If I only write a prompt, I would not consider a character generated by AI to be my IP. I could possibly consider the prompt to be my IP if it was really unique, but that’s it. I see AI more like a tool or maybe inspiration or help in creating something, it doesn’t have to do everything for me. I think if the output of AI generation is the end result, that’s not really your work. So this doesn’t really bother me. I think it’s kind of right and even beautiful - if you use AI that is trained on huge datasets basically produced by all humanity, then the copyright is not yours, but instead it should belong to everyone. I am quite OK with laws matching this idea. It’s nice. Sharing art the same way we share FOSS… If you use AI to inspire you and then you create something, or you create something and then let AI improve parts of it and then put a lot of your own work on top of that, that’s different. And I think whatever copyright laws currently are, it’s going to get really confusing really fast when real world AI use cases begin. I don’t think the law is going to work really well here. I mean even current copyright laws don’t really work half of the time…
Adobe has recently been showcasing Firefly, their ‘ethical’ AI tool that only draws from their own massive library dataset. I can see the other big companies following this lead.
I imagine the photo stock use terms would be like:
- 2020: You agree that the use-terms might change in the future.
- 2023: You agree that your image perhaps (or perhaps not) could be used for machine learn training.
- 2033: If you have agreed to all of the previous, you have already agreed that you will automatically agree, to what the userterms are, without having to agree at all.
There would still be the issue of licensing deals of content before the age of AI. Sure, the companies may own the rights for certain works, but that was before AI could mass produce stuff based on previous works, which is a whole new ball park. Imagine having freelanced for a big company in the past just to have said company use those works to make whatever without compensation or ability to work for them again. Sure, they would own the rights to the old works, but contracts will have to be looked over and re-negotiated if artists are to have any way to defend their line of work to be able to make a living.
Using your own built libraries are going to be more viable than the current versions of AI for sure, but there are still some issues to be sorted out before I at least would consider it “ethical”.
I always thought it was weird how a relatively niche kind of software like a 3D editor would become so powerful, but GIMP and the like never really caught up with the present. Just looking at the user interest it should be the opposite. Maybe the closed source history of apps like blender and godot has something to do with it.
Krita has a little leapfrog advantage, making it preferable over other software in some areas. but it still is behind the performance and ease of use of comparable apps in a lot of other ones.
This could have to do with the user base, 3d software is pretty technical and attracts technically minded people. The more users that have at least notions of programming the more potential to collaborate there is.
I would venture to say that there are less 2d artists with interest in coding than 3d artists.
There is another factor I can think of which is competition in the open source world.
In commercial software competition drives progress, commercial software has a need to keep up, with stand out from the competition.
Open source projects are collaborative and there are only so many people capable of improving code. In that sense the more concentrated that collaboration is the better it will be.
Blender does not have a serious competitor in the open source world so all the people who are willing to collaborate converge in one project. Blenders “competitors” are commercial software.
In 2d software there are more open source options Gimp, Krita, Mypaint etc. This dilutes the collaborative effort.
Libre office is another example that has little competition in the open source world therefore it has also concentrated the effort.
Adobe are advertising this as being built on datasets that they have 100% rights to. The AI can only draw from the dataset that it is fed, in this case limited to Adobe’s own library.
Don’t forget about Plasticity 3d being such a W for 3d artists, specially due the versatile and one of the best cad engines (parasolid/niemens nx) and many more and it’s reasonable indie price, and no subscription!
It’s new paid fake CAD software without real purpose. You cannot even use constraints or make drawings with it and geometry coming out of it is ugly and requires fixing - so more work. Just a modeller. I hear it is fun to use but the way I see it it doesn’t compete with anything yet and it’s not open source. It doesn’t seem too relevant in this context. …I mean… In my opinion. 
For me, it really just has to do with the Adobe ecosystem. None of the software individually is perfect, and neither is the group of them, but overall it’s an incredibly strong and flexible package there’s all very well integrated.
I don’t have to deal with exporting and converting things all day, I can just double click on a layer in premiere, edit in after effects, not even render, and move on with my life.
Same with InDesign and illustrator, and illustrator and Photoshop.
FOSS simply offers nothing like this.
3D software has always existed somewhat outside of this; I don’t require it to link to with everything else. So that opens up the options.
Here’s my ten pence. Annual subscription for Photoshop is 240 USD, so being conservative, The Gimp is worth 10 dollars? 30? Who here, myself included, has ever paid even a single dollar towards the cost of The Gimp’s development cost? It is evident that if you don’t HAVE to pay for something, you don’t.
Fake is what sense wtf are you talking about?? Constraints is something the developer avoided afaik, nobody wants that that shit is for industrial design at least, its more meant to use of hard surface, concept, surfacing and more for quick and direct modelling (destructive workflow for the moment but non destructive is planned for upcoming versions) production of assets.
Drawings you mean 2d curves or somekind of SVG vector drawing? Yes but using illustrator or other app with SVG format that can be imported
Ugly geometry oh common where did you get that idea? If you mean that for surfacing it still need for more advanced tools and operators for the upcoming versions but what we can do currently is decent, and do you even know how complex surfacing can be, it takes a while to get a nice clean surface in any cad software such as atlas (more advanced for that tho), moi or fusion. Or if you mean for exporting cad to polys geometry then that’s logical to give you some unclean yet good shaded model for exporting, it not some magical button and it doesn’t have some kind of Zbrush remesher don’t you know??
Sure ! Blender is an exception in FOSS CG software, in some use-cases it’s a better alternative to paid counterparts, and it offers some unique features like grease pencil. Most CG software are far from that.
Indeed software like adobe suite is quite unlikely to be replaced soon by a better open source alternative.
However, in some cases open source software can be good enough if you don’t need all the bang and whistles commercial software provide.
In my case I use Gimp and Krita for quick image manipulation or to prepare textures and they’ve got all I need. For my simple video editing needs I can use shotcut or kdenlive or olive.
Just like with my compositing needs, I manage with Blender or Natron.
This is great since for my personal projects I’m more than happy to avoid paying too much licences fee.
But if things start to get more challenging especially in the professional world, I probably start using Fusion or Nuke instead of Natron .
OSS software doesn’t have to be better to be useful, especially for hobby or educational work.
Open-source software has won
To me it won since a long time because it’s now really part of software culture and live in non-OSS ecosystem too, with libraries like EXR, VDB, USD, Alembic, OIIO … Or softwares like blender.
Years ago when linux was getting usable there was a law project (about patent IIRC) that would completely kill opensource, while this law project was cancelled at that time, today no-one will ever think of making something like that.
Now that opensource is part of our culture it’s up to us to decide what it’s going to be, and also
how much we want to support the ideas behind it. Like everyone should have access to software , knowledge, how to define liberty in the “virtual” world, how about the real one ?
Are we using FOSS software just because it’s convenient and cheaper or do we also want to promote that philosophy by using these software and build upon the ideas behind ?
I’m using blender because it’s absolutely the best 3D software for the type of work I want to do. The fact that it is available for free is a big bonus, but I do support the foundation, buy add-ons, etc.
As to the philosophy and intrinsic ideals of open source, I could not care less.
I did say ‘in my opinion’… “Fake CAD” in a sense that it’s not really CAD as you cannot do precision modelling with constraints like in Fusion and cannot get technical drawings out of it. It is a CAD-like modeller for artists working with surface models?.. but then it doesn’t produce surface models with clean topology… I think it has potential and it’s probably psychologically more pleasant to work with it if compared to regular poly modelling… And when you need clean topology you have to do extra work so at the end of the day (from my perspective at least) it’s the same as to just model it right in the first place. And it’s paid. So like a toy. You don’t really need it, but you can buy it because it’s fun to play with it. I probably sound more against it than I really am. It’s great and fun. I just don’t see the point of it. But it’s very new, so who knows what it’s going to be in the future, maybe it’s the next ZBrush. I remember first ZBrush versions… Anyway, why are we talking about it in the context of open source software or FOSS battle with commercial software?.. It has little to do with the topic the way I see it.
blender can do those things, but where it falls short is that it’s not adopted by big studios in their pipelines. Techniccally there is not much reason not too,
- studios rather pay 100k a year for support, than use the free blender with no support.
- Ton made it quite clear in last years blendcon opening that he doesn’t like the big boys so that wont help either.
- blender moves so fast, that there was concern from the VFX industry when they abondonded the VFX platform, which is often way behind current tech. but used by all big studios as a guideline. they kinda backtracked on this but a bit half hearted with the build it yourself approach.
i’ve only seen Blender adopted studio wide in smaller studios, or if used in the bigger ones it’s usually an artist using blender, but then in the end exporting to max or maya. More exceptions are starting to appear though which is great.
i really hope to see more AAA studio adoption since Blender is improving so much. and the FOSS community would bring so much new tech to the table compared to autodesks slow release cycle
this post might not make me many friends on a blender forum 
A good software developer always balances design, usability and functionality. And writes useful/understandable manuals.
And why? Because this reduces my support times to a minimum…
And doesn’t defend an obvious bug or flaw with a reply of “that’s a feature request, not a bug.”
In practical terms, is that design and usability needs lots of testing and iterations. To be done correctly. In terms of functionality everything is calculated mathematically with test cases. But the the design is a matter of psychology that is most useful to the common consensus of the most people.
Then you have an open source developer (volunteer), who shall be flooded with design-usability requests. Not only design is not his preference. He is not get paid at all on top of that for spending such time. 
True. I‘m developing software for the industry since 1996, mostly with operators alongside,
stopping me designing creepy interfaces
Most developers are kept away from the customer, which is a big mistake imo.
And this is the plus for Blender, most of the developers are using Blender by themselves.
Sure, the financial resources are limited if you compare the budget with the AAA corps, but these guys know that Blender is also their baby …
I‘m watching the Blender progress some years now, after dropping modo due their crappy licensing system, and I did not regret this move.