111–120 of 152 posts
A friend of mine works in their European HQ in Dublin and told me that their AI leadership are basically missing, leaving the office leaderless in favor of promoting themselves at tech conferences.
Hopefully Getty makes the necessary changes, because there are lots of good engineers in Shutterstock beholden to lots of bad management.
Earlier quoted context omitted.
Do you have a test you like? I just took one at https://sightengine.com/ai-or-not?version=2024Q3 and got 18/20 correct, and I'm not zooming in on details or anything, I'm just using some basic discrimination based on what I've generated and seen generated in the past. I would do even better at this if we limited it to pictures of "realistic" settings.
I think we might be talking about two distinct cases. If you're actively thinking whether an image is AI or not, you're already biased to it potentially being AI-generated. That improves your recognition of slop-finding. As I mentioned, I definitely agree how it's fairly straightforward to spot the slop if you're looking for it. I'm not even sure how we could implement a real-life test without bias. Maybe if there wa…
Bringing it back to the topic of stock photography: A large percentage of stock photos are of real things, people, scenery. So, when someone says I'll have a hard time spotting generated stock photos, I kinda go uhh, well, no, not generally, because stock photos are very often of people and real life scenes, the thing that is the easiest to spot as a generation.
Earlier quoted context omitted.
Survivor bias. In that, you're reacting only to the images you assume are AI. It could be you're really good at spotting them, or they're really bad. But it could also be you spot a tiny proportion, or even misidentify real images as AI. Without knowing the real rate, it tells us nothing about whether picking AI images over stock images is a good tradeoff or not.
Oh, please. I've generated many, many images. They are not hard to spot.
How many more do you think would get past you if the person running the hypothetical campaign was someone with a similar experience at picking images to you spending the same amount of time they would picking stock photography on ruling out any picture that looks like it's AI-generated to them, or editing them to remove things that'd tip you off?
Earlier quoted context omitted.
I think we might be talking about two distinct cases. If you're actively thinking whether an image is AI or not, you're already biased to it potentially being AI-generated. That improves your recognition of slop-finding. As I mentioned, I definitely agree how it's fairly straightforward to spot the slop if you're looking for it. I'm not even sure how we could implement a real-life test without bias. Maybe if there wa…
Right, and even there I think we might need to get specific about categories of images. Images that are supposed to be photo realistic are far easier to spot than "battleship in outer space" generations. Bringing it back to the topic of stock photography: A large percentage of stock photos are of real things, people, scenery. So, when someone says I'll have a hard time spotting generated stock photos, I kinda go uhh,…
We still don't know whether or not you're good or bad at picking out AI images used in actual campaigns, because we have every reason to assume at least a reasonable proportion of AI images used in actual ads will have been through an editorial process that'd rule out a lot of the easily recognized shlock, and so a test that does not use images that have been through the same selection process is meaningless.
I have no doubt you can recognize some. You may well be able to recognize all of them perfectly for what I know. The point was not to argue you can't, but that your impression can't reliably tell you, because you'd be likely to think the same whether your accuracy is high or low.
Anecdotal, but I haven't bought a stock image since Stable Diffusion was released. Edit: with Flux, you can't even tell the difference: https://blackforestlabs.ai/
Earlier quoted context omitted.
I think we might be talking about two distinct cases. If you're actively thinking whether an image is AI or not, you're already biased to it potentially being AI-generated. That improves your recognition of slop-finding. As I mentioned, I definitely agree how it's fairly straightforward to spot the slop if you're looking for it. I'm not even sure how we could implement a real-life test without bias. Maybe if there wa…
Right, and even there I think we might need to get specific about categories of images. Images that are supposed to be photo realistic are far easier to spot than "battleship in outer space" generations. Bringing it back to the topic of stock photography: A large percentage of stock photos are of real things, people, scenery. So, when someone says I'll have a hard time spotting generated stock photos, I kinda go uhh,…
Earlier quoted context omitted.
So how do we fix it? Better search/aggregator engine and unified payment scheme, but photographers get the money directly and simply pay 1 cent per purchase that came via the aggregator, rather than having to sign away their rights and getting pennies from a centralized platform? Wondering if photographers can't already do this with regular search engine's image search, which (speaking for myself) is what I use when…
It's crazy after all this time we still don't have low friction small transaction capability on the western world's web. When I was in China way back in 2014 it seemed like they had an ability to this person to person from your phone, so why can't we get it for the web? Maybe there's enough out of work developers someone can go after this seemingly low value but wished for since forever payment space.
Everyone with a bank account can transfer money online, merchants just need to accept it and not try to use dumb schemes that charge extra fees on top of the bank fees to "support more payment methods", that's my problem...
Anecdotal, but I haven't bought a stock image since Stable Diffusion was released. Edit: with Flux, you can't even tell the difference: https://blackforestlabs.ai/
There are plenty of businesses that think the same way and every time I see an ad with an image that's clearly AI-generated I steer clear of it. It looks cheap, hits the uncanny valley and is often a good sign of lowest effort possible.
Frankly speaking they are getting so good I can hardly tell by first glance
Earlier quoted context omitted.
They also make money by chasing down people who use their images without paying a license (fair) by "extortion". Once my co-founder used an image downloaded from Google (bad!) for the company website, GettyImages noticed that and threatened our company to legal actions (C&D) unless we pay the price of the license for the stock image, which magically became "premium" (or whatever their top tier is) for the occasion. T…
How are they suppose to do that without coming across shady?
All in all, as stated in the original comment, I believe it's in their right to do so (because the copyright infringement happened), but they take advantage of this in a shady / scammy way