AI Art Generation Notes from MoniGarr

• monigarr

AI Art Generator Use Cases

Folks are using AI Generators to create graphics, animations, stories, music, motion graphics, graphic novels, comic books, story plots, characters, video games, icons, illustrations, 2D references for 3D projects and so much more.

I will edit these notes as needed to include the most recent use cases and links.

  • Artist being replaced & erased by AI Art Generators
  • AI Short Film: The rebirth of magic
  • Brainstorming: Niche AI Generators compensating curated Artists with Royalties & Similar
  • Clairvoyance:
  • Kris Kashtanova Graphic Novel Zarya of the Dawn
  • MoniGarr uses their own art (visual, audio, text) as an input for AI generators to output more reference art in the form of visuals, audio and text.  The reference art is then used in an XR production pipeline for humans to create new art with various tactics including software, real world media and multi-disciplinary art as digital animations, films, games and real world performances, fashion and more.
  • Org to protect Artists in AI era.
  • Sky Ships by MoniGarr:  some info about how we are using AI Generators in our production pipelines.
  • TheZakMan: Stable Diffusion & Photoshop (img2img) Workflow: 
    1- 5 min Doodle in Photoshop
2- SD "img2img" input + prompt
3- Paintover in Adobe Photoshop
4- I added the finished image in photoshop and re-inserted it into "img2img" to get new ideas and experiment with variations
PS:All in less than 25min
PS2: Added a 4th image to show the continuous variation process, inputting the last image obtained from Photoshop into img2img again.

AI Art Generation Links

Side Paths, Accessories & Helper Links

LIST OF GITHUB REPOS

COLAB NOTEBOOKS

AI Links

LIST OF NICHE AI GENERATORS

Copyrights, Legalities, PHILOSOPHY

Reciprocity 101

I am concerned that the artists’ who’s work is being scraped to train the data models that the AI Generators learn from, are often not being credited, not receiving royalties nor any type of reciprocity for their contributions to commercial AI Generators, while a few folks that build the tools to scrape data / train and create new data models are being compensated to benefit from artists’ original works.

I do use my own original art, haikus, text, ideas, audio as my input prompts, but the trained data models for all public access / commercial AI Generators are still receiving All the compensation, credits and benefits from artists’ original works while providing zero reciprocity to the original artists.

Folks can train our own data models with our own teams of complimentary skills to curate original art works while providing credit, respect, royalties & a wide variety of compensation / reciprocity to the whole team that is involved with the production and distribution of AI Generators (original artists, software architects / engineers, marketing, legal, etc. – the whole team).  One of the traits of technical colonization is a few folks (often call themselves pioneers) take all credit, compensation, accolades, rewards, support when they willfully exploit & extract from individuals, communities, cultures that they then attempt to erase, destroy and replace with themselves.  Every one colonizes to varying extents in colonial societies willfully and unwillingly – this is an observation and not meant to cause further harm to anyone for any reason.  I do believe strongly in acknowledging our own individual actions / choices so that we can figure out healthier ways to do our work, live our best lives and just be the best we can be at any time with what we have to work with.

The teams that are providing reciprocity to the original artists that they extract data from to create their trained data models – that is truly sincerely appreciated.  I hope that more of the under-represented artists in the tech industries can figure out how to create our own AI Generation tools that also have traits of Onkwehonwehneha AI with reciprocity and respect for all of creation too.

Terminology

  • Algorithm: is an extended subset of ML that tells the computer how to learn & operate on its own. It continues to gain knowledge to improve processes & run tasks more efficiently.
  • Algorithmic Bias: is a phenomenon that occurs when an algorithm produces results that are systemically prejudiced due to erroneous assumptions in the machine learning process.
  • ANN: artificial neural network is based on a collection of connected units or nodes called artificial neurons, which loosely model the neurons in a biological brain.
  • Bias: see Algorithmic or Data Bias and Societal AI Bias.
  • Data Bias:  is a phenomenon that occurs when an algorithm produces results that are systemically prejudiced due to erroneous assumptions in the machine learning process.
  • Deep Learning: is a type of machine learning and artificial intelligence (AI) that imitates the way humans gain certain types of knowledge. Deep learning is an important element of data science, which includes statistics and predictive modeling.
  • Diffusion: refers to generative models used to make data that looks like the data they were trained on.
  • GANS: generative adversarial network
  • Inference: the process of making predictions, using trained neural network models. AI training refers to creation of models or machine learning algorithms using training datasets. Inference, training and data engineering are key stages of a common AI workflow.
  • ML: machine learning is the use and development of computer systems that learn & adapt without explicit instructions when using algorithms & statistical models to analyze & draw inferences from patterns in data.
  • NLP: natural language processing is a branch of AI in computer science that focuses on helping computers understand the way humans communicate naturally (speak & write).  MoniGarr’s projects regarding ‘Mohawk Language XR‘ and ‘Onkwehonwehneha AI’ have been creating NLP solutions since the early 1990s.
  • RNN: recurrent neural network is a type of ANN that uses time series or sequential data.
  • Societal AI Bias: happens when AI behaves in ways that reflect social intolerance and/or institutional discrimination. The algorithms and data itself might appear unbiased, but the output reinforces societal biases.