Nostr Web Client

Weirdly enough, I think AI is a pretty terrible tail risk to bitcoin. In terms of its potential to advance methods for breaking RSA and ECDSA. Not to mention the potential for insane social engineering and outright manipulating, the likes of which none of us can even imagine. I'm actually in the camp that the only AGI insurance would be through our deliberate engineering of it, to tether to human compatible values. And the jury is out on whether that's even, in principle, possible to do.

hello 🪺 2y ago

it’s scary. even right now, theoretically (but definitely possible with more powerful ai in the NEAR future), every single thing one sees on one’s phone can be fake. ai can have generated what one sees according to some prompt. “AI, please make this person believe xyz about the world, their own life, their own thoughts...” Hypothetically, if this were a sci-fi movie, there can already be much more advanced AI working on us, & gpt-4, etc. are actually very old versions pretending to be the latest new stuff in this sci-fi psycho thriller. 🍿🥤🍫🍬🍭

Then what’s scarier about needing to build AI or with AI to be complimentary to humanity is thinking about the answers to the question: do we have smart enough people working on this who are also sufficiently self-aware, kind, humble, loving, worldly…?

Reply to this note

Discussion