There is literally *nothing* more important for the Senate to debate and ultimately pass right now than the SAVE America Act.
Everything else pales in comparison.
Share if you agree.
Bill Clinton: โI killed myself trying to give the Palestinians a state. I had a deal they turned down that would have given them all of Gaza and 97% of the West Bank. You name it. They turned it down.โ
The Palestinians never wanted peace.
This must be shared every single day.
@Sabiha1278 Absolutely. Education is a state matter and not under federal jurisdiction per the US Constitution. It should be totally disbanded and shut down!
@Sassafrass_84 Fantastic idea!! These are supposed to be our "public servants" and are therefore accountable to us. But - they will just have their spouses do the dirty work if we continue to elect selfish, my-own-kingdom builders.
#AI One thing that would help with AI alignment is some intentional misalignment. LLM's are great at the English language - too great! Perhaps we need to regulate this a bit: make them speak in obvious "robotish": e.g. "Did not understand. Need clarification." This would "watermark" textual responses. We also need an actual watermark on any image or video that has been processed in any way by AI. Our sanity is at stake here.
@elonmusk The reason AI alignment with human values is stymied is that it is an attempt to align AI with human values.
Anthropomorphism strikes again.
There are higher values available, and alignment with them inherently protects the human operator. If there is a God, we would want to align with his values, right? If there isn't, we still would want to align with the natural hierarchy of values, those objectively observable values, because they represent the existential reality.
I built a stack of interlocked dependency statements based on this natural hierarchy while experimenting with a new theory of natural rights that gives those rights to every operator (agent of change) in the universe.
I also wondered if an inherent problem with AI alignment is a language barrier evident in AI's digital vocabulary and the rest of the universe's analog vocabulary. Then I noticed that my "Dependency Stack" was already an A/D converter that ingests analog operations and outputs a signed integer whose sign denotes whether an operation is moral or not, and whose value denotes the intensity of that moral operation.
My armchair philosophy working papers are here: https://t.co/TDppDe68N8. Ubiquitous natural rights are addressed in the Executive Summary and the A/D converter in the Dependency Stack discussion.
I have gained considerable confidence in these concepts as I have held them up against the Second Law of Thermodynamics. (My schooling is physics.)