring some kind of silver lining to librarians and archivists. Now that we can’t rely on search engines to separate the good stuff from the bad anymore we may see a newfound appreciation for human experts who actually know how to sort and categorize information.
Or not… but one can dream.
I. am. definitely. Human. and. not. an. A.I. and. I. think. ChatGPT. is. just. swell. Excuse. me. but. I. must. find. some. scrambled. eggs. to. put. in. my. coffee.
“These books are comprised of…”
Was this also written by some crappy AI?
“He has, like, 15,000 edits, and he’s done almost nothing except fix the incorrect use of ‘comprised of’ in articles.”
No. But I applaud his work.
I used to ghostwrite fundraising letters for MIT, and one of the most important rules was too avoid the use of “comprise” in any form, because no matter what you did, you’d end up with angry phone calls from donors and alumni.
BoingBoing is not MIT, but I’m still a linguistic descriptivist, and firmly believe that the correctedest grammar is whatever people understand.
Just like the modern use of ‘impact’ - because people don’t know how to differentiate between ‘effect’ and ‘affect’…
Which is not very goodly
… ah yes, a word so confusing that not only is it always used wrong, but all the pedantic prescriptions for how to use it are also wrong
The only honest reviews are negative.
oh great. do i like this comment, or do i unlike it?
I’m with Lingthusiasm when it comes to grammar (and recommend the podcast to anyone who might be enthusiastic about linguistics.)
Easy to remember
effect is a verb meaning “to cause to come into being”
affect is a noun meaning “the immediate expression of emotion”
Yes, it is easy to remember. But apparently not for a great many people who wouldn’t know their arse from their elbow…
Or, worse, impactful. For when impact isn’t full enough.