5 ms·

That seems to depend on your use case. Frankly, I don't have much use for either of them but Midjourney was much closer. I've twice spent a couple of hours uns

That seems to depend on your use case. Frankly, I don't have much use for either of them but Midjourney was much closer. I've twice spent a couple of hours unsuccessfully trying to generate a simple background image that would be blurred out when rendering 3D models. SD out-of-the-box was far worse, but Midjourney still was not up to the task. It's incredible how well they can generate images of nearly any subject/object and make some changes to the style and placement, but trying to precisely achieve critical broad-stroke things like like perspective, sizing, lighting direction/amount/temperature, etc. was far too cumbersome. Prompt refining is just like having a program with a bunch of nebulous undocumented menu entries that you just have to click on to see what they do rather than just giving you the tools to make what you need to make. Was that the right entry or the wrong entry? Who knows! Maybe just try it again to see if it works better! There's a fundamental disconnect between professional-level and consumer-level tools. Consumer tools must be approachable, easy to use, quickly yield appealing results, affordable, and require little maintenance. Professional tools need to be precise, reliable, capable of repeated results with the most demanding tasks, and easily serviceable into perfect working order. These are consumer-level tools. If you merely need a cool picture of a space ship done in such and such style with such and such guns blah blah blah (that for some reason always looks 10%-50% Thomas Kinkaid,) these tools are great, but they abstract away the controls that really matter in professional work. Novices who get overwhelmed by all of those factors love it because they don't understand, and probably don't care about what they're giving up. For serious work, aside from getting inspo images or maybe generating deliberately weird bits of whatever, they're hit-or-miss at best. Without exception, doing a rough mock-up in a modelling program took FAR less time than trying to wrangle exactly what I needed from one of those generators. I'm sure they'll get there someday but right now they're miles away from being professional-quality image generation tools.

chefandy 3y ago
I will say though that low-effort higher-volume professionals (e.g. mobile game mills, Fiverrrr designers) will likely profit from these tools once they can out-compete cheap online assets from stock images/models/etc. but they're so not there yet.
jstarfish 3y ago
> Without exception, doing a rough mock-up in a modelling program took FAR less time than trying to wrangle exactly what I needed from one of those generators. I think a lot of people have unrealistic expectations of the tech-- they think they can get exactly what they want if they are articulate enough in describing it with words. Feed your rough mock-up to img2img (or use inpaint sketch) and you'll land much closer to where you're trying to go. It's a power tool. It will do tedious manual work (producing art) very quickly. The difference between professionals and consumers in how they use it is that the professional asks the machine to "finish what I started," whereas the consumer tells the machine to "do all of the work for me."
- chefandy 3y ago
  I tried img2img. It will do rough finishing work but it won't take some lighting vectors and match my lighting. It won't shift the viewpoint by 18 degrees. It puts a smooth sheen on rough work with broad stroke needs and that's valuable in some cases, but it is not a general-purpose professional tool. Canva competently satisfies most non-professional needs but it only satisfies a narrow slice of professional needs. Trying to use it for most professional work takes vastly more time and effort than using a proper professional tool. LaTeX fits academic paper publisher's needs and can pump out formatted and laid-out text far quicker than someone using InDesign but you'd go crazy trying to assemble a modern magazine or high-end book. It doesn't need polish or sheen. It needs something fundamentally structurally different. I'm both a professional digital artists and a long time back-end software developer. This slice of time has really opened my eyes to what it must be like for most non-developers to speak to developers: constantly oversimplifying your use case and assuming some algorithmic approximation will do without really understanding the problem.
  - jstarfish 3y ago
    Fair enough. Your 3D modeling needs might be a bit advanced for the current state of things. It works pretty well for flat graphic design, stock photo or illustration purposes. I'm holding out for an instruct-based model that will take explicit instructions, or at least layered prompts. Mutating the original prompt along with the picture (or changing the entire thing to only describe certain parts, a la inpainting) is frustrating to me.
jamilton 3y ago
ControlNet helps a lot with composition and lighting (https://sandner.art/create-atmospheric-effects-in-stable-diffusion/ https://sandner.art/create-atmospheric-effects-in-stable-dif...). It requires more work than just entering a prompt, but probably less work than doing it manually once you get used to it. I think there's a number of StableDiffusion clients in development that are trying to make this easier.
- chefandy 3y ago
  Ha... I accidentally replied to the wrong comment. Anyway, thanks! That's pretty neat. The sample images look a bit overwrought like a lot of other AI images do but I'll bet they're doing that to follow the trend rather than it being a technical limitation.
cubefox 3y ago
Classical professional tools like Photoshop have a lot less potential though. They are very precise, but (I assume) they have barely advanced in the past decade. Tools based on generative AI will probably improve massively over the next few years. Most such tools seem currently based on Stable Diffusion, and apparently OpenAI/Midjourney/Google have zero interest in supporting such tools. But this could change soon, e.g. when Adobe tries to compete with the SD ecosystem. We already now see deepfakes (e.g. of Trump or the Pope, recently even videos) that a far beyond what we saw in the years before, indicating that the old professional tools weren't so powerful after all. Now if we extrapolate this a few years into the future...
- chefandy 3y ago
  You assume wrong. What these tools offer is constantly churning. Far faster than ever before and Photoshop has been around for over 30 years. They release updates constantly. Photoshop got AI filters like detail enhancement for zooming a few years ago. Automatic object detection, content-aware delete, etc. etc etc. a few years before that. That's only what I can recall off the top of my head for Photoshop alone, but it's such a giant environment that even most of their own product people probably couldn't tell you off the cuff. In areas like video compositing, tools like Nuke are developing tools with these capabilities even more quickly... and they better when the cheap license costs $3500/yr. As I mentioned in another comment, so much of this hype is based on developers assuming they understand something that they don't. I've indulged in this hubris as a developer but straddling both sides of this line has been illuminating.
  - cubefox 3y ago
    Well, Adobe at least seems headed to fully embrace the generative AI hype now: https://www.adobe.com/sensei/generative-ai/firefly.html https://www.adobe.com/sensei/generative-ai/firefly.html This sounds all very similar to the Stable Diffusion tool chain, though probably cloud based and with a more intuitive UI on top.
    - chefandy 3y ago

chefandy 3y ago

I will say though that low-effort higher-volume professionals (e.g. mobile game mills, Fiverrrr designers) will likely profit from these tools once they can out-compete cheap online assets from stock images/models/etc. but they're so not there yet.

jstarfish 3y ago

> Without exception, doing a rough mock-up in a modelling program took FAR less time than trying to wrangle exactly what I needed from one of those generators. I think a lot of people have unrealistic expectations of the tech-- they think they can get exactly what they want if they are articulate enough in describing it with words. Feed your rough mock-up to img2img (or use inpaint sketch) and you'll land much closer to where you're trying to go. It's a power tool. It will do tedious manual work (producing art) very quickly. The difference between professionals and consumers in how they use it is that the professional asks the machine to "finish what I started," whereas the consumer tells the machine to "do all of the work for me."

chefandy 3y ago
I tried img2img. It will do rough finishing work but it won't take some lighting vectors and match my lighting. It won't shift the viewpoint by 18 degrees. It puts a smooth sheen on rough work with broad stroke needs and that's valuable in some cases, but it is not a general-purpose professional tool. Canva competently satisfies most non-professional needs but it only satisfies a narrow slice of professional needs. Trying to use it for most professional work takes vastly more time and effort than using a proper professional tool. LaTeX fits academic paper publisher's needs and can pump out formatted and laid-out text far quicker than someone using InDesign but you'd go crazy trying to assemble a modern magazine or high-end book. It doesn't need polish or sheen. It needs something fundamentally structurally different. I'm both a professional digital artists and a long time back-end software developer. This slice of time has really opened my eyes to what it must be like for most non-developers to speak to developers: constantly oversimplifying your use case and assuming some algorithmic approximation will do without really understanding the problem.
- jstarfish 3y ago
  Fair enough. Your 3D modeling needs might be a bit advanced for the current state of things. It works pretty well for flat graphic design, stock photo or illustration purposes. I'm holding out for an instruct-based model that will take explicit instructions, or at least layered prompts. Mutating the original prompt along with the picture (or changing the entire thing to only describe certain parts, a la inpainting) is frustrating to me.

jamilton 3y ago

ControlNet helps a lot with composition and lighting (https://sandner.art/create-atmospheric-effects-in-stable-diffusion/ https://sandner.art/create-atmospheric-effects-in-stable-dif...). It requires more work than just entering a prompt, but probably less work than doing it manually once you get used to it. I think there's a number of StableDiffusion clients in development that are trying to make this easier.

chefandy 3y ago
Ha... I accidentally replied to the wrong comment. Anyway, thanks! That's pretty neat. The sample images look a bit overwrought like a lot of other AI images do but I'll bet they're doing that to follow the trend rather than it being a technical limitation.

cubefox 3y ago

Classical professional tools like Photoshop have a lot less potential though. They are very precise, but (I assume) they have barely advanced in the past decade. Tools based on generative AI will probably improve massively over the next few years. Most such tools seem currently based on Stable Diffusion, and apparently OpenAI/Midjourney/Google have zero interest in supporting such tools. But this could change soon, e.g. when Adobe tries to compete with the SD ecosystem. We already now see deepfakes (e.g. of Trump or the Pope, recently even videos) that a far beyond what we saw in the years before, indicating that the old professional tools weren't so powerful after all. Now if we extrapolate this a few years into the future...

chefandy 3y ago
You assume wrong. What these tools offer is constantly churning. Far faster than ever before and Photoshop has been around for over 30 years. They release updates constantly. Photoshop got AI filters like detail enhancement for zooming a few years ago. Automatic object detection, content-aware delete, etc. etc etc. a few years before that. That's only what I can recall off the top of my head for Photoshop alone, but it's such a giant environment that even most of their own product people probably couldn't tell you off the cuff. In areas like video compositing, tools like Nuke are developing tools with these capabilities even more quickly... and they better when the cheap license costs $3500/yr. As I mentioned in another comment, so much of this hype is based on developers assuming they understand something that they don't. I've indulged in this hubris as a developer but straddling both sides of this line has been illuminating.
- cubefox 3y ago
  Well, Adobe at least seems headed to fully embrace the generative AI hype now: https://www.adobe.com/sensei/generative-ai/firefly.html https://www.adobe.com/sensei/generative-ai/firefly.html This sounds all very similar to the Stable Diffusion tool chain, though probably cloud based and with a more intuitive UI on top.
  - chefandy 3y ago