Uncanny Valley

04/07/2023

Inspiration für neue Klamotten. KI sind nicht nur die Meister der Mustererkennung, sie können auch interessante Muster erzeugen.

30/06/2023

Ein paar erste Versuche mit Stable Diffusion XL, unbearbeitet. Kaum noch Ausschuss. Hoffentlich lässt sich das bald lokal installieren, sodass man damit arbeiten kann. Parameter einstellen wäre auch gut. Und am wichtigsten: Image to Image!

Die Entwicklung ist soooo rasant. Kaum vorstellbar, was in einem Jahr alles möglich sein wird.

29/06/2023

https://clipdrop.co/stable-diffusion
Sehr hohe Bildqualität und Auflösung. Nicht schlecht für eine Umsonst-Seite, die online läuft.
Leider gar keine Parameter zum Einstellen.

Remove background, cleanup pictures, upscaling, Stable diffusion and more…

08/01/2023

Here an example in 2:1 (1260px x 640px upsampled to HD width 1920px)

08/01/2023

Using different aspect ratios, for example here 21:9, is difficult in art generation, since all pictures are based on seeds that are square (here: 512x512 pixels). Larger Images tend to double their content, which can be seen in this video.
Some images and videos just need the width of a cinema screen. But then it's just a lot more trial and error. :)

08/01/2023

https://www.codingem.com/best-ai-art-generators
Overview of AI-art generators:

This is a comprehensive guide to finding the best AI art generator software. The tools in this list use AI to convert text to images.

07/01/2023

01/01/2023

a draft for the video „glitches“

01/01/2023

The glitches appear, because we trained a model on ai generated fotos. This recursive process seems to amplify any glitches inherent in generated art.

Learning from introspection amplifies your misconceptions. Go out there and act, if you want to learn.

30/12/2022

A point of virw on the dangers of AI development

AI gets smarter, more capable, and more world-transforming every day. Here’s why that might not be a good thing.

17/12/2022

https://chat.openai.com/chat
We asked the OpenAI chat to describe itself:
„I am an artificial intelligence trained by OpenAI. My primary function is to assist users by providing information and answering questions to the best of my ability. I do not have personal experiences or the ability to interact with the world in the same way that humans do. Instead, I rely on the knowledge and instructions that have been programmed into me by my developers. I do not have the ability to form personal opinions or make decisions, and I do not have access to personal data or records. I am simply a collection of algorithms designed to process and generate text based on the prompts and questions that I receive.“

A conversational AI system that listens, learns, and challenges

17/12/2022

https://experimental-epistemology.ai/the-red-pill-of-machine-learning/
Very good article on epistemology and ai:

Reductionism is the use of Models. Holism is the avoidance of Models. Models are scientific models, theories, hypotheses, formulas, equations, naïve models based on personal experiences, superstitions (!), and traditional computer programs

16/12/2022

https://www.timothybrooks.com/instruct-pix2pix
Next level of image editing:

We propose a method for editing images from human instructions: given an input image and a written instruction that tells the model what to do, our model follows these instructions to edit the image. To obtain training data for this problem, we combine the knowledge of two large pretrained models---...

05/12/2022

Comprehensive List of AI-Tools

heres the tools you needed for your research

Looking to get started with AI art? A good place to start is one of the popular apps like DreamStudio , midjourney , Wombo , or NightCafe . You can get a quick sense of how you can use words and phrases to guide image generation. Read up on prompt engineering to improve your results. T...

24/11/2022

https://futurism.com/the-byte/clip-interrogator-mean-ai

CLIP Interrogator is the hot new AI generator on the block, but instead of generating images, this one very rudely does the reverse.

08/11/2022

Mubert creates Music with AI. It is hardly less boring then most stuff you hear on the radio.

Mubert - The new royalty-free music ecosystem for content creators, brands and developers 🔥 Come See How Our High-Quality Music Can Elevate Your Content ⏩

07/11/2022

https://jalammar.github.io/illustrated-stable-diffusion/
This guy tries to explain how Stable Diffusion works for people without a PhD in Computer Science. He fails, but he fails less then most others.

AI image generation is the most recent AI capability blowing people’s minds (mine included). The ability to create striking visuals from text descriptions has a magical quality to it and points clearly to a shift in how humans create art. The release of Stable Diffusion is a clear milestone in thi...

05/11/2022

https://infiniteconversation.com/

An AI generated, never-ending discussion between Werner Herzog and Slavoj Žižek.

21/10/2022

https://diffusionbee.com/
Stable Diffusion for M1 Mac, easy to install, lots of features, but somewhat limited in the range of options. But a good and easy point to start:

DiffusionBee is the easiest way to run Stable Diffusion locally on your M1 Mac. Comes with a one-click installer.

14/10/2022

Sometimes things don't turn out as expected. The following weirdness emerged from the same settings:

14/10/2022

A good case of the human brain's obsession to produce sense even from randomness:

10/10/2022

03/10/2022

Intelligence is all about recognizing patterns. So is artifical intelligence. We asked our AI to produce some patterns on its own.

30/09/2022

The next step:

Meta's giving us more efficient video generation to bring motion to latent space.

23/09/2022

Artists are free to create fantasies or nightmares with unrestricted image generator Stable Diffusion, but some fear a flood of AI-made horrors.

22/09/2022

Based on a simplistic drawing of Dora Laetitia Asemwald, a AI (Stable Diffusion 1.4) tried to make some illustrations. Here are some results.
Music written and produced by Martin Zentner

19/09/2022

Flowers and Tentacles
Music/Video: Martin Zentner

03/09/2022

Try it yourself? Here is an easy way for the first steps:
https://huggingface.co/spaces/stabilityai/stable-diffusion

Discover amazing ML apps made by the community

03/09/2022

Our first try on producing a video out of generated pictures. The music isn't yet generated, it is still produced the old way.