That will be your conscious choice because it will still be marked as fake meat. You are denying your readers the chance to make that decision.
‘I didn’t want to tell people it was fake meat because it may trigger their prejudies against eating fake meat’ would get short shrift from the Food Standards agency…
Do you really think my work is fake and fraudulent?
I don’t mind if you do!
I’ve tried not to personalise this, Clarke. I don’t doubt your sincerity at all or your knowledge or credentials in your field. My concern is purely with the transparency.
I think the failure to tell your readers openly the extent to which you used AI (which goes beyond the initial help with thinking and spell checking that I think both of us are happy with) denies them a legitimate choice in a currently very controversial area, and that should this come out subsequently, you risk tainting in retrospect what you have written for those (increasingly many) people who are worried about AI and the activities of its corporate pushers.
On a personal level, I believe that the core skill of an author lies in the ability to express their thoughts, so that I would view needing AI to write the text as a personal failure, but that’s for me. I understand that others may see things differently. As long as they let the reader know, that’s fine: it’s the lack of transparency I find troubling.
I suspect that the way the world is going such concerns will be swept aside, but I do think they’re worth registering while we still have time.
OMG now I think you think I’m a failure! Ha ha. Only kidding, I don’t really think you think that.
Seriously, it’s good and generous of you to share what you really think.
But, also seriously, I reckon you’d change your mind if you spent some serious time thinking and writing with AI. I did. It’s been like an ebike for my mind. And I’m expressing my thoughts, in my way, and the thoughts are better because I’ve bounced them around a lot with an intelligent coworker.
It doesn’t. I tried a black bean “burger” once. It tasted like charbroiled cardboard. 
Ha ha. I can only imagine!
It’s not principally the ‘discussion’ which causes me personally not to use Gen AI, it’s the ‘writing’.
If you find that the initial process helps you to refine your thoughts, then I’m not seeking to stop you, of course.
But as for writing, you clearly envisage more than ‘expressing my thoughts, in my way’:
They’re better at writing than most people. So that’s a win - who doesn’t what to read stuff that’s well written? That said, you might find that you don’t like the writing, so you use it as a first draft and clean it up.
I am not just not convinced that tweaking something produced by an algorithm constitutes ‘writing’ in any meaningful sense, and it is not something I wish to do.
More generally, I am not particularly happy with the way the models have been trained, almost certainly on ‘stolen’ data from living authors who have not been recompensed for the use, at great environmental cost. Of course, the rush to shove AI into everything in the vague hope that it will make money means that I am subject to it all the time, but when I do have a choice, I choose to avoid it, and as far as I can see, I am not alone in this.
Hence the need for writers to be honest about their use of AI upfront so that those who wish to opt out can, or if they continue to read, can do so knowing what they’re reading.
This seems to me to be such an obvious point that I’m intrigued by your unwillingness to adopt it, given that you’re very open in this forums about the process (and I genuinely do thank you for that). Why is it such a big step to include a short paragraph in the introduction or acknowledgements when the information will benefit the reader?
As an aside, some publishing houses now require their authors to guarantee that they have not used Gen AI in writing their book.
I must respectfully differ. I am someone who does study the work of other artists for inspiration. There’s a big difference between learning to create with your own voice through the study of someone else’s work and mimicry. What you are doing is much more akin to hiring an illustrator to mimic the work of someone else to give you the kind of image you’d like to accompany your blog post—except no one is being compensated.
PS: You could have downloaded this image for free from Unsplash, given the photographer, Nataliee Shell, credit, and perhaps lead others to her work. It took me less than five minutes to find this image on Unsplash.
Fair enough! Good luck with your retirement.
Interesting, ChatGPT 4o now refuses to make “Miyazaki-style” images due to “content rule violations”. Probably a good thing to not be so explicitly derivative.
It will accept “use a style reminiscent of popular 21st century Japanese animated film”.
Also, this long introduction to image generation in 4o is helpful.
Katie
I use both free and paid stock image services and have found it takes so long to find the exact image I needed for a specific use that I found myself regulary using AI image generation instead.
I am a photographer/hobbyist, but for business use taking a photo myself is too time consuming and usually not cost effective.
I am, however, encouraged by what Adobe Stock library is doing (and I assume others too): They have added the ability to manipulate stock images with AI commands to adjust the image to be more suitable.
I don’t have the legal background to evaluate whether Adobe’s AI claims of using licensed work and not scraping/stealing data is truly different than other AI image generation services, but I do have a few tiny checks from Adobe Stock service paying me real dollars (or should I say cents) for use of my stock images that I have contributed in the past when I opted-in to allow them to be used for AI purposes.
In the meetings I remember, nearly everyone’s head was somewhere else.
So I did, @Clarke_Ching
The book is just fine to me, it’s a short and easy read. Granted, it will not win any prizes and I’d say sometimes it seems written for younger people, it serves the purpose as a terse introduction to the topic (couldn’t resist thinking about process mining).
But getting to the point: the book feels like written by a human, and in my opinion delivers the promised value. I don’t care if GenAI was used to write it and I believe the provenance acknowledgment was not necessary but is a fine touch of honesty.
That’s for reading it - especially since it’s not the sort of thing you’d normally read.
Glad that you picked up the “written for children” vibe. I based it on some kids books from the 1970s or 80s called Encyclopaedia Brown - little detective books. One day - probably 18 months ago, I think - I whimsically asked chatGPT to write a chapter about a kid bottleneck detective based on those books, and it did a terrible job, but it kinda got me the vibe I wanted, and I did an enormous amount of rewriting since then.
