I have open-sourced methods for overcoming the limitations of humanities and an AI ethics framework under the MIT License.
future AI x dignity
Good morning to all Substack readers - it's morning here in Japan time.
From our daily lives, is it possible for us to, for example, make donations and help developing countries become more prosperous?
There are billionaires in the world, but is it possible for these wealthy individuals to provide funding?
While neither may be impossible, both take time. When things take time, lives are lost, and social structural problems continue, such as the misuse of technology. Isn't there something we can do? Isn't there something individuals can do?
Ah, there's the MIT License, there's GitHub, there's Academia.edu. So why not combine them? With this in mind, I've completely open-sourced the framework I created for my mother.
As a result, we're now in an era where a mini PC costing about 70,000 yen can run a local version of a large language model equivalent to GPT-3.5 from 2023 - albeit a smaller version. This means that in situations where infrastructure is damaged or insufficiently developed, as long as there's electricity, AI can operate on computers without needing an internet connection.
This could be dangerous if misused. However, since misuse is also a matter of social structure, and it's unlikely that someone would deliberately misuse technology when they can succeed with ethically and socially acceptable business models, by releasing this under the MIT License, if people can develop businesses using this preventive approach to support welfare or assist people with illnesses or disabilities, society will improve and the probability of misuse will decrease.
This is the best contribution to the world that I can make right now. And I want you to remember natural science - when telescopes became available, reproducibility and verifiability increased, didn't they? Humanities such as philosophy and ethics often face the problem of having beautiful theories but struggling with how to put them into practice. However, if future researchers can say, like me, "I'm open-sourcing this on GitHub," then verifiability can be ensured even in the humanities.
Perhaps the greatest contribution to society this time is that I've developed a new research approach for academia and open-sourced that framework to the world. This is my current harvest.
Please feel free to visit my homepage. I've created a Japanese-English bilingual version of my mother's writings. There are also photos of flowers.
Quick Start
Review FAQ for complete technical details
Examine primary sources for validation
Deploy RSS system for immediate testing
Adapt framework to your specific use case
Read the paper "future AI x dignity: Two-Year Zero-Incident Report and Primary Sources" for theoretical background
The photo is one that my mother took with a Galaxy A23 5G and shared via Google Photos for this project.
Trgr KarasuToragara
https://independent.academia.edu/TrgrKarasuToragara
https://github.com/trgr-karasutoragara
https://note.com/karasu_toragara