Machine learning, artificial intelligence

[Kova V]

Anyone else have any interest in the things of this nature, predictive analytics, machine learning, natural language processing, nerdy stuff? I just thought it would just be interesting to see if anyone else has an interest in the same thing. Come on girls and boys, where my people at?

[ImagineKate]

I would love to learn that stuff. I am an EE with a minor in math and I am intrigued by AI and related subjects. Do you do this professionally?

[Dee Marshall]

Nearly went to grad school for it 30 years ago. A professor once told me "there will never be artificial intelligence. Every time we succeed with a technique they use it for something else and remove it from the definition." I think he was wrong though, some day we'll get it. But tell me this, when we finally make a truly intelligent machine would it be ethical to treat it like a machine or would that be slavery?

[Lyric]

Come on girls and boys, where my people at?

I'm going to take a wild guess and say they're easier to find at technology-oriented forums.  About all I know about this subject is a bit from reading Ray Kurzweil.

[Kova V]

Nearly went to grad school for it 30 years ago. A professor once told me "there will never be artificial intelligence. Every time we succeed with a technique they use it for something else and remove it from the definition." I think he was wrong though, some day we'll get it. But tell me this, when we finally make a truly intelligent machine would it be ethical to treat it like a machine or would that be slavery?

I think there was a Star Trek Voyager episode about this topic... Eventually it boiled down to "rights" for the smart ones and none for the drones. This is assuming its an embodied AI with single agency - no colonies of AIs with one agent or one AI with numerous agents. I'm writing a paper on this topic in my spare time.

I would love to learn that stuff. I am an EE with a minor in math and I am intrigued by AI and related subjects. Do you do this professionally?

I'm not getting paid for it, so no. I'm at the amateur level right now, but I'm planning on doing something with this professionally when I find the appropriate opportunity. There aren't very many companies around here that do this. The ones that do graph theory or can do calculus in there sleep. Id rather partner with someone brainy and start a business from scratch. I can dig up some online lecture series on different topics if you'd like. I have a few favorited somewhere on my computer (I'm Susan-ing on my phone right now)

I'm going to take a wild guess and say they're easier to find at technology-oriented forums.  About all I know about this subject is a bit from reading Ray Kurzweil.

That's cool, I was curious if anyone here shared this hobbie    I haven't read any of Kurzweil's stuff, I've only see a handfull of his lecture videos. Is it any good? What did you think about it?

[Dee Marshall]

An MMO I play has intelligent robots and the writers put this very good point in the mouth of one of them. Paraphrased, the robot said that full human rights aren't appropriate for machines. They don't need the "pursuit of happiness" because what makes them happy can be anything that's programmed in, and voting is a bad idea because any machine could duplicate itself enough times to alter the results of any election. Assuming that an AI can be copied this argument is valid. A true artificial intelligence, with the appropriate freedom would be just as entitled to be a bigot or an a$$hat as any human. It wouldn't even have to start with the machine. Imagine some marginalized political group creating and manufacturing robots that believed, wholeheartedly, that group's political philosophy.

[Kova V]

Today's AI's aren't that smart. They look at different corpuses of "knowledge" and do some pretty straightforward processing to rationalize an outcome. They don't quite have an opinion.

Chatbots _kind of_ have an opinion, or more appropriately they have the regurgitated opinion of whom the people it interacts. Theres nothing around context though, it's usually simple probability with a little part-of-speech and a library of pre-canned responses.

You have an interesting point though. Assuming they are "strong" AI's with one operating system per robot, they have "free will" to make desisions and are able to learn and adapt, what if we made the robots live (operate) in society for 18 years prior to allowing them to vote? If they were copies, they would diverge at some point and develop unique "personalities". I think we all know how it will play out though. All the robot cars would vote to increase taxes for replacing roads. Cars would also have to pay taxes and would be forced to charge their owners fees. Taxis would go out of business and there will be a resurgence in horses and horse drawn carriages.

Random question, has anyone seen the movie Her? This movie blew my mind with the personal context issues that may happen we build intimate relations based off "human" social rules. It's one of those easy to watch romance movies where a guy and an anvanced AI fall in love.

[Lyric]

I haven't read any of Kurzweil's stuff, I've only see a handfull of his lecture videos. Is it any good? What did you think about it?

Kurzweil's Age of Spiritual Machines made a strong impression on me, for sure. I suspect any of his books could inspire a dozen interesting sci-fi stories. Kurzweil is best known for popularizing the concept of AI "singularity", which is the point at which that matches human intelligence. He thinks we're pretty close. My theory is that if it happens it will use the Internet to amass large amounts of computer power and might even emerge from a computer virus. Let's just hope it turn out as creepy as the version on Person of Interest.