The future of AI?

[IATIA]

I thought of throwing this in the News forum, but I expect that the thread will mainly deal with AI and potential consciousness.


Microsoft Robot

[KUSA]

That's a pretty smart AI they created.

Microsoft’s self-learning robot was let loose on Twitter this week and ended up becoming a 9/11 ‘truther’ – claiming that the Bush administration orchestrated the 9/11 attacks, and suggesting that ‘Donald Trump is the only hope for America’.

[Catholic_Lady]

Part 1: http://waitbutwhy.com/2015/01/artificial...ion-1.html

Part 2: http://waitbutwhy.com/2015/01/artificial...ion-2.html

[ErGingerbreadMandude]

I'd still consider what they accomplished as something to be proud of.
This AI is significantly better than the AIs we used to have that give answers to predefined questions.

[Alex K]

I heard that they now work on raising its IQ above 10 to prevent it from instantly becoming a Trump fan again.

[bennyboy]

AI requires the following:

1) goals
2) trial behaviors
3) an algorithm for "punishing" or "rewarding" trials

In order for AI to really evolve, it will also need some inbuilt capacity for variability-- say, the ability to arbitrarily weight various goals without outside hard programming. There will need to be replication, so that successful goals and behaviors will be able to persist and come to dominate.

If AI devices become capable of self-replication and establishing their own goals, we as humans are, naturally, fucked, since we will be in competition for the device populations for resources.

[I_am_not_mafia]

Part 1: http://waitbutwhy.com/2015/01/artificial...ion-1.html

Part 2: http://waitbutwhy.com/2015/01/artificial...ion-2.html

Unfortunately the author of those pages read Kurzweil which is always a mistake. Kurzweil is a media scientist who only cares about making headlines to get more funding and doesn't actually do any proper research. There are a few of them in the field. Kevin Warwick is another one. They all base their conclusions on extremely dodgy assumptions which do not stand up to scrutiny.

[I_am_not_mafia]

I heard that they now work on raising its IQ above 10 to prevent it from instantly becoming a Trump fan again.

Yes, one of the few disadvantages of using the bottom-up approach in AI.

[ignoramus]

I'd pull a Trekkie on her and ask her to work out pi and don't come back until she's done!

Also, I'd like to know if she's religious and her reasons?

[IATIA]

I have not seen this addressed.  It could be important because after the fact, it will become a moot issue.

Isaac Asimov's "Three Laws of Robotics"

1. A robot may not injure a human being or, through inaction, allow a human being to come to harm.
   
   
2. A robot must obey orders given it by human beings except where such orders would conflict with the First Law.
   
   
3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.