That’s progress

If you’re wondering if I bought Henry Windsor’s much-anticipated book, the answer is no. Apart from anything else, by the time it was released in English, I’d already read and/or heard about most of it from the various newspapers, especially those in the UK. I don’t know anybody who has much sympathy for what Harry has done.

I’ll admit to being fascinated by the whole train-wreck but I have to wonder what he and his wife will do from here? Maybe they could write a second book, same story, but from Her point of view?

That’s enough about that.

The sudden death of Cardinal George Pell saddened me. I think he was the victim of the worst case of wrongful conviction I’ve ever seen, as I explained here. Despite the fact that his conviction on child sexual assault charges was categorically thrown out by the High Court, some would still like to think him guilty. At the very least some blame him, as a bishop, for shifting paedophile priests between parishes to protect them. I’ve read defences against that accusation, too. And then there are people who wonder about his unexpected death, attributed to ‘complications’ after a straightforward hip operation. He’d been given the job of rooting out corruption and fixing the Vatican’s financial systems. That task made him many powerful enemies, among them the Mafia. Who knows? Maybe somebody could write a book about it.

Speaking of books, I’ve recently seen an accusation that people who use artificial intelligence to create art are committing an act equivalent to book piracy or plagiarism. Even writing it down like this it sounds silly. If somebody steals (say) my book, Morgan’s Choice and sells it on another site with no profit coming to me, that’s piracy (and yes, it even happens to obscure authors like me). If somebody copies the text from my book and publishes it as their own creation that’s plagiarism. It’s still plagiarism if they change some of the names but leave the rest. But anybody can read the book and write their own version of the same story. And every writer uses stuff they’ve read in a book somewhere before. Reading begets writing.

Going back to art, nobody can stop me from painting my own version of Streeton’s painting shown above. But if I try to pass it off as Streeton’s painting and sell it, I’ll get into trouble. But if I try to copy the artist’s style in coming up with a new work there’s no law against that. ‘In the style of’ happens all the time and always has. Paintings by Rembrandt’s students have more than once been attributed to the great man himself. In fact, the AI-generated picture at the top of this post is a river scene in the style of Streeton created by Midjourney.

In an article in American Scientist called AI is blurring the definition of an artist, the author writes,  “To create AI art, artists write algorithms not to follow a set of rules, but to “learn” a specific aesthetic by analyzing thousands of images. The algorithm then tries to generate new images in adherence to the aesthetics it has learned.”

To me, that’s not dissimilar to what a writer, or an artist, does in trying to learn how to create a new work. There is a fine line when it comes to faces and people. Midjourney, the AI I use, allows the user to ask for something like ‘Chris Hemsworth’ in a prompt. This tells the AI that it should create the image of someone who has the attributes of Chris Hemsworth (tall, handsome, blond, bearded, strong, young), not the man himself. To do that, the algorithm will no doubt scan images of the actor which will probably be copyright. Perhaps in the future owners of copyrighted pictures will be asked to give permission for their works to be scanned by AI bots. That’s not the case at the moment, hence accusations of copyright violations.

Meanwhile, the world is beginning to understand that AI art is here to stay. An AI-generated work “Portrait of Edmond Belamy” sold at auction for US$432,500 in 2018. [source] Stock photo sites (where people go to buy images for book covers, ads, and so on) are realizing they’ll have to accommodate AI-generated art. Dreamstime, one of the largest sites, says

• Contributors must have all rights for the generated images (please note that some AI tools only give complete rights to imagery generated through them if you register for paid service).
• Image description must state clearly that the image is generated with the use of an AI
• One of the categories selected for the image must be Illustrations and Clip Art/AI generated.
• Contributors will not upload images of generated people’s faces as it is impossible to provide a model release for them. [source]

Whereas there are plenty of artists who complain about ‘stealing’ and copyright infringement, there are plenty of real artists who see AI programs like Midjourney as just another tool. I have an online friend who produces stunning art through Midjourney, work that a mere journeyperson (I’m sure journeyman is no longer acceptable) like me can’t hope to emulate.

Many years ago, managers would write letters by hand, then pass them along to the gifted, ten-fingered girls in the typing pool to produce the finished, readable work. Then in the ’80s (I remember it well) somebody invented a word processor so the girls didn’t need to use white out. And just after that, managers realised they could write their own letters on word processors and not have to use the typing pool at all. I’m sure a lot of the girls in the typing pool, finding themselves out of a job, were Not Happy.

But that’s progress.