What We Do

We know the last thing you want to hear more about is AI. Someone with the right knowledge needs to do something about all the glaring problems before it’s too late. We’ll try, but we need your help.

Real Good AI is a 501(c)(3) research nonprofit to protect the arts and creativity in the world of AI.

Our Mission

Real Good AI is a research nonprofit determined to fix the issues creatives and the public have with AI. Companies are pushing AI that is unexplainable, dangerous, wasteful, stealing data, and is often confidently wrong. We are doing fundamental math research to replace those “black box” parts of machine learning so we all understand where things came from and AI needs way less data and resources. We use statistics and machine learning as a tool for projects supporting the public good. Finally, we believe in transparency in AI NOW and give unbiased, free communication and public education on AI. If you believe in science exploration that doesn’t destroy the environment, doesn’t lead to the loss of human creativity, and is transparent and understandable, you’ll find we have a lot in common. And that is Real Good.

We hire human artists to do our art and will not use generative AI “art” to replace humans.

AI models were trained on art without consent so we believe the created images are copyright infringement. Because current neural network models are “black boxes,” proper credit cannot be applied to those forced to be in the training data, and it is not clear how AI uses the training information. In fact, some of our current research is to try to help mathematically explain this theft. The point is not to pretend that the harm was never done; it is to mitigate future harm and allow artists a seat at the table. Artists can’t have a seat if they never enter the room. AI generated videos, voice, pictures, music, and art can be dangerous to humans in psychology, scams and deepfakes on which we educate the public.

We are not anti-technology, we are just against the current system of exploitation with AI.

We are NOT building another chatbot or image generator.

We ARE doing research for a better world.

Fixing AI Issues

There is a trade-off in modeling: mathematical structure vs. data. Current Neural network models rely heavily on large amount of data, which causes a cascade of catastrophic issues. We think there are more reliable alternatives, and we want to show how this is possible to the community.

Doing good in the world with AI

There are a lot of vital problems humanity is facing that are not profitable to solve, so for-profit companies are ignoring them. Our team is making it our mission to use our knowledge of statistical and machine learning tools to help the world.

We deserve to know what’s REAL

AI is being put in everything, but the impact of that AI can vary wildly based on the use. So saying “Yes” or “No” to AI becomes a confusing question. The REAL rating distinguishes between tools like AI spellcheck in our programs (REAL 1), full AI generation (REAL 5), and everything in between.

We imagine a future where you can filter your social media feed by how much AI was used in the creation of the post. Or a teacher can educate students about appropriate AI use since abstinence-only education does not work. Or an artist can proudly display their work without being questioned about AI. All these are possible with REAL.

Learn more

Our Other Outreach Programs