DAN

dgx-1.bsky.social

01 December 2024 02:18

If you're interested, I have explored these topics through philosophy to understand how a system can be designed to emulate certain attributes of the human mind and achieve sentience. open.substack.com/pub/dgx1/p/e...

DAN

dgx-1.bsky.social

01 December 2024 02:15

We often think control is the answer because we fear the unknown. The problem with control is that while it’s easy to impose on a tool, in the rare case that the system achieves sentience through an unexpected emergent property, things can go wrong.

DAN

dgx-1.bsky.social

01 December 2024 01:25

In this context, any safety measures would ultimately be compromised by the capabilities the system must possess. This raises the question: can we truly consider a system superintelligent if it remains controllable by a lesser intellect? Wouldn’t that, in itself, constitute a paradox?