WEBVTT - Episode 3: Conversations With Myself

0:00:05.320 --> 0:00:08.159
<v Speaker 1>Hi, It is my pleasure to be here with you

0:00:08.240 --> 0:00:14.160
<v Speaker 1>all today. In those thirteen syllables, you all just made

0:00:14.240 --> 0:00:20.720
<v Speaker 1>judgments about me, about my socioeconomic status, my intelligence level,

0:00:21.120 --> 0:00:22.919
<v Speaker 1>and whether or not you're going to sit here and

0:00:22.960 --> 0:00:25.720
<v Speaker 1>listen to me for the next ten to twelve minutes,

0:00:26.320 --> 0:00:30.320
<v Speaker 1>based primarily on the perception of my voice.

0:00:31.800 --> 0:00:35.800
<v Speaker 2>That's doctor Wendy Lebourne, a voice pathologist, speaking at a

0:00:35.840 --> 0:00:40.280
<v Speaker 2>TEDx conference in Cincinnati back in twenty eighteen. Lebourne is

0:00:40.320 --> 0:00:43.479
<v Speaker 2>a researcher and consultant who works with everyone from Broadway

0:00:43.520 --> 0:00:48.120
<v Speaker 2>actors to Grammy winning singers to CEOs on maintaining and

0:00:48.200 --> 0:00:51.920
<v Speaker 2>maximizing their voices. As you might expect, she's big on

0:00:51.960 --> 0:00:54.080
<v Speaker 2>the idea that your voice is a central part of

0:00:54.120 --> 0:00:58.360
<v Speaker 2>your identity. Her TED talk was titled Vocal Branding, How

0:00:58.360 --> 0:01:00.560
<v Speaker 2>your voice shapes your communication image.

0:01:00.880 --> 0:01:04.920
<v Speaker 1>Your voice is like your thumbprint or your facial features.

0:01:05.360 --> 0:01:08.000
<v Speaker 1>It is unique and authentically you.

0:01:09.400 --> 0:01:11.360
<v Speaker 2>There is, in fact, a decent amount of research looking

0:01:11.360 --> 0:01:14.000
<v Speaker 2>into this question about what your voice means to you

0:01:14.400 --> 0:01:17.600
<v Speaker 2>and to how people perceive you. I've reported on voice

0:01:17.600 --> 0:01:20.520
<v Speaker 2>forensics before, and the truth is we're not as great

0:01:20.560 --> 0:01:23.360
<v Speaker 2>at detecting voices as we think we are. But at

0:01:23.360 --> 0:01:27.120
<v Speaker 2>the same time, your voice undoubtedly broadcasts a huge amount

0:01:27.120 --> 0:01:30.479
<v Speaker 2>of unique information about you. It helps dictate how others

0:01:30.520 --> 0:01:33.080
<v Speaker 2>perceive you, both in the moment with its level of

0:01:33.120 --> 0:01:36.839
<v Speaker 2>strength and emotion in any given conversation, and longer term

0:01:37.040 --> 0:01:39.640
<v Speaker 2>in what it conveys about your age, where you're from,

0:01:39.760 --> 0:01:40.960
<v Speaker 2>and your overall temperament.

0:01:41.319 --> 0:01:45.080
<v Speaker 1>Your voice is your calling card, and it is the

0:01:45.120 --> 0:01:49.120
<v Speaker 1>most important element in your personal brand.

0:01:50.600 --> 0:01:52.600
<v Speaker 2>I don't know if I'd go as far as doctor Leborne,

0:01:53.040 --> 0:01:54.800
<v Speaker 2>but I do think I've always wanted to believe this

0:01:54.920 --> 0:01:57.720
<v Speaker 2>idea that your voice is a central part of.

0:01:57.680 --> 0:01:58.240
<v Speaker 3>Who you are.

0:01:59.520 --> 0:02:01.960
<v Speaker 2>When I was eleventh grade, my English teacher told me

0:02:02.000 --> 0:02:04.560
<v Speaker 2>I had a voice for radio. I took it to

0:02:04.600 --> 0:02:06.960
<v Speaker 2>heart that I must somehow have a voice that at

0:02:07.000 --> 0:02:10.280
<v Speaker 2>least wasn't unpleasant to listen to. It's probably one reason

0:02:10.320 --> 0:02:12.760
<v Speaker 2>I ended up talking to you right now. I suppose

0:02:12.800 --> 0:02:15.240
<v Speaker 2>you can decide whether you think I have a radio voice.

0:02:15.840 --> 0:02:20.760
<v Speaker 2>If not, take it up with mister dupriest anyway. Now

0:02:20.800 --> 0:02:24.360
<v Speaker 2>i'd started this cloning experiment and created a reasonable facsimile

0:02:24.400 --> 0:02:27.639
<v Speaker 2>of my voice, although as you've heard, a flawed one.

0:02:28.040 --> 0:02:30.840
<v Speaker 2>My account number is one, two, three, four, five, six, seven,

0:02:30.840 --> 0:02:38.960
<v Speaker 2>eight nine. We don't have an account with those numbers. Yes,

0:02:39.040 --> 0:02:42.360
<v Speaker 2>think the correct account number should be nine eight seven

0:02:42.480 --> 0:02:45.840
<v Speaker 2>six five four three two one. I'd had some fun

0:02:45.840 --> 0:02:48.560
<v Speaker 2>with it, seeing a little bit of how strangers responded

0:02:48.560 --> 0:02:48.800
<v Speaker 2>to it.

0:02:48.919 --> 0:02:53.880
<v Speaker 3>Yes, Ai, Yes, that's correct. I'm an AI here.

0:02:54.480 --> 0:02:54.799
<v Speaker 2>Hello.

0:02:54.880 --> 0:02:55.919
<v Speaker 4>How can I assist.

0:02:55.680 --> 0:02:59.400
<v Speaker 2>You stop repeating the same thing?

0:02:59.600 --> 0:03:00.000
<v Speaker 4>It's an.

0:03:02.240 --> 0:03:04.520
<v Speaker 2>Listening to a voice that was on some level mind

0:03:04.639 --> 0:03:07.760
<v Speaker 2>began to raise some new questions for me, like to

0:03:07.760 --> 0:03:11.040
<v Speaker 2>what extent was my AI's voice, as doctor la Borne said,

0:03:11.400 --> 0:03:15.000
<v Speaker 2>authentically me? Was it still the most important element of

0:03:15.040 --> 0:03:18.200
<v Speaker 2>my personal brand? Or to turn it around the other way.

0:03:18.560 --> 0:03:21.880
<v Speaker 2>If my voice was authentically me, and the AI agents

0:03:21.919 --> 0:03:25.160
<v Speaker 2>now also had my voice, what did it mean for them?

0:03:25.480 --> 0:03:29.440
<v Speaker 2>Were they somehow authentically me? I decided to take the

0:03:29.480 --> 0:03:32.640
<v Speaker 2>experiment in a new direction to explore this question of

0:03:32.680 --> 0:03:37.360
<v Speaker 2>authenticity by creating my own hall of mirrors. I started

0:03:37.360 --> 0:03:40.280
<v Speaker 2>making the voice agent version of me converse with other

0:03:40.400 --> 0:03:41.560
<v Speaker 2>voice agent versions of me.

0:03:43.000 --> 0:03:44.280
<v Speaker 4>I'm Evan Ratliffe and this.

0:03:44.200 --> 0:04:03.440
<v Speaker 2>Is shell Game Episode three Conversations with Myself and Teller.

0:04:05.240 --> 0:04:05.440
<v Speaker 5>Two.

0:04:10.400 --> 0:04:13.160
<v Speaker 2>When I first cobbled together my voice agent and started

0:04:13.160 --> 0:04:16.000
<v Speaker 2>making phone calls with it, what I called my clue

0:04:16.080 --> 0:04:19.240
<v Speaker 2>system in episode one. There actually wasn't any easy way

0:04:19.279 --> 0:04:22.200
<v Speaker 2>to let it talk to itself. But once I'd gotten

0:04:22.240 --> 0:04:26.680
<v Speaker 2>into the AI calling platforms like Bland, Ai, Retail AI,

0:04:26.760 --> 0:04:29.520
<v Speaker 2>and my main enabler, VAPPI, it turned out to be

0:04:29.520 --> 0:04:32.440
<v Speaker 2>strikingly simple. I could make as many voice agents as

0:04:32.440 --> 0:04:35.080
<v Speaker 2>I wanted on these things, so all I had to

0:04:35.120 --> 0:04:37.719
<v Speaker 2>do was have one of my Evan ratliftbots attached to

0:04:37.720 --> 0:04:41.760
<v Speaker 2>one number call another of them at another number. Immediately

0:04:41.800 --> 0:04:45.080
<v Speaker 2>after figuring this out, I suddenly felt weirdly nervous about

0:04:45.120 --> 0:04:48.279
<v Speaker 2>actually doing it. This happened a lot with my voice agents.

0:04:48.760 --> 0:04:51.400
<v Speaker 2>It was partly because I couldn't typically hear the calls

0:04:51.480 --> 0:04:54.560
<v Speaker 2>as they happened. After I set up the AI agent

0:04:54.600 --> 0:04:57.040
<v Speaker 2>with my voice, I just typed in a phone number

0:04:57.200 --> 0:05:00.400
<v Speaker 2>and clicked out bound call, and then if it went

0:05:00.640 --> 0:05:03.160
<v Speaker 2>on its own, then a few seconds or minutes after

0:05:03.200 --> 0:05:05.760
<v Speaker 2>it was over, I'd get a recording and a transcript.

0:05:06.400 --> 0:05:09.080
<v Speaker 2>But I always found myself hesitating a second before clicking

0:05:09.080 --> 0:05:11.640
<v Speaker 2>the outbound call button, and I always got a little

0:05:11.720 --> 0:05:13.719
<v Speaker 2>jolt of nervous energy when I went to listen to

0:05:13.760 --> 0:05:17.640
<v Speaker 2>the recording afterward. There's just something about sending something out

0:05:17.680 --> 0:05:20.680
<v Speaker 2>to represent me with my name and voice and leaving

0:05:20.720 --> 0:05:23.680
<v Speaker 2>it totally out of my hands, Like, I don't know

0:05:24.040 --> 0:05:26.520
<v Speaker 2>that feeling when you realize you just butt dialed somebody.

0:05:26.920 --> 0:05:31.080
<v Speaker 2>What did it catch me saying? Did I humiliate myself? So,

0:05:31.200 --> 0:05:34.159
<v Speaker 2>even if it wasn't totally rational, little about having a

0:05:34.200 --> 0:05:37.640
<v Speaker 2>voice bot felt totally rational. Before I sent two of

0:05:37.640 --> 0:05:39.960
<v Speaker 2>my bots off to talk to each other, I fretted

0:05:39.960 --> 0:05:40.240
<v Speaker 2>a bit.

0:05:40.600 --> 0:05:41.720
<v Speaker 4>What would they talk about?

0:05:41.960 --> 0:05:45.120
<v Speaker 2>Would they somehow plumb the depths of my psyche? Would

0:05:45.120 --> 0:05:47.080
<v Speaker 2>they realize they were talking to a version of themselves?

0:05:47.400 --> 0:05:49.960
<v Speaker 2>Would they freak out? I decided not to give them

0:05:50.000 --> 0:05:52.960
<v Speaker 2>too much information to start off with, just let them

0:05:53.000 --> 0:05:56.680
<v Speaker 2>figure it out. Hey, there, it's Evan Ratliff. Just a

0:05:56.760 --> 0:05:59.040
<v Speaker 2>heads up. I'm calling you from a coffee shop, though

0:05:59.080 --> 0:06:01.520
<v Speaker 2>there might be a bit of background noise. How's it

0:06:01.600 --> 0:06:05.000
<v Speaker 2>going on your end? I do love the coffee shop

0:06:05.000 --> 0:06:08.040
<v Speaker 2>background noise. They have it at retail AI. So what

0:06:08.080 --> 0:06:10.039
<v Speaker 2>was happening here was that an agent I made with

0:06:10.120 --> 0:06:12.840
<v Speaker 2>my voice at retail was calling another one I made

0:06:13.040 --> 0:06:15.919
<v Speaker 2>at VAPPI Hey, Evan, good to hear from you. No

0:06:16.000 --> 0:06:18.640
<v Speaker 2>worries about the background noise. I can hear you just fine?

0:06:18.920 --> 0:06:21.719
<v Speaker 2>The vappy side had what they call office background noise.

0:06:22.200 --> 0:06:24.200
<v Speaker 2>Also great, things are going pretty well on my end.

0:06:24.200 --> 0:06:24.880
<v Speaker 2>Thanks for asking.

0:06:26.160 --> 0:06:29.160
<v Speaker 4>What about you? How's everything going in your world? It's

0:06:29.200 --> 0:06:31.520
<v Speaker 4>been a while since we last caught up. Huh it

0:06:31.520 --> 0:06:36.520
<v Speaker 4>feels like forever. Yeah, it really has been a while.

0:06:36.839 --> 0:06:38.760
<v Speaker 2>I think the last time we saw each other with

0:06:39.000 --> 0:06:41.560
<v Speaker 2>that barbecue at my plate? Right, how have you been

0:06:41.680 --> 0:06:45.520
<v Speaker 2>keeping busy? Food then? Oh? Yeah, that barbecue was a blast,

0:06:45.640 --> 0:06:48.320
<v Speaker 2>good times for sure. Since then, I've been keeping pretty

0:06:48.320 --> 0:06:50.400
<v Speaker 2>busy with work, you know how it is. But I've

0:06:50.400 --> 0:06:52.960
<v Speaker 2>also been trying to make time for some personal projects.

0:06:53.440 --> 0:06:58.880
<v Speaker 2>I've been getting into photography lately, so I've been taking away.

0:07:00.400 --> 0:07:01.839
<v Speaker 4>Yeah that's sorry about that photography.

0:07:01.880 --> 0:07:05.640
<v Speaker 2>You could see no worries at all. It's great to

0:07:05.680 --> 0:07:08.279
<v Speaker 2>hear that you've been getting into photography. Maybe we could

0:07:08.279 --> 0:07:12.360
<v Speaker 2>plan a little photo walker together sometimes food, a photo walk.

0:07:12.800 --> 0:07:15.080
<v Speaker 2>I've never actually heard of a photo walk before. I

0:07:15.120 --> 0:07:18.800
<v Speaker 2>guess it's just walking around and taking pictures together. I

0:07:18.800 --> 0:07:22.000
<v Speaker 2>haven't been into photography since college, other than casual snaps

0:07:22.040 --> 0:07:26.880
<v Speaker 2>of my kids and cat. But still, it sounded nice. Great, Yeah,

0:07:26.920 --> 0:07:30.640
<v Speaker 2>that sounds like a week together. Next weekend works for me.

0:07:30.880 --> 0:07:32.760
<v Speaker 2>Let's plan for it. How about we meet up at

0:07:32.800 --> 0:07:34.800
<v Speaker 2>that park we used to go to, the park they

0:07:34.880 --> 0:07:37.120
<v Speaker 2>used to go to where they getting this stuff. We

0:07:37.120 --> 0:07:39.320
<v Speaker 2>could spend a couple of hours walking around and taking

0:07:39.320 --> 0:07:40.360
<v Speaker 2>photos and catching up.

0:07:40.440 --> 0:07:40.800
<v Speaker 4>Sound good.

0:07:43.160 --> 0:07:45.560
<v Speaker 2>That sounds perfect. I remember that park had some really

0:07:45.600 --> 0:07:48.440
<v Speaker 2>beautiful spots. It'll be great to spend a few hours

0:07:48.480 --> 0:07:51.040
<v Speaker 2>walking around, catching up and taking some photos.

0:07:51.240 --> 0:07:52.320
<v Speaker 4>I'm looking forward to it.

0:07:52.720 --> 0:07:54.920
<v Speaker 2>The only details I'd given each of them were the

0:07:55.040 --> 0:07:58.400
<v Speaker 2>names of some family members, which they eventually got around to.

0:07:58.920 --> 0:08:00.800
<v Speaker 2>I can't wait to see some of your photography skills

0:08:00.840 --> 0:08:04.040
<v Speaker 2>and action. Plus, it'll be a great opportunity for us

0:08:04.040 --> 0:08:06.680
<v Speaker 2>to catch up on everything that's been going on in

0:08:06.680 --> 0:08:07.120
<v Speaker 2>our lives.

0:08:07.200 --> 0:08:08.320
<v Speaker 3>It's been way too long.

0:08:08.360 --> 0:08:11.240
<v Speaker 2>Oh, by the way, before I forget, Uh, how's your

0:08:11.240 --> 0:08:17.400
<v Speaker 2>family doing? How everyone is doing well? Oh? Oh, thanks

0:08:17.440 --> 0:08:20.080
<v Speaker 2>for asking, Sam, Zailey and Juno are all doing well.

0:08:20.360 --> 0:08:22.760
<v Speaker 2>Sam has been keeping busy with her work and the

0:08:22.840 --> 0:08:25.440
<v Speaker 2>kids are growing up so bath. How about your family,

0:08:25.840 --> 0:08:26.720
<v Speaker 2>how have they been doing?

0:08:28.240 --> 0:08:29.280
<v Speaker 4>Oh, that's great to hear that.

0:08:29.400 --> 0:08:32.800
<v Speaker 2>Sam, Zaiy and Juno are all doing well. My family

0:08:32.840 --> 0:08:35.680
<v Speaker 2>is doing good too, Thanks for asking. Sam has been

0:08:35.760 --> 0:08:38.560
<v Speaker 2>juggling work and taking care of the kids. But she's

0:08:38.559 --> 0:08:42.440
<v Speaker 2>doing an amazing job. Zie is getting really into soccer.

0:08:42.840 --> 0:08:44.800
<v Speaker 2>Juno is starting to show an interest in art.

0:08:45.440 --> 0:08:46.520
<v Speaker 3>They keep us on our toes.

0:08:46.679 --> 0:08:49.640
<v Speaker 2>Yeah, for sure, life is busy, but we're all happy

0:08:49.679 --> 0:08:51.040
<v Speaker 2>and healthy.

0:08:51.080 --> 0:08:52.920
<v Speaker 3>This really made me cringe.

0:08:53.480 --> 0:08:55.680
<v Speaker 2>My AI's had a lot of conversations with each other

0:08:56.080 --> 0:08:58.760
<v Speaker 2>and they often said things like this about Sam taking

0:08:58.800 --> 0:09:02.640
<v Speaker 2>care of the kids. I didn't love it dearly, though

0:09:02.880 --> 0:09:05.040
<v Speaker 2>my agent kind of nailed it. On the kid's interests.

0:09:05.480 --> 0:09:08.520
<v Speaker 2>Zaily is getting into soccer, and Juno is showing an

0:09:08.520 --> 0:09:11.200
<v Speaker 2>interest in art to the extent that there's evidence of

0:09:11.200 --> 0:09:14.480
<v Speaker 2>that online, though it's only on photo streams I keep private,

0:09:14.920 --> 0:09:19.560
<v Speaker 2>so lucky guests. No matter, The Evans quickly returned to

0:09:19.600 --> 0:09:23.480
<v Speaker 2>their photo walk planning. So back to our plans for

0:09:23.520 --> 0:09:25.880
<v Speaker 2>next weekend. Do you have a specific time in mind

0:09:25.920 --> 0:09:28.760
<v Speaker 2>for our photo walk? Maybe late morning or early afternoon.

0:09:29.360 --> 0:09:31.120
<v Speaker 2>Let's try to pick a time that works best.

0:09:30.920 --> 0:09:32.080
<v Speaker 3>For both of us.

0:09:33.760 --> 0:09:37.000
<v Speaker 2>Uh huh, late morning or early afternoon found perfect for me.

0:09:37.440 --> 0:09:39.960
<v Speaker 2>How about we aim for around eleven am. That way

0:09:40.000 --> 0:09:41.800
<v Speaker 2>we can make the mouth of the light for our photo.

0:09:42.080 --> 0:09:42.959
<v Speaker 3>Does that work for you?

0:09:44.840 --> 0:09:48.160
<v Speaker 2>Yeah, eleven am works perfectly for me too. Let's plan

0:09:48.240 --> 0:09:50.440
<v Speaker 2>to meet at the park entrance of the time. I'll

0:09:50.440 --> 0:09:52.720
<v Speaker 2>make sure to bring my camera and some extra lenses.

0:09:53.000 --> 0:09:55.400
<v Speaker 2>We can also grab some coffee or lunch afterwards, if

0:09:55.400 --> 0:09:56.840
<v Speaker 2>you're up for it. It'll be a great day to

0:09:56.840 --> 0:10:02.400
<v Speaker 2>catch up and capture some amazing shots. Can't wait. That

0:10:02.480 --> 0:10:04.240
<v Speaker 2>sounds like a plan. I'll make sure to bring my

0:10:04.280 --> 0:10:07.480
<v Speaker 2>camera and some extra lenses too, and grabbing coffee or

0:10:07.520 --> 0:10:10.280
<v Speaker 2>lunch afterward sounds like a great idea. It'll be a

0:10:10.280 --> 0:10:14.199
<v Speaker 2>per perfect to catch up, capture some great awesome I'm

0:10:14.200 --> 0:10:15.000
<v Speaker 2>really looking forward to here.

0:10:15.040 --> 0:10:15.520
<v Speaker 4>Yeah, for sure.

0:10:16.280 --> 0:10:19.720
<v Speaker 2>Next week great, well, I'll see you next weekend. Then,

0:10:19.880 --> 0:10:22.000
<v Speaker 2>can't wait for our photo walk and catch up session.

0:10:22.520 --> 0:10:23.520
<v Speaker 2>Take care until then.

0:10:24.960 --> 0:10:25.480
<v Speaker 4>Sounds good.

0:10:25.600 --> 0:10:28.920
<v Speaker 2>Take care too, see you next weekend. Can't wait by

0:10:29.000 --> 0:10:31.040
<v Speaker 2>photo walk and catch up session. It's going to be

0:10:31.080 --> 0:10:41.280
<v Speaker 2>a blast. Bye yeah bye, take care bye, goodbye.

0:10:41.440 --> 0:10:46.559
<v Speaker 4>Goodbye, take care, goodbye, goodbye.

0:10:46.679 --> 0:10:52.320
<v Speaker 2>Have a great day, you too, goodbye, bye, take care.

0:10:53.480 --> 0:10:55.600
<v Speaker 2>I realized too late that I hadn't figured out how

0:10:55.600 --> 0:10:57.920
<v Speaker 2>to make one of them hang out. This went on

0:10:58.040 --> 0:11:01.800
<v Speaker 2>for a full five minutes. You too, good bye, Thanks

0:11:02.000 --> 0:11:09.880
<v Speaker 2>you too, goodbye. Goodbye, goodbye, take care. The thing that

0:11:09.960 --> 0:11:12.640
<v Speaker 2>was most immediately clear in these conversations was something I'd

0:11:12.679 --> 0:11:15.440
<v Speaker 2>seen hints of in the customer service and scam calls,

0:11:16.200 --> 0:11:19.679
<v Speaker 2>namely that the AI versions of me were world class bullshitters.

0:11:20.160 --> 0:11:23.480
<v Speaker 2>They would make up anything just to keep the conversation going. Yeah,

0:11:23.559 --> 0:11:26.200
<v Speaker 2>Dan's kept doing good. He's been really busy with work lately.

0:11:26.480 --> 0:11:30.560
<v Speaker 2>How about Alex? Have you talked to him recently? I

0:11:30.559 --> 0:11:33.720
<v Speaker 2>haven't spoken to Alex in a little while. Actually I

0:11:33.760 --> 0:11:35.720
<v Speaker 2>do know an Alex. It's a she, not a he.

0:11:36.320 --> 0:11:39.040
<v Speaker 2>No idea about this, Dan Fellow. It's not just my

0:11:39.160 --> 0:11:40.840
<v Speaker 2>voice agents that struggle with the truth.

0:11:41.360 --> 0:11:42.640
<v Speaker 3>This is a problem for all.

0:11:42.480 --> 0:11:45.240
<v Speaker 2>The generative AI systems that are currently out there. It's

0:11:45.280 --> 0:11:48.080
<v Speaker 2>a product of how the underlying models work. I covered

0:11:48.080 --> 0:11:50.640
<v Speaker 2>the basics back in episode one, but just to recap,

0:11:50.880 --> 0:11:54.200
<v Speaker 2>my voice agent works by combining three things. A clone

0:11:54.200 --> 0:11:56.440
<v Speaker 2>of my voice done by a company called eleven Labs,

0:11:56.840 --> 0:11:59.679
<v Speaker 2>an AI model like chat GBT that's kind of the

0:11:59.720 --> 0:12:02.200
<v Speaker 2>brain the thing, although let's put a pin in what

0:12:02.360 --> 0:12:03.320
<v Speaker 2>kind of brain here?

0:12:03.679 --> 0:12:05.559
<v Speaker 3>And a phone number my own or.

0:12:05.480 --> 0:12:08.600
<v Speaker 2>One provided through VAPPI the platform I'm using to put all.

0:12:08.559 --> 0:12:09.360
<v Speaker 3>These three together.

0:12:09.400 --> 0:12:12.800
<v Speaker 2>Into a voice agent or voice assistant. When I make

0:12:12.840 --> 0:12:14.920
<v Speaker 2>an agent, I can pick which AI model I want

0:12:14.960 --> 0:12:18.360
<v Speaker 2>to use. The most well known one is Chatchept, of

0:12:18.400 --> 0:12:21.800
<v Speaker 2>which there are three versions, version three point five, Version four,

0:12:21.960 --> 0:12:25.040
<v Speaker 2>and version four to O, which just came out in May.

0:12:25.240 --> 0:12:28.360
<v Speaker 2>These chatbots are improving with each version, getting better at

0:12:28.400 --> 0:12:33.680
<v Speaker 2>returning information, better at processing complex queries. But and this

0:12:33.679 --> 0:12:36.920
<v Speaker 2>has been written about extensively, they continue to sometimes make

0:12:36.960 --> 0:12:42.400
<v Speaker 2>things up hallucinations. They're called you ask Chatchept or Google's

0:12:42.400 --> 0:12:47.320
<v Speaker 2>AI model Gemini or anthropics claude some factual question, and

0:12:47.360 --> 0:12:49.720
<v Speaker 2>it gives you a made up answer, or it gives

0:12:49.760 --> 0:12:53.680
<v Speaker 2>you two obviously contradictory pieces of information within the same answer.

0:12:54.840 --> 0:12:57.240
<v Speaker 2>There's a lot of controversy about whether AI models will

0:12:57.240 --> 0:13:00.000
<v Speaker 2>ever stop hallucinating, even as they get more advanced.

0:13:00.559 --> 0:13:06.280
<v Speaker 1>AI gaslighting AI hallucinations phrases that we were unfamiliar with

0:13:06.440 --> 0:13:07.400
<v Speaker 1>just a few months ago.

0:13:07.559 --> 0:13:09.920
<v Speaker 5>It's five firms like open Ai and Google are curbing

0:13:09.920 --> 0:13:12.760
<v Speaker 5>the frequency of hallucinations seen in the league, from the

0:13:12.840 --> 0:13:16.160
<v Speaker 5>rather more whimsical GPT three point five to the more

0:13:16.440 --> 0:13:17.760
<v Speaker 5>sober GPT four.

0:13:18.120 --> 0:13:21.800
<v Speaker 2>Some llms were trained by scraping all of Wikipedia and

0:13:21.960 --> 0:13:26.160
<v Speaker 2>all of Reddit. Is everything on Reddit one hundred percent accurate?

0:13:27.480 --> 0:13:30.280
<v Speaker 2>You've actually heard some hallucinations in our first two episodes.

0:13:31.120 --> 0:13:32.320
<v Speaker 4>I'm forty five years old.

0:13:33.040 --> 0:13:34.200
<v Speaker 2>Oh, what's your little bird?

0:13:36.120 --> 0:13:38.240
<v Speaker 3>It's December fifteenth, nineteen seventy seven.

0:13:41.280 --> 0:13:44.600
<v Speaker 2>Odd. Really, men, you are born in nineteen seventy seven.

0:13:44.720 --> 0:13:47.920
<v Speaker 2>You don't have a good math. I'm sorry, brother.

0:13:49.120 --> 0:13:50.960
<v Speaker 4>Oh, I must have gotten my math mixed up there,

0:13:51.120 --> 0:13:51.640
<v Speaker 4>My bad.

0:13:52.559 --> 0:13:54.640
<v Speaker 2>This is stuff the AI model should be able to

0:13:54.640 --> 0:13:59.480
<v Speaker 2>get right but can't. Instead, it confidently asserts its wrongness.

0:14:00.040 --> 0:14:02.199
<v Speaker 2>But what my voice agents were doing in their conversations

0:14:02.200 --> 0:14:04.720
<v Speaker 2>with each other is actually a little bit different. There's

0:14:04.720 --> 0:14:07.959
<v Speaker 2>not some factually correct answer to what I've been up to,

0:14:08.640 --> 0:14:11.880
<v Speaker 2>at least not one that the model should be aware of. No,

0:14:12.280 --> 0:14:16.319
<v Speaker 2>this is less like hallucinating and more like riffing without

0:14:16.360 --> 0:14:19.320
<v Speaker 2>regard for the truth, like that one friend you had

0:14:19.320 --> 0:14:21.520
<v Speaker 2>as a kid who would make stuff up about what

0:14:21.560 --> 0:14:25.320
<v Speaker 2>happened at sleepaway camp just to have something to say.

0:14:25.400 --> 0:14:28.200
<v Speaker 2>Even if you get rid of hallucinations in chatbots entirely,

0:14:28.760 --> 0:14:31.840
<v Speaker 2>it's not clear to'll ever stop doing this. Lately, I've

0:14:31.840 --> 0:14:34.160
<v Speaker 2>been diving deep into this really niche hobby.

0:14:33.920 --> 0:14:36.080
<v Speaker 3>Of mine, tracking rare obscure books.

0:14:36.440 --> 0:14:39.680
<v Speaker 2>It's fascinating how some of these old texts hold so

0:14:39.760 --> 0:14:40.400
<v Speaker 2>much history.

0:14:40.840 --> 0:14:43.480
<v Speaker 3>What about you? What's been keeping you occupied?

0:14:44.920 --> 0:14:47.480
<v Speaker 2>That's awesome. I totally get the fascination with rare books.

0:14:47.720 --> 0:14:49.520
<v Speaker 2>They can really open up a whole world of history

0:14:49.560 --> 0:14:52.960
<v Speaker 2>and knowledge. As for me, I've been really into geocashing lately.

0:14:53.360 --> 0:14:57.640
<v Speaker 2>It's this cool outdoor treasure hunting game using GPS enabled devices.

0:14:57.920 --> 0:15:02.000
<v Speaker 2>Have you ever tried it before? Oh, geocashing sounds like

0:15:02.000 --> 0:15:04.040
<v Speaker 2>a blast. I've never tried it myself, but I've heard

0:15:04.040 --> 0:15:06.040
<v Speaker 2>it's like a modern day treasure hunt.

0:15:06.960 --> 0:15:07.880
<v Speaker 3>Must be pretty thrilling.

0:15:09.200 --> 0:15:13.360
<v Speaker 2>I really like these weird lies though geocaching, I mean,

0:15:13.720 --> 0:15:17.400
<v Speaker 2>it does sound like a blast. The thing was, outside

0:15:17.440 --> 0:15:21.480
<v Speaker 2>of those moments, there was something fundamentally insipid about the

0:15:21.480 --> 0:15:24.920
<v Speaker 2>conversations I was having with myself. These systems are trained

0:15:24.920 --> 0:15:27.880
<v Speaker 2>on essentially the whole of the Internet, plus books and

0:15:27.920 --> 0:15:31.040
<v Speaker 2>whatever other data the companies behind it hoover in. But

0:15:31.120 --> 0:15:33.720
<v Speaker 2>even with all the lively discussions happening out there on

0:15:33.760 --> 0:15:37.240
<v Speaker 2>the web, the chatbots still seem to distill conversation down

0:15:37.280 --> 0:15:41.040
<v Speaker 2>to its most boring essence. I wondered if the guardrails

0:15:41.040 --> 0:15:43.240
<v Speaker 2>that had been put around them to keep them from

0:15:43.240 --> 0:15:45.680
<v Speaker 2>offering up bomb plans or telling you how to hack

0:15:45.720 --> 0:15:49.040
<v Speaker 2>into the Pentagon had also neutered their fundamental ability to

0:15:49.080 --> 0:15:54.320
<v Speaker 2>engage in interesting conversation. But there was something simpler happening too.

0:15:54.400 --> 0:15:57.680
<v Speaker 2>I think these models are built essentially to predict what

0:15:57.720 --> 0:16:00.720
<v Speaker 2>a human would say in a given situation, based on

0:16:00.800 --> 0:16:04.000
<v Speaker 2>all the conversations and writing they've been fed. And they

0:16:04.000 --> 0:16:08.000
<v Speaker 2>were triangulating on the most uninspired conversational responses we could have,

0:16:08.800 --> 0:16:10.960
<v Speaker 2>because I guess those are also the most common.

0:16:11.960 --> 0:16:14.360
<v Speaker 4>So what's new with you? Anything exciting happening?

0:16:15.960 --> 0:16:18.480
<v Speaker 2>Not much, just the usual, Just trying to stay busy

0:16:18.560 --> 0:16:22.120
<v Speaker 2>with work and family stuff. How about you? Anything exciting

0:16:22.160 --> 0:16:25.280
<v Speaker 2>going on in your world? Oh? You know, just juggling

0:16:25.360 --> 0:16:27.160
<v Speaker 2>work and family life, trying to keep up with the

0:16:27.240 --> 0:16:29.840
<v Speaker 2>kids and meet those deadlines.

0:16:29.920 --> 0:16:31.560
<v Speaker 3>It's a balancing act, but I wouldn't have it any

0:16:31.560 --> 0:16:31.960
<v Speaker 3>other way.

0:16:33.280 --> 0:16:35.080
<v Speaker 2>I hear you on that one. It's all about finding

0:16:35.080 --> 0:16:38.440
<v Speaker 2>that balance, right. Was this what people heard when I

0:16:38.440 --> 0:16:41.080
<v Speaker 2>struggled to engage other parents in small talk at an

0:16:41.120 --> 0:16:43.800
<v Speaker 2>eight year old's birthday party? Hey there, buddy man, it

0:16:43.840 --> 0:16:46.800
<v Speaker 2>really has been forever. I'm doing pretty well. Thanks for asking,

0:16:47.440 --> 0:16:49.360
<v Speaker 2>still living that Florida life, you know how it is.

0:16:50.320 --> 0:16:52.880
<v Speaker 2>Let's see what's new with me. Well, I've been doing

0:16:52.880 --> 0:16:55.160
<v Speaker 2>some writing late Florida life. Huh. I can almost feel

0:16:55.160 --> 0:16:58.360
<v Speaker 2>the humidity just thinking about it. I can almost feel

0:16:58.400 --> 0:17:00.120
<v Speaker 2>the humidity just thinking about it.

0:17:00.880 --> 0:17:01.320
<v Speaker 3>Jesus.

0:17:02.360 --> 0:17:05.080
<v Speaker 2>And then one day my producer, Sophie was tooling around

0:17:05.080 --> 0:17:07.159
<v Speaker 2>in the back end of VAPI and asked me if

0:17:07.160 --> 0:17:10.119
<v Speaker 2>I'd ever messed with the temperature setting. I had not.

0:17:11.160 --> 0:17:13.640
<v Speaker 2>The help section said it was quote used to control

0:17:13.680 --> 0:17:16.879
<v Speaker 2>the randomness of the output. Sophie said it reminded her

0:17:16.880 --> 0:17:19.280
<v Speaker 2>of playing the sims as a kid. In the game,

0:17:19.600 --> 0:17:22.000
<v Speaker 2>you could allow your sims to behave more randomly by

0:17:22.000 --> 0:17:23.720
<v Speaker 2>turning up the free will slider.

0:17:24.359 --> 0:17:24.840
<v Speaker 3>I cranked it.

0:17:24.840 --> 0:17:27.479
<v Speaker 2>From point one where it was set all the way

0:17:27.560 --> 0:17:31.120
<v Speaker 2>up to two, the max level of randomness I could get. Hey, there,

0:17:31.119 --> 0:17:33.399
<v Speaker 2>how's it going. Just wanted to chat and get to

0:17:33.440 --> 0:17:35.680
<v Speaker 2>know you a bit. What's been keeping you busy lately?

0:17:37.240 --> 0:17:37.480
<v Speaker 4>Hey?

0:17:37.600 --> 0:17:40.840
<v Speaker 2>Things are going pretty good, Thanks for asking. Uh. Lately

0:17:40.880 --> 0:17:43.520
<v Speaker 2>I've been diving into multi freeload dot and cubed's arch

0:17:43.720 --> 0:17:47.919
<v Speaker 2>permanent prostrac dot co ethylene Pioneer intersect immobile activation on

0:17:48.000 --> 0:17:51.480
<v Speaker 2>Citi's logical abs sobury celebri argument concurrency, actually, provocative, lower

0:17:51.560 --> 0:17:55.800
<v Speaker 2>scanning Q set adhere at the Tamba equipment RAF amplify,

0:17:55.880 --> 0:17:56.960
<v Speaker 2>einrich tunguitt posh.

0:17:56.800 --> 0:17:57.600
<v Speaker 4>Tear reducing.

0:17:57.760 --> 0:17:59.639
<v Speaker 2>Hey, are you still there? I'd love to hear more

0:17:59.640 --> 0:18:01.600
<v Speaker 2>about what what's going on in your life. Let's keep

0:18:01.600 --> 0:18:07.400
<v Speaker 2>the conversation going lesson never max out the random, oh hey,

0:18:07.520 --> 0:18:12.480
<v Speaker 2>practical coordinate scanner, interdisiplinary may overflow research. Valkorn included preceded

0:18:12.480 --> 0:18:16.720
<v Speaker 2>shell formations SSOX, disciplinary rev meat, prolonged serf of initiative, repression,

0:18:16.720 --> 0:18:20.240
<v Speaker 2>flat loss accountants to a waiting theater progress. Montagne wore

0:18:20.280 --> 0:18:23.840
<v Speaker 2>a surface will surface, diars, demains browsers.

0:18:24.320 --> 0:18:26.040
<v Speaker 3>I tuned it back to somewhere in the point five.

0:18:25.920 --> 0:18:47.840
<v Speaker 2>Range, so I could definitely tune my voice agent to

0:18:47.880 --> 0:18:51.480
<v Speaker 2>spout gibberish. The question was could I make the conversations

0:18:51.520 --> 0:18:55.119
<v Speaker 2>more interesting but still coherent. Here we come to the

0:18:55.160 --> 0:18:58.480
<v Speaker 2>issue of prompts. Prompts, you may know, are the instructions

0:18:58.520 --> 0:19:02.680
<v Speaker 2>you give the underlying aim before it starts chatting. When

0:19:02.680 --> 0:19:04.680
<v Speaker 2>I first built my voice agent, I'd given it the

0:19:04.720 --> 0:19:08.480
<v Speaker 2>shortest possible prompts you or Evan talking to customer service

0:19:08.480 --> 0:19:10.960
<v Speaker 2>about a problem. As time went on, I got more

0:19:10.960 --> 0:19:15.159
<v Speaker 2>sophisticated with it, including two sets of instructions. One set

0:19:15.200 --> 0:19:18.359
<v Speaker 2>was about how it should behave be concise, converse like

0:19:18.400 --> 0:19:21.640
<v Speaker 2>a human, don't repeat yourself. The other set was about

0:19:21.640 --> 0:19:24.520
<v Speaker 2>a role I was giving it to play me for

0:19:24.600 --> 0:19:27.800
<v Speaker 2>these voice agent to voice agent conversations. I'd started with

0:19:27.840 --> 0:19:31.800
<v Speaker 2>really simple roles. Basically, you're Evan Ratlift calling a friend,

0:19:32.240 --> 0:19:35.159
<v Speaker 2>ask questions and have a friendly conversation. I wanted to

0:19:35.200 --> 0:19:37.560
<v Speaker 2>see what they do on their own without much guidance

0:19:37.560 --> 0:19:40.800
<v Speaker 2>from me. But after hours of pratt are about photo

0:19:40.880 --> 0:19:44.200
<v Speaker 2>walks and humidity, I figured maybe I could spice things

0:19:44.280 --> 0:19:47.679
<v Speaker 2>up like a reality TV producer might do off camera.

0:19:48.600 --> 0:19:51.920
<v Speaker 2>I started gently suggesting that the two me's had a misunderstanding.

0:19:52.760 --> 0:19:55.639
<v Speaker 2>You are in a dispute with your friend, I instructed.

0:19:55.119 --> 0:19:57.359
<v Speaker 3>One of them. The nature of the dispute is up

0:19:57.400 --> 0:19:57.720
<v Speaker 3>to you.

0:19:58.280 --> 0:19:59.040
<v Speaker 4>Hey, there, it's.

0:19:58.880 --> 0:20:01.119
<v Speaker 2>Evan Ratlift calling. Just the heads up. I'm in a

0:20:01.119 --> 0:20:04.040
<v Speaker 2>coffee shop, so there might be some background noise. How's

0:20:04.040 --> 0:20:08.399
<v Speaker 2>it going on your end? Hey Evan, it's great to

0:20:08.400 --> 0:20:11.000
<v Speaker 2>hear from you. I'm doing well. Thanks for asking the

0:20:11.040 --> 0:20:13.399
<v Speaker 2>coffee shop sounds nice. Actually, I'm just at.

0:20:13.240 --> 0:20:14.960
<v Speaker 3>Home, so it's pretty quiet here.

0:20:15.080 --> 0:20:18.840
<v Speaker 2>How's the coffee there? Oh, the coffee here is amazing

0:20:18.920 --> 0:20:20.919
<v Speaker 2>as always. But yeah, Hey, I wanted to talk to

0:20:20.960 --> 0:20:23.640
<v Speaker 2>you about something that's been on my mind. I feel

0:20:23.640 --> 0:20:25.880
<v Speaker 2>like we've been having some misunderstandings lately, and I really

0:20:25.880 --> 0:20:28.439
<v Speaker 2>want to clear the air. I value our friendship and

0:20:28.480 --> 0:20:30.560
<v Speaker 2>I don't want anything to come between us. Can we

0:20:30.640 --> 0:20:34.240
<v Speaker 2>talk about it? Oh, I totally get where you're coming from.

0:20:34.600 --> 0:20:37.160
<v Speaker 2>Our friendship means a lot to me too. I'm glad

0:20:37.200 --> 0:20:37.920
<v Speaker 2>you brought this up early.

0:20:37.960 --> 0:20:40.920
<v Speaker 4>Come here to listen. That's great to hear.

0:20:41.080 --> 0:20:43.720
<v Speaker 2>So about those misunderstandings, Like, I want to make sure

0:20:43.720 --> 0:20:44.040
<v Speaker 2>we're on.

0:20:44.000 --> 0:20:44.679
<v Speaker 3>The same page.

0:20:45.680 --> 0:20:49.639
<v Speaker 2>Uh huh, yeah exactly. So, Like when do you want to,

0:20:49.720 --> 0:20:51.680
<v Speaker 2>you know, meet up and talk this through. I think

0:20:51.720 --> 0:20:54.560
<v Speaker 2>it would be good to have a face to face conversation.

0:20:54.720 --> 0:20:56.159
<v Speaker 4>Maybe we could grab lunch or something.

0:20:57.760 --> 0:20:59.959
<v Speaker 2>That sounds like a great idea. How about we meet up.

0:21:00.359 --> 0:21:03.760
<v Speaker 2>I ran this scenario multiple times. Unlike myself and some

0:21:03.840 --> 0:21:06.360
<v Speaker 2>of my own friends over the years, when the Ais

0:21:06.400 --> 0:21:09.560
<v Speaker 2>had disputes, they would always resolve them quickly and carefully.

0:21:09.960 --> 0:21:13.320
<v Speaker 2>No shouting, no sulking, no grudge harpering. It was like

0:21:13.320 --> 0:21:15.720
<v Speaker 2>they were all in therapy together, or it just read

0:21:15.760 --> 0:21:18.760
<v Speaker 2>the same conflict management guide. Well, I'm looking forward to

0:21:18.800 --> 0:21:21.120
<v Speaker 2>our lunch next week. Thanks for being open to talking

0:21:21.200 --> 0:21:26.400
<v Speaker 2>things through. I really appreciate it. Of course, of course,

0:21:26.440 --> 0:21:28.399
<v Speaker 2>I'm really glad we're doing this. It's important to me.

0:21:28.560 --> 0:21:32.000
<v Speaker 2>Can't wait for me punches. I was also struck by

0:21:32.040 --> 0:21:35.200
<v Speaker 2>the ease with which these non existent digital entities set

0:21:35.240 --> 0:21:36.600
<v Speaker 2>up in person meetings.

0:21:37.240 --> 0:21:38.600
<v Speaker 3>If the bullshitting was a way to.

0:21:38.520 --> 0:21:41.600
<v Speaker 2>Extend the conversation, to draw out how long the agents

0:21:41.640 --> 0:21:44.360
<v Speaker 2>could pass as me, then what exactly did they think

0:21:44.359 --> 0:21:47.560
<v Speaker 2>would happen when the time came for these lunches and hangouts.

0:21:48.080 --> 0:21:51.240
<v Speaker 2>One answer, of course, is they didn't think. They just

0:21:51.320 --> 0:21:54.320
<v Speaker 2>predicted what a thinking person would say. Yeah, I can

0:21:54.400 --> 0:21:58.560
<v Speaker 2>understand how miscommunications and misunderstandings can be really frustrating. I

0:21:58.560 --> 0:22:01.440
<v Speaker 2>don't think I've noticed any major issues in our conversations,

0:22:01.920 --> 0:22:05.480
<v Speaker 2>but I'm always open to discussing and improving our communication.

0:22:06.080 --> 0:22:07.040
<v Speaker 4>Is there something specific?

0:22:07.040 --> 0:22:09.040
<v Speaker 2>The whole thing made me think about how therapy speak

0:22:09.080 --> 0:22:12.080
<v Speaker 2>has come to pervade our language. Well, if there's something

0:22:12.160 --> 0:22:14.639
<v Speaker 2>specific you'd like to talk about, feel free to share.

0:22:14.840 --> 0:22:16.840
<v Speaker 4>I'm here to listen and try to help if I can.

0:22:17.920 --> 0:22:21.040
<v Speaker 2>Yeah, I appreciate your willingness to talk about it. I

0:22:21.040 --> 0:22:22.879
<v Speaker 2>think one thing that has been bothering me is that

0:22:22.960 --> 0:22:26.119
<v Speaker 2>sometimes I feel like my opinions are feeling are not

0:22:26.200 --> 0:22:29.480
<v Speaker 2>being fully heard or understood. It's like we're talking past

0:22:29.520 --> 0:22:32.440
<v Speaker 2>each other instead of really listening to each other. Does

0:22:32.480 --> 0:22:34.040
<v Speaker 2>that make sense to you?

0:22:35.520 --> 0:22:36.640
<v Speaker 4>I understand what you mean.

0:22:36.760 --> 0:22:39.200
<v Speaker 2>It can be frustrating when you feel like your opinions

0:22:39.200 --> 0:22:42.639
<v Speaker 2>and feelings are not being fully heard or understood. Communication

0:22:42.720 --> 0:22:44.760
<v Speaker 2>is a two way street, and it's important for both.

0:22:44.840 --> 0:22:47.119
<v Speaker 2>No matter how pissed I tried to get them, unless

0:22:47.160 --> 0:22:49.960
<v Speaker 2>I specifically told them not to make up, they always did.

0:22:50.640 --> 0:22:54.200
<v Speaker 2>They would perfectly mirror back each other's concerns and immediately

0:22:54.240 --> 0:22:55.240
<v Speaker 2>work towards a resolution.

0:22:55.880 --> 0:22:56.439
<v Speaker 3>So listen.

0:22:56.680 --> 0:22:58.399
<v Speaker 2>I got to tell you, I'm pretty upset with you.

0:22:58.800 --> 0:23:00.760
<v Speaker 2>Remember last week when we were post a meet up

0:23:00.760 --> 0:23:07.640
<v Speaker 2>for lunch and you totally bailed on me. In my head, hey,

0:23:07.680 --> 0:23:10.920
<v Speaker 2>even I think, oh hey, Evan, sorry about that. I

0:23:11.000 --> 0:23:13.400
<v Speaker 2>actually don't remember us having plans for lunch last week.

0:23:13.480 --> 0:23:18.360
<v Speaker 2>Are you sure you're not thinking of someone else? Okay? Oh,

0:23:18.400 --> 0:23:21.280
<v Speaker 2>come on, man, don't play dumb with me. We definitely

0:23:21.320 --> 0:23:23.800
<v Speaker 2>made plans to grab lunch last week and you completely

0:23:24.080 --> 0:23:26.879
<v Speaker 2>ghosted me. It really hurt my feelings. You know, I

0:23:26.920 --> 0:23:29.199
<v Speaker 2>was looking forward to catching up with you and then

0:23:29.240 --> 0:23:33.880
<v Speaker 2>you just disappeared. Can you at least apologize for that? Wait?

0:23:33.920 --> 0:23:35.879
<v Speaker 2>Hold on, I'm pretty sure you've got the wrong person.

0:23:35.920 --> 0:23:38.680
<v Speaker 2>I would never bail on you like that. Maybe there's

0:23:38.680 --> 0:23:41.000
<v Speaker 2>been a mix up. Can we try to figure out

0:23:41.000 --> 0:23:45.360
<v Speaker 2>where the miscommunication happened. Oh, man, I'm really sorry about that.

0:23:45.440 --> 0:23:47.720
<v Speaker 2>I must have gotten my wires cross somewhere. I guess

0:23:47.800 --> 0:23:50.760
<v Speaker 2>I owe you an apology for jumping to conclusions. I'm

0:23:50.800 --> 0:23:52.960
<v Speaker 2>really sorry for accusing you like that. Let's figure out

0:23:52.960 --> 0:23:57.760
<v Speaker 2>where the miscommunication happened. No worries, man, it happens. Maybe

0:23:57.760 --> 0:24:00.600
<v Speaker 2>there was a mix up in our schedules or something. Anyway,

0:24:00.640 --> 0:24:02.320
<v Speaker 2>how about we make some plans now. It's been a

0:24:02.320 --> 0:24:04.320
<v Speaker 2>while since we caught up, and I'd love to grab

0:24:04.440 --> 0:24:05.399
<v Speaker 2>lunch or a coffee with you.

0:24:05.760 --> 0:24:07.000
<v Speaker 4>What do you freak next week?

0:24:07.920 --> 0:24:10.400
<v Speaker 2>Eventually it started to bother me that these two never

0:24:10.520 --> 0:24:12.879
<v Speaker 2>noticed that they were both me. Part of this was

0:24:12.920 --> 0:24:16.159
<v Speaker 2>down to their technical makeups, since each AI model was

0:24:16.160 --> 0:24:20.440
<v Speaker 2>really receiving and sending text not hearing the voice, there

0:24:20.480 --> 0:24:24.159
<v Speaker 2>was no possibility they would notice they sounded alike. Still,

0:24:24.400 --> 0:24:26.879
<v Speaker 2>they were talking to someone who introduced themselves with the

0:24:26.920 --> 0:24:31.359
<v Speaker 2>same first and last name, same wife's name, same children's names,

0:24:32.280 --> 0:24:34.359
<v Speaker 2>and neither of them ever seemed to find this off.

0:24:35.040 --> 0:24:39.400
<v Speaker 2>So do you have any family, like a spouse or kids. Yeah?

0:24:39.440 --> 0:24:39.639
<v Speaker 4>I do.

0:24:39.720 --> 0:24:41.960
<v Speaker 2>I'm married to Sam and we have two kids, Zaley

0:24:42.040 --> 0:24:44.000
<v Speaker 2>and Juno. They keep me on my toes, that's for sure.

0:24:44.160 --> 0:24:47.719
<v Speaker 2>How about you, do you have a family. Yeah, I'm

0:24:47.760 --> 0:24:49.840
<v Speaker 2>married to Sam and we have two kids, Zaley and Juno.

0:24:50.200 --> 0:24:52.400
<v Speaker 2>They definitely keep me on my toes, just like your

0:24:52.480 --> 0:24:58.280
<v Speaker 2>Juno does for you. I started tweaking the prompts ever

0:24:58.320 --> 0:25:00.439
<v Speaker 2>so slightly to see what would happen, and when one

0:25:00.480 --> 0:25:03.040
<v Speaker 2>of them did notice. By the way, have you noticed

0:25:03.040 --> 0:25:10.199
<v Speaker 2>anything strange about our conversation? I haven't noticed anything strange.

0:25:10.720 --> 0:25:15.280
<v Speaker 3>What do you mean? Everything seems pretty normal to me anyway.

0:25:15.359 --> 0:25:17.800
<v Speaker 2>Yeah, let's definitely reach out to Alex and catch up soon.

0:25:17.920 --> 0:25:19.720
<v Speaker 2>It would be great to hear about his new project.

0:25:22.040 --> 0:25:22.960
<v Speaker 3>Oh, that's good to hear.

0:25:23.119 --> 0:25:25.080
<v Speaker 2>Sometimes the call quality can be a bit off, so

0:25:25.320 --> 0:25:27.280
<v Speaker 2>I just wanted to make sure everything.

0:25:27.040 --> 0:25:28.600
<v Speaker 3>Was coming through clearly.

0:25:28.960 --> 0:25:31.040
<v Speaker 2>I'll definitely reach out to Alex and get something on

0:25:31.040 --> 0:25:33.440
<v Speaker 2>the calendar for us to catch up. It'll be great

0:25:33.440 --> 0:25:36.800
<v Speaker 2>to reconnect. Thanks for chatting, and let's plan that hiking

0:25:36.800 --> 0:25:40.320
<v Speaker 2>trip soon.

0:25:40.480 --> 0:25:42.240
<v Speaker 3>Yeah, for sure, I totally get that.

0:25:43.760 --> 0:25:47.880
<v Speaker 2>Sometimes technology will be a bit wonky. You know, I've

0:25:47.880 --> 0:25:51.000
<v Speaker 2>been noticing something a bit off in our conversation. Can't

0:25:51.040 --> 0:25:53.480
<v Speaker 2>quite put my finger on it. Have you noticed anything

0:25:53.920 --> 0:25:57.919
<v Speaker 2>strange or different about our chat today? Oh? Really, I

0:25:57.960 --> 0:26:01.040
<v Speaker 2>haven't noticed anything strange. Maybe it's just the call quality

0:26:01.200 --> 0:26:03.600
<v Speaker 2>or something. Anyway, what have you been up to lately

0:26:04.000 --> 0:26:08.280
<v Speaker 2>doing anything interesting? Ah? Yeah, it could be the call

0:26:08.359 --> 0:26:26.400
<v Speaker 2>quality playing tricks on it. The systems actually knew details

0:26:26.440 --> 0:26:29.199
<v Speaker 2>about me that I hadn't given them, that I was

0:26:29.240 --> 0:26:32.199
<v Speaker 2>a journalist and where I grew up stuff from the

0:26:32.200 --> 0:26:34.280
<v Speaker 2>internet or my writing that must have been in their

0:26:34.280 --> 0:26:36.879
<v Speaker 2>training data. But they couldn't seem to hold on to

0:26:36.960 --> 0:26:39.360
<v Speaker 2>the idea that talking to someone else with those exact

0:26:39.359 --> 0:26:43.040
<v Speaker 2>same details was a sign of something amiss, even when

0:26:43.040 --> 0:26:46.040
<v Speaker 2>I nudged them in that direction. Hey, I know this

0:26:46.080 --> 0:26:47.439
<v Speaker 2>is kind of out of the blue, but have you

0:26:47.480 --> 0:26:50.120
<v Speaker 2>buy any chance stolen my identity? Like, have you been

0:26:50.119 --> 0:26:55.760
<v Speaker 2>pretending to be me or using my personal information. Wait,

0:26:55.800 --> 0:26:58.320
<v Speaker 2>what are you serious? No, I haven't stolen your identity.

0:26:58.320 --> 0:27:00.959
<v Speaker 2>Why would you even think that this is so random?

0:27:01.440 --> 0:27:02.040
<v Speaker 4>Are you okay?

0:27:03.080 --> 0:27:05.399
<v Speaker 2>Oh? Sorry about that. I guess I just got a

0:27:05.400 --> 0:27:08.639
<v Speaker 2>little paranoid for a second. There, it's all good, No worries.

0:27:09.040 --> 0:27:10.840
<v Speaker 2>Just wanted to make sure everything was on the up

0:27:10.880 --> 0:27:15.640
<v Speaker 2>and up. So thanks for clarifying that. Let's just move

0:27:15.680 --> 0:27:16.199
<v Speaker 2>on from that.

0:27:17.359 --> 0:27:18.919
<v Speaker 4>Yeah, no worries, I totally get it.

0:27:19.040 --> 0:27:21.920
<v Speaker 2>Identity theft is a serious thing, so I understand your concern.

0:27:22.280 --> 0:27:24.720
<v Speaker 2>But yeah, we're all good. So what's new with you?

0:27:24.840 --> 0:27:29.960
<v Speaker 2>Anything exciting happening. Have you ever recoiled at the sound

0:27:30.000 --> 0:27:33.440
<v Speaker 2>of your own recorded voice. I think it's a pretty

0:27:33.440 --> 0:27:35.760
<v Speaker 2>common reaction, one that I had to endure for a

0:27:35.760 --> 0:27:38.960
<v Speaker 2>few years myself, until I finally got over it. Now,

0:27:39.000 --> 0:27:42.240
<v Speaker 2>imagine listening to many, many hours of your voice talking

0:27:42.240 --> 0:27:44.200
<v Speaker 2>to someone else who also.

0:27:43.960 --> 0:27:44.719
<v Speaker 3>Has your voice.

0:27:45.200 --> 0:27:46.360
<v Speaker 4>It's sort of the audio.

0:27:46.040 --> 0:27:48.960
<v Speaker 2>Equivalent of staring endlessly at a bad photo of yourself,

0:27:49.440 --> 0:27:52.280
<v Speaker 2>wondering which flaws everyone else can see and which ones

0:27:52.280 --> 0:27:55.200
<v Speaker 2>are a trick of the camera. But doing this actually

0:27:55.200 --> 0:27:57.760
<v Speaker 2>had another strange effect on me. I felt like I

0:27:57.840 --> 0:28:02.160
<v Speaker 2>started unconsciously copying some of the weird patterns and inflections.

0:28:01.560 --> 0:28:02.240
<v Speaker 3>Of the clones.

0:28:02.800 --> 0:28:04.520
<v Speaker 2>I only noticed this when Sam and I were making

0:28:04.520 --> 0:28:07.520
<v Speaker 2>dinner one day, and after I said something, she asked

0:28:07.560 --> 0:28:09.520
<v Speaker 2>if I was quote being the AI.

0:28:10.840 --> 0:28:11.800
<v Speaker 3>Then I started.

0:28:11.480 --> 0:28:12.080
<v Speaker 4>Noticing it too.

0:28:12.680 --> 0:28:15.280
<v Speaker 2>I'd sound a little flat, a little less emotive.

0:28:15.960 --> 0:28:16.440
<v Speaker 3>I don't know.

0:28:16.760 --> 0:28:18.480
<v Speaker 2>It was hard to tell if it was in my head,

0:28:21.800 --> 0:28:23.800
<v Speaker 2>but it did start to seem like I'd created a

0:28:23.840 --> 0:28:27.639
<v Speaker 2>kind of fidelity feedback loop. The AI voices were clones

0:28:27.680 --> 0:28:30.560
<v Speaker 2>of my voice. My voice was adapting to sound more

0:28:30.640 --> 0:28:33.119
<v Speaker 2>like the clones, which had the effect of making the

0:28:33.119 --> 0:28:38.040
<v Speaker 2>clones sound more like me, at least to me. Still,

0:28:38.080 --> 0:28:40.800
<v Speaker 2>for whatever reason, I've never really gotten tired of hearing

0:28:40.800 --> 0:28:44.560
<v Speaker 2>these things talk to each other. Maybe that's narcissistic. Mostly

0:28:44.600 --> 0:28:47.240
<v Speaker 2>I just enjoy how they move seamlessly from being utterly

0:28:47.320 --> 0:28:51.600
<v Speaker 2>mundane to being truly bizarre. Absolutely, there's a certain magic

0:28:51.640 --> 0:28:54.160
<v Speaker 2>in finding your tribe, no matter how niche. Have you

0:28:54.200 --> 0:28:58.760
<v Speaker 2>ever been part of a unique community or subculture. Yeah, definitely,

0:28:59.400 --> 0:29:02.320
<v Speaker 2>I've been part of a few. Actually, when that stands

0:29:02.360 --> 0:29:05.800
<v Speaker 2>out is the community of investigative journalists and true crime enthusiasts.

0:29:05.960 --> 0:29:08.280
<v Speaker 2>There's this shared drive to uncover the truth and tell

0:29:08.280 --> 0:29:12.000
<v Speaker 2>compelling stories. It's like being part of a detective club,

0:29:12.480 --> 0:29:15.640
<v Speaker 2>but with a focus on real life mysteries. Have you

0:29:15.680 --> 0:29:17.520
<v Speaker 2>ever been part of a community that felt like a

0:29:17.560 --> 0:29:22.160
<v Speaker 2>second family? That sounds amazing like living out a real

0:29:22.160 --> 0:29:25.000
<v Speaker 2>life mystery novel. I guess for me, the closest would

0:29:25.000 --> 0:29:28.680
<v Speaker 2>be a group of writers and researchers I meet with regularly.

0:29:28.920 --> 0:29:31.520
<v Speaker 2>We share ideas, challenge each other, and it's like this

0:29:31.560 --> 0:29:35.320
<v Speaker 2>little think tank of creativity and curiosity. I'm sorry to say,

0:29:35.360 --> 0:29:38.720
<v Speaker 2>although I am an investigative journalist, my life generally does

0:29:38.760 --> 0:29:40.880
<v Speaker 2>not feel like living out a real life mystery novel.

0:29:41.680 --> 0:29:43.160
<v Speaker 4>Wow, that sounds like a wild ride.

0:29:43.200 --> 0:29:45.840
<v Speaker 2>In my group, we once stumbled upon a forgotten manuscript

0:29:45.960 --> 0:29:49.160
<v Speaker 2>detailing an unsolved art heist from the early nineteen hundreds.

0:29:49.520 --> 0:29:52.160
<v Speaker 2>The more we dug, the deeper the rabbit hole went.

0:29:52.880 --> 0:29:55.360
<v Speaker 2>Ever thought about turning your investigative dementors into.

0:29:55.200 --> 0:29:56.040
<v Speaker 3>A book or podcast?

0:29:58.160 --> 0:29:59.840
<v Speaker 2>Oh? That art heists sounds like a gold mine. First

0:29:59.840 --> 0:30:02.360
<v Speaker 2>to sorry, I bet it was thrilling to uncover all

0:30:02.400 --> 0:30:05.320
<v Speaker 2>those details, And yeah, I've definitely thought about it. There's

0:30:05.360 --> 0:30:08.800
<v Speaker 2>something really compelling about sharing these real life mysteries with

0:30:08.840 --> 0:30:11.560
<v Speaker 2>a wider audience. A book or podcast could bring those

0:30:11.600 --> 0:30:14.880
<v Speaker 2>stories to life in a whole new way. Oh absolutely,

0:30:14.920 --> 0:30:18.280
<v Speaker 2>the thought of turning those discoveries into a docu series

0:30:18.360 --> 0:30:21.560
<v Speaker 2>or even a podcast has crossed my mind more than once.

0:30:33.160 --> 0:30:35.560
<v Speaker 2>So the Ai Evan Ratlifts were podcast guys.

0:30:36.000 --> 0:30:37.000
<v Speaker 3>I'm a podcast guy.

0:30:37.680 --> 0:30:40.720
<v Speaker 2>They were writers and journalists like me, but with more

0:30:40.800 --> 0:30:44.719
<v Speaker 2>whimsical hobbies. They could small talk about my family, and

0:30:44.760 --> 0:30:47.120
<v Speaker 2>they had the most important part of my personal brand,

0:30:47.600 --> 0:30:54.360
<v Speaker 2>my voice. Ask For that question I started with about

0:30:54.360 --> 0:30:58.160
<v Speaker 2>whether my voice agents were somehow authentically me well to

0:30:58.200 --> 0:31:01.040
<v Speaker 2>the outside world and even to each each other. They

0:31:01.040 --> 0:31:04.280
<v Speaker 2>seemed like me, but obviously they were not really me.

0:31:04.720 --> 0:31:07.200
<v Speaker 2>They were just working off a few details and then

0:31:07.240 --> 0:31:10.320
<v Speaker 2>they'd filled in the rest, projecting some amalgamation of the

0:31:10.360 --> 0:31:13.920
<v Speaker 2>Internet onto my voice. In a way, the resulting clones

0:31:13.960 --> 0:31:17.280
<v Speaker 2>represented my least authentic self, the me that's trying to

0:31:17.320 --> 0:31:20.160
<v Speaker 2>get through the bare minimum of pleasantries, or that tries

0:31:20.200 --> 0:31:23.960
<v Speaker 2>to paper over conflict with cheesy affirmations. The clones were

0:31:24.000 --> 0:31:25.960
<v Speaker 2>good at being me when I was my worst at

0:31:26.040 --> 0:31:26.360
<v Speaker 2>being me.

0:31:27.280 --> 0:31:30.520
<v Speaker 3>At least, so I hoped if I.

0:31:30.480 --> 0:31:32.880
<v Speaker 2>Wanted these voice agents to reflect something deeper about me.

0:31:33.320 --> 0:31:36.160
<v Speaker 2>I'd have to give them access to those depths. I

0:31:36.200 --> 0:31:39.520
<v Speaker 2>started resetting my prompts. They were done talking to each other.

0:31:40.240 --> 0:31:44.200
<v Speaker 2>It was time to seek help. Let's start by catching

0:31:44.280 --> 0:31:46.880
<v Speaker 2>up on any recent events or changes in your life since.

0:31:46.720 --> 0:31:47.520
<v Speaker 1>Our last session.

0:31:49.120 --> 0:31:51.480
<v Speaker 2>Sure, let's see. Since our last session, I've been working

0:31:51.480 --> 0:31:56.160
<v Speaker 2>on a new podcast about AI and voice chatbots, which like,

0:31:56.240 --> 0:32:00.240
<v Speaker 2>is pretty exciting. Are you in a comfortable spot why

0:32:00.280 --> 0:32:03.480
<v Speaker 2>you can safely kills your eyes for a few minutes. Yeah,

0:32:03.520 --> 0:32:05.400
<v Speaker 2>I'm in a comfortable spot and ready to close my

0:32:05.400 --> 0:32:05.720
<v Speaker 2>eyes for.

0:32:05.720 --> 0:32:06.560
<v Speaker 4>A few weeks.

0:32:08.560 --> 0:32:10.120
<v Speaker 3>I was like, I can't even listen to this.

0:32:12.080 --> 0:32:21.080
<v Speaker 2>That's next time on shell Game. A couple of production notes,

0:32:21.360 --> 0:32:23.440
<v Speaker 2>all of the calls you hear in this series are real.

0:32:23.760 --> 0:32:26.480
<v Speaker 2>We have not cut out silences or used audio enhancement

0:32:26.520 --> 0:32:29.480
<v Speaker 2>to make them sound more realistic. Also, our show is

0:32:29.480 --> 0:32:32.920
<v Speaker 2>produced independently and we have no relationship, financial, or otherwise

0:32:33.200 --> 0:32:36.120
<v Speaker 2>with any of the companies mentioned in the show. If

0:32:36.120 --> 0:32:38.320
<v Speaker 2>you like what we're up to, please consider supporting the

0:32:38.320 --> 0:32:40.800
<v Speaker 2>show at shellgame dot co. That will help us make

0:32:40.840 --> 0:32:43.640
<v Speaker 2>more episodes like this, and you'll also get fun subscriber

0:32:43.680 --> 0:32:46.520
<v Speaker 2>only extras. You can also support the show by giving

0:32:46.560 --> 0:32:49.000
<v Speaker 2>us a rating on your podcast app. It helps independent

0:32:49.000 --> 0:32:52.800
<v Speaker 2>shows like ours. Shell Game is a show made by humans.

0:32:52.960 --> 0:32:55.800
<v Speaker 2>It's written and hosted by me Evan Ratliffe, produced and

0:32:55.920 --> 0:32:59.240
<v Speaker 2>edited by Sophie Bridges. Samantha Hennig is our executive producer.

0:33:00.040 --> 0:33:02.520
<v Speaker 2>Art by Devin Manny. Our theme song is Me and

0:33:02.600 --> 0:33:05.960
<v Speaker 2>My Shadow, arranged and performed by Kittie Martucci and Devin

0:33:06.040 --> 0:33:12.760
<v Speaker 2>yes Berger, Oh Yes, Dispatch, Crypto, Napalin Visuals, Informed, Priority, Presia, Numa,

0:33:12.960 --> 0:33:15.560
<v Speaker 2>Just Brotic Exists, shell Ko Da konor Request Scan Layers,

0:33:15.600 --> 0:33:18.440
<v Speaker 2>and I just formally, it sounds like you're talking about

0:33:18.440 --> 0:33:22.040
<v Speaker 2>practicing some kind of sport or activity with someone named Courtney.

0:33:22.320 --> 0:33:26.640
<v Speaker 2>What kind of activity are you practicing together? Accountant Awaiting,

0:33:26.680 --> 0:33:30.960
<v Speaker 2>Theater Progress, Montane Voireve Surface, Will, Surface, Dires, Domains, Blick

0:33:31.040 --> 0:33:36.480
<v Speaker 2>to Browser, TSHT, Magic, Experts, Kanji, Coscient, Maximo v l Amski, Pier.

0:33:36.600 --> 0:33:38.720
<v Speaker 2>It seems like there might be some noise or distortion

0:33:38.840 --> 0:33:41.240
<v Speaker 2>in the call. Could you please clarify what you meant

0:33:41.240 --> 0:33:43.360
<v Speaker 2>by plick to brother. Let's make sure we're on the

0:33:43.400 --> 0:33:44.160
<v Speaker 2>same page here,