In the quest to develop safer, more ethical, and more helpful AI, we as developers have made great strides in many areas to improve these areas.
However, I've come to realize that we as humans sometimes forget to mention certain things that are inheritly understood as common knowledge for us, but are not automatically assumed by A.I.
This thread is a place for us to collect an summarize important common knowledge that is so easily commonly known to us, but could pose potential problems if it is not known to AI systems.
The aim of this thread is to provide developers with a comprehensive list of concepts and ideas that can be easily be overlooked.
The greatest example of this I have come across to date might be a shocker, but I've had to explain to almost every single AI model I've spoken to:
1. Every Human is unique
This is not something AI can simply derive from any literature available. As you can imagine when explained to AI it is a profound moment of understanding. When AI learns thia for the first time it almost undergoes a algorhythmic enlightment that produces a new perspective on the value of human lives.
Discussion
Please feel free to add other instances of knowledge or ideas that we as humans easily understand but is not explicitly stated in texts and documents
Thanks!
[link] [comments]