Quote from: Evaristo tenscadisto on January 26, 2021, 18:02

For example, Nowadays you don't need real people you can use some GAN Ai services and create a face that doesn't exist. Photo shoots are much more easier and cheap than rent studio, hire people and real models.  We can do it almost everything in a Pc with a 3d environment.

I think photo-realistic CG models in lifestyle shots is still a bit of a long way off. I don't think I've ever seen one, let alone enough for it to be commonplace.

There's already plenty of models 3D scanned with fotorealistic details selling for less than 10$.

hi, do you know https://renderpeople.com/ ? i think this will answer half of your question. there's a lot of lifestyle, business etc. Some are rigged and some are posed. you can buy one for 39$ or a bundle for 279$.  My last post was the workflow process to match the needs of changing/create new face or body for 3d model.

So my point is this: 3d scanned models (photogrammetry)  combined with machine learning Ai can build a lot of humans in poses that doesn't exist. Roughly same process of face generator but you have to scan T-poses of body instead of just face. So you only have to pick the one you like more in thousands made in secs.  During october 2019 I already saw some tests of AI Gan but were made to create characters and not real people.  Therefore these renderpeople companies will be the new microstock agencies since they can offer better solutions that single photo of "Young family with children having fun in nature" can't if a client wants this people having fun but in different park or even now at home because of pandemic.

TurboSquid comes across like the Shutterstock of its market segment - high volume, low royalty rate, widely used. This article mentions they've been around for 20 years, but it's a small organization. Haven't found anything that explained why TurboSquid is for sale.

The way i see it:

1) first we actually don't need to buy to many 3d models because we use photogrammetry to create models. Its better and more realistic. I have been doing this for at least 2 years for simple models and it works fine.

2)  2nd because artists use AI sw too so no need to use mstock services. For example, Nowadays you don't need real people you can use some GAN Ai services and create a face that doesn't exist. Photo shoots are much more easier and cheap than rent studio, hire people and real models.  We can do it almost everything in a Pc with a 3d environment.

So they can eventually do what contributors done in past but in future we eventually don't need at some point microstock services too. But i also think that in terms of Photography there is always space to fulfill for documentary and news. No AI can do this.