Honest question: Shouldn't Facegen and similar programs also be considered AI? (CLOSED)

Hey all,

So, I'm not trying to create an controversy or start a fight, honestly that's not my point here. But after interacting in the new thread about how AI products should be handled here on Renderhub that made me thinking...  I know it's not exactly the same thing, but when we use programs like Facegen, Headshop and other similiar programs that create morphs and textures from real photos, that shouldn't be considered a form of AI generated content? 

I don't have a license for this programs because they are a bit expensive for me, but I already tested Facegen's trial version and with a few clicks you have a complete face with textures that you can use in your renders or even sell if you have the permissions, licenses and etc.

So, if you didn't made all that from zero, shouldn't we consider this programs and characters created by it as a smaller but existent form of AI? I know that it's not like making a prompt, pressing enter, download, and uploading a brand new product, but honestly I was thinking today...using, allowing and even buying those kind of tools and then later arguing and complaining about AI on our lives doesn't sound a bit "wrong"? 

Again, I'm not trying to start a fight or make people scream at each other..  I just want your honest opinion about this because I'm genuinely confused!
I'm not saying we should then "allow everything" but I'm honestly start to thinking we already entered in a gray area with AI in 3D and programming that we can't exactly argue about it a lot actually... for example, a lot of porgrammers today use ChatGPT to help the to fix code lines or even improve them, right? Should they say that their final product also used AI? Should they put "AI was used in this code"? I don't know....but that's just my opinion. 

What you guys think?

Sorry about the errors, English is not my main language and we can't edit the posts, RH please add this soon

    I'm hoping folks work out what part of "AI" is offensive.  In the Art community generative "AI" gets the most flack and IMO the offensive part is art theft used to train the "AI".  If you train an "AI" off things you own (your likeness, photos of you, art generated by you) then IMO that should grant you an exception.  But the loudest voices focus on the term "AI" with no consideration for the details and break out the pitchforks.     My concern for Facegen (and similar) is that it can be used to generate the face of a real person (via images).  While our community is mostly benign, the potential for deepfake, NIL/Copyright violation, and non-consensual intimate images are a concern when models are made from real people.  So, that's a long winded way of saying I would not automatically flag it for AI unless it's an obvious copy of someone else's work; but I would flag it for something more concerning.

Facegen and similar programs don't use AI, they employ a mathematical procedure to derive a diffuse texture and some level of height information from one or more photos. At best you could describe it as 'computer assisted'.

If you want to go down the purist route of 'if you didn't make it yourself, then you only did part of the job' you would have to ditch the computer entirely, and create your art from real paint, or carved from wood/stone, even then you would be using paint that you didn't make yourself, or tools that you didn't make.

Personally, for most of us, I'd say use what you want to use if it gets you to the end result your looking for.  I tend to go for the "Iray Real" style as I like to call it. 5 year olds draw better stick figures than I can and I simply don't have the "artistic eye" to go into photoshop and play with layers to get the look my imagination was presenting. My fix was to learn the environmental and tone mapping settings along with camera settings and the basics with the surface tab settings for character skins.

Overall, I think we all have different backgrounds, reasons and skillsets for whichever 3D application we're using and should use whatever side applications we have access to or are comfortable with to achieve whatever image and or vision we have running around in our head.

I tried it once and, well, the results were extremely suboptimal. I do not think it uses AI and, like others said, is just trying to match depth and UV wrap an image around its assumption of depth. Maybe for a quick, background and out of focus character it is a shortcut to consider, other than that, I wouldn't use it for any other reason. 

But, having said that, it is yet another reason to avoid uploading your images to social media and such. 

I looked at the FaceGen website and there is no mention anywhere of the use of AI in the software; the first version was released in 2001, did AI exist at that time?

PSA: I hope my comment does not cause any misunderstanding or conflict. This is only my personal opinion based on more than 34 years experience in coding and 3D development.

I feel that much of current confusion is coming from mixing AI as a supporting tool and AI as replacement of authorship. These two should be considered separately.

If we decide that "any software which generate content automatically should be forbidden," then logically we must also stop using many existing tools such as Substance Painter, Photoshop, Blender smart UV functions, photo projection, scan data, procedural shaders, and similar technologies. However, it seems that nobody is willing to abandon these tools and return to older, but still very efficient, workflows used by Japanese game development studios in early 2000s. Those who carefully study UV layout and topology of PS1/PS2 era Final Fantasy titles (8, 9, 10, and especially 11) will understand how strict the consistency and pipeline discipline was at that time.

Regarding AI, unless one is training own model from beginning, most users are simply using software-as-a-service that automate certain tasks. In this sense, it is not fundamentally different from using Photoshop or other advanced tools. Interestingly, when creators train models using their own data, AI suddenly seems to become less problematic in discussion.

Therefore, the more important question may not be "was AI used somewhere in the production pipeline?", but rather: was the content mass-generated without sufficient human involvement or responsibility? In such case, I can understand the concerns being raised.

We are working in this industry in order to earn living. Tools that reduce production time are not necessarily bad; in many cases, they are essential.

Twenty years ago, when I was working full-time as programmer, what is now called "AI" was simply called Stack Overflow. Clients never asked how solutions were found; they only cared that their applications worked correctly and stably for many years.
In fact, I developed an application 16 years ago which is still in use today, and even now some large SSII-level solutions have not matched its efficiency. Results were important at that time, and they remain important today.

AI will likely impact many jobs in future, this is difficult to deny. However, while ideological position is understandable, it unfortunately does not pay the bills.