Phillip Winn @pwinn@qoto.org

Since a previous post alerted me to the idea of an AI art model generating images of Buddha statues without having "Buddha" as an input, I decided to test it, so I prompted MidJourney with "cree, pasadena, ish, informal, preview, lindo, kids, potential, statue, calcutta, phenomenal, sigma, chero, heh, âľĵ, kier, bourdain, anjali, ori, displa"

The first time I forgot I still had the "anime" filter turned on, so I switched to MidJourney model v4 and re-ran it.

I'm not seeing Buddha in my results, although I do see statues. Then again "statue" is one of the words in the prompt, so nothing magical there.

I'm including both results because I guess I want to spend a long time writing image descriptions. 😀

#AIart #midjourney

This site presents "both sides" of at least one aspect of AI art by displaying the content of a current lawsuit over AI art on the left, and via fair use, a response or rebuttal on the right.

http://www.stablediffusionfrivolous.com/

(Replace frivolous with litigation in the URL to get the original site.)

I'm obviously a fan of AI art, and occasionally post it here. I'm also very far from a copyright maximalist, although I'm *not* in favor of abolishing copyright altogether. I agree with the right side of the site on at least two points:

1. Copyright law doesn't apply here.
2. The suit dramatically misrepresents, I think because Butterick doesn't understand, how latent diffusion works.

That said, while I overall agree with the critique, I think I'm more sympathetic to the complaint than the rebuttal writers are.

There's something truly disconcerting about adding an artist's name to an existing prompt and seeing the result change to be much closer to something in line with previous works of that artist.

I can understand that statements about how latent diffusion works fall flat after seeing that. No, there are no images stored in the model, not even textual representations of images. And yet, there's something, right? It's not nothing. There's some sort of minimal association of given names with a given set of characteristics, more so for some than for others, and the effect is often uncanny. The example of what seems like an arbitrary series of words consistently producing Buddha statues is a good start to counter that, but it doesn't quite seem adequate.

Overall, I think there are serious ethical issues with how art is being sourced for training the current generations of AI art models. However, I don't think there are legal issues; copyright certainly isn't being violated.

More to the point: the primary users of these tools currently seem to be artists. In that sense, it really does seem like the advent of photographs, or the advent of photoshop, and it will likely be quite a while before we will know how these issues settle. In the meantime, anybody using these tools to try to mimic a particular artist is probably missing out.

#blog #AIart

The CSV spreadsheet is usually the one that's a day ahead, so I guessed it was the one using GMT. Except... every now and then, it's the one that's a day behind. The inconsistency rules out GMT shenanigans, I think, but also, I don't ever supply a time when marking a book as read. I select a year, month, and day. Which suggests it's arbitrarily storing a time, and the fact that it's off so often suggests it's storing a time within five or six hours of midnight. And apparently inconsistently.

Another amusing bug anecdote is that books I finished reading on Oct 1, 2022, are just listed on the site as having been read "Oct 2022." No date shown. The CSV spreadsheet has a date, of course.

It seems like there's an entire class of date/time bugs on GoodReads, before we even get to the missing ISBNs and badly-corrupt data from 2017.

#GoodReads #Amazon #nerdery #blog #Bookwyrm

I'm not sure which is worse, exactly: GoodReads exporting my library with no "Date Read" filled in, despite having a "Date Read" on the site, or GoodReads exporting my library with the wrong "Date Read" filled in, one that is a day off from the one on the site.

How does that even happen? Given that it's off by a day when it's off, I'm guess it's a time zone issue, with either the site or the export function using GMT and the other using my local time, six hours removed.

That doesn't explain the many rows in which the date is just missing. Nor the missing ISBNs.

It's been clear for a long time that Amazon doesn't actually care about making GoodReads a GoodSite, but this is really, really bad.

What I've settled on doing is opening the export CSV in a spreadsheet in one pane, sorted by reverse "Date Added," and looking at "My Books" on GoodReads in another pane, sorted the same way.

The Arc browser from https://arc.net/ is working pretty well for this.

#GoodReads #Amazon #nerdery #blog #Bookwyrm

As I continue to explore GoodReads with an eye to exiting that site, I am amazed at just how terrible it is. I exported my library, intending to delete everything that wasn't read in the last five years, but the resulting CSV file is missing so much!

There is no "Date Read" for many books that are only present in the file because I read them, and a surprising number of books are missing ISBNs, too. I'm talking about books in print, not ebooks or audiobooks that might not even have ISBNs.

It is clear that Amazon does not prioritize accuracy in export files, and as I explain previously, they don't prioritize accuracy in their own database either, although the effect is seen in different ways.

The long and short of it is that I'm apparently going to need to manually edit the CSV file in order to get anything like accurate data, and giving up the convenient GoodRead mobile app for a BookWyrm server seems more than ever like a fair trade. I'll happily pitch a few bucks a month toward @bookrastinating for their help getting things set up, and to avoid propping up Amazon's less-than-lackluster efforts with GoodReads.

This seems like a good way to spend a day off work, right?

#GoodReads #Amazon #nerdery #blog #Bookwyrm

When I decided to jump into the fediverse with both feet, I signed up for a BookWyrm account at @bookrastinating. When I tried to import my history/library, there were two big issues.

The first was timing. I joined and started an import at about the same time many, many, many other people did, and the server promptly fell over. The owner of Bookrastinating was helpful and friendly and eventually the queue started moving again, so that's not an issue anymore.

The second was not actually a problem with BookWyrm at all, but with GoodReads! It turns out that a bunch of books I'd previously tagged as read on GoodRead were no longer the books they had been. No offense to Marc Blake, author of “How to Be a Sitcom Writer: Secrets from the Inside,” but I’ve never read that book. When I “shelved” it on January 9, 2017, I assigned it “baroque, cycle, fiction, hardback, series, read,” so it seems very clear that what I actually shelved was a book by Neal Stephenson, and checking now shows that GoodReads only knows I've read the second book in the tagged "baroque cycle" series, but it has lost the first and third.

I'm not sure when the GoodReads database was corrupted, and in a review of my Reading Challenge book lists for 2022, 2021, 2020, 2019, 2018, and 2017, everything seems probably-correct. But my 2017 Reading Challenge also does not include the book I mentioned above, the book that caught my eye as clearly-incorrect.

When I look at books *shelved* in 2017, I see many that are definitely not right, although they don't have "read" dates. I'm sure <cite>After the Martian Apocalypse: Extraterrestrial Artifacts and the Case for Mars Exploration</cite> is a perfectly fine book, but it definitely wasn't what I tagged with "fiction paperback series dragon king trilogy" on January 9, 2017. <cite>Love on the Dotted Line</cite> could be a fantastic romance novel, but it's not what I tagged "paperback anthology science fiction" on January 9, 2017. A dual-language collection of Italian women's poetry is definitely not what I tagged as a volume of the "Writers of the Future" science fiction anthology series. And so on. The ones I notice most easily are the titles I would never read, but the date January 9, 2017, stands out. Sorting by date added, I can see that some of the books added on that date seem correct. I know the books, and the tags match the books. Most do not.

I'm sure there's some irony in an import failure on the fediverse alerting me to serious corruption on Amazon-owned GoodReads, and the result stopping me from actually migrating.

I clearly cannot trust GoodReads, as they've broken the first and second rules of a database: they've lost data, and represented data falsely to belong to me when it doesn't. I'm not sure which of those is the first rule and which is the second, but both seem bad.

I also cannot import my entire library from GoodReads into BookWyrm, because I don't want to start with bad data. I think it's time to let most of the past go, and create a cleaned-up import file with just my reading history from 2017 onward.

Good thing I have the day off tomorrow!

#GoodReads #amazon #nerdery #blog #Bookwyrm