Hello, everyone! How are we this afternoon? Anxious? Same, me too!
It’s just been a lot, lately. I’m relishing in whatever goods news I can.
My good news: I get to celebrate three years with my wonderful partner this weekend and I have the opportunity to do some interesting writing work in a couple weeks. I also have some really great cheese in my fridge; it has a rind washed in balsamic. Yum!
What’s your good news?
…
Big thanks to Gail to sending this one over and apologies for not sharing sooner. There is a Kickstarter for a photo book, highlighting groundbreaking romance authors. The campaign is due to end in around 12 hours.
…
The Trans Rights Readthon begins this Friday, March 21st!
…
Diversity in historical romance discourse is making the rounds again, after a particularly disastrous set of comments surfaced in a romance-related Substack. I feel like we revisit this topic every couple years.
…
Sarah: Dr. Peter J. Barclay from the School of Computing at Edinburgh Napier University sent me some information about their research into AI written books, and reached out based on my write-up of all the AI narrators. Dr. Barclay has a quiz to see if people can distinguish a human-written text with an AI generated text.
…
Don’t forget to share what cool or interesting things you’ve seen, read, or listened to this week! And if you have anything you think we’d like to post on a future Wednesday Links, send it my way!
That quiz was interesting . My score was 11. 11 correct or 11 incorrect, it didn’t say. Many of the samples included Hercule Poirot.
14 out of 20. It seems like all of the passages were taken from Agatha Christie, which is fascinating, as a Christie buff. I don’t know if that made it easier or harder!
I got a score of 14. I wish it had said what the sources of each was in the end. They all felt fairly old-fashioned, which may be because it’s easier to use excerpts of out-of-copyright books without permission, but styles have changed. Ironically, since a lot of AI is likely trained on Old books out of copyright, AI tends to have a stodgy feel with more telling than showing.
It might be entirely different if we were asked to choose between modern writing and AI. Or an even better question might be, “What do you think is good writing?’ Plenty of people write as poorly as AI, so I wouldn’t automatically assume mediocre writing is AI, even if I judged it as something I wouldn’t continue reading.
When my eyes pass across unavoidable AI “what readers are saying” on my way to actual reviews, the sentences are always short and choppy. That’s what I was looking for in the quiz.
11 out of 20. I had no idea what they were, AI or human written. I’d like to know what results obtained if someone chose all AI or all human.
I scored 13 out of 20. I assume that means 13 of my answers were correct.
I think all the human ones were from Agatha Christie mysteries (probably early Poirot and Tommy and Tuppence would be my guess) and the AI-generated ones were probably imitations of her style.
So if you’ve read Christie recently it would probably really help you in picking out the passages actually composed by Christie versus the AI-generated ones. But it’s been years since I’ve read any Christie and I never really thought that her prose style was all that great. It was how she plotted her mystery that was the best aspect of her writing, not how cleverly she put words together. And smart plotting is not going to be revealed in such short passages.
Are other folks doing the Trans Rights Readathon?
I’m going to read Second Chances at New Port Stephen by TJ Alexander for transmasc rep, The Sapling Cage by Margaret Killjoy for transfemme rep, Lakelore by Anna-Marie McLemore for non-binary rep, The Bruising of Qilwa by Naseem Jamnia for ‘outside my experience’ (because my Persian pal just celebrated Nowruz, so a Persian-inspired non-binary fantasy seemed like a good pick), and Love After the End edited by Joshua Whitehead for Two Spirit/Indigiqueer rep.
Happy Anniversary Amanda!
And lol, did Dr. Barclay really write that quiz or did he use his AI? I scored 15.