You have no way of “ensuring” that a computer cannot Do what you can do.
You have ZERO control over how the rest of mankind will respond to what an AI creates.
I imagine the Chemical engineers at Eastman kodak had similar confidence in their craftsmanship in the face of digital photography…. and where are they now?
The great “unwashed masses” of lay people do not care about your craft they only want instant gratification and these AI art generators provide exactly that.
That’s the kind of “journalism” that I wholeheartedly dislike. Why did they link a thread where it is mentioned that the model was “nerfed”? Why did they claim in the title the ability to copy artist styles was removed?
How about asking the people who trained the model? Did they actually remove that? (Most likely not!) Did they check what caused this issue? Was this a consequence that many images were removed? Do they think this will stay like that in the future?
Why bothering doing something like this when we can likely draw more interesting conclusions without actually spending time and researching a topic?
Any digital creation like 3D design is part art and part tedious clicking it together.
It will be a few years until AI can generate relevant ideas for complex ideas based on lots of human data. Take a design competition for a new bridge that works well with the history of a town, vs a model of a medieval farm.
The creation of models, textures, scenes and animation based on detailed instructions will be there much sooner.
I don’t know if you are all artists or sometimes work for corporations who just want a product in a specific setting, I’m the latter, AI might do 90% of my job in less than 10 years. I’ll be selecting from a grid of generated AI ideas 
Despite the recent advancements in AI and its ability to create art, there are still many compelling reasons to learn 3D modeling. For one, the field of 3D modeling is constantly evolving and expanding, and there are many exciting opportunities for those with the skills and expertise to create high-quality 3D models.
One of the biggest advantages of 3D modeling is its versatility. 3D models can be used in a wide range of fields, including architecture, animation, gaming, and product design. This means that there is always a demand for skilled 3D modelers, and that learning 3D modeling can open up a variety of career opportunities.
Another reason to learn 3D modeling is that it is a highly technical skill that requires a deep understanding of 3D geometry, anatomy, and the principles of 3D design. This means that mastering 3D modeling can be a challenging and rewarding process, and it can provide a solid foundation for a career in the creative industries.
Furthermore, even though AI is capable of creating impressive art, it is still limited in many ways. AI art is typically generated using algorithms that are designed to mimic certain styles or techniques, but it is not yet capable of truly creative expression. This means that there will always be a need for human artists who can bring their own unique vision and perspective to their work.
In short, while AI is an impressive technology that is capable of creating art, it is still in its infancy, and there are many reasons why learning 3D modeling is a valuable skill. Whether you are interested in a career in the creative industries, or you simply want to explore your own artistic interests, 3D modeling is a skill that can open up many exciting opportunities.
That’s what ChatGPT told me about this subject 
Just wait until the ‘AI art paradox’ kicks in. 
These machines are ‘stealing’ human artwork as inputs, so when eventually we’re all out of a job, the AI will have no more human art inputs and will eventually feed itself purely on machine ‘art parody’ until everything is just one single image…to infinity.
Exactly that. And as time progresses all different art AIs will start to converge into the same big neural network and beyond the point where they squeezed human art out completely, what they define as art will start to drift apart from what we as humans find artistically interesting, pleasing or stimulating. Consequently a new branch of AI comes up, that starts simulating crowds of people discussing, liking and consuming all kinds of AI art. 
Sports died in a similar way, since they invented a robot with humanlike extremities that could run faster, jump higher and broader. At that time it was funnily totally negleted that the human factor was the central most important aspect of sport events, so robots finally replaced humans in all sportbased competitions aswell. It simply looked so similar to what humans did, but it was much less cost intensive in the long run.
From what I see from this machine processing - I will never call it art because a machine cannot create art, merely a parody - it’s already starting to look very similar. You can tell it’s a machine mashup by looking at it. This technology has no place in art. They should put these machines to work on creating a decent retopology algorithm or UV unwrapper, or rigging tools. Leave the actual art to the humans.
I keep hearing artists say ‘but it’s great for getting concept reference…’ seemingly oblivious to the fact that they are merely heaping fuel onto their own funeral pyre.
The current image generation tools are just toys where researchers combine different state of the art machine learning techniques to find out what those are capable of. One of the goals is clearly to get a better understanding about how those can be improved and be applied in new fields. If they are good enough, there are certainly people who commercialize them, but I am sure the current ones are not yet profitable.
When it comes to retopology, uv unwrapping and rigging, you are essentially in a territory where almost no research takes place and where the gained knowledge could hardly be used for other areas. You would pretty much need to start from scratch, build datasets in a very niche domain.
Even though this argument seems to pop up all the time at the moment, it literally makes no sense to me.
Autodesk and Adobe are already doing R&D on this. It makes perfect sense, I would have thought: Feed the machine with ‘X’ examples of the optimal goal.
I see it the same way. I think currently there’s a big mistake in perception. Its a bit like looking at a wooden egg a machine made and say now finally machines have superseded chickens. There is no life in it. Those machines are workinghorses of image recombination heavily relying on stolen art. Its legal situation is somehow new but questionable IMO. And no matter how perfect the images are, its just a lifeless complex recombination of work done by artists. There’s no real intent or soul in it. So appreciating it as art is indirectly disparagement of the human mind IMO.
They already have 50 papers published
I feel the very same. It’s an esoteric argument that has been made above I’m sure, but creating art, imo, requires inspiration from the subconscious section of the human brain - whether from memory, feeling, dreams, emotions, etc. A machine isn’t capable of any of that. What it creates will always be a cold parody based purely on mathematics.
The small issue when it comes to retopology is: What is an appropriate input and output format for the mesh (works for an arbitrary number of vertices)? That’s an unsolved problem at the moment as far as I know. What is an optimal goal for retopology?
I know that people might get the impression that with enough data, everything is possible today. That’s just not true.
I glanced over their publications:
https://www.autodesk.com/research/publications
I could not find a paper which would help with machine learning based retopology, uv unwrapping or rigging. Their machine learning papers are as expected very similar to what others are doing. When dealing with 3d objects in machine learning, they are usually represented as point clouds (or voxels). Pretty much all their papers are about what others have been doing, with a twist towards what they might be able to commercialize, which makes a lot of sense.
If there was a somewhat solid foundation for dealing with meshes in machine learning, they would be on it. But there is no sign they are…
After just watching a recent AI animation, I am officially feeling a little threatened by the awesomeness of it.
Answer is NO! and YES! … in short.
If you are young and want to make professional career in CG, answer is NO!
Not this day, not next year, but in near future human “operators will become obsolete”. Job market will shrink tremendously. There will be place for only couple percent of today workforce.
So, do you feel lucky today punk 
If you want to deal with CG as hobby and maybe sometime do some paid gig, answer is YES.
… everyone who have illusions, must look in history. Not some “ancient” history. Just from "50 to these days there are tons and tons professions which become obsolete. Do you think that CNC machines exist 100 years ago? Or 3D printers. Before machines humans do such highly precise jobs. They go through long and expensive training, and they were paid very well, and they were respected just like today top artists. Oh, these professions exist these days, some old grandpa shops here and there with just few customers which deal with handmade stuff more because eccentricity than real need.
You watched an AI animation? Please tell us more about this.
Trying to be clever? Keep trying.
I am actually curious about what you mean.
Is it 3d character animations, generative techniques like Stability AI for video or something else?