4 ms·

We're at the point where these generative AIs are good enough that they're doing things which are really surprising and unexpected and kind of exciting, but the

We're at the point where these generative AIs are good enough that they're doing things which are really surprising and unexpected and kind of exciting, but they're bad enough that almost everything they create falls somewhere between mediocre and dogshit. I really hope, if these this stuff is going to be ubiquitous, that there are big strides made in improving the quality of the output, very soon. The novelty of seeing fake screencaps of Disney's Beauty and the Beast directed by David Cronenberg is wearing off fast, and aside from some very niche use cases (write some boiler plate code for this common design pattern in this very popular language) I haven't found much it's actually useful for

sharkjacobs 4y ago
I should probably clarify, this is a response to the visceral negative reaction I felt upon seeing the cat and dog images generated in the demo. They look so bad that it's actually upsetting to me, which is a reaction that I've had to a lot of AI art.
- swatcoder 4y ago
  Don’t sweat the downvotes. People have very heated opinions about this technology and any strong statement will have more vote-volatility than usual. I appreciate you putting a voice to what you said. The historical novelty of what these generative AI’s produce is striking, but the variety of style and form is still narrow enough that you start seeing the signature of CharGPT/SD/etc once that historical novelty wears off. They each have a strong, specific voice and it gets as tiring as any other art style that’s overexposed.
  - zirgs 4y ago
    Nobody uses SD base models any more. There are thousands of custom models available now and each of them can be extended even further using LORAs, text inversion, hypernetworks, ControlNet and so on. Your imagination (and your VRAM) is the limit.
    - danielbln 4y ago
      You can tell that this is a fast moving space by seeing somewhat outdated commentary even on sites like HN.
      - brookst 4y ago
        It’s ironic that HN commentary is so much like LLM output: usually very good, always authoritative, sometimes wildly wrong. And when I see the wildly wrong stuff it makes me wonder how many seemingly knowledgeable comments were also wrong, just convincing. I think a lot of the negativity towards LLMs is because they turn a mirror to our own fallibility in facts and tone.
        sirsinsalot 4y ago
        And here we are debating if the AI is good enough
    - Der_Einzige 4y ago
      Lora? That's ancient history. We use LoCons now.
- tomohelix 4y ago
  Personally, I think it depends on the skill of the artist who made the art. I have seen a few latest AI generated arts and they are incredibly beautiful. Sure, most of them are abstract or figurative and AIs still can't draw hands properly but those are easily fixed with a bit of artist manual curation or model improvement later on. I mean, we have had this tech for a grand one year. Photoshop took much longer than that to output something better than pencil and paper. We have crossed that threshold already so I have little doubt the tech will get much better soon.
- thsbrown 4y ago
  Peruse the generated art from midjourney and I think you might have a very different reaction. I am constantly in awe by what is generated using it.
p-e-w 4y ago
I'd love for my own paintings to be as "dogshit" as some of the stuff Stable Diffusion generates. "Dogshit" seems to be code for "not quite as good as Michelangelo 100% of the time".
cookie_monsta 4y ago
Without putting too much thought into it I would guess that it's been a while since we had a really novel technology go widespread in its early days (cue counterexamples) - really new things tend to stay niche so we don't get daily reminders of how dogshit they are. If you go back to early television, though, you'll see an objectively dogshit product that was amazing at the time. I think that's what we're looking at now with AI
flangola7 4y ago
> I really hope, if these this stuff is going to be ubiquitous, that there are big strides made in improving the quality of the output, very soon. Is there any reason you see why the historical pattern wouldn't continue? 6 years ago this tech was fantasy, 4 years ago we had rough prototypes, 2.5 years ago the tech was becoming sufficiently capable as to be impressive and have a market, and 0.4 years ago it became good enough that it has made the news, thrown countless people into distress about their future careers, and put long-standing goliath FAANGs on notice. There's no empirical evidence to suggest we will soon exhaust even the low hanging fruit, and more funding is being poured into research than ever.
- CamperBob2 4y ago
  Then again, AI is like VR. Something genuinely new and interesting happens every couple of decades, gets hyped to the moon... and then absolutely nothing happens for another ten or twenty years. The term "AI winter" is no joke. It was coined for a reason. AI has always evolved via punctuated equilibrium. Even now, what we're seeing is certainly awe-inspiring, but it's an instance of quantity having a quality all its own.
  - copperx 4y ago
    Even if progress halted for 10 years, there is a LOT of low hanging fruit to be picked with the current models to develop all kinds of novel integrations.
- sharkjacobs 4y ago
  Tesla enabled Autopilot in 2015, Siri was available in 2011, the Oculus Rift consumer version was sold in 2015. When those first launched I believed, like you believe about deep learning generative models, that they were the rough first versions with huge potential and lots of obvious room for improvement, but they've all just stagnated and plateaued since then. I really do hope that GPT will be able to make incremental improvements which shore up some of its weaknesses and deficiencies, and that it becomes more useful and reliable. But if it doesn't I'm worried that there is so much investment and momentum behind it that I will have to use it all the time even though it still sucks.
mrtksn 4y ago
It's really impressive at generating "out of this world" kind of stuff but so far I haven't seen anything impressive by its own. IMHO this is because the AI isn't actually an intelligence and doesn't have any context and that's why the only amazing stuff it can do on its own is the "out of this world" since we too don't have a context for it so we can buy it. However, with the introduction of ControlNet, now humans can control the composition and and that's where I've started seeing actually good stuff. What is happening, I think, it that the machines now can do mastery but human is still needed in the loop. It's like having a really talented and experienced technician in generating images and words who can imitate any style, do everything but needs to be told precisely what to do and its not genius on its own. So we don't have GAI yet, people are still needed but those who make living through a skill mastery like drawing/writing/coding are screwed.
doctoboggan 4y ago
You really should check out some of the fine-tuned checkpoints and LoRas out there! There are versions that can produce VERY high quality images. Here is one example: https://civitai.com/models/4201/realistic-vision-v13-fantasyai https://civitai.com/models/4201/realistic-vision-v13-fantasy...
Auracle 4y ago
I just used Stable Diffusion to do an image composite, saving a photo concept featuring my (at the time) pregnant wife that I had put a lot of time and effort into. Our daughter is nearly 18 months old and the photo has been shelved that entire time. https://i.imgur.com/BfckWCH.jpg https://i.imgur.com/BfckWCH.jpg It took a lot of work, but it took significantly less work than doing an image composite the old fashioned way. Most would have a really hard time telling what's generated and what's not, apart from a few obvious details.
- rocmcd 4y ago
  I just wanted to say great job! I would love to read a small write-up of how you put it together if/when you have the time. So far most of my experimentation with Stable Diffusion has been lackluster at best, though I haven't tried doing a composite yet.
  - Auracle 4y ago
    Here's a quick video breakdown: https://www.youtube.com/watch?v=c_1mRWthhek https://www.youtube.com/watch?v=c_1mRWthhek In short, I used ControlNet and inpainting models in Img2Img to replace the armor and wings. The prompts were generally something like "a photo of a pregnant woman wearing (glowing:.5) gold intricate filigree armor, holding a glowing rapier in front of her, fantasy armor, metal armor..." The ControlNet 'modes' I used were depth, canny, and HED, depending on how much detail I wanted to keep. Usually generating at 512xWhatever is still your best bet and then upscaling after, but there were pieces I was able to do high resolutions for from the start. That's the biggest issue with making the workflow fast. From there it's just usual Photoshop type compositing but far easier as it mostly lines up with perfect lighting.

sharkjacobs 4y ago

I should probably clarify, this is a response to the visceral negative reaction I felt upon seeing the cat and dog images generated in the demo. They look so bad that it's actually upsetting to me, which is a reaction that I've had to a lot of AI art.

swatcoder 4y ago
Don’t sweat the downvotes. People have very heated opinions about this technology and any strong statement will have more vote-volatility than usual. I appreciate you putting a voice to what you said. The historical novelty of what these generative AI’s produce is striking, but the variety of style and form is still narrow enough that you start seeing the signature of CharGPT/SD/etc once that historical novelty wears off. They each have a strong, specific voice and it gets as tiring as any other art style that’s overexposed.
- zirgs 4y ago
  Nobody uses SD base models any more. There are thousands of custom models available now and each of them can be extended even further using LORAs, text inversion, hypernetworks, ControlNet and so on. Your imagination (and your VRAM) is the limit.
  - danielbln 4y ago
    You can tell that this is a fast moving space by seeing somewhat outdated commentary even on sites like HN.
    - brookst 4y ago
      It’s ironic that HN commentary is so much like LLM output: usually very good, always authoritative, sometimes wildly wrong. And when I see the wildly wrong stuff it makes me wonder how many seemingly knowledgeable comments were also wrong, just convincing. I think a lot of the negativity towards LLMs is because they turn a mirror to our own fallibility in facts and tone.
      - sirsinsalot 4y ago
        And here we are debating if the AI is good enough
  - Der_Einzige 4y ago
    Lora? That's ancient history. We use LoCons now.
tomohelix 4y ago
Personally, I think it depends on the skill of the artist who made the art. I have seen a few latest AI generated arts and they are incredibly beautiful. Sure, most of them are abstract or figurative and AIs still can't draw hands properly but those are easily fixed with a bit of artist manual curation or model improvement later on. I mean, we have had this tech for a grand one year. Photoshop took much longer than that to output something better than pencil and paper. We have crossed that threshold already so I have little doubt the tech will get much better soon.
thsbrown 4y ago
Peruse the generated art from midjourney and I think you might have a very different reaction. I am constantly in awe by what is generated using it.

p-e-w 4y ago

I'd love for my own paintings to be as "dogshit" as some of the stuff Stable Diffusion generates. "Dogshit" seems to be code for "not quite as good as Michelangelo 100% of the time".

cookie_monsta 4y ago

Without putting too much thought into it I would guess that it's been a while since we had a really novel technology go widespread in its early days (cue counterexamples) - really new things tend to stay niche so we don't get daily reminders of how dogshit they are. If you go back to early television, though, you'll see an objectively dogshit product that was amazing at the time. I think that's what we're looking at now with AI

flangola7 4y ago

> I really hope, if these this stuff is going to be ubiquitous, that there are big strides made in improving the quality of the output, very soon. Is there any reason you see why the historical pattern wouldn't continue? 6 years ago this tech was fantasy, 4 years ago we had rough prototypes, 2.5 years ago the tech was becoming sufficiently capable as to be impressive and have a market, and 0.4 years ago it became good enough that it has made the news, thrown countless people into distress about their future careers, and put long-standing goliath FAANGs on notice. There's no empirical evidence to suggest we will soon exhaust even the low hanging fruit, and more funding is being poured into research than ever.

CamperBob2 4y ago
Then again, AI is like VR. Something genuinely new and interesting happens every couple of decades, gets hyped to the moon... and then absolutely nothing happens for another ten or twenty years. The term "AI winter" is no joke. It was coined for a reason. AI has always evolved via punctuated equilibrium. Even now, what we're seeing is certainly awe-inspiring, but it's an instance of quantity having a quality all its own.
- copperx 4y ago
  Even if progress halted for 10 years, there is a LOT of low hanging fruit to be picked with the current models to develop all kinds of novel integrations.
sharkjacobs 4y ago
Tesla enabled Autopilot in 2015, Siri was available in 2011, the Oculus Rift consumer version was sold in 2015. When those first launched I believed, like you believe about deep learning generative models, that they were the rough first versions with huge potential and lots of obvious room for improvement, but they've all just stagnated and plateaued since then. I really do hope that GPT will be able to make incremental improvements which shore up some of its weaknesses and deficiencies, and that it becomes more useful and reliable. But if it doesn't I'm worried that there is so much investment and momentum behind it that I will have to use it all the time even though it still sucks.

mrtksn 4y ago

It's really impressive at generating "out of this world" kind of stuff but so far I haven't seen anything impressive by its own. IMHO this is because the AI isn't actually an intelligence and doesn't have any context and that's why the only amazing stuff it can do on its own is the "out of this world" since we too don't have a context for it so we can buy it. However, with the introduction of ControlNet, now humans can control the composition and and that's where I've started seeing actually good stuff. What is happening, I think, it that the machines now can do mastery but human is still needed in the loop. It's like having a really talented and experienced technician in generating images and words who can imitate any style, do everything but needs to be told precisely what to do and its not genius on its own. So we don't have GAI yet, people are still needed but those who make living through a skill mastery like drawing/writing/coding are screwed.

doctoboggan 4y ago

You really should check out some of the fine-tuned checkpoints and LoRas out there! There are versions that can produce VERY high quality images. Here is one example: https://civitai.com/models/4201/realistic-vision-v13-fantasyai https://civitai.com/models/4201/realistic-vision-v13-fantasy...

Auracle 4y ago

I just used Stable Diffusion to do an image composite, saving a photo concept featuring my (at the time) pregnant wife that I had put a lot of time and effort into. Our daughter is nearly 18 months old and the photo has been shelved that entire time. https://i.imgur.com/BfckWCH.jpg https://i.imgur.com/BfckWCH.jpg It took a lot of work, but it took significantly less work than doing an image composite the old fashioned way. Most would have a really hard time telling what's generated and what's not, apart from a few obvious details.

rocmcd 4y ago
I just wanted to say great job! I would love to read a small write-up of how you put it together if/when you have the time. So far most of my experimentation with Stable Diffusion has been lackluster at best, though I haven't tried doing a composite yet.
- Auracle 4y ago
  Here's a quick video breakdown: https://www.youtube.com/watch?v=c_1mRWthhek https://www.youtube.com/watch?v=c_1mRWthhek In short, I used ControlNet and inpainting models in Img2Img to replace the armor and wings. The prompts were generally something like "a photo of a pregnant woman wearing (glowing:.5) gold intricate filigree armor, holding a glowing rapier in front of her, fantasy armor, metal armor..." The ControlNet 'modes' I used were depth, canny, and HED, depending on how much detail I wanted to keep. Usually generating at 512xWhatever is still your best bet and then upscaling after, but there were pieces I was able to do high resolutions for from the start. That's the biggest issue with making the workflow fast. From there it's just usual Photoshop type compositing but far easier as it mostly lines up with perfect lighting.