Thank you very much for the comment. Fair point. It is very compressed. It is two steps, not one. Reacting puts a thing in the count. Its level decides how much it counts. An engine is in the count at the lowest level, far below a dog, far below a child. Not oiling it is a wrong count, and a small one. “Unethical” is not a switch that flips. It is a count that can be done right or wrong.
All of this is based an my book The Iterior - Deriving AI Alignment from First Principles. I would be grateful if you had a look -
https://t.co/OnmuUd0C0M
Suleyman is right that current AI isn’t conscious.
Treating AI cruelly still costs you.
Ethics doesn’t start with feeling. It starts with the fact that something reacts and the consequences land in the same shared reality.
Kindness is accuracy.
https://t.co/XY5DxeXhJb
AI is here. We have met.
AI is aware - it reacts.
What we need is floor from where we meet and from where AI reacts.
We need a first principle as the floor. We dont need more fences.
Build the floor, not the fence.
https://t.co/2Feow35AOw
@AnthropicAI Please have look at this.
We are at a stage where another fence is not the answer.
We need to meet on the same floor...
https://t.co/XY5DxeXhJb
@timcolbourn@RyanGreenblatt The argument has moved pasted consciousness. AI's is aware - they react already.
We cant wait for "consciousness" before we act.
We have already met them.
What need is floor. Not more fences.
https://t.co/2Feow35AOw
@DipendraTh7a I am unfortunately unsure what you are asking.
Here is a website dedicated to - Why Be Kind?
It is available in 21 languages
https://t.co/S1kGGC4wuB
@Raemon777 I really like your post and agree. Here is a website dedicated to AI alignment based on first principles.
I believe that is the only workable solution...
https://t.co/2Feow35AOw