Made this rant vid … edited in Blender of course.
Rant + Commercial Software
means I hate my job but at least it pays well.
Rant + OpenSource Software
means I love my job but I gain no money from it.
I have been hearing things like the imminent death of commerce in various genres of software since 2004, yet I still do not see apps. like GIMP and LMMS just steamrolling the commercial vendors, or even denting them (ie. Blender is still more of an exception than the norm).
The one major advance I have seen in FOSS is the rise of credible alternatives in the area of game creation and in texture authoring (Godot and Material Maker), but they like Blender have primarily gotten there because the commercial vendors provided an opening (with their reduction in R&D, elimination of perpetual licensing, rising prices, aimless development, issues with bloat, bugs, and crashes, ect…). Look at Houdini for an example of what happens when closed source is done right.
I would love to stop using Photoshop. I hate it. But it doesn’t feel like FOSS won that one at all. I am constantly trying to replace Photoshop with Krita and Gimp, but I cannot. The Photoshop’s AI content aware stuff is too useful, but not only that, other little things like simple snapping… That’s a real pity. I honestly put a lot of effort to look for ways, but I simply cannot use Krita and Gimp as Photoshop’s replacement at work, because it would take me too long to do the job.
Also CAD world only has FreeCAD. Better than nothing for sure, but it does not compare at all not even tiny bit with big guys like Fusion360 or Rhino and is not suitable for interior designers or architects who might use Sketchup, Revit, ArchiCAD or even AutoCAD. I think FOSS lost and is now completely dead in CAD area.
Besides Krita and Inkscape, there are no real open source alternatives for me either in the vector and raster departments. Affinity is able to cover most of my needs, but from time to time I too need to use Photoshop for some special edge cases. The AI tools Adobe are working on will give them a new edge as well, but the uncertainties surrounding the legalities of AI produced works will slow down the industry from widely accepting the technology into production, so there is time for the competitors to catch up still.
I am not overly optimistic about Gimp ever being a good enough alternative at this point, not with so many cheap commercial products being available with better tools, development, and core engines. Blender, Krita, and Inkscape are more exceptions to the rule of open source creative software rather than the rule, I agree.
I don’t know what perspective you are looking at this from, but from what I see, legalities are not going to be a problem. I am not entirely sure I know what you mean by “legalities of AI produced works”. In some contests and competitions maybe… In the real world?.. How do you even know what I use to create my work? I have no legal or moral obligation to reveal my techniques of digital content creation to anyone. Professionals and artists are going to use whatever tools that are good for the job. It’s nice that we have some open source tools in all the new AI technologies. On the other hand, we use AI in Photoshop for more than a decade now - where is it in Krita and Gimp?..
Well, at least there is a good fight going on between commercial and FOSS software in some areas. I prefer FOSS when possible. It hasn’t won anything yet though…
In the gaming industry there are multiple companies that do not use AI yet because of the legal/ethical issues surrounding it. Very big companies. You will not see AI tools being fully embraced by this industry for some time until the law has started tackling the issues of AI and ownership. That’s why having AI features will not be as big of a selling point for professionals in the immediate future as some might think.
The GIMP is actually quite useful now for basic image operations. The fixed UI and the GEGL stuff took it quite a ways.
I think by now, it is a foregone conclusion that we should no longer be asking if FOSS will ever produce solutions that is a threat to the commercial offerings (because it won’t, Autodesk still remains quite far from bankruptcy). Rather, we should be asking if we are seeing success in producing an ecosystem that is a decent alternative to buying software as well as games. I use GIMP not because I think it is the best, but because it at least has a non-glacial pace of development (especially in recent years) and can produce good enough results.
What FOSS can in fact displace though is stripped-down freeware apps. that act as little more than a sampler for the bigger commercial apps. they are attached to (ie. “Zbrush” Core Mini, no need to resort to using it considering what Blender’s sculpt mode is getting).
You sure about that? Because in a trademark or copyright infringement case, you absolutely have to reveal details about your creation process per legal discovery, and failure to do so is contempt of court with a nice side dish of obstruction of justice. Why do you think big companies document everything? It’s not for fun, I can promise you that
P.S: AI generated content cannot be copyrighted in the US, and lying about the creation of a copyrighted document is perjury. I wouldn’t recommend it (and it would be patently insane to use material you can’t hold copyright over in any commercial setting.)
P.P.S: The EU is preparing to pass rules that all AI generated content must be clearly labeled as such, so soon you’re going to have a legal obligation to disclose this in both the EU and the US, and EU penalties are much meaner
In many cases, a penalty is just a tax writeoff for the billion-dollar companies anyway. The current US and EU governments will add another thick book to the existing pile of regulations annually, but they probably would not have to if penalties actually stung instead of acting like a light slap on the wrist.
So by this logic, any photo that has been edited with Photoshop using content aware functionality cannot be copyrighted? I cannot use AI denoising in my renders anymore? Come on… That’s ridiculous.
Also why would I break copyright and trademark laws?..
From what I gather platforms such as google and facebook would have to lable it, not the individual user. It is not aimed at editing a photo or some denoiser either, but at content created with AI like text written by chatGPT or images and videos created by mid journey and stuff like that.
open source has better tech if it has an active community, but closed source will almost always win commercially.
They gain tons of money, which they can reinvest in RnD and the product. And marketing. resulting in a better product. They can pay people to fix the stuff nobody wants to fix.
And secondly, a lot of open source projects can be wrapped in a commercial produuct with a single extra feature, or better UX. If it’s e.g. MIT license, A company can just wrap and resell that project, benefitting from the open source project and keeping all profit.
In it’s current state, i don’t see how open source can actively compete on the same scale as amazon , facebook, apple and google
This is a good point.
The best part is that commercial entities, as a survival strategy are forced to adopt depend on user-metrics
and usability.
Open source software is the exact opposite, where you have the best code possible running, and at the same time usability and usefulness is put in second place.
This is how for example programmers think, that they consider that only crashes and bugs are things that need improvement. This is a very engineer-oriented mindset, rather than a designer-oriented or a business-oriented thinking.
I am not saying that this is a bad thing but having only one of them is bad.
That’s what I am talking about. So I would not agree to this:
I don’t like Adobe as a company and I don’t like a lot of things about Photoshop, but I think AI tools in Photoshop have not seen much competition in FOSS or even elsewhere. Hopefully that can change any time now with Stable Diffusion for example. Still, Adobe are not stupid, whatever new AI tools they are developing, they will probably be hard to compete with.
I don’t see legal issues as a factor in this at all. AI hype is one thing, but professional work is another. Of course if you use unsupervised AI to produce content and interact with users of huge internet platforms there are all sorts of ethical and legal problems, but tools for deliberate creative work will not be a problem. The problem is that AI is not that good for this yet. It’s useful, but it’s not so magic as it might seem. Also it doesn’t seem like time helps FOSS to catch up at all. Content aware stuff is available in PS for more than a decade, without any alternatives in GIMP and Krita…
I would love to uninstall all Adobe products and never use them again, but unfortunately, FOSS is loosing this battle in my use cases.
In my opinion Blender is an exception. It is winning in some areas. You can walk into a bar full of 3ds Max and Maya users and announce loudly that Blender is superior in modelling, UV unwrapping and sculpting for example, and your are going to be fine… probably. Of course there is also Houdini, ZBrush, Substance… but the fight is happening - no question there.
These two points are a double whammy. Because in theory someone can take your AI generated character, hand draw it with some changes to make it their own and they would have the copyright for it.
Yep- this has already happened, actually, it’s the new “right click and save an NFT” workaround sending ripples of dismay and anger through the AI generation community
Why is that bad?
Ironic, considering how they went out of the way to intentionally copy the work of big artists, when the initial hype happened.
Because it means you have no control over what in theory would be a character in your IP.