Top-level takeaway of tonight's catch-up reading on ML... I still find it hard to take AI Risk seriously as a special problem here. There are definite risks but they don't seem qualitatively different from other kinds of engineering risk.
maybe I'm misunderstanding the term "grey goo"? thought that referred to a nanoscale von Neuman replicator.
-
-
Metaphorical use, original replicator increases entropy of physical environment - consumes without producing anything interesting Badly organized weak AI might similarly eat all human interest and emotional vulnerability while driving nonsensical action.
-
the paperclip maximizer then. I see that as almost entirely a thought experiment though. our systems are not so integrated at the global scale where this matters to us. the paperclipper ends at the factory driveway.
-
if a "clickbait optimizer" gets loose on the internet then what happens? it has limitations too. how much budget for ad buy does it have? how much ad inventory does it have access too?
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.