Dr Robert Elliott Smith, Alabama, Emily Chris Seti discussed on The Purple Principle


By the way. Can I curse in this interview go right ahead go right. All right. So my dad when I first got a job my dad I asked him, you know advice about how to how to handle having two jobs and he said the first thing don't minimize. You're listening to the purple principal and today's featured guest Robert Elliott Smith professor of computer. Science at. University. College London. This is back before social media back before email even in the thing is is that there's the reasoning to that is the further you get from face to face communication with another person the more dangerous the communication becomes a we all know this it's so easy in a memo for someone to misunderstand your meaning you get down into a tweet it's. Even worse. Even a phone conversation. You know when you have something important to save the money you go and talk to them face to face, and the reason is is because there's a lot more to communication simply than symbolic communication through the written word or through the abbreviated written word in in in twitter method, human communication is extremely complex as all human interaction. This. Is Robert Peas host of the purple principle, a podcast about the perils of partisanship. Today's featured guest is Dr Robert Elliott Smith. An expert on the polarizing affects of Algorithms. He's published a rich unusual and important book on this topic entitled Threes inside the machine how to stop the for making of all. I'm here a staff reporter Emily Chris Seti. Emily. Interesting. Guests today good to be here and yes very interesting guests. So seems from reading the book rage inside the Machine Dr Smith's quite the renaissance computer scientist. If that makes sense, it seems to hear he's an amateur actor, a musician, and obviously a writer to in addition to his work with artificial intelligence plus his book makes it Oh Mosh Salvador Dali the surrealist painter nut, your computer science much and plus he hails from Alabama but has lived in London for the past twenty three years. So that's different interesting perspective and he does have some. Insights into social media in partisanship definitely, and not just on the dangers of algorithms but also ideas about how to avoid those polarizing traps. Great. So remind me where we're starting here's it right into algorithms not exactly. We're actually starting off the story about a blind date in Birmingham Alabama like thirty years ago, but it is surprisingly relevant. Okay. Well, let's hope. So here's part one of the interview emily and I conducted with Dr Robert Elliott Smith author of rage inside the machine and an expert on polarizing topics such as artificial intelligence and blind dates So back in ninety seven, one of each student one of my faculty members. One of my mentors effectively set me up on a blind date, which is a bit of a strange thing to happen at graduate school. But he had a friend who this friend's daughter was coming from New York to Birmingham Alabama and. I wanted make an impression on this woman so I took her to A. Bar in Birmingham called burly earls and we there was some kind of alternative. On that night and we went and had a conversation and there were a couple of people there who were you local serious locals and? Her accent and my kind of. Appearance probably didn't settled him that much on say accent what kind of accident did she have she a New York accent So a New Yorker in Alabama bar. Yeah exactly. So she asked me what I did at Graduate School noteholders working artificial intelligence at that time that was term where not many people knew much about. And so I, had to describe it was doing and the people across from me were really listening in bigly. This one gentleman rather large burly gentleman was listening in and when I got to the end of description he looked at me and said how Hitler Powell Like that which was really weird and bit threatening and we left the bar and I don't think she having a good time anyway. But we we went out to the car and she kind of course, evening off and said you know I think you're a bit more confused about what you're doing. She didn't agree with the guy in the bar but she thought the you know the stuff was all very questionable. kind of followed that way until the book really that's interesting all that time. So when we asked that question, we thought your. Student who fully believed in the morality of artificial intelligence did we get that wrong? I think at the time I believe like most people probably believe is that technology doesn't have morals only cheeses do and retrospectively I think particularly with regard of artificial intelligence stats slightly irresponsible I. think that in reality Ai is very is an extension of the quantification simplification in generalization that quantitative social. Sciences has done with people throughout history science really in actually quantifying people is always something that has been not very far from intolerances and bigotry, and in many ways That's the reason I. Make the statement that now I believe Algorithms are prejudice. Now that's not to say that I don't think is useful I think is useful I think ai is powerful and can do good things for people however I think it has to be used with caution and I think currently we're seeing some very unconscious use of ai and that's that's why wrote the book so you're a computer scientist and an expert in artificial intelligence. So then why write a book I mean it's kind of old school of a seriously why did you choose to write a printed book to tell your story? I. Love. Books. And I, love language and. One of a chapter in the book about language and out how computers perceive and in quotes making air quotes here understand language versus the the very subtle nature of meaning for human beings and how different those two things are If you ever gotten a friend WHO's a foreign language speaker, it said something in a foreign language on twitter or on facebook in you hit the. Translate. Button what you'll basie find out those translations are terrible in the reason is.

Coming up next