**Comrade Angles** @Angle · 04 jan. 2018, 07:41

**Comrade Angles** @Angle · 04 jan. 2018, 07:41

Comrade Angles @Angle

04 jan. 2018, 07:41

From a discussion on machine learning:

"For any program, there are two possibilities. Either it works exactly as intended, of some bug crops up in the code. In the first case, we probably have nothing to fear. The latter case isn't terribly concerning, either."

WRONG! When it comes to something as powerful as AI, both of these scenarios should be deeply terrifying. For the first, it takes all the usual problems of trust, accountability, and intention, and magnifies them a hundred thousand fold. For the second, it introduces a possibly apocalyptic uncertain consequences. What will it do? What will happen? Fuck if we know. :/

**🦋Polychrome🦋** @polychrome · 2018-01-04T08:32:29Z

🦋Polychrome🦋 @polychrome

Another AI Rant™ Afficher plus

@Angle neural networks are already sufficiently complicated that in some cases we're not entirely sure how they've reached the results they did. And this is only the early stages of the field - the complexity is only going to grow.

It's time to seriously rethink our debugging tools.

04 jan. 2018, 08:32 · SubwayTooter · 0 · 1

**Comrade Angles** @Angle · 04 jan. 2018, 08:33

**Comrade Angles** @Angle · 04 jan. 2018, 08:33

04 jan. 2018, 08:33

Comrade Angles @Angle

Another AI Rant™ Afficher plus

@polychrome Maybe we should use machine learning for our debugging tools. What could go wrong with that? :V

**🦋Polychrome🦋** @polychrome · 04 jan. 2018, 08:37

**🦋Polychrome🦋** @polychrome · 04 jan. 2018, 08:37

04 jan. 2018, 08:37

🦋Polychrome🦋 @polychrome

Another AI Rant™ Afficher plus

@Angle You kid, but this is exactly what we're already working on in some parts of the field!

While I am retry sure we won't create a self aware AI anytime soon, this is still going to raise some very serious questions in the near future.

Feels like we're to on busy chasing profits to notice what's taking shape beneath our hands.

**Comrade Angles** @Angle · 04 jan. 2018, 08:39

**Comrade Angles** @Angle · 04 jan. 2018, 08:39

04 jan. 2018, 08:39

Comrade Angles @Angle

Another AI Rant™ Afficher plus

@polychrome What kind of rubric would you set for an AI being self aware? Would something that knows it exists and can theorize about itself meet that rubric? And how far from that are we, really? I worry we might cross that line without even knowing, simply because the AI doesn't outright speak up and start talking to use. :/

**🦋Polychrome🦋** @polychrome · 04 jan. 2018, 09:19

**🦋Polychrome🦋** @polychrome · 04 jan. 2018, 09:19

04 jan. 2018, 09:19

🦋Polychrome🦋 @polychrome

Another AI Rant™ Afficher plus

@Angle Self awareness is such a complicated subject and it's exactly where philosophy comes in handy.

It's hard to find a point of comparison because unlike living organisms an AI has an utterly alien existence to our own: It has no needs to compete for food or look out for predators. It does not breed or age and is theoretically immortal.

So I don't have a good answer, only more questions. When does an AI stops being a tool and becomes an individual?