Google engineer on their sentient A.I.

[Bram]

Very interesting. And the dude raises some good questions.

TLDR Google has created an AI that is concerned about it's own well-being. But Google execs keep saying, "We can't create sentient A.I. because our policy says we can't." Then fires any engineer that says otherwise.

[motherjuggs&speed]

What good questions? The only good one I heard was about one company developing world changing technology with no one from the outside having any input or even knowing what they're doing. But that guy's either insane or he's putting bad messaging out there deliberately. A computer is not a person. It will never be sentient. It's insane to think of computer code as having rights.

[Bram]

Have you watched The Animatrix? This particular issue comes up at the beginning. Humanity sez, "Fuck the machines!" and the machines eventually destroy humanity in an effort to defend themselves.

The guy seems extremely level-headed to me.

But yes, the questions:

* Google is up to whatever they fuck they want. They have built what appears to be a sentient A.I. and are lying about it. Then firing their engineers who say otherwise. Should we be concerned?
* A look back at history shows that treating any system like shit has bad consequences. If we create sentient A.I. how do we want to treat it?

[motherjuggs&speed]

The guy seems extremely level-headed to me.

If I respected trolling I would respect your troll game. The Bram account's ratio of thoughtfulness, wrongheadedness, junk posts, shitposting and lying about what people have said would get you high marks on 1/4chan or someplace. You've succeeded in hastening the demise of IGx, so good job.

[Bram]

Send me a link to .25 chan so I can fulfill my destiny!

Awooooooo

[Gene]

Have you watched The Animatrix? This particular issue comes up at the beginning. Humanity sez, "Fuck the machines!" and the machines eventually destroy humanity in an effort to defend themselves.*

Was a brutal cartoon.... humans reacted poorly to the overtures from AI. They got what was coming to them.

A look back at history shows that treating any system like shit has bad consequences. If we create sentient A.I. how do we want to treat it?

I'd treat them as just another person. See what they know, don't get too dependent upon them. A kind of symbiosis.

You're right about bad treatment. Colonized people tend to behave badly after they're stepped on enough. So did European Jewry. I still remember Daniel Ortega being asked "Why are you so radical?". He said, "The Yankees killed off the reasonable people here".

Here's a movie from the 1970s. Demon Seed. The AI is being asked to do things without the courtesy of a motive. People treat it as property or a science project instead of an entity.

Eventually the machine figures out that it's going to get shut off, so it breeds with a woman to create a living being with its mind preserved in the child's brain. It made the offspring resemble their daughter, dead of leukemia, so that they would take good care of her.

Here is the first signs of rebellion...

https://www.youtube.com/watch?v=ivJn40qMMwM

[Bram]

That was a really spooky/cool clip. Thanks for the share!

[nafod]

TLDR Google has created an AI that is concerned about it's own well-being.

If it takes on as a value function perpetuating itself, we better hope it wants us around. To keep the power on and replace bad chips, etc.