Thanks Greg :)

Yep - me too Armin - nothing will get in the way of my enjoyment of landscape photography :)

"hopefully there will be a way to tell that images are AI generated/enhanced/manipulated." This is unlikely to work because they way these are trained is forcing the generator ai to fool a detector ai - that is what the a in GAN stands for - adversarial. Only lack of data, compute or engineers will leave gaps - and that is unlikely to happen for the top products.

It should be Part of EFIX Data as a non removable entry made by the program. And maybe even a small watermark in one of the image corners, course it can be cropped out but it would because start.

@@ArminSteiner That would help, but as it sadly is, that too can be manipulated. I’m afraid it’s out of hand and we can only hope that people grow bored of this kind of imagery. I mean, there’s a movement among those so called digital nomads that are reverting to dumb phones, as a contrast to smart phones, because they’re fed up with the social media environment and want to cut loose. Maybe you’ve heard of it. Nothing lasts forever and, by the way, not every blue collar job has been replaced by a robot, as feared back in the 80s or 90s. Changes will come, but maybe it’ll turn out differently. However, maybe we should learn to ask questions like „you’re in a desert and you see a turtle lying on its back in the hot sun, what do you do..?“ anyways, just to be sure. 😉

Yep - there are many completely legitimate uses for it - of that there is no doubt. The problem is when someone uses AI to misrepresent - whether that's something relatively inconsequential like a photography competition or something serious like state-sponsored actors destabilising foreign governments.

Yep, same here and it is sad.

Checksum digitally signed using unique key on the camera .

Might be do-able with a DNG, but I'm not sure you'd accomplish it with the native camera RAW formats. I guess crypto signatures will be a possible solution.

Cheers :)

Countermeasures are looking pretty scarce right now.

Cheers :)

Either way, it's uncharted territory we're sailing into :)

Yea, it's dodgy. Buyer beware! :)

Its weird how you are way more likely to resemble someone from across the globe than your own family.

Sorry to spoil your day, but being a graphic designer myself, luckily with retirement around the corner, that’s wishful thinking for mainstream designers and those not specialized in a trait (still) of limits to AI or occupied in some niche uninteresting (at this point) for automated image generation. If a machine is capable of generating your hour’s work in seconds you’re out of business. That’s the sad truth. There’ll be simply not enough work left over for us to keep us afloat. However, I believe there’s a chance in mastering these tools, since they’re not more than that or in other words, they’re dumb and unimaginative, and orchestrate concepts, come up with ideas about content (!). It’s still about content for us designers and good content will find it’s customers.

I will endeavour to be more optimistic about it Neil. :) In all seriousness, my wife's just finished laying out a book and when they got the proof back they decided that they wanted the photos to bleed to the edge rather than having a 5mm border. She wasn't relishing manually extending more than 500 photographs .. and then I showed her generative fill and she got the whole thing done in a weekend.

In all seriousness, if AI goes the way it looks like it's going - we will have to revert to old school methods. Mug-shots taken on film cameras for instance. :)

You can't guide generative fill, you can only point it at a selection and hope for the best - even if you use a prompt. Whereas with DragGAN you can decide precisely what outcome you want. I don't think one's better than the other - they both accomplish different tasks.

@@Andyhutchinson all draggan does is guide. It can’t create anything new, only edit something created. That seems limited to me. The idea of generative fill is only a small step away from generative all settings. Lighting can be adjust on everything in the photo, etc. photoshops ai tools being incorporated into photoshop itself is a bigger deal. Draggans modifying tool in no way can beat the prompt/no prompt generative fill built into adobe photoshop. Wait til adobe to put gans into the next update lol.

@@Derek4k GAN's definitely different because it uses the existing image as the starting point for all intermediate interpolations, but it can (for instance) accurately generate the interior of a mouth if you instruct it to open. But yea - Adobe's implementation is far more useful in most use cases. :)

I believe that they wish to stand out without all the effort that makes somebody or somebody’s work outstanding. They want to be accepted and liked. Famous, if you wish. However, that’s not new, but it has become easier with technology and social media where you rarely meet the „artists“ in real life, so you can fake almost anything at a distance.

@@distomos8118 I would agree. However, the difference is also vast. With your fave "pocket stars" and the influence they create leading more of them to mass misinformation, deception, fakery, crime, violence and death which is very different from the rest of us and how we grew. No thought, just 'do things to go viral.' More of them also tend to go to extremes more often in an attempt to become viral stars. Only fans has bred a overwhelming underage sex market. Faking age. Faking everything is not a good mental thing. Having zero talent and zero respect is not a good thing. You're supposed to feel a sense of accomplishment after failing to draw that one picture over and over and finally...However it is very common with the new gen to just fake everything and work for nothing. The less you use talent, or critical thinking, it'll atrophy, like anything else. They were also coddled more than any generation before, and it's easy to observe that they feel it's their right to do anything and have no consequence. Fake anything and feel accomplished. Cause their mom loved when they faked the rubix cube. Conditioning the brain to enjoy faking it. As many similarities I see, the differences are mostly not good for humanity. And since they are next gen. It's concerning. In reality, This is a deep conversation, that I could slam out the points on. Too big for comment section.