Adobe Photoshop Beta with Generative AI

[pschlute]

As a tool in the marketing advertising etc toolset I have no issue with use of this application, mostly because I am a cynic and don't look in general, for artistic 'authenticity' or truth, with advertising/marketing these days!)

I am playing "devil's advocate" with my reply here to some degree.

"Authenticity" or "truth" in a photograph is to me an irrelevant concept. What we see in a picture (print or digital) is never the same as we see with our own eyes, our brains are far smarter than any film emulsion or paper or digital processing software, and an image often has to be edited to more correspond with what we actually saw.

If I take a picture and clone out telegraph wires, have I negated the "truth" in the image ? Possibly, but if the final image is more pleasing as a result, then I have achieved my aim. I do not view my camera as some sort of digital photocopier.

I realise the images presented by Manfred in this thread are at the extreme end of manipulation, but as long as they are not presented as accurate records of what he saw when he pressed the shutter button I see nothing wrong with them.

They will not be to everyone's taste (even Manfred himself). But photography in most of it's forms is an art, not a science, and terms like truth and authenticity I do not consider relevant.

[DanK]

I think there is an additional issue: whose work is it?

For a long time, digital editing has allowed people to use other people's work. At one extreme are things that many people would find acceptable, even if they don't do it themselves, e.g., buying someone else's Lightroom presets or paging through Nik filters to find one you like. This stuff is an order of magnitude more substantial because it is using other people's actual images.

I think some of the AI-based stuff is qualitatively different. E.g., I think of content-aware fill as more like buying a different sort of paintbrush: it's just a more powerful tool for accomplishing the intent of the photographer.

[Manfred M]

I think there is an additional issue: whose work is it?

For a long time, digital editing has allowed people to use other people's work. At one extreme are things that many people would find acceptable, even if they don't do it themselves, e.g., buying someone else's Lightroom presets or paging through Nik filters to find one you like. This stuff is an order of magnitude more substantial because it is using other people's actual images.

I think some of the AI-based stuff is qualitatively different. E.g., I think of content-aware fill as more like buying a different sort of paintbrush: it's just a more powerful tool for accomplishing the intent of the photographer.

I completely agree Dan, but I find some of the nuances even more interesting.

People have been doing sky replacements for a long time and it took a lot of time and effort. The modern software really speeds up this operation and often the result is stronger than doing it by hand. This software often comes with a prepackaged" of skies one can use. They often allow one to use images of skies instead of the prepackaged one; you can use a shot of a sky that you have taken, you can use a stock sky that someone else has taken or you can even use Generative AI to create a new sky and blend it in (like I did). If you use one of your own skies, it is clearly your own work, but what about some of the other scenarios?

The content-aware functions are just as interesting. The spot-healing tools can be considered a fairly rudimentary generative product that does not use words to describe the change to the image, but the software does attempt to calculate an acceptable fill that looks realistic. No one really says much when we use this functionality to remove sensor dust on our image, but what about more major surgery? Removing some power lines, a tree, part of a building, a person, etc. The main difference is that this functionality uses our image to create the replacement, rather than being generated by a large set of images that have been fed to "teach" ChatGBT or Adobe Firefly. My understanding is that ChatGBT 4 (which is not generally available for public use, is far more talented than ChatGBT 3 that we are now using).

The other issue are the unexplored legal aspects of this technology. I seem to remember reading that the US Patent and Trademark office has issued a preliminary opinion that Generative AI output is not covered by copyright laws and works produced this way cannot be copyrighted. I believe that goes for images, text, etc.. None of this has been tested in the courts yet...

[nijolcreative]

I largely agree with Bill.

There's AI, and then there is AI. It doesn't make sense to me to drop into the same bin everything that has been produced by machine learning. Street drugs and legitimate drugs rely on the same chemistry, but we don't treat them as similar in effect or desirability.

I have generally have no problems with machine-learning-based techniques that simply allow one to perform an action more efficiently. I use Adobe's subject select frequently, for example. It makes editing candids of kids far faster. Based on my experience so far, I expect I'll use the AI-based noise reduction often as well.

However, this generative fill is inserting someone else's content, modified by the software to fit the specific shape you have to fill. Those clothes presumably came from the training set, which is other people's images. The software had crossed that line some time ago. Sky replacement using one's own images of skies is analogous to subject selection, IMHO. It's also analogous to a painter who decides to paint a sky differently than they see. But substituting a sky from the set provided by the software is using part of someone else's image. It's like directly copying another painter's sky.

I posted before a comment that I think bears repeating. A friend of mine who spent his career in tech and was dabbling in travel photography said numerous times that what would serve many people better than a real camera is a device that reads location via GPS and presents the "photographer" with a library of images take from that spot that are better than what they can do themselves. That's pretty much where we stand now: we can "create" images that are increasingly not our own.

That's a path we actually started on without AI. Programs like Nik allow people to drop on styles that someone else created. You can just page through them and decide after the fact what you like.

There are plenty of aspects of modern life where that's true and where it doesn't bother me. I'm old enough that I used to tune my own car and do routine maintenance. Timing was simple to describe and not that hard to set. Now I can't do anything at all to maintain my current car, a VW ID.4, and in fact, I often don't actually know what it's doing. I frequently find out that there are additional functions going on under the hood that I didn't know where there. That's fine: I am not spending my free time as a car designer or mechanic. If it's reliable, safe, and enjoyable to drive, I'm a happy camper. It's the result I'm interested in. In contrast, I want the satisfaction of creating my own images, and this sort of software will erode that.

I agree with your nuanced perspective on AI in photography. It's about maintaining creativity while embracing efficiency.

[billtils]

I think there is an additional issue: whose work is it?.

Absolutely spot on Dan. There are two different issues in the AI "thing". The 3 authorities that govern what's acceptable in a competition image here in Scotland all agree that every element in an image must be the photographer's own work but using something like Topaz AI noise removal is acceptable. I find that entirely reasonable.

However not everyone who owns and uses a camera - including a mobile phone - is interested in that degree of rigour and as long as they do not try to pass something off as "all my own work" that's fine - it might even contribute to the survival of camera manufacturers