I'm not sure what to make about this, so I wanted to hear some of your opinions

Photojournalist uses Midjourney to create intense new photo series
www.digitalcameraworld.com/news/photojournalist-uses-midjourney-to-create-intense-new-photo-series-is-it-cheating

NFTs are pure grift to begin with. It's certainly not photography and it'd be a stretch to call it journalism. I'm sure it'll drum up some outrage and online discourse which is probably the main goal.

Agree, I find this part insulting:"The photorealistic quality now possible with AI"

Beam me up, Scotty !

I think as long as he's up front and very clear about what he's doing -- and clearly so, not down in some small print somewhere -- then I don't think it's unethical.

I also think that many of the example images in the article are really not good, with unnatural and downright bizarre artifacts that are not uncommon with AI generated images. If we're worried about people not being able to recognize AI images as such they're going to need to be better than these.

As far as NFTs go - same deal, as long as he's completely transparent about what it is, it's fine. The only reason I can imagine anyone would want to buy one is that they expect it to go up in value so they can make a profit selling it, and that's fine, but I don't see why anyone would think any of these are going to have any value...

Moody Landscape created with Midjourney AI
Image: www.flickr.com/photos/jae_at_wits_end/52279110981

“This is not an image editor.
You can get images based on your photo and prompts. But you are not going to get a post-processed version of your photo.”
expertphotography.com/midjourney/

I'm with chd up there. As long as it's disclosed, anything goes. It's when people don't disclose it, the issues begin. We can all manipulate and create something that wasn't there, always be open about it though IMO and there is no problem.

Danny.

I should add that I also do not consider it journalism, and he should be clear that it is not.

I find tools such as Midjourney and Stable Diffusion very enjoyable to play with and they can output pretty impressive images.

But that's a very different thing than photography. I'd rather prefer the resulting images to be described as digital art, or maybe in case of images with photographic style like these, synthetic photographs rather than "post-photography" which sounds frankly a bit pretentious in this context.

Not journalism. Not photography. Not photo series.

To me, the purpose of photography is documentation. If the scene never existed, it's neither photography nor documentation, and it has no purpose if it's not "eye candy", which itself has very limited value.

Seems no different setting up a scene, or pose for a portrait. Depiction is as valid as documentation. Same goes for cinema, with actor, script, costume, set design, CGI, etc. As a photographer, I never touch or alter the scene out in the field, but not all photography has to be like that. AI just another dimension.

The problem is the sheer number of people who will take the "photos" as truth, disclosed or not. Especially when the "photos" are distributed on social media, citing the "photographer" as the source, but not disclosing the nature of the "photos". Especially when this is done by people with an active agenda.

Like any technology, nothing inherently wrong with what he's doing (assuming full disclosure), but as the past few years have shown all too clearly, truth and honesty play less and less of a role. Still, once the genie's out of the box, there's no putting it back, so I guess it's just one more thing we'll all have to live with. In short, deal with it. : )

It's not, as it never happened.

Nothing wrong with creating it, but to assign some sort of copyright protection to images that are themselves created in part from automated plagiarism is not valid, is arguably unethical, and for the photographer involved, somewhat embarrassing.

Ai is nothing more than a shortcut. And if that's how you want to live your life, fine, but where do you draw the line in your short cut journey through life?

Those images are fine storybook illustrations, but have no value or authenticity and are infinitely copyable without assigning any credit or financial compensation.

One human experience, it would be such a waste of life to short cut or fake it.

100% agree, to be lived with. Plus the person spending a few seconds typing in a few words shouldn't expect to be respected as a creative, shouldn't expect to be financially compensated. Ai is a shortcut, nothing more. It's just eye candy. It's creators should be focusing on curing cancer not providing a way to create generic text and images. But we're selfish money-first species now aren't we...

Wow, a lot of judgmental stuff in here to be sure, and have you thought through how this works as a general case?

Copyright protection is a question of law. I don't imagine you are a copyright lawyer, so your opinion on that isn't going to change much of anything.

"images that are themselves created in part from automated plagiarism [...]"

Is it the plagiarism part that you have a problem with, or the automation of it? Or both? When exactly does imitation become plagiarism?

So what? We all take shortcuts, even you I dare say. Do you post-process anything digitally? Well isn't that just a shortcut compared to doing it with chemicals and lights and gels and timers in a darkroom? Explain to me why this is worth less. If you refuse to use modern tools just because you're old school you're only putting yourself at a disadvantage. I've done a fair amount of audio production and I'm pretty good at editing audio tape with a razor blade and a splicing block, but I'm very happy to never have to do that again.

I draw it a bit short of recognizing your right to judge my choices. I will however stay off of your lawn.

Exactly.

I like it for the bold experiment that it is, and being clear upfront that they are not real images. The risk does exist of people quoting / copying / circulating those images and them being assumed as real photographs. Technology with counter-measures is needed sooner than later.

He did use words like photo and reporting for lack of better terms. I would like a new word coined for images created with AI. They are not photos created by "photography" in the sense of using light, lens-like mechanism and a medium to capture the light. It is like painting without ink or colors. It is a new form of creating images that we may not agree with in all circumstances but is interesting in some situations.

I can call them AIGI, for "AI Generated Images". It would be even more useful to give them their own file type or extension, like "men-on-boat.aigi" and "sinking-boat.aigi". Additional technical features can be added to clearly distinguish them as AIGI and prevent their copying or misrepresentation as a natural photograph.

The words experiment and illustration are the main theme in his post. If viewed from that point of view, it is interesting what the possibilities are. For example, one thing that always comes to my mind is practically unlimited resolution. Just like we see online maps at one level and zoom in, AIGI could be zoomed into. One small piece of the original image could be explored to a wall-chart size!

That's my first thought anyway. I am sure it will change (for better or worse!) as I see more use cases (or misuse) of the technology. I cannot stress enough the need for technical measures to prevent misrepresentation or confusion of AIGI with actual photographs.

I am not a copyright attorney, but I have one. In the publishing world, AI generated text is not copyrightable, and publishers are amending their contracts across the industry to address that. The way publishers protect their IP is through copyright, so they will not accept AI generated submissions, as they cant protect it.

I wonder if there will be an uncanny familiarization. I remember the first time I saw The Matrix and the Star Wars Prequels, the CGI looked photorealistic to me. Yet nowadays it looks obviously fake, because we learned to read the clues that give it away.

I suspect we'll see a similar thing with AI generated images. In fact, I can already feel the process beginning to take hold, with modern cell phone images taking on a dissonant feeling, and something about the AI generated textures feeling subtly off.

One can also sense it in the photographic vernacular, seemingly walking back from HDR towards a less stylized rendition, of the kind that's difficult to generate convincingly, relying on the subtleties of lighting and texture that we are highly attuned to.