Kiteet
@kiteet.bsky.social
Neo-romantic anarcho techno environmental scientist craftsman.
With AI, I have worked with painters, theatre & musical and NGOs in innovative projects.
I also write electronic music. For π΅ I work with sustainability in cryogenics. Also a gardener πΏπ±
With AI, I have worked with painters, theatre & musical and NGOs in innovative projects.
I also write electronic music. For π΅ I work with sustainability in cryogenics. Also a gardener πΏπ±
This is very interesting π€
October 1, 2025 at 7:20 PM
This is very interesting π€
#gemini did a surprisingly good job of creating a composite of two old #midjourney v4 images. The more the images are iterared, the more the features change, but what is very good about Gemini (or #nanobanana actually, I think) is that the automatic segmentation Is really smart.
October 1, 2025 at 4:18 PM
#gemini did a surprisingly good job of creating a composite of two old #midjourney v4 images. The more the images are iterared, the more the features change, but what is very good about Gemini (or #nanobanana actually, I think) is that the automatic segmentation Is really smart.
#veo3 is finally an #aivideo tool that creates good and somewhat predictable result easily. I gave #geminipro some old #midjourney v4 images from way back to create composites, and Veo animated those better than I expected, really.
October 1, 2025 at 4:13 PM
#veo3 is finally an #aivideo tool that creates good and somewhat predictable result easily. I gave #geminipro some old #midjourney v4 images from way back to create composites, and Veo animated those better than I expected, really.
I saw this tacky giant print at a shopping centre today, looks like Stable Diffusion XL + stock ESRGAN. Why not tiled ControlNet or just any tiled upscaler? Who knows, but it could have been so much more detailed π§ #sdxl #aiart #esrgan #digitalart #aiart
September 6, 2025 at 8:47 PM
I saw this tacky giant print at a shopping centre today, looks like Stable Diffusion XL + stock ESRGAN. Why not tiled ControlNet or just any tiled upscaler? Who knows, but it could have been so much more detailed π§ #sdxl #aiart #esrgan #digitalart #aiart
Phew, so much going on nowadays π
But that seems a lot of fun and I hope could be bent in bizarre ways. It needs videos I assume? Well, going to look for some resources...
March 9, 2025 at 9:14 PM
Phew, so much going on nowadays π
But that seems a lot of fun and I hope could be bent in bizarre ways. It needs videos I assume? Well, going to look for some resources...
Wait, it can be finetuned? π§
March 9, 2025 at 5:19 PM
Wait, it can be finetuned? π§
A "spot the AI" game by the media awareness NGO Faktabaari is out. I made the images. What did I learn?
Red Panda would be terrifying if it's VAE was not weak.
Flux Dev is sometimes better than Pro Ultra. The new inpainting is bad.
SD3.5 π
Ideogram, Photon, etc π
peli.faktabaari.fi/en
#media
Red Panda would be terrifying if it's VAE was not weak.
Flux Dev is sometimes better than Pro Ultra. The new inpainting is bad.
SD3.5 π
Ideogram, Photon, etc π
peli.faktabaari.fi/en
#media
January 29, 2025 at 7:23 AM
A "spot the AI" game by the media awareness NGO Faktabaari is out. I made the images. What did I learn?
Red Panda would be terrifying if it's VAE was not weak.
Flux Dev is sometimes better than Pro Ultra. The new inpainting is bad.
SD3.5 π
Ideogram, Photon, etc π
peli.faktabaari.fi/en
#media
Red Panda would be terrifying if it's VAE was not weak.
Flux Dev is sometimes better than Pro Ultra. The new inpainting is bad.
SD3.5 π
Ideogram, Photon, etc π
peli.faktabaari.fi/en
#media
Does the audio glitch a lot on your end? I think BS broke the compression. There's a bit of natural CPU overload on the track also, but on my end the whole uploaded video crackles a lot.
January 25, 2025 at 8:44 AM
Does the audio glitch a lot on your end? I think BS broke the compression. There's a bit of natural CPU overload on the track also, but on my end the whole uploaded video crackles a lot.
Are you sure AIs threaten our cognitive function? I use them a lot in my dayjob for drafting boring texts and expanding my notes for circulation. So far I've found that I have a feeling of considerably lighter cognitive load when I can both relax and concentrate more on the interesting stuff.
January 21, 2025 at 7:01 PM
Are you sure AIs threaten our cognitive function? I use them a lot in my dayjob for drafting boring texts and expanding my notes for circulation. So far I've found that I have a feeling of considerably lighter cognitive load when I can both relax and concentrate more on the interesting stuff.
In a better world, we'd have more pulp #adventure movies.
Some recent test images from #FluxDev finetunes.
Some recent test images from #FluxDev finetunes.
January 20, 2025 at 7:52 PM
In a better world, we'd have more pulp #adventure movies.
Some recent test images from #FluxDev finetunes.
Some recent test images from #FluxDev finetunes.
Oh, it's just their own life and looking through a dirty window π
January 14, 2025 at 6:30 AM
Oh, it's just their own life and looking through a dirty window π
I love that overgrown city. I really love organic forms which vaguely, but not exactly resemble something real.
January 11, 2025 at 10:49 AM
I love that overgrown city. I really love organic forms which vaguely, but not exactly resemble something real.
January 11, 2025 at 10:07 AM
LoRA mix of my own models. Nice perspective, often AI characters tend to be flat by default.
#aiart #characterconcept #ai
#aiart #characterconcept #ai
January 11, 2025 at 10:05 AM
LoRA mix of my own models. Nice perspective, often AI characters tend to be flat by default.
#aiart #characterconcept #ai
#aiart #characterconcept #ai
More test renders with a mix of some recent Flux LORAs I've trained.
I'm really particular to organic shapes which almost but not quite resemble existing things. This fascination has been with me long before these tools.
#aiartcommunity #fantasyart #flux1dev
I'm really particular to organic shapes which almost but not quite resemble existing things. This fascination has been with me long before these tools.
#aiartcommunity #fantasyart #flux1dev
January 10, 2025 at 10:44 PM
More test renders with a mix of some recent Flux LORAs I've trained.
I'm really particular to organic shapes which almost but not quite resemble existing things. This fascination has been with me long before these tools.
#aiartcommunity #fantasyart #flux1dev
I'm really particular to organic shapes which almost but not quite resemble existing things. This fascination has been with me long before these tools.
#aiartcommunity #fantasyart #flux1dev
I'm also a fan! It's a shame, really π’
January 10, 2025 at 10:38 PM
I'm also a fan! It's a shame, really π’
Playing around with various mixes of LoRAs I've trained recently for Flux. 1 Dev. It's amazing how well it trains, considering how bland the default model is. It's almost like if they hid all of the good vectors π
#flux1dev #aiart #illustration
#flux1dev #aiart #illustration
January 10, 2025 at 11:35 AM
Playing around with various mixes of LoRAs I've trained recently for Flux. 1 Dev. It's amazing how well it trains, considering how bland the default model is. It's almost like if they hid all of the good vectors π
#flux1dev #aiart #illustration
#flux1dev #aiart #illustration
Funny thing is, even normal photos look all like experimental vintage photography β¨
January 8, 2025 at 6:34 AM
Funny thing is, even normal photos look all like experimental vintage photography β¨
I trained a LoRA on 60s silk screened posters, but the effect spread everywhere, which is just great π
I made five runs at 15,000 steps (n=200), two without prompts and three with them. I have the initial feeling that the ones with prompts feel slightly better, but I'll need to test more.
I made five runs at 15,000 steps (n=200), two without prompts and three with them. I have the initial feeling that the ones with prompts feel slightly better, but I'll need to test more.
January 8, 2025 at 6:33 AM
I trained a LoRA on 60s silk screened posters, but the effect spread everywhere, which is just great π
I made five runs at 15,000 steps (n=200), two without prompts and three with them. I have the initial feeling that the ones with prompts feel slightly better, but I'll need to test more.
I made five runs at 15,000 steps (n=200), two without prompts and three with them. I have the initial feeling that the ones with prompts feel slightly better, but I'll need to test more.
But in general it's insanity. Anyone can run the calculation on, let's say, five thousands years of growth at 2-3% π¬ So luckily it's not a matter of opinion, we will hit the ceiling sooner or later.
December 28, 2024 at 9:04 AM
But in general it's insanity. Anyone can run the calculation on, let's say, five thousands years of growth at 2-3% π¬ So luckily it's not a matter of opinion, we will hit the ceiling sooner or later.
Yes, I completely agree. I've been thinking this about a lot regarding for an example the company I work for. Of course we could share the profits, but currently they're almost completely put into further investments. So if we were a co-op, the company should trust the owners to lend this.
December 28, 2024 at 9:03 AM
Yes, I completely agree. I've been thinking this about a lot regarding for an example the company I work for. Of course we could share the profits, but currently they're almost completely put into further investments. So if we were a co-op, the company should trust the owners to lend this.