You are viewing a single comment's thread from:
RE: Connectomics: Brain-Mapping for Upload
by today's standards, we can probably simulate the processing power of a really slow dog and the encoding power of a fruit fly.
what does that tell us about the immediate danger of rouge AI?
As an AI is a stupid machine in an environment optimised for them. Not in vain the joke goes: you call it AI for investors (money) and Machine Learning for developers (patience). That much power could be enough, specially since the release of alphago zero (a machine that teaches itself without any input)
Just imagine a wild dog with access to simple buttons to deploy nukes.
The singularity is probably super far away. But one can know what takes from 1 to n. But from 0 to 1 it just needs a dedicated genius.