Hi there! I know it’s been a bit, and I am working on something I hope you’ll really like, but I felt like I should take a break from that to spotlight the real harm being done in the process of training generative AI systems, by sharing the voices of those who are being harmed.
First, here’s an extensively researched (sources included) piece by Regula Ysewijn, food historian and cookbook writer:
Others who I follow who found themselves in the searchable database are:
There are plenty of other cookbook authors whose names appear in the database, whose books have been used without their permission. While the conventional wisdom is that you “can’t copyright recipes”, you can definitely copyright the headnotes, cooking directions, photos, footnotes, research, and cohesive product that is YOUR cookbook. When a food writer is inspired by, or adapting someone else’s recipe, the proper etiquette is to give them credit in the headnotes. Obviously, no AI training system is doing anything like this.
Another food writer and coach who has been pirated by this system is:
As it happens, Dianne recently published a couple of fascinating articles about “recipes” being generated by AI. Here are some links to those:
If it isn’t clear to you, food writing is work. Recipe development is even more work. At its best, it is creative, original, personal, an expression of unique humanity and culture. AI-generated recipes and food writing are not any of that, but the exact opposite, sucking all the vitality out of other peoples’ creations, and trying to appear human. Overly dramatic, maybe. But if I had a published book in that database, that’s how I would feel. Like AI was a vampire in the night.
You can do what I did, and sign up for a 1 month free trial of The Atlantic, and play around with the database, to find out which of your favorite authors’ lifeblood has been sucked up by AI. Be prepared to be alarmed.