Some of us are using AI to help create new things (software, sites, etc.). AI can be very good at known patterns, but without a good driver won’t generate anything new. Indeed, trying to force AI to create something that hasn’t been done before can be frustrating.
It’s right to be concerned about where things are headed. The big tech AI companies have no idea themselves. They’re scrambling to find monetary traction, mostly using the old Silicon Valley 2.0 strategy (free until you figure it out, by then you have a big enough base to sustain).
The economist in me (PhD work in New Technology Economics and Management at the Kelley School at IU) says the current back-of-envelope calculations don’t add up. The infrastructure investments being made won’t pay off before being obsolete. That would be my biggest worry. My second biggest worry is that Amazon/Google/Meta/Microsoft/SpaceX begin enshitifying what they’ve done once they have a clue on how to truly monetize it.
This is way too generalized. Basically condemnation with a broad brush. Take it point by point:
- Treat customers better — True. This is an area that Adobe has historically been bad at and hasn’t really fixed. Ultimately, it will impact my third point, but it hasn’t yet.
- Produces worse software, doesn’t innovate — Wouldn’t agree. The worry was that the shift to Creative Cloud would make Adobe lazy. That hasn’t happened on the photographic side (Photoshop, Lightroom), and I’m not sure you can say it has happened on anything except perhaps the page layout, PDF, and web tool side. So I don’t think you can make the broad argument here.
- Generally better businesses — Businesses need customers and specifically money from customers to stay alive and progressing. Exactly which company are you saying does this better than Adobe?
Can Adobe do better? Absolutely. If I were running it, I’d be paying a lot of attention to that, as ultimately in tech you’re only as good as your next product (or update). As I note in my second point, there are areas where Adobe does okay at that, others where they don’t.
This is one of the interesting aspects of AI used as an agent (e.g. via MPC). Those that understand the program they’re working with do quite well using AI to add, control, change, and improve. We currently still need creators that know the tools to do the best work via AI.
Uh, I’m not so sure about that. AI is pattern matching (technically token relating in a 31-dimensional space). As I and others have already proved, you can get AI to create things that already exist. So if you want to develop a Web site that looks like all other Web sites, sure, tell your AI to “create a site like X.”
But this is actually where Elements (and the others you mention) comes into play: I don’t want a site that looks like all other sites. I want my own branding, I want to try new looks, styles, relationships, etc. This is why I was so excited about Elements when @dan first described what Realmac intended to do. In essence, they’ve created a Lego set manipulator. They’ve provided us with a number of “base” Legos, but we’re free to create new ones, modify the existing ones, and control how they interlock. All this is about “enabling.”
Software that enables will be around for some time, I believe. But it can’t be a locked black box that only the software maker can extend. As @MichaelDroste and others have shown, enabling software coupled with AI is a big, big win for creators. Microsoft should be embarrassed by the way they tried to hammer Copilot into a narrow hole they controlled. Apple has clearly the better idea there.
The majority of Adobe upates over the past 18 months or so have been integrating AI into their software not really the best progression when you look at it. All they are doing is hoovering up peoples creative work and taking it under the AI T&Cs and calling it innovation.
Apple, Blackmagic are doing way better at actually giving users a little more of what they want and need. Yes they are slowly integrating AI in their software but not as aggressively.
Ever tried to resubscribe to one of Adobe’s plans - its a right PITA and they make it as difficult as possible. I don’t know any other company who treats its customers like that.
The majority of Adobe’s software is build and continues to be build on legacy code so there will always be problems as things move on. Apple had the right idea a few years ago to build for Apple Silicon and its paid off.
The debate will rumble on and will be affected by both personal experience and internet noise.
The one thing that will happen soon will be the big tech companies will be crying for wider legislation on AI specifically in the copyright area because soon they won’t have a business left. If everyone has the ability and the technology to clone services then they will crumble. I doubt that is going to happen because it will already be rumbling on behind the scenes so they can safeguard their businesses.
Started using Affinity when Photoshop CS6 would no longer run on my Mac.
I am now happy to use the free Affinity Studio because I have no need for Canva’s paid add ons.
I looked at Photocraft. It is good so far but not there yet.
If ArtCraft can get away with this, I think software developers might be wanting to look over their shoulders. It’s great that Realmac is integrating all this AI stuff into Elements but what happens when someone clones the app? Does Elements become free with paid AI?
Let’s look at one program’s additions (Lightroom) during those 18 months:
- Not AI additions: Landscape masking, Album sharing, enhance without creating DNG, point color, batch rename, color labels, adaptive presets, blemishes retouch, webp support, film presets, natural language search, duplicate detection, interactive histogram, render to DNG, crop improvements, photo-to-video animation
- AI additions: distraction removal, culling, stacking, dust removal, snow mask, firefly support, topaz gigapixel, improved subject masking, topaz sharpening, feather/edge mask sliders, generative expand
Again, Adobe’s customer handling leaves a lot to be desired. However their product improvements, both AI and non-AI have been quite constant.
As for Copyright, I was an expert witness in multiple such cases for software, and there is quite a bit of decided law out there. ArtCraft has probably run afoul of the rulings in Lotus vs. Paperback software, among others. One problem AI and AI use have is that expression of ideas is Copyrightable, repeating expressed ideas tends to violate someone’s intellectual property.
To me - too complex to create - I’m doing little safari extensions, trumpet apps, games and my first PDF app…
It’s many iterations - never comes out perfect the first time…
An app like elements - although possible - time and money would be better spent on purchase than on tokens - imho
I personally don’t see software going away for 15 or more years (if ever)
This is a super interesting thread, and something I could write for hours about… but in answer to your question:
AI is changing software really fast, and established software developers should be looking over their shoulder (including us), I think we’re all in for a wild few years.
There’s nothing stopping anyone from cloning Elements, or building a competitor. Elements is not a simple app, so it’s going to be harder to clone well compared to a simple utility app. Still, if someone wanted to put in the effort (and money), I’m confident AI could clone Elements.
I have a feeling AI is going to end up killing off a lot of developers as they’ll be a deluge of AI built apps competing with them.
We’re going to continue to invest in Elements and make it as kick-ass as possible. Fingers crossed we can stay in business!
Agreed, most normal users don’t want to vibe-code an app and then deal with updating and maintaining it.
However, with more of the technically inclined users building thier own apps with AI, will that leave enough users willing to pay for a service or app? I guess we’re going to find out…
What really kills off developers is not solving user problems and continuing to iterate that as both tech and problems evolve.
I was amused by WordCraft. Not only is it a clone of Word, but it’s a literal clone in the sense that it can’t create/edit files over 2GB in size. In other words, it was cloned as 32-bit, not 64-bit. Plenty of other Word challengers have broken the 2GB barrier by just doing the right thing. Microsoft deserves to die by their apps the way they’ve handled them.
Again, it gets back to user problem solving. Some problems are big, complex, interlocking, and require real applications with a lot of thought behind them. Others don’t. It definitely is easy to vibe code a TI calculator imitation, but as Elon has discovered, not so easy to drive a car.
Then there’s the unsolved problems. The first four small apps I’m doing are all essentially tackling smallish problems that the Big Companies didn’t, mostly because they didn’t see the market size needing the solution being large enough to devote resources to. Vibe coding and vibe design will definitely be filling those niches, though often by someone who knows what they’re doing with an AI tool, not individual users pulling up a chat session. But frankly, that’s no different than it was when personal computers first came out. They all had BASIC built in. How many bought solutions as compared to making them themselves?
Inherently, we’re all lazy. It’s the old buy-versus-build thing in just another form. If the build is too complex, too costly, too long a process, while the buy solution is easier and faster, then buy customers will. This is one reason why I’d encourage you to build out more sample projects for as many niches as possible. “Bring your content and a chat engine” becomes remarkably good marketing if you solve 80% of the problem for a modest cost.
That’s what I think is kind of inevitable. But soon after AI will be switched off due to it either draining too many of the earths natural resources or simply not being profitable enough for whoever runs it and we will be back to the 1970’s.
I am at the opposite end of the AI spectrum to many on here. Which reminds me, I must dig out that old ZX Spectrum, could be back in fashion when we are thrown back to the DOS days.
Getting back to Adobe, I was a happy user of Photoshop for many years but when they changed to a cloud based subscription model I moved on.
I am happy with the direction of Elements without AI.
What I think would be hilarious, if true, is after the 2 to 3 years it takes to build a data center the technology has advanced so fast that data center is obsolete at completion.
I, too, am pleased with the direction Elements has taken—steering clear of AI.
To be honest, I’m not really a fan of AI. While I think it can be helpful to some in the short term, in the long run, it will likely impoverish our collective knowledge. We see it already: rebuilding a website using AI is the easy way out, but where is the creativity ?
That’s just a thought, not a judgment.
Also, let’s not forget the environmental impact of every single query made. Yet, we dive right in…
And getting back to Adobe: once everything has been copied—once all the software has been replicated—what will be left ?
Knowledge and creativity are two different things. AI models are essentially a form of knowledge (they’re trained) in 31 dimensions.
Where people keep getting confused in application of AI.
Would you say people shouldn’t go to college (or school at all) because that keeps them from being creative? You can be creative using AI, or you can simply ask for a new form of the old thing. Your choice. AI is directed by someone, both at the AI company (via training and the chat/agent interaction), and by a user.
The popular notion that mainstream media is promulgating that AI is going about hacking on its own is incorrect. Somewhere in the chain of events is human direction (or lack of direction). Hollywood tends to get it right more often than not: malevolent actors directing AI create the apocalypse or dystopian state.
At this point you can’t pretend AI is going away. Ignoring it or refusing to use well-considered versions of it is Luddite. For me, I’m more productive (and still creative) with AI. You need to consider that in rejecting AI: there will be people who use it that are more productive (and still creative) than you. Where’s that put you competitively?
https://tenzen.studio To get back to an Adobe alternative, this is very interesting: I’m trying out Photon. It uses offline models and can be connected to an assistant; it’s the closest thing I’ve found.
Assuming one wants to be competitive.
Speaking of a non-AI coded Photoshop alternative (so far), I use Pixelmator Pro. Does all I need and it’s mature, and owned by Apple now. 
Apparently it only gets updated under subscription.
That’s only if you are using it under the Apple Creator Studio subscription.
You can purchase it separately per the below link: