It’s interesting that we’re monitoring AI for “self awareness”, when we haven’t fully mastered it in ourselves. 

99% of humans (myself included included) are not fully self aware: we rarely understand our real motivations, we take thoughts for granted and we are tremendously prone to suggestion and bias.

I feel like you might enjoy this book (or his lecture series at Stanford).

Operator turned investor. Work in TradFi. Exploring the future of social investing @crossroads.  🟪

Productizing RWAfi @plumenetwork. Letting you earn from all RWAs on Plume using https://nest.credit. Hosting /rwa, /plume, /nest and https://palmer.earth

I feel like you might enjoy this book (or his lecture series at Stanford).

https://www.goodreads.com/book/show/31170723-behave?ac=1&from_search=true&qid=N7VaXm0a6Q&rank=1

I think we should be putting a lot more work into monitoring AI for manipulation

“AI makes mistakes”. That’s fine, I’ll take iterating fast with GPT 4 over parsing stack overflow with a mix of random hacks and occasionally the right solution