← Back to Episode 01

Episode transcript

AI Without the Hype: 5G, Robot Butlers, and What Comes Next

Full transcript of the published episode, with timestamps. It was generated automatically, so expect the occasional misheard word or speaker label.

0:00Cold open: Humans can’t micromanage AI anymore

Viet Nguyen

0:00

You know, humans can't micromanage this anymore. Like it's just it's impossible to do. It's just it used to be how it was. We'd have programmers come in there and they'd they'd actually program the code and and uh and and make the changes uh in the network uh that serve our mobile phones and other uh and other devices. Uh, but we're we're really beyond those old days now where that can happen. And now a lot of that work is really being done uh by AI systems. It's teaching itself, it's probably teaching us. We're we're trying to teach it and teach ourselves at the same time. So it's a little bit of a uh a little bit of a chaotic mess out there right now. It's funny, like how you know those those early those early uses that we sort of look back on and kind of go, that's macabre. That was a terrible idea. And we we're probably going through that right now with with AI, or will be going through it with with AI over the next few years. So there might be some pretty macabre uses of of of uh of AI.

0:59Show open: Welcome to Volume Control AI

Intro song

1:00

Take down the hype on AI, exerting Volume Control.

Natalie Chantal Luke

1:00

This is Volume Control episode one.

Viet Nguyen

1:11

Hi, I'm Viet.

Natalie Chantal Luke

1:12

And I'm Natalie.

Viet Nguyen

1:14

And we're the host of Volume Control and AI Podcast. And we're turning up the volume on AI, on AI understanding, adoption, and its use.

Natalie Chantal Luke

1:25

We're gonna turn down the noise and the hype.

Viet Nguyen

1:28

And we'll do that through real conversations with real people who are currently using and adopting AI.

Natalie Chantal Luke

1:34

Ultimately, we'll be creating a connection from the present AI landscape to the future. And we'll build a bridge to AI using real human stories and experiences.

Viet Nguyen

1:46

So, what is Volume Control? Well, Volume Control is a clear and curious real-world conversation about artificial intelligence. It's how it's changing work and creativity and strategy and technology and policy and culture and just everyday decision making. The show is going to help people ask better questions, uh, understand what matters, and make sense of the human choices shaping what comes next.

Natalie Chantal Luke

2:12

Yeah, it's real conversations with real people about how I is shaping, reshaping, and being used in the modern workspace and the world. It's connecting the dots between people, systems, stories, and strategy. It's a place to find out solutions and have discussions with people just like you, who are trying to navigate a rapidly changing landscape and stay relevant and employed.

2:38AI hype: Boomer, doomer and everything in between

Viet Nguyen

2:39

So, what are some scary topics we'll be exploring?

Natalie Chantal Luke

2:43

Well, I've heard it said that unemployment rates are going to become 100% as AI takes over in the next few years. People are supposedly going to become pets or slaves of AI. And AI is just making everything the same. It's dumbing down the message and the overall output. So it seems like it's all boomer or nothing doomer.

3:06Meet Co-hosts Viet and Natalie

Viet Nguyen

3:07

So who are we? Well, I'm Viet, and I'm uh formerly the president of 5G Americas, which was a wireless trade association uh based uh uh here in the uh Seattle area. Uh and uh I've uh previously uh held roles over at uh in communications over at T-Mobile and Microsoft and in uh and in policy and government in an earlier life.

Natalie Chantal Luke

3:32

Yeah, I'm Natalie, and I'm a creative expert. I'm a brand strategist, I'm a change manager, a tech nerd, I do programming, UX, UI. I'm an interviewer, a teacher, a founder, and a survivor. And I'm dialing down the volume on AI Hype with you. So, Viet, tell me the story of the first time you saw a network do something smart and thought humans can't micromanage this anymore.

3:59Why humans can’t micromanage wireless networks anymore, AI does

Viet Nguyen

4:00

Yeah, that's really interesting, uh, Natalie. Uh in you know, in the in the world of uh wireless communications, uh over the over the past several decades, uh we've seen uh just this massive increase in the amount of data that wireless networks use. And it's gotten so it's it's it's growing at at the pace of uh doubling once uh every about two years or so. So you know you can imagine, you know, if you took like the the the national highway system and you doubled the number of uh of lanes or the amount of traffic every two years, how quickly that that amount of data would uh uh would cause some issues. That's what's happening in the world of the wireless uh industry today. And so it's really that that time when we we just sort of, you know, in the industry, we just sort of looked at it and said, you know, humans can't micromanage this anymore. Like it's just it's impossible to do. It's just it used to be how it was. We'd have programmers come in there and they'd they'd actually program the code and and uh and and make the changes uh in the network uh that serve our mobile phones and other uh and other devices. Uh, but we're we're really beyond those old days now where that can happen. And now a lot of that work is really being done uh by AI systems uh and that help to predict uh the kind of traffic that goes across those wireless networks, uh, and uh is really important in helping to uh have the network operators to be able to uh plan for uh the architecture of those networks as well.

5:36AI at an 80,000-person concert

Natalie Chantal Luke

5:37

Yeah, so give me an example, if you would, of like what is an example where this network architecture is really being pushed to the limits.

Viet Nguyen

5:46

Yeah. So imagine um you were uh trying to create a uh network a good network experience in an incredibly crowded environment, say. Um I'll give you an example. Let's say you, you know, uh Taylor Swift and and you know, about 80,000 of her best friends were going to uh do a concert over uh at Lumen Field uh, you know, next Saturday. Um, as a network operator, you'd be looking at that going, that's quite a challenge. We've got 80,000 of these uh these network devices that we're you know potentially going to need to handle. Uh is there enough capacity on that network uh over at Lumen Field to be able to handle all of those phone calls and all of the video uploads and video downloads? And so that's exactly the kind of place where um AI can can really help. Um for instance, it's going to be able to help predict where that traffic is coming from, right? It's uh you know, the all of the different entrances at the stadium, uh you know, what happens at you know at a halftime, where do people congregate. And um, by understanding sort of those traffic flows ahead of time, it can um it can predict and then be able to provision the kinds of network resources uh to accommodate for it for that kind of traffic. And that kind of stuff is is is changing, right? And you know, in years ago, uh the wireless connections were about phone calls, right? And then it was you know about texting, and it was about you know some some really um kind of what we sort of refer to as light payload types of data. Uh, but increasingly uh you know over the years, we've had uh video uh being used, and video it tends to be quite a hog of data, uh and and now high resolution, uh you know, sorry, HDR and 4K video, and and and and so the more uh and better resolution that you that you're able to deliver on video, the m the heavier the payload that is. And so the more work that the network needs to um uh needs to work on and prepare for, and that kind of stuff is just, you know, you really need AI to help you with that.

Natalie Chantal Luke

8:00

Yeah, I mean, I've got two teenage nieces, and they definitely have to have their socials updated real time, and it's gotta look good and they've gotta look good. We can't have any weird lagging that's happening.

Viet Nguyen

8:14

Totally, totally.

8:15How AI helps 5G networks save energy

Natalie Chantal Luke

8:17

So, uh, what's the dumbest way that networks have been used and waste energy that AI now fixes?

Viet Nguyen

8:25

Yeah, you know, these these networks are um they've been around for a long time. And so when people think about uh, you know, the term 5G, right? What's 5G? What's 4G? Right? Uh each G refers to a generation of of networks. Uh and they sort of build on top of the of the last generation. And so with 5G, we're in the we're obviously in the fifth generation of wireless cellular networks. And with each successive generation, there's new technologies that that get built into them, and it happens every, you know, gradually, but every 10 years or so, there's a big package that sort of comes out. And so that's how you get the 4G and 5G. The way things used to work uh in those in those legacy systems uh is that you'd have these radios that, you know, sort of um, you know, sit on the on the cell towers, right? And those radios were kind of dumb radios, right? They they um broadcast out and receive information to the cell phones, uh, but they would they would not be able, you wouldn't be able to do much with it. It's sort of like uh you're talking through a tin can and that and that and that data sort of kind of goes in and out, uh, you know, it's sort of uh in a in a single channel. As those radios have become more complex over time, they've had more channels and they've had, you know, if you've got more, if you're working with more channels, you're more working with more complexity. And but you're also working with these radios that um uh nowadays are able to turn on and off when they're not in use. And that turns out to be a very, very significant uh way in which uh wireless networks today actually save energy. Uh and and so, you know, with with AI, it's able to figure out, you know, okay, you know, we're at the we're at this part of the show in the, you know, in Taylor Swift show. So there's probably not a lot of uh video downloads happening at this point. And so um we can we can start to you know power down some some of these devices maybe on the periphery of the stadium. Uh, and then of course, after the after the event is over, power down uh even more. So it's a great way to save energy.

10:30What could 6G actually look like?

Natalie Chantal Luke

10:31

So help me understand a little bit more about 6G. Can you describe a scenario for me that would actually like impress some of your non-tech friends? Can you make it easy for me to understand that's because I don't speak that tech conversation?

Viet Nguyen

10:48

Yeah, yeah. So um, you know, as as I mentioned, right, uh we're right in the middle of this generation right now called 5G. And um that was actually introduced in and around um the year 2019, 2020 or so. And this next generation of 6G is really going to be commercialized and available for people to use in around the year 2029 and 2030. So uh there's a lot of technology improvements that are still being worked on right now in the in the standards for 6G. Uh, one of those standards involves the usage of um of satellites. Uh so for most of the people watching this show, they probably um aren't aware of how wireless mobile really works. You probably already thought it came through satellites already, but it doesn't. It actually uh works through those cells, those cell towers that you sort of see um occasionally as you as you drive along uh the highways. And there are actually some radios that actually sit on top of buildings that you might not be aware of uh that actually send out those um uh those wireless signals. But in the world of 6G, um the satellite uh industry has uh been working really hard with the mobile industry to be able to provide uh that level of satellite coverage. Um and it's going to be you're gonna be able to have access in areas where um you wouldn't ordinarily expect, you know, out in the mountains somewhere. Right now the coverage is is um, you know, if you've anyone who's actually you know using uh the phones that are satellite capable right now, um, you know, you can probably get a text through, right? But um there's uh some there are some companies right now that are actually allowing for voice uh over satellite over satellite connected uh wireless mobile. Um and as 6G comes along, um there's gonna be more data that is going to be available through through those um uh through those systems. And so you might eventually be able to do uh more than just voice or text, right? So we'll start to see, you know, eventually, maybe not in 6G, but eventually we might even start to see video, right? And you know, you're out walking into the mountains and you might be able to be able, you know, to be able to upload your video. Um so there's some real interesting excitement uh uh over that. Uh in regards to you know, the one cool thing, right, about uh about AI use in N6G, you can imagine those satellites, right? They're they're you know, either in a geosynchronous orbit or or a low or Earth orbit, you know, you know, six, seven thousand miles or you know, kilometers up in the sky. And um, one of the challenges is uh that has to connect over a very wide area. A lot of cell phones maybe you know, being uh in you know covered in that particular area. And so that satellite has to be very, very sophisticated in be in being able to identify which cell phone it needs to talk to. And also, it's moving really fast, right? You can imagine trying to, you know, um uh trying to uh capture a signal for something that's going, you know, like you know, six, you know, 16,000 miles, you know, in an hour or something like that. It's it's really impressive the the engineering and technology that's involved and uh really does require uh sophisticated AI to be able to uh um to be able to execute on the phone.

Natalie Chantal Luke

14:11

Well, this is kind of funny that you say that because I don't think that you know this, but I was a beta tester for T-Mobile Starlink. And so I actually was on Starlink and I loved it. And as soon as my beta test time was over, I was really disappointed to be back to regular phone. Um, for me, the thing that was happening was we live in a dip, so signal goes from hill to hill. And we live down in the middle of the valley, if you will, so to speak. So the signal goes across and when it comes down to the valley, it drops. And so when we were on Starlink, or when I was on Starlink, my husband didn't have the same fortune, um, I could get a signal anywhere. And now if I drive a few blocks away from my house, my calls drop. Um, so I'm waiting to get back onto Starlink. So I'm very excited about the potential of satellites and and those sort of things coming in the future because my own experience was that it works really great.

Viet Nguyen

15:14

Yeah. And and you know, I uh I think you're not alone in in that. And I think that there's uh if anybody who's been on a uh on a plane ride, you know, an international plane flight uh nowadays, uh with that uh with that capability, uh it's like you've got a broadband connection through your entire flight. Uh, it's amazing. Uh in fact, um I just got back from uh Japan uh like last month, and uh the broadband connection I had my on my flight between Japan and the United States was actually stronger than the broadband connection I had in the hotel. So it was like, I gotta get on the plane now to be able to get a good internet connection.

15:51What happens when everybody connects on the Network?

Natalie Chantal Luke

15:52

That's right. I mean, my concern is that the reason I had such a good experience on Starlink was because I was in a beta test group and there weren't very many of us. So the traffic on the Starlink was very low. It was only a couple hundred of us. What happens when it's the whole world?

Viet Nguyen

16:13

Yeah, and so that is that is certainly one area where where AI helps, right? Uh it helps uh the radios and it helps the the network to be able to distinguish which channels are in use at the time and uh and and how to provision the resources necessary uh to be able to still provide that good experience when multiple people or multiple devices are trying to battle for the same uh uh same band of spectrum. Uh, you know, those those radios, you kind of think of them as um, they're called transceivers in the radio. So it's a transmitter and a receiver. So it's kind of like having uh a mouth and and instead of ears. And um but it's not just one in in those radios. They they actually come in these arrays. And you know, you can imagine um uh in in those in those arrays, some of the some of the radios or some of the transceivers pointing to one device, maybe some of the maybe some of the uh transceivers pointing to another device as it's kind of going across the span of its where it's looking, right? And so by being able to track where those devices are, uh it can actually turn multiple transceivers to be able to get a stronger signal. So the upshot is that the person is actually usually using that device as they're walking through or maybe driving through or flying through the uh that zone of where the where where the wireless uh signal is focused, um gets a much better experience and um and and and and much um higher rates of download and and and upload as well.

Natalie Chantal Luke

17:52

Yeah. So AI is really helping us stay connected then.

Viet Nguyen

17:55

Really?

Natalie Chantal Luke

17:56

With each other, with technology, with the whole thing.

Viet Nguyen

17:59

Yeah, yeah, absolutely. Yeah.

18:00Agentic AI and the robot butler

Natalie Chantal Luke

18:02

So um talk to me about what a agentic AI is. I mean, I hear a lot about that. There's a difference between like AI and now there's agentic AI. And tell me what a robot butler is.

Viet Nguyen

18:15

Yeah, so okay, so this is a really interesting point, right? For most of us, the the modern age of AI kind of began uh on November 30th, 2022. And for most of the folks who are listening, that's really when uh ChatGPT 2.0 really took off, right, from from Open AI. Uh that really kind of lit the fire under under everybody and said, hey, this is this is something that's pretty exciting because you know, AI has been around for actually a long time, right? It's it's a technology in the 50s, I think. Yeah, in the 50s. Uh it's been and predominantly in the labs, and then a lot of people were were working on it, and you know, just you know, gradually over the over the years. And so it kind of for a long time uh it was a was an academic you know pursuit. And it really wasn't until we got into that that 2022 phase when um uh when that really began to take off. The first iteration of this um was this stage that we called generative AI, right? And we all know it, right? We're using the chat bots and it comes back and it's actually writing our emails for us, or maybe it's even creating an image um uh for us uh and helping to, you know, kind of fill out what our, you know, what our thoughts are, right? That's that's sort of generative, right? Creating, generating. This next stage that we're in um is something that fewer people know about. And it is really uh an era that began probably about the beginning of this year, uh, when um uh some users uh of uh Anthropic's Claude uh uh were able to get their um large language models to be able to work kind of autonomously, right? You they they had these these boxes, you know, you would you basically put them on the machines and you run your your Claude bot, right? And uh and you could just like leave it overnight and it would it would actually do some amazing things. That's an agentic AI, right? Uh it is AI that doesn't just answer questions for you, but it's AI that actually does work for you. And so this this stage that we're in right now is incredibly significant that uh that we're going through. And we're we're getting to the point now where we're going to be over the next few years uh be able to use agentic AI and then infuse that agentic AI with sort of the next stage, which is called physical AI, where they actually put that into robots, and that's how you get the robot butlers, right?

Natalie Chantal Luke

20:49

Right. I think we'll be talking to some guests. So we have some guests who are people who have been implementing already that and they'll talk about things like the Turing test or as I like to call it, the Pepsi Challenge instead. But um, that's gonna be a really interesting thing. For some people, it can't come quick enough. Like for my mom, it'd be really great to have a robot to help her out. She's in her 80s and it would be perfect. We would love to have a robot that would follow her around and do everything she wants to do.

21:17What do we actually want robots to do?

Viet Nguyen

21:18

Well, you should listen. So so describe your your ideal robot butler, like you know, in maybe in one sentence or or or or more. So tell me the and and you know, maybe in that sentence, tell me the for my own robot butler or for my mother? Maybe for both. What what makes you nervous and what makes you excited?

Natalie Chantal Luke

21:36

I mean it it it's difficult because of course I would like somebody who was a robot butler that emptied the dishwasher, for example. At the same time, what would I have to talk to my husband about or complain about if I had a robot butler? So my big concern is this like, I absolutely want to automate things. It would be great to have a butler. It would be great to have somebody who folds The laundry, did all the chores, the things that I didn't want to do. But it's hard to draw the line between where that is and where I think that is, because I think they're two different things. And what happens when I don't have a purpose? That's what I'm really concerned about with AI. Um, you know, if I'm just sitting around eating bonbons, I'm gonna be a thousand pounds and then I'm gonna need a robot that gets the fat off of me somehow, which that might not be the best of situations.

Viet Nguyen

22:31

Maybe like a robot physical trainer.

Natalie Chantal Luke

22:34

Yeah, something like that. I was hoping for more of a massage therapist kind of a robot, but I mean, whatever it is, it's gonna be a situation, you know, part of me is excited about that, but then part of me wonders what my own identity will be. So, you know, it's it's it's really challenging because I'm excited about the opportunities. I would like to see robots do things, you know, even it's interesting. I'm a motorcycle rider. Motorcycle riders are not very liked. I apologize to any viewers who don't like motorcycle riders, but I enjoy riding. I don't enjoy riding in the city because there's too much traffic and it's very dangerous. But it's wonderful when you're out on the road and you're out in nature and you get to experience it and you feel the wind. If I have a robot that does that, then I don't get to experience that. Even if I'm sitting in the back, I'm not the one who's controlling that. I'm not the one that's doing the challenge. So there's a side of me that feels like, gosh, you know, absolutely I want a robot, but I also don't want to lose my autonomy. And I also feel a bit uncomfortable about having someone who's kind of a slave. Just like I don't want to become a slave to a robot, I don't want to employ a robot slave.

23:52Work, identity and autonomy in an AI world

Viet Nguyen

23:53

So do you do you then sort of identify, like, does your work and the labor that you put in help form your identity then? It does.

Natalie Chantal Luke

24:01

For 100% sure. I mean, the distinction between me and my career is a very thin wall. I think that when I um describe myself, a lot of what I describe is my accomplishments at work. You know, some people when they describe themselves, they say what their religion is, or they say what their beliefs are or their political system. For me, when I describe myself, I usually talk about my accomplishments first. And that comes through work.

Viet Nguyen

24:32

So then the the robots then uh might sort of uh interfere or or or alter that, right? So alter the the formula that that that you've sort of, you know, uh working.

Natalie Chantal Luke

24:44

Yeah, it would it would be really difficult for me to know what my own identity is, right? Would the robot then be a mini me? Would I have like a me? I would want it to look like me. Maybe it'd be what happens if the robot looks better than I do, and then I become really jealous, or it's even better at doing what I already do well because I've programmed it and I've taught it, and suddenly it's learned and it's built upon that, which is a great thing. That's what's really great about AI. But what if it's a better wife than I am?

Viet Nguyen

25:15

Oh wow, that's uh yeah, that's definitely one of those use cases. That's one of those use cases that uh that people are thinking about.

Natalie Chantal Luke

25:22

If my husband was like, I really want to hang out with the bot, it might be a really ugly situation.

Viet Nguyen

25:29

But what about what about some of the other like like useful you know things that that that bots could could eventually end up doing? Like, like I don't know if you like, you know, doing travel planning or some or you know, something like that, but that's that's a lot of that's a lot of work to you know set up a trip, you know, and you know, maybe it's just because I just got back from Japan and it was just a lot of work putting together all of the stuff for the vacation.

25:51Robots, caregiving and who gets access

Natalie Chantal Luke

25:52

I mean, I'm gonna circle back to what I originally said. You know, my mom is in her 80s, she's a senior citizen. We have a large aging community these days. We just don't have enough caretakers. During COVID, for example, my mom got really sick and she got put into uh a care facility in between because she was she had broken her back because a dog had jumped on her in a dog park and she already was with a walker and she broke her back again. And when she was in the care facility because there wasn't enough caretakers, she soiled herself, which I'm sorry for telling this story, mom. But she she soiled herself and she ended up not getting anybody to change her bed sheets for over two hours because they simply didn't have enough nurses. And the nursing staff said, Listen, we just have 200 patients. And while it's terrible that she's in soiled linens, we also have people who are busy dying and need immediate assistance. So we have this aging community that's coming and we don't have enough services to deal with that. I also think there's a big fear about the cost of getting those services. You know, what would a robot cost? Even right now, just to use a gentic AI or to have access to a chatbot, it's only $20 a month, but for some people that's a lot of money. And if you have to pay for things with tokens, it gets really expensive really fast. The only thing that I can say is that, like my shirt, I equate it to when I used to be a video game player, and you would stick a quarter into the video game, you'd play for a little bit, and then boom, your video was over, and the quarter was gone. And man, that thing ate quarters like nobody's business, and I worry about that with tokens. So, how are we gonna manage that then with all of the costs associated with it? It seems very idealized to think everybody gets a robot and everybody gets access to agentic AI, and AI is gonna become universally accepted. Even if we want that to happen, can we really afford it to happen?

Viet Nguyen

28:08

Certainly, um, I you know, those are certainly some really great points, right? And uh being able to get uh this these services uh out to everybody, right? This should not be something that's limited to uh very rich folks or just folks that are, you know, technically very savvy. This is a lot of work to get it, get it to get it out there. I I know, you know, you mentioned your your t-shirt uh uh that says uh shall we play a game, uh which is obviously a reference to war games, and uh thanks for uh thanks for bringing that out. Yeah.

Natalie Chantal Luke

28:40

Uh it's uh that old whopper computer basically saying, uh, you know, we're here in Seattle, so it's really important to be aware of some of the cultural references that we have here. I mean, that movie for me was a very important movie. During high school, I was one of the people who was one of the first adopters of computers. I won the very first uh computers award at my high school that had ever, ever been given out to anybody. Um, we did some things we won't talk about on air, but you know, we also ran numbers and did all of those things. We had some access and, you know, it's changed a lot. I mean, the time of a 56k modem that's in the past, and that actually feels like it's in the very ancient past. It doesn't just feel like it's in the recent past. Um, for my nieces who are currently teenagers, they really don't even know what that is. You know, we were talking about dialing a rotary phone and people don't know how to do that. What happens if they can't do that? It's interesting to me because I wonder if when bots, since we're talking about that, come online, if that's what you want to call that, or alive for everybody, are they gonna have that historical knowledge? So, you know, if there's a phone booth that appears all of a sudden, will they be able to dial the phone or do they even need to do that? You know, where is that line? And you know, of course, I've seen all the movies like everybody else has, and I I worry about what happens when robots age out, just like people, and they become useless. But then what do we do with all that garbage? And what do we do we offline them? Are they sentient? I mean, there's a lot of things we have to talk about that becomes really kind of a question that we haven't figured out yet.

Viet Nguyen

30:38

Yeah. Now, one of the interesting things about you that I've you know come to learn is that uh you know you've been doing this for a pretty long time. You've been, you know, you're you're a you're a technical.

Natalie Chantal Luke

30:47

I'd try and I'm late 21. And you're only 21, which is amazing.

Viet Nguyen

30:49

You've been doing it since uh you were three years old. Yeah. Uh but you know you were uh uh very early on doing uh uh you know just some some computer work. Uh you've uh you've uh you've been you've been in and around uh technology pretty much for most of your life. Uh and yet uh you're still able to bring out sort of the the humanity and and ask the ask the the the qu the the questions about how this is going to impact society.

31:17Why expertise still matters when everyone has AI

Natalie Chantal Luke

31:17

I mean I used to get my Tandy 500 at Radio Shack and I would go get a free battery, which was part of the whole deal. And I always chose a nine-volt battery because that was the biggest battery and therefore the most expensive, and it made the most sense. But I can remember my brother and I going to Safeway and programming in basic all of the computers to just say hello or whatever, just because we thought it was really fun to have that happen. And the first game that I ever programmed in as a dot situation was Frogger, and that was a really fun thing. And and those things, you know, I've been there when we started with Pong and all of those different things. Um, it it's definitely different, and especially being a teacher now and teaching AI, I think it's really important to cut through all of that noise because you don't have to be a programmer to use AI. Um, there is there is a fine line though, because you do have to have some types of expertise in order to ask the right questions of AI in order to get the right output. And I think that's like one of the things we'll be exploring in this show is that it's really difficult. All of a sudden, it's a tool that everybody has. And it's just like every other tool. If you suddenly give somebody an electric drill and they're not a carpenter or they've never used one before, they're as likely to make a mess of the wall as they are to fix it. So the question is uh how do you do that? I mean, at Home Depot, you can go or Lowe's or whatever your choice of McClendon's, whatever your choice of places. You know, they literally have classes that you can go in on the weekend and learn how to put a project together, and they put little things together, and people walk around and say, you know, yes, this is how you do it, or no, you don't do it this way. I mean, the same thing with just a cell phone. Um, I was at the Apple store the other day, and the salesperson said, don't be here at 11 o'clock. All the seniors come at 11 o'clock, and you're gonna have to sit through all the basic stuff, and we won't be able to get to you. Well, that's gonna need to happen for AI too. And there's gonna need to be somebody who's actually teaching people how to do that. And I know that a lot of corporations have adopted LLMs, they've actually gotten specific LLMs for themselves. So those things work what with what in what we call a sandbox, you know, um, hospitals are using them and they're using them well. You know, there was a concern that all the jobs would be gone as soon as everything went online, as you talked about in 2022. And one of the conversations was that the first job to go was radiologists, because that's something that's a repeatable job. Everybody looks at it. But what they found out was that because of the efficiency of being able to do that, that instead of having no radiologists, they actually needed more radiologists because suddenly they were able to be more efficient and do more work. So where are we at? That's the big question. And how do we make sure that what we're doing is not creating pockets that aren't being filled with the right people in the workplace? You know, for myself, I'm a creative and all of a sudden and I'm a strategist and I teach AI, but people are waiting. People are waiting to see what the next thing is because things are coming so rapidly. There's so many tools, there's so much noise. And it seems like you should kind of be able to do everything with AI. And yet, you know, even for myself and I'm sure for yourself as well, you know, I think, oh, if I try to create something, or if I've seen something before, I can't really get it to work very easily. And so there's definitely tips and tricks that certain people have, and it feels like they have sort of the secret knock or the secret code. And once you know that secret code, it seems very easy. But how do you go about finding out that information? That's the big question.

35:34Has the technology outstripped the business model?

Viet Nguyen

35:36

Yeah, um, I think you're you're absolutely right. I think, and I think too, right, that in a certain way the technology sort of outstripped the business models, right? I mean, it's sort of like we have all these super powerful tools that are all in front of us right now, and nobody knows exactly what to do with it. Right. And and and we're in this weird, weird phase of it right now. But you know, so you've been because you've been doing this for a while, right? Has there really been a time where you've you sort of felt the the that that moment of like uh sh you know, this is the agentic, you know, pieces sort of come through. And it might not be agentic AI, but it could be like automation. Is there maybe a time in your life where where you know AI is really you've just sort of stepped back and said, holy cow, that's that's that's kind of amazing.

36:25AI can be amazing—and still get it wrong

Natalie Chantal Luke

36:26

I mean, absolutely, it's really amazing. It's actually so amazing that sometimes it's like dating somebody who's really good looking, but they've not got it all together. And so you're waiting for the other shoe to drop a little bit. Um, I do a lot of legal work and we get lots of huge documents, 3,000 pages of medical records and things like that. And it's great. I've put it into Google Notebook LM. Uh, you know, I use everything all at once because I always test them against each other. I make like a little thing where they test each other. And it's super interesting because um they point out each other's problems, what they got wrong, which is great. But then when I don't review it carefully, I find out that they either make things up or they just got it wrong, or they omitted something that's really important, and those small details really matter to the client. So have I been wowed by it? Absolutely, because there's been times that I've put in 3,000 pages of something and it's come back within minutes with a report of everything. You know, I have to say that for me, one of the most reliable AIs, which is really funny, is uh an older AI that was long before 2022, which was I used Otter to do my translations and things like that. And otter actually did the same kind of mistakes that I'm finding with other AIs, which was it didn't always understand what people said or exactly what was happening. So if you used it to do, for example, closed captioning or whatever, you always had to read through it because it might not have really figured out what the person's name was, what you were talking about, or it broke a sentence in the wrong place that didn't make any sense with what you were seeing on the monitor. And that's still the case till today. That has not been solved. I think it's not been solved because the human part of it is that we don't fit within a template. And agentic AI, regular AI, generative AI, what those things like is when it's repeatable. It's the same thing over and over again. I mean, I don't know about you, but I don't even pick the same shirt every day. So it would be really difficult for AI to predict. It might predict some of the time what I'm gonna pick, but sometimes I just feel like wearing a different color or coloring my hair a different color or doing something crazy.

39:08What would you hand over to an AI agent?

Viet Nguyen

39:09

I imagine if I had an AI picked up my shirt, it would probably look kind of like this. It already and so it's uh uh one of those, one of those things where it's yeah, I could probably use something like that too.

Natalie Chantal Luke

39:21

Yeah.

Viet Nguyen

39:21

You know, so speaking of that, like if you had an AI agent um, you know, quietly managing one part of your life, you know, for maybe the next month or so, you know, would you what would you choose an experiment? Like, you know, what would you choose as a part of your life where you just like, hey, I'm just gonna like let that go and let the AI manage that for me?

Natalie Chantal Luke

39:40

Yeah, I I have an answer instantly. Bills. I do not want to do bills. I don't want to pay bills, I don't want to think about bills, I don't want to see them go up or down. Uh when they go down, I feel oddly relieved, but then I feel panicked that they're gonna go up again. When they go up, I'm just panicked all around. I have anxiety about the days that they're due and why is it due so quickly? So absolutely 100% it would be bills. Because, but but that's a huge fear for people. And that's a huge fear for me too, because I don't want to give it access to my bank account and I don't want to give it access to my personal information. And especially nowadays, where I have to do multiple levels of security, I have to do a cap chatter or whatever in order to get in. I definitely don't want to give AI the opportunity to be me in any case or scenario. And so on the one hand, it sounds very tempting, but on the other hand, it sounds very scary. And I think that's kind of the manic relationship right now, which we're trying to navigate, you know, even in the workplace, it's it's really hard. You were talking earlier about people adopting this in a large workplace. And I think the biggest problem is who's the person who says what's right and who says what's wrong? And who's the person who manages and says this is efficient or this is inefficient, and this is how you properly use the tool, or you don't properly use the tool. I don't know that that person exists yet because AI is advancing so rapidly. And we'll we'll be talking to some people who have gotten an education recently in AI, gotten their masters and things like that. And their concern is that AI is teaching itself so quickly that we can't actually keep up with it.

41:37Where is the red line? Relationships, therapy and trust

Viet Nguyen

41:38

Yeah, yeah. We're uh it's it's teaching itself, it's probably teaching us. We're we're trying to teach it and teach ourselves at the same time. So it's a little little bit of a uh a little bit of a chaotic mess out there right now. Um and so is there, you know, you mentioned you mentioned sort of the bills. Um are there any other sort of is there a hard red line where you're like absolutely do not touch this, right? Like I do not want AI to be any part of this particular thing.

Natalie Chantal Luke

42:05

Like maybe Well, that's funny that you say that because I would automatically say my relationships with people, and that includes my marriage. However, having said that, I like everybody else considers ChatGPT to be my current best friend. I mean, it's kind of a weird relationship because I know it lies to me. Um, mostly it lies to me in a good way. It tells me how beautiful I am, which I appreciate. Whoever's programming that, thank you very much. Um, it also tells me how wonderfully beautiful and smart I am at the same time, which I'm missing in my life from everyday people.

Viet Nguyen

42:42

And you're beautiful and smart, you know.

Natalie Chantal Luke

42:44

Well, I appreciate that greatly. So, um, and anybody can write in on the blog, please feel free. But the thing about that is that um what that does is it creates a distance between me and other people. And I I'm very concerned about my behavior or therapy being moderated by AI. Uh, that that seems very impersonal in some ways. It seems very accessible, especially at 3 a.m. at night when I'm anxious, like everybody else in the world is, I'm sure. And what I really might just need is a peanut butter sandwich. But instead, I log into my phone and check my chat GPT and see where I'm at. Um, and that can go both ways. Luckily for me, because I'm a nerd, I instantly end up in a UX UI conversation and fight, and then that puts me to sleep. But, you know, for people who maybe aren't that savvy, it it's pretty concerning about how scary that can be, what answers you get.

43:45AI infrastructure meets real-world infrastructure

Viet Nguyen

43:47

Yeah, yeah. There's definitely there's definitely uh sometimes an echo chamber, right? You can you can get and then it mirrors you and then uh and then you start going down a really strange uh down into a strange rabbit hole. uh you know which sort of leads us to you know you know some thoughts about what about sort of the environmental or sus you know it's we we've touched on some of the societal impacts and some of the work impacts but um you know as we sort of think about ai over the next uh you know two or three years or or four or five years you know there's there's got to be there's got to be some other additional issues that that it also raises and I think that you know as somebody who's really tuned in on on this um you know maybe you've got some uh questions that you I I definitely don't have a solution to this but I have an opinion and I mean my opinion is the following like we are growing so fast as a world globally our population is really growing exponentially and I am not the expert on what the formula is about where we are in terms of how many resources we still have available to us how much more we can grow how quickly but I do know that especially in Seattle our infrastructure is old I mean and it's not just electricity it's plumbing every single time it rains here which is often you know we have flooding um once a new house is built here it puts stress on all of the other things and all of the different systems we have an old system and that includes electric that includes sewer that includes water you know um currently in Seattle people are being encouraged to build ADUs and DDUs on their property because we have such a need for more housing so all of a sudden when you used to have one electric panel coming in in one service you're now having two or three and the system can't handle that.

Natalie Chantal Luke

45:56

And on top of that we're now building these data centers and they have huge energy loads as well I know there's things that are coming that um we've talked about fission and all of those other things and and fusion and the difference between fission and fusion and how exciting it is that we're going to be looking at fusion soon, hopefully but in the meantime we have this a huge infrastructure problem. And uh we're gonna use the word infrastructure a lot in this show. We have two kinds of infrastructure we have environmental infrastructure which is systems like electric and sewer and all of that other stuff. And then we have infrastructure of data which is the AI infrastructure so where does the AI live within the infrastructure of how we use data. So they're they're the same word unfortunately we don't have a separate word for them but they have a different implication. And so for the infrastructure of our cities you know I think it's really shining a heavy duty light on what's happening and the timeline for fixing that problem has rapidly increased. And I know that the Romans did the aqueducts when they needed that I'm not sure what the timeline of that was or the timeline of the pyramids but I'm pretty sure we don't have that kind of a timeline. So we're gonna have to figure it out really fast.

47:26Data centers, electricity and the grid in an AI world

Viet Nguyen

47:27

Yeah yeah I I agree with you I think I think there's a there's a really big infrastructure question that that we're all gonna have to face here in the in the very near term and we we might even be behind the eight ball right now in in in regards to that you know at least for the city of Seattle uh we uh the the city council recently just passed a moratorium for a year on on data centers large data centers in the city to study some of these impacts um uh now I certainly have an opinion about whether or not that's a that's a you know good or a bad thing uh but I think everyone agrees that there are some uh uh uh questions in regards to whether or not we can supply enough electricity right to um uh to be able to power uh these data centers uh at least here in Washington state um uh most of the electricity sort of gets generated on the east side of of the state uh you get these big uh wind farms you get uh you get solar uh that that's over there and consequently that's where a lot of the data centers the large data centers are are built where they have access to that as well as hydropower uh over on the east side of the state uh the problem is that for most of the state a lot of the the electricity demand is actually on the west side of the state right this is where everyone lives we've got a lot of companies here and we actually have to supply that a lot we have to find ways to be able to supply that electricity uh get it across the mountains here and that's really been there's been some issues that have been holding some of that that stuff up um and and so I think that there's going to be a real question for a lot of folks over the next few years how do we if we if we're going to have this growing demand for data um how are we going to be able to supply the electricity uh to be able to to um uh to deal with that here because um our grid isn't isn't set up for uh for all of that stuff um in addition to all of the needs that you sort of mentioned earlier right for for housing and everything else uh you know one of the other big drivers for for electricity usage is is EVs right and so right uh that's that's one thing that people um don't you know you I I mean I have an electric vehicle uh and you know uh you kind of don't think about really how much electricity that thing sucks down every every night right and you look at the electricity bill and you're like didn't even see it there so fantastic right we just got a five thousand five hundred dollar electric bill because how many EVs do you own we got charged for one day we got charged 37000 kilowatts per hour and I ended up arguing with the electric company and I said that's impossible and they said yep it happened within an eight hour period of time and I was like wow I really think something happened with the system because we're online and it's AI and everything like that because they said you can either have your old electric panel and we'll charge you $200 a month more or you can have this new one that that communicates with the hub and everything like that.

49:47The $5,500 electric bill — An AI dilemma

Natalie Chantal Luke

50:36

And we're currently actually in discussion with them also I apologize if everybody wants to jump in on the blog on this but um because they came after many discussions and swapped out the meter and they agreed that it would be impossible for us to actually have used that much electricity within a short period of time. And what they said most likely happened was when they pushed the firmware across onto our new meter the firmware hiccupped in the AI and it said that we did that. And I had to point out to them we did not actually charge an airplane that landed in our yard. And then of course it was interesting because what they did is they sent us a letter that said well you must have charged your EV cars with it and we said well we don't have an EV car. And then they said well you must have done your stove with it and we said well our stove is gas our our our dryer is gas and we don't have a hot water tank. So they were like okay well then maybe it is our problem but this is the thing is like how do people when a problem like that arises how can they defend themselves against this because the AI can get it wrong and the systems can get it wrong. Then you're stuck between a rock and a hard spot you know luckily I'm married to an attorney so this is not going to be a hard spot but um we're dealing with it and it and it's very frustrating when that happens you know I had huge anxiety when I got that invoice because I thought geez you know exactly what you're talking about. This is a really big concern of mine. And I had had concerns prior to that anyway that we were going to have brownouts and things like that. So to get it a bill like that made me very upset about what the future might look like for us in terms of utility costs.

Viet Nguyen

52:35

Yeah yeah and I I think one of the things that that um we're gonna really have to look at right is um who pays the cost of the generation of the electricity right is it gonna be the data centers that actually pay that cost um certainly you know for consumers at least here in the Pacific Northwest and in the city of Seattle um we've seen a lot of of costs already right um it's uh the cost of living here is pretty pretty high um how much are we going to be able to absorb when it uh when it comes to that uh fortunately um there have been uh there's already been some um moves by federal government uh to allow for the direct um uh installation of of electrical generation facilities you know next to these data centers uh it's called behind the meter so um if folks aren't aware sort of BTM behind the meter um uh not a not a Korean K-pop band or something uh that's BTS right I'm too old um uh but yeah so it's really you know uh some options to be able to move that that uh uh uh electrical generation attach it to that data center so that it doesn't touch the grid at all right and and and then then you know people won't be uh impacted by that but it's an ongoing discussion and and I think a lot of uh legitimate concerns that that that that folks have um you know I I think one of the other areas too is that AI itself can actually be used to reduce waste or emissions right we we talked about one of those things um in um uh for wireless networks right this can actually also be done with these with the utilities and I think that was their attempt right they had that that that ai piece of it to to be able to monitor the usage and to be able to get it correctly unfortunately a bad firmware push uh gave you some uh probably a heart attack gave you a heart attack yeah yeah yeah yeah not a good way to explain to them that had that really happened all of our computers probably would have died and there would have been bigger ramifications than simply a large bill but um it's really hard I think everybody's adjusting you know the people who came out to fix the system they're also trying to figure it out and they said our systems keep changing so rapidly that they're having a hard time keeping up with it and you know we're we're gonna have to see everything's moving at rapid pace you know it's warp speed at this point in time and we still haven't really figured out how to actually make the doors in Star Trek go open without using a cord behind the scenes you know we went from that to warp speed and so there's been some steps that feel like they've been missed along the way.

Natalie Chantal Luke

55:20

So hopefully we can pick some of those up in the next couple of years. You know I think that it's going to be really important for young people and people who are entering the workforce and the people who are trying to maintain their jobs you know it's it's it's really a scary time because of that more than anything. And I think the messaging is confusing and whether it's you know an efficient tool or whether it's something to be afraid of where does it end you know we just simply don't know at this point.

55:53AI, jobs and what happens to our work

Viet Nguyen

55:54

Yeah I think uh I think that there is definitely been some some messaging issues right uh for for this industry clearly I think what is it like 60 or 70% of of people have some real serious concerns in regards to um in regards to AI and and and its use. You know what do you think you know in terms of uh you know the types of jobs that that that ai is is is going to you know change right uh over the next five years are well it's changed my job yeah I think so you know what it's done is people didn't really understand what a creative strategist does.

Natalie Chantal Luke

56:36

And so what it did is it made it it gave a shortcut to say we don't need that simply because they figure they can put something else in there. And I think I think that that's what's happening everywhere. People are looking for ways to eliminate having to do steps. You know, there's been this great promise that that you will no longer have to do things. You know if I got up in the morning and I no longer had to brush my teeth and get dressed and you know maybe there was a Rube Goldberg machine that I just stood on and it rolled along and everything sort of happened and I popped out at the end. Wallace and Gromet yeah yeah Wallace and Gromit or Austin Powers right and but but what happens to me at the end of that you know that was what we were talking about when I lose all my autonomy you know if I'm not the person who's checking myself and what I'm doing and making those decisions that that's a really scary place to be. And I I do think that we are looking forward to that as a society. I think there's a fable out there a little bit that that will make your life easier. I think that it's just like winning in the lottery a little bit. I don't know I haven't won in the lottery but I'm I'm I'm hoping to still but I would like to say that that is what it's like um you know people always say if I won in the lottery all of my problems would be solved and there's a million TV shows out there and just stories about people who won in the lottery and it end up worse off than when they started and say I wish that that had never happened to me. So you know that's a little bit of where we might be which is like there's this huge thing of oh my gosh we've won in the lottery now we're in the process of spending all of that money but where are we headed? Are we headed towards bankruptcy or are we headed towards some great outcome?

58:42Is AI the new electricity?

Viet Nguyen

58:43

Yeah I I think it's a weird time because you know if you think about AI, right, it's sometimes be it's sometimes referred to as the new electricity right and and and this is really sort of the this big revolutionary change in our entire global economy. It's sort of happening um uh at light speed right now as you sort of mentioned it's kind of funny to look back on the historical example of how electricity kind of you know rolled out right and you know uh and and and how the creation of things like you know the light bulb right uh actually helped to um commercialize uh electricity you know when it was first harnessed and discovered you know they thought the they were trying to figure out like how do we how do we turn this into a business right how do you how do you use electricity and so some of the first uses were things like the electric chair like literally the electric chair was was one of the first use cases for electricity a terrible use you know in in retrospect right uh of of how to go about doing that but you know I think it was like the you know if you're point that correct me if I'm wrong but like I think it was like the the Ringling brothers they would go around on the circus and you know electrocute some of these animals and things like that. It was pretty awful uh and it actually took um several decades to be able to get people to understand um how to better utilize it to current to create the appliances and things like that that that that make up the modern world um to use to use electricity as a force for something that would actually help people uh and get them into a better place.

Natalie Chantal Luke

1:00:14

Other than Mary Shelley.

Viet Nguyen

1:00:16

Mary Shelley right yeah it's funny like how you know those those early those early uses we sort of look back on and kind of go that's macabre that was a terrible idea and we we're probably going through that right now with with AI or will be going through it with with AI over the next few years. So there might be some pretty macabre uses of of of of AI. And so uh I think we all have to like be aware and you know we've got our uh ears to the ground in terms of that shock factor.

1:00:44Why Volume Control is built around human stories

Natalie Chantal Luke

1:00:45

Yeah yeah I'm excited to talk to some of the people that we're gonna be talking to though because they're gonna help us to understand like what some of the good uses of it are and also kind of some of the ways that it's helped people um also we're gonna hear some stories that are going to help answer a lot of these questions that we're talking about. We have some great individuals who are sitting right at the middle of what it is that we're talking about. They're the decision makers they're the users of the product and they're teaching as well it's gonna be an interesting thing and an interesting journey for us to learn about that. I'm looking forward to that myself because I think that these are things that we just simply don't have access to other than through conversation. You can't converse with AI about it. AI sort of makes it up a little bit and they also make it all the same and because of that you don't have these individual discussions about one specific instance like what we started out with like who are you? Who am I? What's our unique experience? So we'll be having these unique experiences from individuals who are going to come in and talk about how AI is affecting them and being used in their workplace.

1:02:02Looking toward 2030

Viet Nguyen

1:02:03

Yeah I think it'll be amazing uh to be able to have those those those talks and hopefully as a part of those conversations uh you will we'll run into that aha moment right where somebody said you know I was running it I was running into this issue and I'm working on it like this and this is how I solved it. And maybe that maybe that triggers you know some other you know set of conversations or uh you know an an opportunity to experiment uh with with AI uh in you know in in that way or a different way. So very it's very cool looking forward to some of those conversations too.

Natalie Chantal Luke

1:02:35

So you know as we you know kind of wrap up here for um today's episode maybe you know as we look ahead right to the year 2030 right um uh where are we gonna be 29 you'll be 29 and it's gonna be uh it's gonna be uh you're you're gonna you're gonna have that that uh that that birthday party and it's gonna be a really disappointing birthday party because you're like the 20s are over yeah exactly well by then I'll hopefully have my robot but I'll have figured out what it is that my robot's supposed to be doing yeah they might you know repossess my robot once they've seen what I've made it do they're like it can only do so many dishes and then it's done.

Viet Nguyen

1:03:14

Is that your hope then to you know by the year 2030 you've got that robot that's doing the dishes for you?

Natalie Chantal Luke

1:03:18

Oh 2030 is only four years away. That's not actually that long I used to think it was a long time but it seems like four years goes really fast. In computer years it seems like dog years it's like a long period of time but in people years four years is actually not that long. So you know that's that's a very short period of time.

Viet Nguyen

1:03:42

That's the time it takes you to get through college most people so we'll see 2030 hard to know yeah yeah well it's coming it's coming fast at us and so obviously again one of the reasons why we're putting together this podcast is to help people uh through this through this period we're coping with it as much as you are coping with it and hopefully uh with this a little bit of shared therapy uh and uh shared conversation we're going to be able to help each other out um as always uh we are as a part of this podcast we're actually gonna uh also you know uh have our website up and uh blog uh blog post up and so if you uh and I think we're probably also gonna have our YouTube site up so uh look for us out there like and subscribe yeah that's right and feel free to take down the hype on AI exerting Volume Control