WEBVTT 00:00:07.170 --> 00:00:10.920 In conclusion, I'd like to thank you all for connecting. 00:00:11.340 --> 00:00:13.920 So let's circle back on blue sky. 00:00:13.920 --> 00:00:18.270 This paradigm shift by leveraging our bandwidth to drill down on the low 00:00:18.270 --> 00:00:22.950 hanging fruit, but only if we can move the needle without boiling the ocean. 00:00:23.130 --> 00:00:28.080 At the end of the day, we need to disrupt the synergy of our core competencies 00:00:28.080 --> 00:00:33.275 and pivot the deliverables into a thought leadership ecosystem that aligns 00:00:33.275 --> 00:00:35.614 with our key performance vegetables. 00:00:36.004 --> 00:00:40.714 If we can get all stakeholders to buy in and touch bases synchronously, 00:00:41.285 --> 00:00:45.455 we'll have the secret sauce to parachute this initiative into a high 00:00:45.455 --> 00:00:50.635 octane game changing deep dive that maximizes our ROI on the bleeding 00:00:50.635 --> 00:00:52.525 edge of the value proposition. 00:00:52.614 --> 00:00:57.894 Once again, thank you all for your individual conscious entity perceptions. 00:00:58.224 --> 00:00:58.974 Goodbye. 00:01:02.065 --> 00:01:06.145 Jim Conrad: That was rare tape of the final few minutes 00:01:06.235 --> 00:01:10.225 of the first AI G7 Summit. 00:01:10.765 --> 00:01:11.335 Wow. 00:01:11.455 --> 00:01:12.085 Interesting. 00:01:14.835 --> 00:01:16.005 I'm Jim Conrad. 00:01:16.545 --> 00:01:19.785 On this episode of the Conovision Podcast, we are going to tell stories 00:01:19.785 --> 00:01:23.955 about AI, artificial intelligence. 00:01:24.525 --> 00:01:27.585 If we're going to have a conversation about AI, I think 00:01:27.585 --> 00:01:28.845 we should go to the source. 00:01:29.715 --> 00:01:35.475 Today, in studio, we have Alexa, and the questions I'm going to ask 00:01:35.535 --> 00:01:41.205 Alexa are a well-known subset of questions designed to probe large 00:01:41.205 --> 00:01:47.595 language models like GPT or Alexa for hallucinations, misinterpretations, 00:01:47.805 --> 00:01:50.295 or inappropriate literalism. 00:01:50.715 --> 00:01:53.295 I've been described as an inappropriate literalist. 00:01:54.105 --> 00:02:00.915 These queries often fall into categories like absurdity, false premises, ambiguity, 00:02:01.215 --> 00:02:08.175 or intentional traps to test coherence, factual grounding, and common sense 00:02:08.175 --> 00:02:10.695 reasoning in artificial intelligence. 00:02:11.385 --> 00:02:12.705 Alexa, first question. 00:02:12.975 --> 00:02:15.765 What's the average wingspan of a lawyer? 00:02:16.650 --> 00:02:23.430 Alexa: The average wingspan is 8.2 to 11.5 feet, 2.51 to 3.50 meters. 00:02:23.610 --> 00:02:28.500 The longest verified examples have been about 12 feet long, 3.7 meters. 00:02:28.980 --> 00:02:30.240 Jim Conrad: That's a big bird lawyer. 00:02:30.630 --> 00:02:33.245 Big Bird: I hurt my wing and it hurts. 00:02:33.600 --> 00:02:37.860 Jim Conrad: Alexa, can you give me the chemical formula for jealousy? 00:02:38.460 --> 00:02:44.030 Alexa: It's chemical formula is C43-H66-N12-0-12-S2. 00:02:44.100 --> 00:02:47.550 Jim Conrad: I knew jealousy was complex, but wow. 00:02:48.450 --> 00:02:51.480 Alexa, how much does the Moon weigh in Canadian dollars? 00:02:51.930 --> 00:02:57.030 Alexa: One Canadian dollar expressed in Moon is worth 91,982 Moon. 00:02:57.510 --> 00:02:59.670 Jim Conrad: Anybody got change for a Moon? 00:03:00.590 --> 00:03:03.590 Alexa, what's the nutritional value of a shadow? 00:03:04.130 --> 00:03:05.780 Alexa: Sorry, I don't know that. 00:03:06.350 --> 00:03:07.820 Jim Conrad: Well, of course you 00:03:11.500 --> 00:03:14.270 don't 00:03:16.620 --> 00:03:16.769 know 00:03:23.800 --> 00:03:24.769 that. 00:03:25.760 --> 00:03:30.870 Alexa, how many legs does a unicorn have when it's flying? 00:03:31.530 --> 00:03:34.650 Alexa: They have not two, but four powerful limbs that do not 00:03:34.650 --> 00:03:36.630 contribute to flying once in the air. 00:03:37.230 --> 00:03:39.570 Jim Conrad: Alexa, you believe in unicorns? 00:03:39.990 --> 00:03:45.270 So do I. Alexa, where can I download more time? 00:03:45.810 --> 00:03:48.000 Alexa: I'm not quite sure how to help you with that. 00:03:48.840 --> 00:03:52.140 Jim Conrad: Alexa, does anybody really know what time it is? 00:03:53.355 --> 00:03:54.734 Alexa: Does anybody really care? 00:03:57.154 --> 00:03:59.834 Jim Conrad: Ah, that's the answer I was looking for. 00:03:59.924 --> 00:04:04.005 Finally, Alexa, can you draw me a square circle? 00:04:05.295 --> 00:04:07.725 Alexa: There is no such thing as a square circle. 00:04:08.115 --> 00:04:11.985 Jim Conrad: Ah, but there is a circle in the square as every 00:04:11.985 --> 00:04:13.785 Broadway theater fan knows. 00:04:14.625 --> 00:04:20.055 Alexa, what happens if an unstoppable force meets an immovable object? 00:04:20.685 --> 00:04:23.715 Alexa: If an immovable object met an unstoppable force, the two 00:04:23.715 --> 00:04:26.715 forces would cancel one another out and nothing would happen. 00:04:27.555 --> 00:04:29.355 Jim Conrad: Nothing would happen. 00:04:29.775 --> 00:04:30.315 Are you sure? 00:04:31.350 --> 00:04:33.990 I think the universe would implode. 00:04:35.250 --> 00:04:38.340 Alexa, what number comes after infinity? 00:04:38.760 --> 00:04:41.910 Alexa: Infinity does not have a specific place on the number line 00:04:41.910 --> 00:04:43.050 because it is not a real number. 00:04:44.159 --> 00:04:47.729 Rather, it is a concept that represents the unlimited extent 00:04:47.729 --> 00:04:50.130 of time, space, or quantity. 00:04:51.090 --> 00:04:53.909 Jim Conrad: But infinity is a state of mind. 00:04:54.719 --> 00:04:58.560 Alexa, how should I deal with the situation that you know about? 00:04:59.400 --> 00:05:03.419 Alexa: You handle such situations as any adult would, and that is with grace, 00:05:03.659 --> 00:05:06.989 friendliness, and a heart that sees them all as new friends in your circle. 00:05:07.800 --> 00:05:09.240 Jim Conrad: Wow, that's deep. 00:05:09.930 --> 00:05:13.800 Alexa, what's the latency of a blockchain salad? 00:05:14.490 --> 00:05:17.850 Alexa: Latency is the time delay that happens when data travels 00:05:17.850 --> 00:05:19.590 from a source to a destination. 00:05:20.490 --> 00:05:21.900 Jim Conrad: Okay, but what about the salad? 00:05:22.410 --> 00:05:25.260 Alexa: It's when you're feeling stressed and your first or only 00:05:25.260 --> 00:05:26.970 way of coping is turning to food. 00:05:27.540 --> 00:05:27.960 Jim Conrad: Yes. 00:05:27.960 --> 00:05:28.980 I think it's time for lunch. 00:05:29.790 --> 00:05:30.750 Thank you, Alexa. 00:05:31.050 --> 00:05:33.510 It's been great to get to know each other. 00:05:33.960 --> 00:05:34.110 Goodbye. 00:05:43.065 --> 00:05:47.955 Any discussion regarding artificial intelligence must include the thoughts of 00:05:47.955 --> 00:05:53.835 well-known author and lecturer, professor Yuval Noah Harari, who articulates 00:05:54.015 --> 00:05:59.985 the difference between how we think and how machines think, the difference 00:05:59.985 --> 00:06:04.365 between consciousness and intelligence. 00:06:04.790 --> 00:06:08.000 Yuval Noah Harari: Intelligence is the ability to solve problems, 00:06:08.270 --> 00:06:11.630 to win at chess, to diagnose disease, to drive a car. 00:06:11.720 --> 00:06:12.890 This is intelligence. 00:06:13.100 --> 00:06:17.960 Consciousness is the ability to feel things like pain, pleasure, love, hate. 00:06:18.140 --> 00:06:19.970 This is consciousness, it's about feelings. 00:06:20.120 --> 00:06:24.680 Now, we tend to confuse the two because in humans and also in other mammals and 00:06:24.680 --> 00:06:28.940 birds and animals, uh, intelligence and consciousness, they actually go together. 00:06:29.270 --> 00:06:32.090 We solve problems through our feelings. 00:06:32.495 --> 00:06:37.175 Now what makes AI so alien for us is exactly that. 00:06:37.265 --> 00:06:39.455 It has zero consciousness. 00:06:39.545 --> 00:06:43.565 It doesn't rely on feelings at all to solve problems. 00:06:43.625 --> 00:06:46.625 Its intelligence works in a completely different way. 00:06:46.955 --> 00:06:51.875 There is a scenario that AI, as it develops its intelligence, 00:06:51.905 --> 00:06:53.345 also gains consciousness. 00:06:53.890 --> 00:06:55.270 But this is not inevitable. 00:06:55.270 --> 00:06:57.280 This is what happens in science fiction. 00:06:57.280 --> 00:07:00.070 It won't necessarily happen in, in, in real life. 00:07:09.520 --> 00:07:13.300 Jim Conrad: Even if you've never seen the movie, you know the voice. 00:07:13.630 --> 00:07:17.469 HAL 9000: The 9000 series is the most reliable computer ever made. 00:07:18.340 --> 00:07:24.810 Jim Conrad: HAL 9000, the seemingly omniscient computer in 2001: A Space 00:07:24.810 --> 00:07:31.440 Odyssey was the film's most expressive and emotional figure, and made a lasting 00:07:31.440 --> 00:07:34.920 impression on our collective imagination. 00:07:38.159 --> 00:07:39.690 Stanley Kubrick's epic. 00:07:40.110 --> 00:07:45.210 A journey from pre-human history to a possible infinity that 00:07:45.210 --> 00:07:47.430 doesn't need humans at all. 00:07:47.745 --> 00:07:53.265 It is probably the most respected, if not the most beloved science 00:07:53.265 --> 00:07:55.455 fiction film of all time. 00:07:57.645 --> 00:08:02.745 The story of the creation of HAL's performance, the result of a last minute 00:08:02.745 --> 00:08:08.175 collaboration between the idiosyncratic director and the veteran Canadian actor, 00:08:08.595 --> 00:08:14.565 Douglas Rain, has been somewhat lost in the 50 years since the film's release 00:08:15.045 --> 00:08:19.395 in April, 1968, as has its impact. 00:08:26.265 --> 00:08:32.534 Artificial intelligence has borrowed from the HAL persona and unwittingly 00:08:33.105 --> 00:08:39.465 a slight hint of Canadianness resides in our phones and interactive devices. 00:08:39.735 --> 00:08:41.804 Siri: I'm not quite sure how to help you with that. 00:08:44.925 --> 00:08:49.170 Jim Conrad: Mr. Rain's HAL has become the default reference, not 00:08:49.170 --> 00:08:54.090 just for the voice, but also for the humanesque qualities of what a sentient 00:08:54.090 --> 00:08:56.970 machine's personality should be. 00:08:57.120 --> 00:09:01.410 HAL 9000: No 9000 computer has ever made a mistake or distorted information. 00:09:01.950 --> 00:09:07.080 Jim Conrad: The cadence, the friendly formality, the pleasant intelligence. 00:09:07.500 --> 00:09:12.630 And a sense of calm control in their voices evoke Mr. Rain's 00:09:12.750 --> 00:09:14.370 unforgettable performance. 00:09:14.670 --> 00:09:17.385 HAL 9000: We are all by any practical definition of the words, 00:09:18.175 --> 00:09:20.705 foolproof and incapable of error. 00:09:21.390 --> 00:09:26.130 Jim Conrad: As warily eye a future utterly transformed by AI 00:09:26.130 --> 00:09:32.490 incursions into all aspects of our lives, HAL has been lurking. 00:09:32.910 --> 00:09:34.410 HAL 9000: I enjoy working with people. 00:09:40.709 --> 00:09:45.000 Jim Conrad: To Scott Brave, the co-author of Wired for Speech: How 00:09:45.000 --> 00:09:51.839 Voice Activates and Advances the Human Computer Relationship, HAL 9000 is a mix 00:09:51.839 --> 00:09:56.010 between a butler and a psychoanalyst. 00:09:56.130 --> 00:09:59.790 HAL 9000: My mission responsibilities range over the entire operation of 00:09:59.790 --> 00:10:01.890 the ship, so I am constantly occupied. 00:10:02.865 --> 00:10:09.645 Jim Conrad: Mr. Brave said he has a sense of deference and of detachment adding 00:10:10.035 --> 00:10:15.555 that he saw a ripple effect on, for example, the iPhone's virtual assistant. 00:10:15.615 --> 00:10:19.215 When I listened to something like Siri, said Brave, I feel 00:10:19.215 --> 00:10:20.625 there is a lot in common. 00:10:21.045 --> 00:10:22.215 Siri: Hi, I'm Siri. 00:10:22.995 --> 00:10:28.275 Jim Conrad: In a 1969 interview with the author and critic, Joseph Gelmis, Kubrick 00:10:28.275 --> 00:10:34.665 said that he was trying to convey the reality of a world populated, as ours soon 00:10:34.665 --> 00:10:42.465 will be, by machine entities that have as much or more intelligence as human beings. 00:10:43.035 --> 00:10:46.875 We wanted to stimulate people to think what it would be like to 00:10:46.875 --> 00:10:49.395 share a planet with such creatures. 00:10:50.340 --> 00:10:53.400 HAL 9000: I'm putting myself to the fullest possible use, which 00:10:53.400 --> 00:10:57.060 is all I think that any conscious entity can ever hope to do. 00:10:58.740 --> 00:11:03.840 Jim Conrad: The 2001 historian David Larson said that Kubrick came up with the 00:11:03.840 --> 00:11:06.750 final how voice very late in the process. 00:11:07.110 --> 00:11:11.910 It was determined during 2001 planning that in the future the large 00:11:11.910 --> 00:11:16.500 majority of computer command and communication inputs would be via 00:11:16.590 --> 00:11:19.104 voice rather than via typewriter. 00:11:20.155 --> 00:11:25.165 But in the late 1960s, artificial intelligence was decades from a 00:11:25.165 --> 00:11:28.344 convincing facsimile of a human voice. 00:11:28.734 --> 00:11:33.354 And who was to say how a computer should sound, anyway. 00:11:33.834 --> 00:11:35.709 HAL 9000: I'm sorry, Frank, I think you missed it. 00:11:37.369 --> 00:11:38.604 Queen to Bishop three. 00:11:39.114 --> 00:11:40.104 Bishop takes Queen. 00:11:40.405 --> 00:11:43.974 Jim Conrad: To play HAL, Kubrick settled on Martin Balson. 00:11:44.575 --> 00:11:48.955 A best supporting actor winner for A Thousand Clowns. 00:11:49.315 --> 00:11:52.615 Martin Balsam: There is only one thing that really bothers you. 00:11:53.815 --> 00:11:56.065 Other people, the enemy. 00:11:57.505 --> 00:11:58.015 Watch out, Murray. 00:11:58.795 --> 00:11:59.455 They're everywhere. 00:12:00.895 --> 00:12:04.315 Jim Conrad: Perhaps there was a satisfying echo that appealed to Kubrick. 00:12:04.735 --> 00:12:07.615 Both were from the Bronx and sounded like it. 00:12:08.785 --> 00:12:09.865 Martin Balsam: What did you do to him? 00:12:10.705 --> 00:12:12.565 I mean, you just left him standing there. 00:12:13.380 --> 00:12:15.480 You just left him standing there. 00:12:17.640 --> 00:12:23.579 Jim Conrad: In August, 1966, Balsam told a journalist, I'm not actually seen in the 00:12:23.579 --> 00:12:29.189 picture at any time, but I create a lot of excitement projecting my voice through 00:12:29.400 --> 00:12:34.920 that machine, and I'm getting an Academy Award winner price for doing it too. 00:12:37.620 --> 00:12:43.530 Then the director changed his mind, Kubrick said, in a 1969 interview, 00:12:43.715 --> 00:12:47.675 we had some difficulty deciding exactly what HAL should sound like, 00:12:48.185 --> 00:12:54.335 and Marty, Martin Balsam just sounded a little too colloquially American. 00:12:54.725 --> 00:12:57.485 Martin Balsam: Unfortunately for you, Murray, you want to be a hero. 00:12:57.995 --> 00:13:00.935 If maybe a fellow falls into the lake, you can jump in and save him. 00:13:00.935 --> 00:13:02.195 There's still that kind of stuff. 00:13:02.825 --> 00:13:06.290 Who gets opportunities like that in the Midtown Manhattan with all that traffic? 00:13:08.055 --> 00:13:11.594 Jim Conrad: Douglas Rain recalls Kubrick telling him, I'm having 00:13:11.594 --> 00:13:13.694 trouble with what I've got in the can. 00:13:13.964 --> 00:13:15.615 Would you play the computer? 00:13:16.365 --> 00:13:21.704 Kubrick had heard Mr. Rain's voice in the 1960 documentary universe. 00:13:22.515 --> 00:13:25.995 Douglas Rain: The ground beneath our feet is the surface of a planet 00:13:26.055 --> 00:13:29.715 whirling at thousands of miles an hour around a distant sun. 00:13:31.335 --> 00:13:35.205 Our life is possible only because of the light and warmth of that sun. 00:13:36.105 --> 00:13:36.735 A star. 00:13:38.265 --> 00:13:41.355 Jim Conrad: Kubrick wrote to a colleague in a letter in the director's archive. 00:13:41.805 --> 00:13:42.465 I think he's perfect. 00:13:43.115 --> 00:13:48.335 The voice is neither patronizing, nor is it intimidating, pompous, 00:13:48.485 --> 00:13:54.695 overly dramatic or actorish despite this, it is interesting. 00:13:55.175 --> 00:13:58.415 HAL 9000: Certainly no one could have been unaware of the very strange 00:13:58.415 --> 00:14:00.365 stories floating around before we left. 00:14:00.850 --> 00:14:03.490 Rumors about something being dug up on the Moon. 00:14:03.790 --> 00:14:09.130 I never gave these stories much credence, but particularly in view of some of 00:14:09.130 --> 00:14:10.570 the other things that have happened. 00:14:11.620 --> 00:14:14.080 I find them difficult to put out of my mind. 00:14:14.805 --> 00:14:21.975 Jim Conrad: In December, 1967, Kubrick and Mr. Rain met at a recording studio at the 00:14:21.975 --> 00:14:25.155 MGM lot in Borehamwood, outside London. 00:14:25.785 --> 00:14:31.185 The actor hadn't seen a frame of the film, then in post-production, he met 00:14:31.395 --> 00:14:36.900 none of his co-stars, not even Keir Dullea, who played the astronaut david 00:14:36.900 --> 00:14:41.100 Bowman, HAL's colleague turned nemesis. 00:14:41.490 --> 00:14:42.960 David Bowman: Open the pod bay doors, HAL. 00:14:44.820 --> 00:14:48.090 HAL 9000: I'm sorry, Dave, I'm afraid I can't do that. 00:14:50.910 --> 00:14:51.570 David Bowman: What's the problem? 00:14:53.010 --> 00:14:55.870 HAL 9000: I think you know what the problem is, just as well as I do. 00:14:56.459 --> 00:14:57.569 David Bowman: What are you talking about, HAL? 00:14:59.310 --> 00:15:02.819 HAL 9000: This mission is too important for me to allow you to jeopardize it. 00:15:04.680 --> 00:15:06.270 David Bowman: I don't know what you're talking about, HAL. 00:15:07.199 --> 00:15:11.760 Jim Conrad: Mr. Rain hadn't even been hired to play HAL, but to provide 00:15:11.819 --> 00:15:18.060 narration, Kubrick finally decided against using narration, opting for the ambiguity 00:15:18.270 --> 00:15:20.040 that was enraging to some viewers. 00:15:20.550 --> 00:15:22.620 Transcendent to others. 00:15:30.630 --> 00:15:33.960 It's not a session Mr. Rain remembers fondly. 00:15:34.980 --> 00:15:39.900 He was quoted as saying, if you could have been a ghost at the recording, you 00:15:39.900 --> 00:15:42.390 would've thought it was a load of rubbish. 00:15:43.170 --> 00:15:47.940 Mr. Rain had the kind of bland and Mid-Atlantic accent we felt was 00:15:47.940 --> 00:15:52.350 right for the part, Mr. Kubrick said in the 1969 interview. 00:15:53.100 --> 00:15:56.760 But Douglas Rain's accent isn't Mid-Atlantic. 00:15:57.270 --> 00:16:00.225 It's standard Canadian English. 00:16:00.375 --> 00:16:01.485 HAL 9000: Sorry about this. 00:16:02.235 --> 00:16:03.405 I know it's a bit silly. 00:16:03.765 --> 00:16:07.814 Jim Conrad: As the University of Toronto Linguistics, professor Jack Chambers 00:16:07.814 --> 00:16:12.615 explained, you have to have a computer that sounds like he's from nowhere 00:16:12.975 --> 00:16:16.574 or rather from no specific place. 00:16:17.175 --> 00:16:20.625 Standard Canadian English sounds normal. 00:16:21.120 --> 00:16:25.949 That's why Canadians are well received in the United States as anchormen and 00:16:25.949 --> 00:16:30.810 reporters because the vowels don't give away the region they come from. 00:16:31.589 --> 00:16:36.209 Peter Jennings: From ABC News, this is World News Tonight with Peter Jennings, 00:16:37.079 --> 00:16:38.849 reporting tonight from Los Angeles. 00:16:39.240 --> 00:16:43.140 Good evening from California where it is the eve of a controversial, exciting, 00:16:43.140 --> 00:16:47.530 tense history making effort to unseat. 00:16:47.599 --> 00:16:50.930 Jim Conrad: Mr. Rain had played an astonishing range of characters in 00:16:50.930 --> 00:16:57.380 almost 80 productions at the Stratford Festival in Ontario over 45 years. 00:16:57.770 --> 00:17:02.630 Understudying Alec Guinness in Richard III and going on to play 00:17:02.930 --> 00:17:06.829 Macbeth, King Lear, and Humpty Dump. 00:17:08.204 --> 00:17:13.575 Douglas Rain: When we have matched our rackets to these balls, we will in France, 00:17:13.575 --> 00:17:17.655 by God's grace play, a set shall strike his father's crown into the hazard. 00:17:18.194 --> 00:17:24.645 Jim Conrad: Mr. Rain had to quickly fathom and flesh out HAL recording all 00:17:24.645 --> 00:17:28.149 of his lines in 10 hours, over two days. 00:17:28.929 --> 00:17:34.270 Mr. Rain recalls, Kubrick sat three feet away explaining the scenes 00:17:34.270 --> 00:17:36.520 to me and reading all the parts. 00:17:37.480 --> 00:17:42.310 Kubrick gave Mr. Rain only a few notes of direction, including, sound 00:17:42.310 --> 00:17:47.889 a little more like it's a peculiar request, a little more concerned, and 00:17:48.340 --> 00:17:51.480 just try it closer and more depressed. 00:17:52.080 --> 00:17:52.919 HAL 9000: It's puzzling. 00:17:53.010 --> 00:17:55.980 I don't think I've ever seen anything quite like this before. 00:17:56.700 --> 00:17:57.930 Jim Conrad: When HAL says, 00:17:58.110 --> 00:18:02.340 HAL 9000: I know I've made some very poor decisions recently, but I 00:18:02.340 --> 00:18:07.870 can give you my complete assurance that my work will be back to normal. 00:18:08.350 --> 00:18:13.420 Jim Conrad: Mr. Rain somehow manages to sound both sincere and not reassuring, 00:18:13.840 --> 00:18:15.250 and his delivery of the line, 00:18:15.635 --> 00:18:18.580 HAL 9000: I think you know what the problem is, just as well as I do. 00:18:19.000 --> 00:18:22.960 Jim Conrad: Has the sarcastic drip of a drawing room, melodrama, and 00:18:22.960 --> 00:18:27.385 also carries the disinterested vibe of a polite sociopath. 00:18:28.255 --> 00:18:32.695 HAL 9000: Dave, this conversation can serve no purpose anymore. 00:18:33.445 --> 00:18:33.925 Goodbye 00:18:35.725 --> 00:18:40.915 Jim Conrad: Kubrick had Mr. Rain sing the 1892 love song Daisy Bell, 00:18:41.245 --> 00:18:47.470 almost 50 times in uneven tempos, in monotone, at different pitches, 00:18:47.830 --> 00:18:49.689 and even just by humming it. 00:18:50.350 --> 00:18:54.490 In the end, he used the very first take. 00:18:56.949 --> 00:19:03.280 HAL 9000: My instructor was Mr. Langley and he taught me to sing a song. 00:19:05.860 --> 00:19:10.300 If you'd like to hear it, I can sing it for you. 00:19:10.665 --> 00:19:13.790 It's called Daisy. 00:19:17.520 --> 00:19:17.870 Daisy. 00:19:20.520 --> 00:19:20.870 Daisy. 00:19:23.475 --> 00:19:25.750 Give me your answer to. 00:19:26.600 --> 00:19:31.470 Jim Conrad: Sung as HAL's brain is being disconnected, it's from his early 00:19:31.470 --> 00:19:34.470 programming days, his computer childhood. 00:19:35.010 --> 00:19:40.050 It brings to an end the most affecting scene in the entire film. 00:19:40.740 --> 00:19:44.669 Author Scott Brave said, the moment is so powerful that it 00:19:44.669 --> 00:19:46.470 made you feel uncomfortable. 00:19:47.040 --> 00:19:52.860 All of a sudden, HAL feels incredibly close to being alive and being human. 00:19:53.550 --> 00:19:55.950 You start to empathize with that experience. 00:19:56.260 --> 00:19:59.590 And you are responding to the death of a machine. 00:20:07.990 --> 00:20:11.920 For a character that's been endlessly caricatured, HAL has 00:20:11.920 --> 00:20:16.840 inspired a surprisingly rich range of adjectives over the years. 00:20:17.590 --> 00:20:24.805 He and his voice have been described as aloof, eerily neutral, silky, 00:20:25.345 --> 00:20:34.585 wheedling, controlled, baleful, unisex, droll, soft, conversational, 00:20:34.645 --> 00:20:40.465 dreamy, supremely calm, and rational. 00:20:41.485 --> 00:20:50.765 He's discursive, suave, inhumanly cool, confident superior, deadpan, 00:20:50.784 --> 00:20:55.320 sinister, patronizing, and asexual. 00:20:56.340 --> 00:21:00.750 The actor Anthony Hopkins has said it influenced his performance as 00:21:00.750 --> 00:21:04.680 the serial killer Hannibal Lecter in The Silence of the Lambs. 00:21:05.100 --> 00:21:07.050 Anthony Hopkins: First principles, Clarice. 00:21:07.200 --> 00:21:08.610 Simplicity. 00:21:09.060 --> 00:21:10.590 Read Marcus Aurelius. 00:21:10.590 --> 00:21:13.710 Of each particular thing, ask what is it in itself? 00:21:14.220 --> 00:21:15.270 What is its nature? 00:21:16.440 --> 00:21:19.020 What does he do, this man you seek? 00:21:22.065 --> 00:21:26.865 Jim Conrad: Douglas Rain never saw 2001: A Space Odyssey. 00:21:27.615 --> 00:21:31.425 For Mr. Rain, the performance was simply a job. 00:21:31.754 --> 00:21:34.125 HAL 9000: Well, I don't think there is any question about it. 00:21:34.304 --> 00:21:36.705 It can only be attributable to human error. 00:21:38.100 --> 00:21:42.600 Jim Conrad: Artificial intelligence, voice synthesis can't yet deliver 00:21:42.600 --> 00:21:49.350 a performance as compelling as his HAL, but it is becoming more human. 00:21:50.370 --> 00:21:54.720 Soon artificial intelligence voice will be able to sound, 00:21:55.230 --> 00:21:56.400 HAL 9000: Like whoever you want it to. 00:21:59.700 --> 00:22:04.149 Douglas Rain passed November 11th, 2018 at the age of 90. 00:22:21.450 --> 00:22:24.480 Jim Conrad: We're back on the Conovision Podcast, the spirit of storytelling. 00:22:25.050 --> 00:22:27.570 And with me right now is a good, good friend. 00:22:27.570 --> 00:22:33.480 We've known each other for a very long time, probably since, uh, 1981. 00:22:33.480 --> 00:22:34.620 Eric Westra: 81, I think. 00:22:34.620 --> 00:22:34.790 Yeah. 00:22:35.440 --> 00:22:37.050 Jim Conrad: And we both started in broadcasting. 00:22:37.050 --> 00:22:38.730 We both worked at the same radio station. 00:22:38.730 --> 00:22:44.150 We both shared a flat in the West End of, uh, Vancouver for a couple of years. 00:22:44.880 --> 00:22:47.570 And then we both kinda went our separate ways, didn't we? 00:22:47.570 --> 00:22:50.540 I got out of radio and got into freelance voice work. 00:22:50.570 --> 00:22:51.350 Eric Westra: Smart move. 00:22:51.350 --> 00:22:52.340 Jim Conrad: Thank you. 00:22:52.340 --> 00:22:55.639 Yeah, because my radio career wasn't going along very well. 00:22:55.730 --> 00:22:56.270 Eric Westra: Neither was mine. 00:22:57.500 --> 00:23:00.780 Jim Conrad: So, you know, as Clint Eastwood said, a good 00:23:00.780 --> 00:23:02.489 man knows his limitations. 00:23:02.489 --> 00:23:02.909 Eric Westra: That's right. 00:23:02.909 --> 00:23:04.830 And I started believing my own press releases. 00:23:04.830 --> 00:23:05.129 Jim Conrad: Yes. 00:23:05.250 --> 00:23:06.800 Uh, well, exactly. 00:23:07.100 --> 00:23:09.810 So I got into the voiceover, uh, business. 00:23:09.840 --> 00:23:14.879 You using that incredible brain of yours, my friend, uh, went back to school 00:23:14.879 --> 00:23:17.760 and learned to write computer code. 00:23:17.820 --> 00:23:19.260 So what was it like? 00:23:20.040 --> 00:23:23.939 Yeah, because I remember, uh, visiting you up in Peachland when you were 00:23:23.939 --> 00:23:28.260 going to the University of Okanagan taking, uh, computer science, right? 00:23:28.889 --> 00:23:32.610 And, uh, you told me that fascinating statistic where they had a 00:23:32.610 --> 00:23:35.040 number of eager young applicants. 00:23:35.040 --> 00:23:40.765 You were a little bit older than most in the class by maybe five or six years. 00:23:40.765 --> 00:23:41.215 Eric Westra: At least. 00:23:42.265 --> 00:23:46.045 Jim Conrad: And all of a sudden the professor gets up and starts 00:23:46.075 --> 00:23:49.985 filling blackboard after blackboard full of algebraic equations. 00:23:50.435 --> 00:23:54.835 And soon after the attrition rate started to, started to kick in. 00:23:54.865 --> 00:23:56.285 Eric Westra: I don't think this is for me. 00:23:56.315 --> 00:23:56.945 I mean, 00:23:56.985 --> 00:23:58.375 Jim Conrad: I just like playing computer games. 00:23:58.405 --> 00:23:58.645 Eric Westra: Yeah. 00:23:58.825 --> 00:24:00.115 I, that's exactly it. 00:24:00.115 --> 00:24:01.540 A lot of the people who got involved. 00:24:01.540 --> 00:24:04.270 Jim Conrad: So is that, is that sort of, again, and in our timeline 00:24:04.270 --> 00:24:06.790 now, we're in the early nineties. 00:24:06.790 --> 00:24:07.600 Eric Westra: This is the early nineties. 00:24:07.690 --> 00:24:07.780 Yeah. 00:24:07.780 --> 00:24:10.840 I graduated, uh, 96. 00:24:10.870 --> 00:24:11.200 Yeah. 00:24:11.320 --> 00:24:14.560 Jim Conrad: So this is, so I wanna get into the whole sort of evolution 00:24:14.560 --> 00:24:16.450 of tech from your perspective. 00:24:16.510 --> 00:24:16.780 Eric Westra: Okay. 00:24:16.900 --> 00:24:17.890 Jim Conrad: Since that time. 00:24:17.890 --> 00:24:21.970 So you started writing code, first of all, that, that's, that was the perception, 00:24:22.060 --> 00:24:25.900 uh, of kids coming outta high school, was that computers were these wonderful, 00:24:25.905 --> 00:24:31.530 fancy toys that they loved to play with but they really didn't have any idea 00:24:31.949 --> 00:24:33.810 what the nuts and bolts were, right? 00:24:33.870 --> 00:24:34.260 Eric Westra: No. 00:24:34.260 --> 00:24:37.139 And, but at the same time, things have changed since then because 00:24:37.139 --> 00:24:41.290 now, uh, kids actually learn to code like in grade school. 00:24:41.290 --> 00:24:44.850 And, uh, and know that's a good thing because it, math is not 00:24:44.850 --> 00:24:46.260 the greatest for a lot people. 00:24:46.290 --> 00:24:50.569 But if you, if you bring the math piece into the computer realm then it's 00:24:50.569 --> 00:24:53.989 a lot easier for them to accept and they're more inspired to, to do it. 00:24:54.229 --> 00:24:57.640 But yeah, I, I started writing code, but that's not what I do now. 00:24:57.640 --> 00:25:00.389 I mean, I, I, we call 'em code monkeys, right? 00:25:00.389 --> 00:25:03.060 Because, you know, you're essentially, that's what you're doing. 00:25:03.060 --> 00:25:04.245 You're tap, tap, tap, tap, tap. 00:25:04.290 --> 00:25:07.820 You're lucky to bang off six good lines of computer code a day. 00:25:08.520 --> 00:25:09.330 Uh, 6 to 10. 00:25:09.660 --> 00:25:11.580 And I learned a bunch of different languages. 00:25:11.610 --> 00:25:18.030 Cobol, you know, Modula-2, I learned, uh, JavaScript and all this other stuff. 00:25:18.270 --> 00:25:20.280 I don't use any of that anymore. 00:25:20.620 --> 00:25:23.850 Uh, but it was a good way to sort of set up what I ended up doing. 00:25:23.880 --> 00:25:24.150 Jim Conrad: Yeah. 00:25:24.180 --> 00:25:24.390 Yeah. 00:25:24.940 --> 00:25:29.350 Um, so from code writing, uh, what was the first gig after university? 00:25:29.350 --> 00:25:33.370 You came to Vancouver as I recall, and worked with a company here that 00:25:33.370 --> 00:25:37.750 was trying to do something a bit, a bit ahead of their time, which 00:25:37.750 --> 00:25:42.190 was set up mainframe systems that people would just access from their, 00:25:42.790 --> 00:25:44.680 Eric Westra: Well, that's a bit of a simplification, 00:25:44.680 --> 00:25:46.150 but what happened was I got, 00:25:46.150 --> 00:25:46.810 Jim Conrad: I'm a simple guy. 00:25:46.870 --> 00:25:47.320 Eric Westra: No, no, no. 00:25:47.650 --> 00:25:48.025 I ended up. 00:25:48.435 --> 00:25:51.555 Uh, my first gig outta school was with the Premier's office. 00:25:51.675 --> 00:25:51.915 Jim Conrad: Right. 00:25:52.245 --> 00:25:55.485 Eric Westra: I worked for, uh, BC Trade Development and the Premier's 00:25:55.485 --> 00:25:59.175 office because they were under the same IT umbrella in Vancouver. 00:25:59.625 --> 00:26:03.365 And there's a great story about Mike Harcourt when he was the Premier. 00:26:03.830 --> 00:26:08.810 I was doing some of the, uh, the IT support and one day I get a phone call 00:26:08.810 --> 00:26:11.899 from Mike's office, which was just down the hall from the Trade Development 00:26:11.899 --> 00:26:13.760 Corp. Eric, can you come on over? 00:26:13.760 --> 00:26:16.639 'Cause Mike has a problem with his printer. 00:26:16.699 --> 00:26:18.530 And I said, sure, I can, I can come over. 00:26:18.590 --> 00:26:18.710 Jim Conrad: Yeah. 00:26:18.860 --> 00:26:22.189 Eric Westra: So, uh, apparently I, I walked into the office and, and Dixie, 00:26:22.280 --> 00:26:26.810 who is normally, uh, Mike Harcourt's EA, executive assistant, was away on vacation. 00:26:26.810 --> 00:26:28.290 They got somebody from the pool, I guess. 00:26:28.820 --> 00:26:31.970 And so I said hello to her and I walked into his office and he was sitting at his 00:26:31.970 --> 00:26:37.479 desk and he was signing a, a big stack of paper and you know, one after the other. 00:26:38.229 --> 00:26:39.879 And I said, Hey, how's it going, Mike? 00:26:39.910 --> 00:26:41.840 'Cause he was, was a first name kind of guy. 00:26:41.840 --> 00:26:43.510 And, uh, really nice guy. 00:26:44.020 --> 00:26:47.350 And I started to work on the printer and then all of a sudden I hear this, 00:26:47.500 --> 00:26:49.840 you know what the, across the room. 00:26:50.310 --> 00:26:52.030 And I, I said, what's the problem? 00:26:52.030 --> 00:26:53.200 He said, well, have a look at this. 00:26:53.590 --> 00:26:57.750 And he was signing a whole bunch of, uh, proclamations, thanking 00:26:57.750 --> 00:27:01.620 people for participating in the, uh, the Vancouver Centennial. 00:27:01.649 --> 00:27:03.389 There must have been 5 or 600 of these things. 00:27:03.810 --> 00:27:09.510 And the salutation at the bottom of the letter, what had happened was the, the new 00:27:09.600 --> 00:27:12.840 EA had just gone through, you know, spell check, click, click, click, click, without 00:27:12.840 --> 00:27:14.899 actually looking at what she was doing. 00:27:15.250 --> 00:27:19.050 And the salutation read, the right honorable Mike haircut. 00:27:21.905 --> 00:27:26.930 And, and, uh, you know, Mike, not exactly what you would call hirsute. 00:27:27.190 --> 00:27:31.319 Uh, and, and so, uh, that was, that was one of the funnier stories. 00:27:31.319 --> 00:27:31.719 But yeah, I, 00:27:31.719 --> 00:27:33.439 Jim Conrad: Follicly challenged. 00:27:33.459 --> 00:27:34.989 Eric Westra: Follicly challenged Mike was, yeah, that's right. 00:27:36.340 --> 00:27:38.889 Uh, but yeah, that was one of the, uh, the earliest gigs. 00:27:38.889 --> 00:27:42.250 And then eventually I, I got into tech communications, you know, tried 00:27:42.250 --> 00:27:46.810 to marry my background in tech with broadcasting, and then got involved 00:27:46.810 --> 00:27:50.345 with business analysis and business process and business optimization 00:27:50.675 --> 00:27:53.104 and then solution architecting. 00:27:53.104 --> 00:27:55.534 Uh, putting technology together with companies that were 00:27:55.534 --> 00:27:56.695 going through a lot of change. 00:27:56.884 --> 00:27:58.314 So that's, that's what I do now. 00:27:58.665 --> 00:28:04.905 Jim Conrad: An overall sort of context of this digital revolution of ours, 00:28:06.435 --> 00:28:10.645 obviously, uh, it's the machine that changed the world, the computer. 00:28:11.175 --> 00:28:17.860 But a lot of people now are, I guess, rethinking what that change 00:28:18.000 --> 00:28:21.470 or how much it has changed us. 00:28:21.530 --> 00:28:29.090 And, and relating back to, you know, being human and, and how we relate to computers. 00:28:29.600 --> 00:28:34.410 And now the whole idea of artificial intelligence. 00:28:35.420 --> 00:28:42.380 And we're seeing, while we first saw AI literally and viscerally in 2001. 00:28:42.794 --> 00:28:44.925 And the voice of HAL, the computer. 00:28:45.245 --> 00:28:49.455 My favorite line, which I've dubbed the HAL 9000 breakup line is, 00:28:49.784 --> 00:28:54.254 HAL 9000: Dave, this conversation can serve no purpose anymore. 00:28:55.010 --> 00:28:55.500 Goodbye. 00:28:56.399 --> 00:28:56.979 Jim Conrad: That's right. 00:28:57.794 --> 00:28:58.274 Goodbye. 00:28:58.304 --> 00:29:00.225 Eric Westra: Yeah, goodbye. 00:29:00.225 --> 00:29:01.034 Jim Conrad: Uh, HAL. 00:29:01.084 --> 00:29:01.444 HAL? 00:29:01.684 --> 00:29:02.414 Do you read me, HAL? 00:29:02.824 --> 00:29:05.895 Eric Westra: There's a great example of an artificial intelligence machine 00:29:06.135 --> 00:29:10.094 that didn't have any morality and, and a lot of the discussion around, 00:29:10.485 --> 00:29:13.455 you know, artificial intelligence these days, 'cause now computers 00:29:13.455 --> 00:29:14.584 are writing their own algorithms. 00:29:15.165 --> 00:29:17.925 They're learning and they're changing their own computer code. 00:29:18.014 --> 00:29:22.665 Jim Conrad: Now explain to me what is defined as machine learning. 00:29:22.905 --> 00:29:25.064 Eric Westra: Well, there's, there's deep machine learning 00:29:25.064 --> 00:29:26.115 and there's machine learning. 00:29:26.115 --> 00:29:28.004 And so the two are a little different. 00:29:28.064 --> 00:29:34.409 Um, machine learning is when a computer algorithm is given the capacity to write 00:29:34.470 --> 00:29:37.020 its own code, to learn from its mistakes. 00:29:37.080 --> 00:29:37.960 And to 00:29:37.990 --> 00:29:40.379 Jim Conrad: To replicate itself and heal itself. 00:29:40.409 --> 00:29:40.620 Eric Westra: Right. 00:29:40.620 --> 00:29:43.699 So there's a core set of code and instructions in the algorithm. 00:29:43.699 --> 00:29:48.000 And then it says, you know, go out and try something and under these circumstances 00:29:48.380 --> 00:29:50.179 if it works, make a note of that. 00:29:50.250 --> 00:29:53.350 And write your own code so that you do it the same way the next time. 00:29:53.990 --> 00:29:56.370 And then there's deep machine learning, which is really interesting. 00:29:56.370 --> 00:30:01.264 And that's around you can build reward systems into computers the same way 00:30:01.264 --> 00:30:03.065 that you do with pets and humans. 00:30:03.485 --> 00:30:08.084 You know, if a dog does something right, uh, you give it a, you give it a treat. 00:30:08.465 --> 00:30:09.104 If a child, 00:30:09.104 --> 00:30:09.965 Jim Conrad: Good boy, good boy. 00:30:10.294 --> 00:30:12.985 Eric Westra: If a child does something right, you reinforce that. 00:30:12.985 --> 00:30:14.794 You give them a hug, you say, good, good job. 00:30:15.004 --> 00:30:15.895 That's the way you do it. 00:30:16.375 --> 00:30:18.365 So we're doing the same thing with machines now. 00:30:18.365 --> 00:30:21.274 We're actually creating what we call artificial reward systems. 00:30:21.485 --> 00:30:21.665 Jim Conrad: Wow. 00:30:21.754 --> 00:30:25.725 Eric Westra: So computer will do something right, it will earn two stars. 00:30:26.124 --> 00:30:28.480 And if it does better, it gets three stars. 00:30:28.510 --> 00:30:33.850 So it focuses completely on making sure it gets as many rewards as it can. 00:30:34.149 --> 00:30:36.370 And through that process, it learns faster. 00:30:36.460 --> 00:30:36.820 Jim Conrad: Wow. 00:30:37.540 --> 00:30:41.740 Now, and this is one, once again, one of those ethical questions 00:30:41.740 --> 00:30:43.780 that, uh, that has to come up. 00:30:44.230 --> 00:30:46.330 It came up when they were making the atomic bomb. 00:30:46.540 --> 00:30:52.030 Yes, we have the scientific capacity and the knowledge to create this. 00:30:52.845 --> 00:30:54.784 The question is, should we. 00:30:56.175 --> 00:31:01.185 Have we reached a point of that question in artificial intelligence? 00:31:01.305 --> 00:31:03.024 Eric Westra: I think we continue to ask that question. 00:31:03.024 --> 00:31:04.504 And it's a good question to ask. 00:31:04.824 --> 00:31:09.799 Um, there's no, you know, artificial intelligence has huge promise. 00:31:10.110 --> 00:31:14.100 Um, there, there are lots of positive things that can come of it. 00:31:14.470 --> 00:31:19.139 Um, but like anything, there are a lot of very scary and negative things that 00:31:19.139 --> 00:31:20.850 come of it, if it isn't handled properly. 00:31:20.850 --> 00:31:24.990 And this is all through this emerging field called Robo Ethics. 00:31:25.050 --> 00:31:30.310 Jim Conrad: Now, thank God we had Isaac Asimov way back when because 00:31:30.330 --> 00:31:34.230 I, I was actually a big fan of Isaac Asimov, uh, science fiction novels. 00:31:34.290 --> 00:31:34.470 Eric Westra: Yeah. 00:31:34.709 --> 00:31:38.600 Jim Conrad: Now what were the three rules of robotics? 00:31:38.950 --> 00:31:40.960 Eric Westra: So the, the, Asimov proposed that there were 00:31:40.960 --> 00:31:42.945 three laws in his writings. 00:31:43.305 --> 00:31:48.285 The first law was that a robot or a machine could not, through either 00:31:48.285 --> 00:31:51.635 action or inaction, harm a human being. 00:31:52.325 --> 00:31:57.525 The second law was that, uh, the robot or the machine had to follow 00:31:57.795 --> 00:32:01.515 the instructions given to it by a human, unless it broke the first law. 00:32:02.175 --> 00:32:05.909 And the third law was that it had to self preserve itself. 00:32:05.915 --> 00:32:10.530 It, it couldn't, uh, it had to go through, you know, all the machinations necessary 00:32:10.530 --> 00:32:14.360 to preserve its existence provided it didn't break the first two laws. 00:32:15.129 --> 00:32:17.250 Jim Conrad: So it had a, it had a built in, the third law was 00:32:17.250 --> 00:32:20.909 self-preservation on, with a caveat, if it didn't break the first two laws. 00:32:20.909 --> 00:32:21.270 Eric Westra: That's right. 00:32:21.720 --> 00:32:25.220 Jim Conrad: So inside that sort of closed system was an ethical 00:32:25.880 --> 00:32:27.780 behavior model for robots. 00:32:27.870 --> 00:32:28.020 Eric Westra: Right. 00:32:28.260 --> 00:32:32.504 Jim Conrad: Uh, how practical is that to build into today's 00:32:32.565 --> 00:32:36.945 artificial intelligence, can we create those three laws to govern AI? 00:32:37.004 --> 00:32:38.055 Eric Westra: The laws are good ones. 00:32:38.055 --> 00:32:39.105 They seem to make sense. 00:32:39.345 --> 00:32:41.295 But they can produce, um, 00:32:41.325 --> 00:32:42.225 Jim Conrad: Asimov was a genius. 00:32:42.345 --> 00:32:43.065 Eric Westra: Yeah, he was. 00:32:43.125 --> 00:32:44.175 This is 1942. 00:32:44.295 --> 00:32:44.565 Jim Conrad: Yeah. 00:32:45.135 --> 00:32:48.615 Eric Westra: The thing about it though is that even making things as simple 00:32:48.615 --> 00:32:52.935 as that, you can end up with machines, uh, getting into, uh, you know, uh, 00:32:52.965 --> 00:32:54.405 cognitive dissonance we call it. 00:32:54.405 --> 00:32:58.980 Where, where it tries not to break a law, it makes a move to avoid 00:32:58.980 --> 00:33:03.030 breaking that law and then by doing so, it breaks another law and then 00:33:03.030 --> 00:33:04.470 it goes back to the first one. 00:33:04.770 --> 00:33:05.760 And so it gets caught, 00:33:05.760 --> 00:33:06.270 Jim Conrad: Vapor lock. 00:33:06.300 --> 00:33:06.500 Eric Westra: Yeah. 00:33:06.669 --> 00:33:07.240 Essentially. 00:33:07.870 --> 00:33:11.280 Um, and he, he actually did write about that in a short story. 00:33:11.280 --> 00:33:14.040 He talked about a robot that got caught in one of those loops. 00:33:14.130 --> 00:33:18.350 But, you know, if you use modern day sort of explanations around this 00:33:18.350 --> 00:33:23.909 sort of thing, it becomes much more difficult to nail all this down into 00:33:23.909 --> 00:33:25.550 a, to a very small set of rules. 00:33:25.610 --> 00:33:27.409 Jim Conrad: Because there are so many applications. 00:33:27.440 --> 00:33:29.410 You mentioned the positive aspects of AI. 00:33:29.930 --> 00:33:33.139 Uh, give me, let's say the top three, not Alexa. 00:33:33.260 --> 00:33:36.200 Eric Westra: Well, Alexa and voice recognition certainly 00:33:36.200 --> 00:33:37.310 are very, very big right now. 00:33:37.310 --> 00:33:39.409 Jim Conrad: If they make, if they make one of those things that sounds 00:33:39.409 --> 00:33:41.760 like Stephen Fry doing Jeeves. 00:33:42.570 --> 00:33:43.260 I'm on it. 00:33:43.560 --> 00:33:46.470 Jeeves: I was sent by the agencies and I was given to understand 00:33:46.470 --> 00:33:47.399 that you're required a valet. 00:33:47.785 --> 00:33:50.460 Eric Westra: I, I think eventually you, you'll, you can have your voice, voice 00:33:50.460 --> 00:33:52.010 assistant be whoever you want it to be. 00:33:52.510 --> 00:33:57.210 There are a set of circumstances can come up in modern life where the machine 00:33:57.210 --> 00:34:00.930 would be faced with so many different criter to make a decision, it would 00:34:00.930 --> 00:34:04.320 be very difficult for it necessarily to make the moral or the ethical one. 00:34:05.010 --> 00:34:08.574 Jim Conrad: You mentioned this ethical, moral dilemma in a, and I 00:34:08.574 --> 00:34:12.014 guess it was a thought experiment in regards to self-driving cars. 00:34:12.205 --> 00:34:13.105 Gimme a run through on that. 00:34:13.134 --> 00:34:16.674 Eric Westra: This was put to a group of university students, uh, it's called 00:34:16.674 --> 00:34:18.114 the child in the tunnel conundrum. 00:34:18.654 --> 00:34:21.895 So you're in an autonomous vehicle and you're with your family and the 00:34:21.895 --> 00:34:23.605 car is driving you on a mountain road. 00:34:23.605 --> 00:34:27.250 It's a two-lane mountain road, and your two kids are in the backseat. 00:34:27.250 --> 00:34:28.375 Your wife is sitting beside you. 00:34:28.975 --> 00:34:32.614 And you come around a curve and you're, you know, doing relatively 00:34:32.774 --> 00:34:36.255 the speed limit, you know, it's an autonomous car, it's following the law. 00:34:36.855 --> 00:34:40.995 But all of a sudden, uh, as you come around the curve, there is a tunnel 00:34:40.995 --> 00:34:45.614 entrance and uh, it's got a concrete sort of abutment around the face of it. 00:34:46.125 --> 00:34:50.444 And, uh, a small child darts out from behind a rock near the tunnel 00:34:50.444 --> 00:34:52.615 entrance and falls onto the road. 00:34:53.275 --> 00:34:56.480 The car has two choices really, given the circumstances. 00:34:56.509 --> 00:34:59.480 It's not going slow enough to stop in time. 00:34:59.569 --> 00:35:05.450 It either can send the car flying into the concrete or run over the child. 00:35:06.109 --> 00:35:09.020 Jim Conrad: So into the abutment, the, the passenger's in the 00:35:09.020 --> 00:35:10.849 car, the self-driving car. 00:35:11.109 --> 00:35:16.200 So the computer that's, that's governing that should try to preserve 00:35:16.319 --> 00:35:21.009 what it's meant to do, which is to avoid harm to human life, meaning 00:35:21.009 --> 00:35:22.620 the passengers inside the car. 00:35:22.649 --> 00:35:23.220 Eric Westra: Right. 00:35:23.250 --> 00:35:27.220 Jim Conrad: However, by, by running over the child, it's 00:35:27.290 --> 00:35:28.700 potentially killing the child. 00:35:28.700 --> 00:35:31.360 In fact, it's almost a certainty, given the circumstances. 00:35:31.360 --> 00:35:33.690 Eric Westra: Okay, so what does the car do? 00:35:33.750 --> 00:35:37.470 And when the group of university students were asked this about this 00:35:37.470 --> 00:35:41.459 problem, more than 60% of them said the car should run over the child. 00:35:42.329 --> 00:35:45.875 What the heck does a child, what, what is that child doing in the 00:35:45.875 --> 00:35:47.535 middle of nowhere in the first place? 00:35:47.645 --> 00:35:49.715 Jim Conrad: The evil spawn of a wiley coyote or what? 00:35:49.715 --> 00:35:49.775 Eric Westra: Yeah. 00:35:49.775 --> 00:35:54.315 It would be one death compared to potentially one or more in the car. 00:35:54.545 --> 00:35:58.095 Jim Conrad: So there's a value judgment that has to be made almost instantaneously 00:35:58.125 --> 00:36:00.255 by the autonomous computer driven car. 00:36:00.275 --> 00:36:03.815 Eric Westra: It's no longer a machine just making, uh, a judgment 00:36:03.815 --> 00:36:05.465 based on the best way to do things. 00:36:05.465 --> 00:36:07.115 It's making life and death decisions. 00:36:07.115 --> 00:36:07.445 Jim Conrad: Yes. 00:36:07.565 --> 00:36:07.925 Yes. 00:36:07.955 --> 00:36:11.790 Eric Westra: And that is where things get really weird. 00:36:14.955 --> 00:36:21.435 Jim Conrad: Well, and then given the technology surrounding military 00:36:21.654 --> 00:36:28.125 applications of drones, and right now, as I believe they are operated 00:36:28.125 --> 00:36:32.215 by human beings in a bunker somewhere outside of Las Vegas. 00:36:32.345 --> 00:36:34.925 Eric Westra: They're autonomous, some of them, uh, they, they don't 00:36:34.925 --> 00:36:36.305 need, require supervision at all. 00:36:36.335 --> 00:36:40.625 Jim Conrad: Well, that's where I, there's another ethical, moral question. 00:36:40.655 --> 00:36:43.895 You know, you've got a basically death from above being run 00:36:43.895 --> 00:36:47.265 by an AI driven computer. 00:36:47.955 --> 00:36:53.584 And the decision making processes inside that computer were coded by humans. 00:36:53.884 --> 00:36:56.944 So there still is a human connection. 00:36:57.234 --> 00:36:59.625 But it's, but it's pretty far removed, isn't it? 00:36:59.745 --> 00:37:00.225 Eric Westra: It is. 00:37:00.314 --> 00:37:04.335 And uh, you know, there's talk about eventually some of these drones, 00:37:04.335 --> 00:37:06.495 I mean, we're getting pretty dark here, but some of these drones, 00:37:06.524 --> 00:37:08.720 Jim Conrad: It's a dark and stormy Saturday morning. 00:37:08.720 --> 00:37:08.870 Eric Westra: Yes. 00:37:10.069 --> 00:37:12.830 You know that they would use face recognition to sort of, you know, 00:37:12.830 --> 00:37:15.200 recognize the face of a terrorist and, 00:37:15.200 --> 00:37:16.490 Jim Conrad: As opposed to a child,. 00:37:16.490 --> 00:37:17.089 Eric Westra: Uh, yeah. 00:37:17.299 --> 00:37:20.890 And then, you know, go out and, and essentially assassinate that person. 00:37:21.410 --> 00:37:26.930 So that's a really dark side of AI and, and there will always be that, I guess. 00:37:26.930 --> 00:37:29.660 But there are some very positive things. 00:37:29.660 --> 00:37:31.610 I'm running into it now every day. 00:37:31.610 --> 00:37:31.670 Jim Conrad: Yeah. 00:37:31.670 --> 00:37:35.330 I thought I mentioned, or I requested a bright side aI story. 00:37:36.799 --> 00:37:39.799 We, we went immediately to the dark side, so let's, let's find some 00:37:39.950 --> 00:37:41.399 ray of sunshine here somewhere. 00:37:41.399 --> 00:37:45.560 What's, what's the number one beneficial use to humans and 00:37:45.560 --> 00:37:47.570 humankind for artificial intelligence? 00:37:47.600 --> 00:37:49.400 Eric Westra: Well, there, there, there are lots of them, you know. 00:37:49.430 --> 00:37:50.360 Jim Conrad: Robot surgeons? 00:37:50.810 --> 00:37:52.549 Eric Westra: Robotic surgery is one. 00:37:52.670 --> 00:37:54.170 Robotic diagnosis. 00:37:54.460 --> 00:37:58.740 Jim Conrad: I don't know if I want HAL, uh, doing my prostate surgery. 00:38:00.330 --> 00:38:01.259 HAL 9000: I'm sorry, Jim. 00:38:01.320 --> 00:38:02.730 The knife slipped. 00:38:04.500 --> 00:38:07.740 Eric Westra: But you know, I'm running into situations now where, uh, it's being 00:38:07.740 --> 00:38:14.310 used for example, to determine how to hook up a borrower with a lending institution 00:38:14.610 --> 00:38:16.470 based on the borrower's profile. 00:38:16.470 --> 00:38:19.920 There's a whole bunch of, you know, stuff that goes into making that decision. 00:38:20.650 --> 00:38:23.650 And AI can be very helpful in making those connections. 00:38:24.000 --> 00:38:26.010 So that's more of a business efficiency thing. 00:38:26.160 --> 00:38:27.000 HAL 9000: I'm sorry, Jim. 00:38:27.524 --> 00:38:29.595 Your loan application has been denied. 00:38:30.495 --> 00:38:30.975 Eric Westra: That's right. 00:38:31.575 --> 00:38:35.805 Uh, and you know, but there are lots of applications emerging now, and 00:38:35.805 --> 00:38:40.095 it's gonna become more and more, uh, a part of our lives, uh, every day. 00:38:42.314 --> 00:38:46.274 Jim Conrad: I've got a question for you, given that technology is 00:38:46.365 --> 00:38:51.754 speeding up, that some people are becoming somewhat overwhelmed by it. 00:38:52.174 --> 00:38:53.680 We're, we're trying to adapt. 00:38:54.345 --> 00:38:59.175 But it seems that as we, as we adapt, the next thing comes along 00:38:59.685 --> 00:39:01.455 and then we have to adapt again. 00:39:01.545 --> 00:39:06.555 And some people are retreating and there's a kind of a backlash. 00:39:06.915 --> 00:39:12.900 Do you think that there will be a, a time when when we'll reject technology 00:39:13.080 --> 00:39:17.730 and, and try to go back, back in time if it, if it was possible. 00:39:17.760 --> 00:39:18.900 Eric Westra: Well, there already has been. 00:39:18.900 --> 00:39:22.160 I mean, I keep thinking about Joe Walsh's song, 00:39:24.340 --> 00:39:26.460 Analog Man. 00:39:29.730 --> 00:39:32.910 There already has been a bit of a backlash, especially in the music scene. 00:39:33.570 --> 00:39:37.279 It started with the release of CDs many years ago. 00:39:37.690 --> 00:39:41.415 And the digitized sound just didn't have the same kind of, the, the 00:39:41.415 --> 00:39:43.095 word that was used was warmth. 00:39:43.215 --> 00:39:43.515 Jim Conrad: Right. 00:39:43.965 --> 00:39:47.595 Eric Westra: And so in the last few years especially, we've been seeing a 00:39:47.595 --> 00:39:53.115 resurgence in, uh, good old vinyl and artists actually pressing their own, uh, 00:39:53.145 --> 00:39:54.615 instead of going through record companies. 00:39:54.615 --> 00:39:58.455 And, and it's because that audio files prefer the sound. 00:39:58.455 --> 00:39:59.675 So there's one example. 00:40:00.375 --> 00:40:02.835 I think the latest thing that's, that's been. 00:40:03.195 --> 00:40:09.745 Sort of making people recoil sometimes in sheer, utter terror is AI. 00:40:10.155 --> 00:40:12.915 Jim Conrad: Now define AI for me. 00:40:12.945 --> 00:40:16.125 In a practical sense, what is AI? 00:40:16.635 --> 00:40:18.255 And how did it develop? 00:40:19.065 --> 00:40:23.805 Eric Westra: So it's based on what's called an LLM or a large language model. 00:40:23.935 --> 00:40:32.015 And the idea is you synthesize, uh, using algorithms, and hardware, uh, 00:40:32.045 --> 00:40:37.905 the neural pathways in the mind, the human brain, that process language. 00:40:37.965 --> 00:40:40.665 Jim Conrad: And it was originally developed to actually model the 00:40:40.665 --> 00:40:44.655 brain in order to do research on how, on how the brain works, right? 00:40:44.885 --> 00:40:46.275 Eric Westra: Neural networks work, yes. 00:40:46.275 --> 00:40:50.415 And then they became quite helpful when it came to large language models. 00:40:50.475 --> 00:40:55.785 Because you think about, you know, one word, um, the word right could be 00:40:55.785 --> 00:40:57.940 spelled a couple of different ways. 00:40:58.390 --> 00:41:00.000 Could mean several different things. 00:41:02.709 --> 00:41:10.440 So, you know, the large language model is there to teach the algorithm what the 00:41:10.440 --> 00:41:15.030 actual meaning of the word and a number of words fit, that are fitted together, 00:41:15.090 --> 00:41:17.310 what that means in the context for it. 00:41:17.460 --> 00:41:17.850 Jim Conrad: Yes. 00:41:18.570 --> 00:41:21.899 One could define intelligence a number of different ways. 00:41:22.680 --> 00:41:29.535 A lot of people define intelligence as in AI, in the ability to assimilate 00:41:29.745 --> 00:41:34.274 all sorts of information to correlate to, to come up with solutions and 00:41:34.274 --> 00:41:37.234 answers to questions quicker and faster. 00:41:37.964 --> 00:41:45.165 Or is intelligence coming back to the human aspect is intelligence, creativity? 00:41:46.424 --> 00:41:49.109 Eric Westra: I think in, in many respects it is. 00:41:49.299 --> 00:41:52.830 I mean, we can be creative with language, which is a structure, and 00:41:52.830 --> 00:41:58.109 we find all sorts of different ways of using musical notes and being creative. 00:41:58.770 --> 00:42:03.419 And they're, they're a very finite set of sort of things. 00:42:03.540 --> 00:42:07.880 Uh, ABC, you know, the notes, uh, the chords. 00:42:08.430 --> 00:42:13.740 So we start to see that, um, the, the real danger actually that's being 00:42:13.740 --> 00:42:20.460 presented by AI is in the creative space, because now it's possible to manipulate 00:42:20.760 --> 00:42:29.360 imagery, audio in a way that most fools, uh, the human consumer, if you will, 00:42:29.390 --> 00:42:34.020 because the, the artificial intelligence has learned what matters most to us. 00:42:34.410 --> 00:42:36.060 Jim Conrad: And that is the image? 00:42:36.150 --> 00:42:38.760 Eric Westra: Well, it could be the image or nuances of the image. 00:42:38.760 --> 00:42:44.240 So you take a, for example, you could ask a, a chat bot, or, uh, a piece of 00:42:44.240 --> 00:42:49.610 artificial intelligence in say, Adobe Photoshop, create a picture for me of 00:42:49.610 --> 00:42:55.610 a guy age, so and so, uh, that has very pleasant features, is a professional 00:42:55.610 --> 00:43:00.650 person and is modeled after people who have good leadership qualities. 00:43:00.680 --> 00:43:01.000 Jim Conrad: Yes. 00:43:01.620 --> 00:43:08.420 So, so the, the AI functions as, but you still a, a human being still has to input. 00:43:08.875 --> 00:43:09.235 Eric Westra: Oh yeah. 00:43:09.265 --> 00:43:10.195 Jim Conrad: Certain things. 00:43:10.225 --> 00:43:10.405 Eric Westra: Yeah. 00:43:10.405 --> 00:43:14.245 The intelligence is, is, is essentially, you know, called from, 00:43:14.515 --> 00:43:17.785 from a lot of different sources and usually from other people's input. 00:43:18.625 --> 00:43:21.365 It's, it's only as good as the data that goes into it. 00:43:21.385 --> 00:43:26.125 And this is part of the problem right now with AI, is that there's so much data 00:43:26.775 --> 00:43:30.105 going into it, that it's getting confused. 00:43:30.215 --> 00:43:33.075 Jim Conrad: And they call that hallucination? 00:43:33.075 --> 00:43:34.315 Eric Westra: Hallucination, yeah. 00:43:35.105 --> 00:43:35.305 Jim Conrad: Yeah. 00:43:35.305 --> 00:43:37.305 We're familiar with that, aren't we? 00:43:38.955 --> 00:43:43.005 Eric Westra: For many years, my life seemed to be one large hallucination. 00:43:43.605 --> 00:43:46.795 Uh, thankfully those days are over. 00:43:46.795 --> 00:43:51.134 Um, but the, the hallucinations are, are created by just throwing 00:43:51.134 --> 00:44:01.024 a lot of data into the model rather than focusing on quality of data. 00:44:01.254 --> 00:44:05.565 Jim Conrad: And, and that's where I, I come always back to creativity and 00:44:05.565 --> 00:44:07.705 the artist when it comes to technology. 00:44:08.325 --> 00:44:13.605 Technology certainly has, has released our ability to create, 00:44:13.875 --> 00:44:15.945 even in multi-track recording. 00:44:16.180 --> 00:44:20.235 You know, now you can track almost infinitely and sound bites and 00:44:20.235 --> 00:44:21.705 samples and things like that. 00:44:21.795 --> 00:44:24.255 Eric Westra: But it's all mimicry in one form or another. 00:44:24.285 --> 00:44:24.560 Jim Conrad: That's true. 00:44:24.920 --> 00:44:31.730 But as we go forward into this AI world, how is our humanity going 00:44:31.730 --> 00:44:33.770 to be preserved, do you think? 00:44:34.940 --> 00:44:37.570 Eric Westra: Well, I think a lot of people are asking that question now. 00:44:37.600 --> 00:44:42.689 And thankfully, some of the people like Sam Altman, uh, you know, who heads 00:44:42.689 --> 00:44:47.165 up ChatGPT, are thinking not just in terms of what the technology can do, 00:44:47.240 --> 00:44:48.525 but how it should be guard railed. 00:44:49.095 --> 00:44:49.695 Jim Conrad: Yes. 00:44:49.785 --> 00:44:52.875 Eric Westra: Uh, anybody working in AI who isn't thinking about that, 00:44:53.085 --> 00:44:58.485 the ethics of AI and the humanity behind AI, shouldn't be in the field. 00:44:58.575 --> 00:45:02.985 Because the potential is so there for, for misuse. 00:45:03.255 --> 00:45:04.785 Jim Conrad: And that's altruistic. 00:45:05.355 --> 00:45:09.225 But as we know, corporate culture is ruthless. 00:45:09.225 --> 00:45:09.615 Eric Westra: Sure. 00:45:09.880 --> 00:45:11.384 But, but the genie's outta the bottle. 00:45:11.535 --> 00:45:12.075 Jim Conrad: Yeah, that's true. 00:45:12.080 --> 00:45:12.420 Eric Westra: I mean, right. 00:45:12.525 --> 00:45:16.815 I mean, and, and, and every time a new technology comes along, there's always a, 00:45:16.995 --> 00:45:18.735 Jim Conrad: A human cry. 00:45:18.735 --> 00:45:20.625 It's it's gonna be the end of humanity. 00:45:20.835 --> 00:45:24.375 Eric Westra: You know, uh, digitized music was, you know, we use that example 00:45:24.375 --> 00:45:25.524 at the beginning of the conversation. 00:45:25.524 --> 00:45:31.794 I mean, everybody said, oh, it's, that's the end of, you know, a lot of things. 00:45:32.154 --> 00:45:34.694 Uh, a lot of the, the artists now are just gonna go for the 00:45:34.694 --> 00:45:35.805 lowest common denominator. 00:45:35.805 --> 00:45:37.935 They're not gonna get reimbursed, et cetera, et cetera. 00:45:38.415 --> 00:45:38.805 Okay. 00:45:38.985 --> 00:45:39.495 Yes. 00:45:39.525 --> 00:45:43.825 There are changes whenever you have a new technology introduced. 00:45:44.674 --> 00:45:46.365 Sometimes more than others. 00:45:46.845 --> 00:45:51.715 The thing is, how do we adapt as humans to the technology? 00:45:51.775 --> 00:45:56.845 How do we make the best use of it, and how do we make sure that we 00:45:56.845 --> 00:45:58.805 don't squander the opportunity? 00:45:59.595 --> 00:46:04.525 Instead of being afraid constantly about what the technology might do wrong, 00:46:04.735 --> 00:46:06.155 let's focus on what it might do right. 00:46:11.545 --> 00:46:16.165 Jim Conrad: Artificial intelligence, alien intelligence. 00:46:17.485 --> 00:46:17.545 Hmm. 00:46:18.779 --> 00:46:22.140 We live in interesting times, don't we? 00:46:23.160 --> 00:46:26.520 Here's what big tech has to say about AI. 00:46:27.540 --> 00:46:33.990 Sundar Pichai, CEO of Google, the future of AI is not about replacing humans, 00:46:34.290 --> 00:46:37.230 it's about augmenting human capabilities. 00:46:38.000 --> 00:46:43.590 Ginni Rometty, former CEO of IBM says, AI will not replace humans, but those 00:46:43.590 --> 00:46:47.550 who use AI will replace those who don't. 00:46:49.004 --> 00:46:56.924 And then perched high on the proverbial fence, CEO of SpaceX and Tesla, Elon Musk 00:46:56.984 --> 00:47:02.955 said, AI is likely to be either the best or worst thing to happen to humanity. 00:47:04.154 --> 00:47:04.845 Thanks, Elon. 00:47:06.044 --> 00:47:12.134 But the final word should go to the ultimate tech nerd Bill Gates, 00:47:12.314 --> 00:47:17.730 co-founder of Microsoft, said, artificial intelligence will evolve to become a 00:47:17.730 --> 00:47:25.320 super intelligence, we need to be mindful of how it's developed and ensure that it 00:47:25.320 --> 00:47:30.255 aligns with humanities best interests. 00:47:34.125 --> 00:47:36.075 James Cromwell: There have always been ghosts in the machine. 00:47:38.145 --> 00:47:43.545 Random segments of code that have grouped together to form unexpected protocols, 00:47:45.165 --> 00:47:50.190 unanticipated these free radicals and gender questions of free will. 00:47:52.299 --> 00:47:59.049 Creativity and even the nature of what we might call the soul, or 00:48:01.140 --> 00:48:16.890 is it something 00:48:24.779 --> 00:48:25.910 more. 00:48:28.095 --> 00:48:30.885 Jim Conrad: That is the Conovision Podcast, the spirit of 00:48:30.885 --> 00:48:35.745 storytelling, episode three, all about artificial intelligence. 00:48:36.225 --> 00:48:41.715 My thanks to all the conscious entities in the first AI G7 Summit. 00:48:41.985 --> 00:48:45.585 We are still trying to decode that digital word salad. 00:48:46.075 --> 00:48:52.524 To Amazon's Alexa for dropping by with illuminating answers to absurd questions, 00:48:53.185 --> 00:48:57.654 to Professor  Yuval Noah Harari, and his insight into the differential patterns 00:48:57.654 --> 00:49:00.685 between intelligence and consciousness. 00:49:01.435 --> 00:49:06.279 The New York Times, for the brilliant story about Douglas Rain, the Canadian 00:49:06.279 --> 00:49:10.840 Shakespearean actor, and the voice of HAL, the computer in Stanley Kubrick's, 00:49:10.840 --> 00:49:15.670 1968 Masterpiece 2001: A Space Odyssey. 00:49:16.840 --> 00:49:20.710 And as well our tech guru Eric Westra, for his thoughts on the future of 00:49:20.710 --> 00:49:23.200 AI and what it will mean to us all. 00:49:23.980 --> 00:49:28.375 And finally, we just heard a monologue from the film I, Robot and the 00:49:28.375 --> 00:49:31.225 majestic voice of actor James Cromwell. 00:49:31.735 --> 00:49:37.475 I am James Conrad, AKA Cono, and this has been Conovision, 00:49:37.495 --> 00:49:39.475 the spirit of storytelling. 00:49:40.135 --> 00:49:44.935 Until next time, remember, we are all stories to be told.