FOR EDUCATIONAL AND KNOWLEDGE SHARING PURPOSES ONLY. NOT-FOR-PROFIT. SEE COPYRIGHT DISCLAIMER.

AGI “Guardrails”? Unfortunately NO. (provable jailbreaks)

Mathematically Provable Containment, Forever? YES. (sole option)

Learn more:

• The Containment Problem (common knowledge)

• The AI Safety Problem (common knowledge)

• We do not understand how these systems work (Russell)

• Natural Selection Favors [Intelligence] AIs over Humans (Hendrycks)

• Provably Safe Systems: The Only Path to Controllable AGI. (Tegmark & Omohundro)

Very good reading from very respected sources!!!

Meanwhile, every day, our p(doom) goes up and up and up…

FOR EDUCATIONAL AND KNOWLEDGE SHARING PURPOSES ONLY. NOT-FOR-PROFIT. SEE COPYRIGHT DISCLAIMER.