Replies

The departure point is when the “intelligent” machines no longer need humans to evolve in their ability to remake the world according to their “goals”. Moore’s law has a time period of about 18 months to double capabilities. If self evolving machines cut that to days, their evolution will be incredible. We will also be unable to understand the changes.

When we do not understand what the machines are doing...well I won’t understand it either. And that is the crux of the situation, we will be observers to the world changers. It presents three possibilities.

They like us. The world is good.

They hates us. Bad things follow.

They do not give a rat’s patootie about us. Interfering may or may not be bad, but I doubt we would really understand how to interfere effectively.

Imagine a MMA opponent that did not have humanly breakable joints, reacts in the millisecond vicinity, has complete access to martial arts types that even include ones there were only videos of, and could hit like a Mack Truck. Would you go into that ring? If it promised not to kill you intentionally would that be important? Then assume the three above scenes.

One it might try to teach you how to win, if you promised to not use it on one of those wonderful other humans.

Two it would kill you...or worse.

Three it may stop long enough to give you that go away stare, but probably would not stop long enough to exhibit you were any more of an impediment than the slug in your garden.

We seem to be falling into the idea that an individual machine body isolates the machine. It really does not. MIT has been working on swarms for years and now a weapon is out. My watch talks to my phone, and then my computer. Firefox targets my adds and Google is classifying everyone’s pictures tagging them with names. We are not noticing the minor steps to that end but it is getting faster.

Too much?