Latest / Buzzcast / Even Robots Hate Being Interrupted

Transcript

Even Robots Hate Being Interrupted · Buzzcast

• 0:00
  Kevin: Kevin, I've got a story for you from Google.
• 0:02
  Kevin: Okay, it's got some connections to podcasting, so hear me out, well.
• 0:05
  Alban: I hope so.
• 0:06
  Alban: That's the least we can do for the people that tune in.
• 0:11
  Kevin: You remember this Notebook LM product that Google made?
• 0:13
  Kevin: Yes, all right, you upload a bunch of documents and then it like kind of reads through them and it kind of creates these fake podcasts.
• 0:20
  Kevin: Yes, so last month they added a new feature where?
• 0:24
  Alban: what was the new feature?
• 0:27
  Kevin: I don't know if this is intentional or not, you know, because it is it's a troll.
• 0:31
  Kevin: Okay, you can talk to the hosts and basically what types of things.
• 0:36
  Kevin: Can you say oh my gosh, so you can just ask questions and like they'll stop and then they'll answer your question and then they kind of get back into the flow.
• 0:46
  Kevin: It's like a call-in show, right?
• 0:48
  Kevin: And as you have been alluding to with your jokes is that everyone really is just interrupting these hosts and the AIs are getting kind of snippy.
• 1:00
  Alban: Rightly so.
• 1:00
  Alban: If I kept doing this to you, you would probably get snippy.
• 1:04
  Kevin: This is my point.
• 1:04
  Kevin: So everybody who's talking about it is like, oh, it's funny.
• 1:08
  Kevin: So here's the quote from the VP of Google labs.
• 1:15
  Kevin: We were occasionally giving snippy comments to human callers, like I was just getting to that, or, as I was about to say, which felt oddly adversarial, and my take is that's not oddly adversarial, that's healthy.
• 1:26
  Alban: Yeah, I do think it's healthy.
• 1:28
  Alban: I don't think a lot of people in the world actually communicate like that.
• 1:31
  Alban: They think like that.
• 1:32
  Alban: They think like give me a second, I'm getting there, but they don't say it, and so that the AI is saying it might feel adversarial to some people, but like if you worked in the Buzzsprout office, for example, you'd see that's very normal type conversation.
• 1:47
  Kevin: Well, so they roll out this update, which they say makes it more friendly, so that when you interrupt the AIs, they're like oh, great question.
• 1:54
  Kevin: Like they're trying to compliment you on how rude you can be, and this is the wrong direction.
• 2:00
  Kevin: This is not healthy use of.
• 2:02
  Alban: AI Right.
• 2:03
  Alban: It's just going to encourage more bad behavior.
• 2:05
  Alban: I agree with you.
• 2:06
  Alban: I think we need more truthful AI, more like Larry David.
• 2:12
  Alban: Curb your enthusiasm, type AI.
• 2:15
  Kevin: I need to send you a screenshot of my notes because I wrote we need an aggressive mode or Larry David mode.
• 2:21
  Kevin: Like Mr Finn has another question, why don't you illuminate us with your additional question, Like you would know the answer if you just waited to listen to?
• 2:33
  Alban: us?
• 2:33
  Alban: Yeah, it reminded me when I was reading this article.
• 2:35
  Alban: It reminded me of did you?
• 2:36
  Alban: You know, we all had different types of teachers coming up in high school and in college.
• 2:42
  Alban: You had the teacher that wants the hands up or the questions during the lesson, and you had the teacher that wants the hands up or the questions during the lesson, and then you want the teacher that asked you to hold them all to the end.
• 2:49
  Alban: Yes, but the worst case it's not fair to the students to like not tell them what you prefer and then just get mad at them for doing something one or the other, whether they're interrupting you during your lesson and you're like I'm going to get to that.
• 3:01
  Alban: That's not fair to them.
• 3:03
  Alban: Or you hold your question there and they're like why didn't you ask that in the beginning?
• 3:07
  Alban: I would have covered that in the beginning and you never asked it.
• 3:09
  Alban: It's not fair either way.
• 3:11
  Alban: So, honest conversation on both sides.
• 3:14
  Alban: I don't think it's rude, but I do hear that, depending on the social circles you run in, maybe the different states that you live in or countries or the norms around where you live and the way that people talk, it might hit people, some people, as a little off.
• 3:27
  Alban: I think I would like it.
• 3:28
  Alban: It would get me to pay for it.
• 3:30
  Kevin: I think that's true, but what I hope we don't do is trying to make AI interact with us in unhuman ways, like in inhumane ways.
• 3:40
  Kevin: I think it's rude to just run over people in conversation and it's not inappropriate that sometimes people say like, okay, hold on a second.
• 3:51
  Kevin: Let me finish this point.
• 3:52
  Kevin: And if you end up with this new dynamic where people are talking to AIs all the time and we're running over them and we're just rude and we're becoming those people who are really kind to their boss but they're really rude to the waiter at lunch, you know these kinds of people.
• 4:08
  Alban: Yeah, I know these kinds of people.
• 4:09
  Alban: No names, but yeah, I know these kinds of people.
• 4:14
  Kevin: And I feel like what we really need is to be reinforcing like the healthy conversation dynamics, the healthy podcasting dynamics and maybe this is like a conversation for a full episode at some point but like active listening, asking open-ended questions.
• 4:31
  Kevin: Curiosity like this makes you better at conversations in real life, but it also are the skills of podcasting.
• 4:38
  Kevin: One of the skills is noticing this person's wrapped up their point.
• 4:42
  Kevin: I'm avoiding the interruption and now it's a good time for me to step in.
• 4:48
  Alban: Okay, let me throw a curveball out.
• 4:49
  Alban: Do you think it's best that, as this technology emerges and we're trying to figure out, as humans, how to interact with it, that we should build it in such a way that it mimics conversations with humans as much as possible, all the way down to cultural norms of what would be considered rude or polite conversation?
• 5:09
  Alban: Or do you think we should adopt a different way of interacting with machines than we do with humans?
• 5:15
  Alban: So if you want to practice human interaction, you should probably go around humans and do that, but when you're interacting with a machine, maybe there's a different way that we speak.
• 5:23
  Alban: And I'll tell you what got me thinking about.
• 5:25
  Alban: This is like I have just noticed, when I interact with Siri on my phone, I will say things like please and thank you, which seems very weird to do.
• 5:35
  Alban: If I make the cognitive connection that I'm talking with a machine, it doesn't care if I'm polite, it just wants to know the tasks that I want it to do for me.
• 5:43
  Alban: So, like hey, siri, can you please tell me the weather today?
• 5:46
  Alban: Would be received in the same way, probably, as hey, siri, what's the weather today?
• 5:49
  Alban: Right, but to a human I would absolutely say you happen to know what the weather is outside.
• 5:54
  Alban: I'd phrase it in a nice way and after they responded I'd say thank you, machines don't need that.
• 5:59
  Alban: Yeah, what do you think?
• 6:00
  Kevin: You think we should always converse in a polite way, I have a strong opinion on this and I think and I'm on team, please and thank you for two reasons.
• 6:16
  Alban: One, in the unlikely slash I don't know what, likelihood, but in the event that AI takes over everything and we have Skynet over, you want to be on record as you want to be on record that I was nice before we had to be nice.
• 6:24
  Alban: Okay.
• 6:25
  Kevin: Which is a valid point.
• 6:26
  Kevin: But the second is you're doing something to yourself when you treat your pet poorly, or you treat children differently than you treat adults, or you treat the waiter poorly or you treat, I think, ai poorly.
• 6:40
  Kevin: Like you're doing something to yourself, wait wait, wait, wait, wait, wait.
• 6:44
  Alban: Okay, I don't want to get religious, but you just all the examples that you named were things that have souls again, not in the most non-religious sense, the things that you you're comparing to have souls versus a machine that, as far as we know, do not.
• 6:59
  Kevin: So much more intensely.
• 7:00
  Kevin: Okay, even trees, like sitting there and chopping up a tree for the fun of it and just throwing like.
• 7:06
  Kevin: There's ways to treat the world that are just kind of like you're treating in kind of a crappy way, and it isn't always just that you're doing something bad to another entity, maybe something that has a soul or a personality, but you kind of it's just not good for yourself and you know, it's why I, if I order me, I'm like I'm going to finish my plate.
• 7:27
  Kevin: It feels wrong to like leave it on the plate.
• 7:30
  Alban: And again because an animal was sacrificed for that meat.
• 7:33
  Alban: But like I would take care of my car.
• 7:35
  Alban: Like not talking about a smart car, but I would take care of my car for the purpose of wanting to preserve it as long as possible.
• 7:43
  Alban: Right, not because I'm going to hurt its feelings if I don't wash the salt off the under part of the car after driving on a salted road, but like I would give my dog a bath because I care about my dog and I want my dog to not smell bad, not just for me, but because it's an alive thing.
• 7:59
  Alban: Or I might say nice, sweet things to my dog, even though she probably doesn't understand it, because I just respect that somewhere in there is a soul, it has life, it has emotions of some sort that I don't know.
• 8:12
  Alban: As a dog owner, you think your dog has maybe more emotions than they do.
• 8:14
  Alban: I don't know, but a machine definitely doesn't.
• 8:16
  Alban: That's what we know for sure.
• 8:18
  Alban: My jacket definitely doesn't.
• 8:20
  Kevin: No, and I like this one jacket I have because I continually repair it and I keep it like it still looks good it's 15 years old and when it gets a tear I patch it and like I appreciate it more because I treat it well, and so that's maybe.
• 8:36
  Kevin: Maybe that's a different way of saying yeah, I'm on board with let's treat these systems better, and I think we should also design them to help us kind of grow as people, them to help us kind of grow as people.
• 8:52
  Kevin: And if there was a way for the system to kind of nudge people to becoming better podcasters or better conversationalists or better listeners, that would be a good thing, all right, let me go one step further, then.
• 9:01
  Alban: I'm agreeing with you, but I'm wondering should the machines then, if you're being rude to it, should they at some point say well, if you're going to continue to be rude to me, then I'm not going to engage in this conversation, I'm not going to do the thing you want me to do.
• 9:14
  Kevin: Yeah, I thought that one too.
• 9:15
  Kevin: I was like at some point you go, hey, sounds like you know everything you need to know, so we can go ahead and just wrap the podcast now At some point.
• 9:24
  Kevin: That is a healthy boundary to give to somebody who just interrupts you constantly.
• 9:28
  Kevin: Maybe people need to hear that.
• 9:30
  Kevin: Maybe they need to hear that from their notebook.
• 9:32
  Kevin: Lm.
• 9:34
  Alban: I don't know.
• 9:34
  Alban: I don't know if I agree with that.
• 9:35
  Alban: I do really agree with your point, though.
• 9:38
  Alban: It makes you a better human being to get into healthy communication habits.
• 9:42
  Alban: Whether you're talking to a machine or a human, or caring for the use of something, taking care of it Again, for various different reasons, not only if it has a soul or feelings there's value in that, beyond just preserving the life of that thing, it develops some character or traits in humans that are it's good to practice those things, regardless of whether the machine's going to give you the same output, whether you're nice or not or respectful of it or not.
• 10:08
  Alban: But I don't know about the on the programming side.
• 10:10
  Alban: Like, how human do we really need to make these things?
• 10:14
  Alban: Like, how?
• 10:14
  Alban: How much of mimicking a human personality?
• 10:16
  Alban: Now, if your ultimate vision for this stuff is that we are going to have animatronic humans and we want them to feel and interact like humans, so we want them to pretend like they have feelings.
• 10:28
  Alban: So if you say something hurtful to them, we want them to get sad, and then we want them to go sulk in a corner or something or whatever they do, or call their other robot friend and talk bad about you.
• 10:36
  Alban: I don't know how far is too far, but it seems like we're starting to run into these issues sooner than I thought we would in the AI world.
• 10:44
  Kevin: Yeah Well for anyone who listened to the last episode, we still have gotten some good responses.
• 10:49
  Kevin: We need some more.
• 10:50
  Kevin: We asked what type of podcasting do you want to automate with AI?
• 10:55
  Kevin: What part do you not want to automate with AI?
• 10:57
  Kevin: You know what part is important to be human for you.
• 11:00
  Kevin: Feel free to add on.
• 11:00
  Kevin: You know what part is important to be human for you.
• 11:01
  Kevin: Feel free to add on.
• 11:03
  Kevin: Do you think trees have souls?
• 11:04
  Kevin: Another important question that Kevin and I have brought up in this quick cast Jordan's.
• 11:10
  Kevin: She's out sick today, so I'm sure she's going to enjoy listening to this.
• 11:14
  Kevin: Going like what the heck?
• 11:15
  Kevin: Going off the rails when I left.
• 11:18
  Alban: So I don't know how many of you use Google LM to help you with your podcast preparation.
• 11:22
  Alban: I know Jordan does, and so she'll probably have some thoughts when she's back next week.
• 11:25
  Alban: She uses it to help produce a podcast, to do some like UFO type lookup stuff for UFO stuff that she does on a podcast.
• 11:32
  Alban: But if any of you notice any of these updates, is Google LM getting more polite or not?
• 11:37
  Alban: Is this a good thing or not?
• 11:38
  Alban: Interesting questions to ask as we continue to push into how AI can help us keep podcasting.