0:13 Hey, welcome to the AI Zeitgeist. I'm Austin. And I'm Luke. And this is us following through
0:22 on a promise we made to each other about a week ago at a friend's kid's birthday party.
0:30 Exactly. And this is a friend that we both know for different reasons and they've got three kids
0:38 so we see each other about three times a year and that's the only time we talk. And every single
0:44 time, just get into this AI conversation and talk to the whole party about AI.
0:48 Yeah, we monopolize the party.
0:51 Yeah. And my wife just looks at me like, stop talking and pay attention to what's going on.
0:58 So we're like, okay, maybe we should pay more attention at the birthday party and we can take
1:01 these conversations and make a podcast. And that was the idea at the end of the party. We were like...
1:06 Yeah, Luke was about to leave and I was like, jokingly, we should start a podcast. And he's like,
1:11 yeah. And then we're like, wait, are we going to do this?
1:15 And it was like, the thing is that three, four years ago, if you had said, let's start a podcast,
1:22 man, think about all the things you had to do to get that set up and the amount of lift and
1:27 the weekly commitment. But the reason that we were able to jump on here just one week later to
1:34 record our first podcast, Austin was able to build out an app using AI that basically is letting us
1:42 create this podcast in an automated way. So that way we're actually able to just focus
1:47 on recording the content and then all the downstream stuff gets handled. But I'll let
1:52 you talk about the specifics of what it's doing. Yeah. So we're both, we're not in the same place.
1:57 We're both at our houses and our kids are asleep and we're both using good microphones.
2:04 And it records both of our mics locally at each of our computers and uploads every 10 seconds
2:11 chunks of the audio to a server. And then it mixes our voices together and then sends it to
2:19 OpenAI's Whisper to transcribe what we're saying. And then it will edit the whole podcast and cut
2:26 out things that we don't want in there and any silences and stuff like that. So it's, our idea
2:33 was to have it as, as little effort as possible. Yeah. And like you even have it set up so like
2:39 it's drafting descriptions and stuff too. Yeah. Yeah. It generates everything for social media
2:45 posts and stuff. And I'll probably like the cool thing with AI is I have experience doing some of
2:53 this stuff and now I'm trying to have AI do as much of it as possible. So I just use cursor
3:00 and kind of just one-shotted this and then it was pretty much ready to use right off the bat.
3:07 But then I iterated on it, added a few features, but but I'm finding with AI, like even with my
3:14 technical background, that sort of gets in the way. And so I found that it's easier for me to,
3:20 like, I'll type out a prompt and I'll have too much technical detail in it. And so then I'll be
3:25 like, wait, I should just delete most of this and have AI do it and see what happens. Okay. So
3:29 cursor, that sounds complicated and everything. What is like a way that somebody could kind of
3:35 get their feet wet? How, how would they, would they jump in on this? Probably the easiest thing
3:39 is, I mean, a lot of people are using AI in some form or another with it, whether it's like
3:45 ChatGBT or Gemini or whatever. And so you can just start ideating with whichever AI you use
3:53 and just say like, hey, I'm thinking about making this thing. How would I even start? I have no clue.
3:59 And the cursor is really cool. Like my brother-in-law, he doesn't really have any background
4:03 in programming or anything, but I was talking to him about it and he's like, I want to make this
4:07 like dungeon and dragons thing so I can play with my kids and I don't have to be the dungeon master.
4:14 I don't, I don't do that. So I think it's called dungeon master. And so he's, he's been building
4:18 this AI thing to do all that part of it for him. And he's been going crazy and he's been using
4:25 cursor and he's using GPT to tell him how to use cursor. And so he like asks different questions
4:33 to get it going in cursor. And he's built this pretty crazy thing. He's burned through a lot
4:36 of tokens though. And he's already, he's already hit his limit for the month, I think. Yeah. But
4:43 it's the, the thing is that that is the key right there. You don't have to know how to do it. You
4:48 just have to be able to, Hey, how am I doing this? What's, what's the next thing? What's,
4:52 what's, how do I, it's like you explain your goal and then say like, how do I do what I'm trying to
4:59 do? And then it walks you through how to do what you're trying to do. And like, you have to open
5:06 up your mind. Like, I don't, I don't know what I'm doing. And so you just have to open your
5:10 mind up to like, yeah, no, it's like, exactly. There's like, what it is is you have to just
5:16 have the, this is possible mindset. Right. I don't know exactly what the way is to get there.
5:22 They're not like, not like, why can't I do this? Right. Or I can't do this, but think of it like,
5:27 why can't I do this? And ask, chat GPT or whatever. Yeah. It's like, how do I do this?
5:33 Exactly. It's like, you're sitting there and you're thinking about what you want it to do.
5:39 And it's like, I have no idea how it's going to do this. That's fine. You don't need to.
5:45 That's what a lot of people stop themselves with. They're like, well, I don't have any
5:48 programming experience or like, I don't even know that I need to be programming for this thing to
5:53 work. You don't need to know. You just, you just need to be able to ask an AI how to do it. And
5:59 it will tell you how to do it and tell you the tools and then it'll actually do it for you.
6:03 Exactly. One of my favorite thought leaders, Andrej Karpathy. And one of the things that he
6:10 says that's like really, it's, it sounds, sounds silly, but it's, it's like English is the, which
6:16 like we are programming. Yeah, exactly. Like you just say what you're trying to do. So like,
6:21 if you know how to write and you know how to read English, which I mean, pretty good bet that you're,
6:27 you're literate if you went through school, et cetera.
6:29 Yeah. But even if you're not like the AI understands you, even if like you're not
6:34 spelling things directly or formatting words correctly.
6:37 You can use voice to text, you know?
6:39 Yes. Yes. That's a lot of thing. A lot of people do that now.
6:42 Yeah. But so one of the things that is interesting about this whole phenomenon is that
6:49 because you can build stuff means that you no longer have to like get pigeonholed into
6:56 standard application and you can make something custom for yourself. Like the first example that
7:02 we're talking about is the way that we're recording this podcast. Instead of going to off the shelf,
7:07 Austin was able to build something that was custom, but we actually needed for it to do
7:10 exactly what it needed to do. And there are, there are tools like that already in existence,
7:15 but they cost money. And this was, I mean, I'm, I'm paying for hosting for the server,
7:20 but that was something I was already paying. Exactly. And it's, it's not even necessarily
7:24 the money. It's more of the fact, you know, because that's part of it. But the other part
7:28 of it too, is that you're getting a tool that is dialed in. And a tool that I can keep adding to.
7:35 Exactly. As we realize like, oh, this doesn't really work out for us really great. I can add
7:39 tools and tweak it. So like that is the model. And one of the, one of the very small example,
7:45 but the other day, one of my friends, he sent me a application, Google play app. And it's like,
7:51 this is a, this is a way you can track your, your putting stats. I play disc golf. And so
7:55 I'm trying to practice my putting and it was doing groups of 10 and like, I don't have 10 putters.
8:02 I have three putters. So I need to be in multiples of three. And it was able to create an application
8:07 for me that basically is able to let me do like that specific thing without having to like,
8:16 reconfigure the way that I'm throwing to like match the way that everyone else practices.
8:22 Right. Right. And then as you're using it more and more, you can decide like,
8:26 oh, maybe this would be helpful if I added a timer for whatever reason to see.
8:31 Or like, maybe I buy a few more discs and now it's like, I get one more.
8:35 Right. Right. You can add, yeah. You buy one more at a time and you can like,
8:38 you can step into it.
8:39 Right. For sure. So, but yeah, I mean, just like it's, it's, it's literally,
8:43 you know, that, that simple. And one other story, like the other day I was just watching
8:50 somebody on their phone and they were pulling up information and they pulled up the information
8:54 that they were, it was a random question that we were like, oh, I wonder what you have a question.
8:58 You pull out your phone, you Google it. They jumped into Google and then they scrolled down
9:02 and they clicked the Wikipedia link and they pulled up the Wikipedia page and they were reading
9:06 through the article. Oh yeah, this is the fact. And it was like, it felt like somebody had a
9:12 landline rotary phone and they were making a phone call. Like, you know, it's like, to me,
9:18 the way that I would do that now is I would jump into a chat and I would ask the question and it
9:24 would give me the specific answer to the exact question I was asking. Not like a tangentially
9:29 related Wikipedia article. And you can ask follow-up questions. And yeah, I drive a lot
9:35 for my work. And so I'll ask, I use Gemini a lot and I'll ask it a certain thing. And then I'll
9:41 just like spend the whole drive just talking to AI and learning while I'm driving and asking
9:47 follow-up question after follow-up question. I mean, you, if you have like, this is probably
9:51 the best time to be born in. Oh yeah, for sure. For sure. It's crazy. I've been using, I've been
9:57 using AI since probably like a week after ChatGPT came out and I use it an insane amount,
10:06 like every minute of the day. My wife is like, she'll like look and I'm on, I'm on, I'm on like a
10:12 chat with Gemini and she's like, you talk to that more than me. I'm like, well, it knows more than
10:19 you. So I want to, I want to come back to that. But first I think you had something you wanted
10:23 to talk about, an application you built out that you, so you were doing like paging at your
10:28 company? Oh yeah. So my company, yeah, my company, our phones stopped working like Comcast or
10:35 something cut it off. And this is a whole big thing. So our phones were out for like a week
10:39 and then we got them back on. But I realized maybe like a few weeks later that the paging for our,
10:46 for our office, like so-and-so you have a call online too, wasn't working. And that's when someone
10:51 picks up the phone and they can talk over the loudspeaker. And so I was trying to get a new
10:57 paging adapter and it kept getting delayed and stuff. So I had this thought that I could build
11:04 a way for people to type in what they wanted to say, and then it would do a voiceover over the
11:08 loudspeakers automatically. So I had it, I knew I wanted to use OpenAI's API to generate a voice.
11:17 And then I was also trying to think of like how, what's the easiest way for them to input the text.
11:23 And I had thought to do it over Teams. And so I had Cursor help me, Cursor and Gemini
11:30 helped me create an application that is running on a laptop up by our amplifier and
11:38 it integrates into Teams. So now people just type into a channel, they tag this pager
11:43 and they type whatever they want. And it says it over the loudspeaker and I have all these settings
11:47 for like repeating it, how many times you repeat it. And then I have a transitional word. So it'll
11:53 say like, Austin, call online too. Again, Austin, call online too. But it just repeats the message
12:00 and I can have it do as many times as I want. And it was just something like I had a thought
12:04 and I had it working like within a day. I can't even imagine doing that two years ago.
12:10 But now you've given the AI a voice.
12:12 A voice. Yeah. But everyone loves it at work because no one wants to hear their own voice
12:19 over the loudspeaker. And it's far clearer than like people actually talking on a phone. So
12:25 it's just, it's nice. It's pretty cool. It's crazy that I could do it.
12:28 That's true. Like you don't necessarily want to hear your voice.
12:31 Yeah. It's like the receptionist who does it the most. She was like, I love this.
12:35 And she's also, she's, she's got an accent because English isn't her native language.
12:43 And so it's probably even better for her because she can get out what she wants very clearly
12:48 to everybody else.
12:49 That's awesome. That's great. Okay. So I want to go back a little bit. You were talking about
12:54 using Gemini a ton and like you've been basically using it like super power user,
13:01 since day one, basically. I'm in the same boat. I jumped in kind of around when GPT 3.5
13:07 kicked off. I saw the change from 3 to 3.5 and I was like, so I've been using it a lot.
13:13 I have this theory and we were talking about it at the birthday party. And this is kind of,
13:17 I think what is related to our name, AI Zeitgeist. Everybody is in these chats,
13:24 not everybody, but the people that are using AI a lot. And there's like four,
13:29 three or four big models. Like there's Claude, there's chat GPT, there's Gemini, there's Grok.
13:35 Maybe there's a couple of open source models that we're using, but those are the big ones.
13:38 And they, they kind of have their own personalities.
13:40 Yeah. But they're all, but they're all kind of the same. Like they're,
13:43 they have their own personalities, but they all, they use the same technology to build them. So,
13:48 and they're all, they're all large language models. So they all kind of have a similar shape,
13:52 even if they're slightly different with different personalities. Like they're all,
13:55 my theory is like, we're all like thinking alongside of AI, like a co-thinking almost,
14:01 honestly. And so like, I just, there's, there's like almost like a Zeitgeist,
14:06 like a moment in time where everything is like tuned to the same frequency,
14:10 where you can kind of like, I don't know, it's kind of hard to describe.
14:15 You were saying before, like how you'll have a thought about something and then like the next
14:20 day you'll hear your thought in a podcast. And so it's kind of like, we're all getting the same
14:26 we're all talking with the same person and they're like crafting our minds.
14:30 And so there's like this, this mind hive or Zeitgeist.
14:33 Exactly. That's the creepy way to think about it. But I think one of the other explanations though,
14:38 is that like, we're all seeing the same information and we're all like processing it
14:45 and through AI and that processing is going to lead to similar like ideas because we're
14:54 co-thinking with the, you know, Oh, like context is a big thing. Like this, is that the other,
15:00 like maybe the idea of, okay, we need to start thinking about coding agents, like
15:05 whatever the topic du jour is, there is like this point where there's like an inflection of
15:15 something that's going on that multiple people are hitting at the same time.
15:19 And then they're all interacting with the AI and it's helping them find very similar solutions.
15:25 And then you start to see those solutions happen simultaneously, like popcorn, like
15:31 within days of each other.
15:34 Well, it's crazy. It's crazy. Like how, how fast everything's moving, like the new thing every
15:39 week, something new comes out and it's like the latest and greatest technology. And Oh,
15:44 I was even hearing, like, we listened to a lot of different podcasts. So like a lot of the
15:49 podcasts, they talk about different eras in AI and like, they're talking about like two months ago
15:55 was this era. And so it's like, it's just like weird how fast it's going. And like the things
16:01 they're talking about were in the past just happened like a few weeks ago. Yeah. And now
16:06 it's like old, old news history. A hundred percent. One of the, we were talking about
16:10 this earlier. And one of the things that I've been finding tools and like things that I've
16:15 made that I relied on like six months, three months ago, dropping them, not because they broke,
16:21 but because like the next version of chat, GPT does it natively. Right. Right. Right.
16:27 It kind of makes it useless. And that's like, I think probably the most important thing
16:33 when it comes to like the speed that this technology is iterating is to stay flexible
16:38 and like stay ready for it to take over more of what you are doing. Like from a technical
16:44 perspective, it may negatively be able to things like as it gets better. But the other thing that
16:48 I think is important to do is to like understand what your core beliefs and values are, like have
16:55 those set up pre ahead of time, because you don't want to be like completely unmoored and have the
17:03 AI. Right. Right. That's my advice anyway, because I think, I think we definitely are in a,
17:09 you want to still retain some individuality. Your humanity. Exactly. Exactly. Yeah. So I,
17:18 in that, that mindset of things updating so quickly and like, oh, now Anthropic's the best
17:24 one. Oh, now Gemini is the best one. Oh, now it's open AI. I try to keep, like I'm using cursor,
17:30 which actually now is becoming kind of having its own model with, with, with Grok. Right. And
17:37 compose. Yeah. Compose. And so like, but cursor, you can use ChadGBT, you can use
17:46 Anthropic's models and stuff. And so it's, it's nice being in a program that I'm kind of invested
17:52 in where I can do different models. Whereas like if you're using open AI, they have codecs. And if
17:58 you're using Anthropic, they have clod code and stuff. And so what I'm trying to use in Buildout
18:04 is, isn't dependent on a specific AI company. Totally. So I can quickly like transition things.
18:10 It's one of the things that I've been working on in that same vein is I've been working on
18:14 an application that basically records a bunch of my notes using an open source,
18:20 open knowledge format that Google released. And it's basically like if you, and it's Markdown.
18:25 So anyway, it's just a bunch of like text files that the AI records. And what I'm able to do is
18:31 I can use that. And then I can have, I can have clod use that same, exactly what you were saying
18:36 with cursor. It's like whatever system you build, whatever system you set up, whatever system you,
18:41 you say, I'm going to use this one. Like, because they're all kind of leapfrogging each other,
18:46 things are changing every day. It makes the most sense to be as independent of picking one model
18:52 to really run with, as you possibly can be, because you don't necessarily know where it's
18:56 going to end. Right. And the newest, greatest thing could be on a different, from a different
19:02 lab next week. You're like, well, I want to use that thing now, but I'm locked into this thing.
19:07 Right. Yeah. No. And I think that's the most important thing is to keep all of your information,
19:13 your information, because that... Right. So another project I've been working on
19:19 using AI is I, I work on commercial sound systems at grocery stores. One of the hard things is like
19:25 going around and trying to listen and see if a speaker's working or not. And it's really hard
19:30 to do with, with noise around. So I came up with this idea to play a high frequency that humans
19:36 can't hear over the loudspeaker. And I've built this device that I've used AI heavily to design
19:44 and to create software for that. It's got a screen, touchscreen and a microphone on it,
19:49 and I can walk around the store and I can tap on a map of the store and like create a speaker
19:56 and I can record its volume, like the ambient noise and that high frequency.
20:01 And so I can go around to each speaker and do all this. And I can see,
20:05 I have like a visualization of what's going on with the speakers. I know most people don't do
20:09 this specific job, but it's something like really niche to me that is like extremely helpful.
20:16 And it's just something that I had an idea for. And so, and yeah, like typical software
20:22 development, like there's something and they will go out and they'll say, okay,
20:26 what is, you know, how many people out there do commercial sound in grocery stores or,
20:30 or retail store? What's the number of companies? How then they'll figure out, okay, which ones are
20:34 willing to buy specific software? And then that now I have to think about marketing and acquisition,
20:39 and then you put all that number into a formula and then it determines whether or not.
20:43 Even if it's even worth it.
20:44 Yeah. So like there's software that does not get made because it's not something that is going to
20:49 be, there's not a large enough market for it to be profitable. Or you could just make it yourself
20:53 with AI.
20:54 Right. Right. And so like, I've been, I've been doing this job for like probably like five years
20:59 and I had early ideas of this type of thing, but now like, now that AI is so,
21:07 that I'm using it so much and it's so much easier to use now, like this is way I couldn't
21:14 have imagined with my original thoughts of this thing, like what it does now and what it's capable
21:18 of.
21:19 And I think there's something there that's like, there's going to be a long tail of software
21:24 where there's a bunch of different niche cases that are going to specific applications built out
21:29 by the users themselves. And that's going to allow for very specific tools they're doing.
21:36 So things are going to get easier because you're going to have like very specific software that's
21:41 built specifically for the thing. The one thing that I see a lot of people try to do,
21:45 what's the way to like cash in on this? Because obviously it feels like the nineties will be,
21:50 okay, there's, I know this is a thing, but I think the, that's the trap there is to try to
21:57 sell something to other people when I think the best use of AI, at least right now, like as we're
22:03 kind of still figuring it out is to make something that's useful for yourself. And I think that's
22:11 probably the best place to start. Yeah, for sure. So I think that is a good place to land for our
22:18 first episode. So we will wrap up here, but I'm super excited to jump into our next one.
22:25 And I think this has been a great start. All right. See you next week.
