00:00Melody: Internet checkpoint.
00:01Henry: That’s what we wanna call it. That’s a good idea.
00:03Melody: Yeah. Henry recently shared this YouTube video with our group chat. People just decided on a random YouTube video. It’s like a Japanese algorithm pool. The algorithm pooled folks, and so it’s this Japanese—
00:22Henry: The video is in Japanese, or the title of it, so you might not know what it is. But there’s an interesting background. It looks like vines or something.
00:32Henry: But it’s actually from Donkey Kong, the video game. So I think there’s nostalgic elements.
00:41Melody: What type of music? Yeah, like that.
00:43Melody: So then all of these people are just in the comments doing an internet checkpoint. Usually they’ll say “internet checkpoint,” and then they’ll share something super vulnerable. A lot of them are anonymous accounts. Some of them are actual accounts too, but they’ll just share something and everyone’s just in there deciding to be in the comments to talk about vulnerable things.
01:09Melody: So that’s kind of the power of the internet.
01:13Henry: What was resonating to you? There’s been many videos made about this as well, and I’m sure there’s multiple videos too. Maybe people feel like the algorithm is bringing them towards brain rot and stuff like that.
01:31Henry: But then they find, even the lo-fi community in general is more positive and it’s just about listening to music and studying or working. And then maybe people feel like they randomly got placed into this video and there’s no reason to find it because it’s in a different language or all these things.
01:52Henry: And then once one person does it, or two people, it snowballs. There’s no other way to find it. Now that we’re talking about it, other people were talking about it, there’s a different reason. But before, it felt kind of serendipitous.
02:07Melody: I love the name. And I think, first of all, we’re eating lunch while doing this. It’s actually very clearly like 3:00 PM.
02:23Melody: But we’re having lunch and so you might hear some eating noises.
02:28Henry: I think that’s okay.
02:32Henry: I just feel like when I want to do a podcast, I don’t want it to be like we have to do something high production. I don’t wanna have to go out of my way to do a podcast, even though obviously you do.
02:46Henry: This is sort of like what I was thinking of with the walking podcast. We could do something else. And then I feel like the podcast is a very sterile environment. You have this production, you’re recording, everything’s nice.
02:58Melody: It’s like a natural conversation.
03:01Henry: Just not real life. It really isn’t. At least you’re doing something else. You can’t make it more real. It’s all fake, but I do feel like when you’re at least doing something else, it’s just interesting.
03:14Melody: I agree. But apologize for the chewing in advance.
03:21Melody: But I think we felt, and also this was kind of a last minute thing. Well, should we tell them what we’re having for lunch? Fried chicken. Great.
03:35Melody: You said it was your first time making it, right? It was my first time making it. It’s a simple recipe. I’ll drop it in the description box below.
Why This Felt Urgent
03:49Melody: So I think we were debating whether or not to even record this. We were just gonna have lunch before meeting up with a friend for coffee, but we decided we’ll do both. So here we are. And I think primarily there’s some sense of urgency. At least for me, I really wanted to capture this moment in time.
04:07Melody: And that’s why, bringing it back to why I resonate with internet checkpoint. I think that’s a great little phrase. If you work in tech or even if you’re tech adjacent, you see the headlines and it’s just an insane time in technology right now.
04:25Melody: Every day feels like light years passing. And so just to be able to have a little timestamp of how we’re feeling at this moment feels like something that I just wanna have. Henry and a few other friends and I are in a group chat primarily talking about AI and vibe coding and that sort of thing.
04:51Melody: And every day there’s something new happening or we’re all building random things. What is this moment in time for you?
05:14Henry: The feeling?
05:15Melody: Yeah. How are you feeling? That’s a good place to start. We can both talk about how it feels.
05:22Henry: I was thinking of another podcast which used to be called Future of Coding, and I liked that they renamed it to Feeling of Computing. Referencing Omar and Tim’s essay, “Computer is a Feeling”. Maybe because people don’t really associate computers and feelings at all, but the history of technology is the history of people. Everything has to ultimately involve people.
05:58Henry: I think maybe a lot of people are feeling overwhelmed. I’m not sure if I feel overwhelmed. I feel excited. And maybe that’s all because of how I’m using it or thinking about it or not working a regular job. So I’m not thinking about losing my job. I already left.
06:20Henry: The last few years it’s always been, “What’s gonna happen?” Maybe it’s easier. I already quit.
06:32Melody: But you do have to explain that people are gonna feel some sense of—
06:39Henry: No, I acknowledge that. That’s real. But however we’ve figured out ways to support ourselves outside of a traditional corporate job.
06:55Henry: And maybe that was like preparation or something. I’m thinking now, if you can make anything and all these thoughts around ideas mattering more than implementation, assuming that it gets better and just how good it is now—what are we gonna make? What’s important? What matters? Everyone’s talking about how there’s gonna be a renaissance. Personal software. Making things for yourself and disposable software. All these different words that you can use.
07:22Henry: But then if you wanna go one step further, you don’t have to make only things for yourself or only things for your company. There’s a huge range of things in between, which is groups, which is your community, which is your family, which is your church. Making things for community. I don’t know if it’ll inevitably happen. I think we have to push for that. It’s really easy to default to making money for a company or doing something for yourself, but how do we actually build for a community?
The Vibe Coding Journey
08:03Melody: 100%. If this is really a timestamp, maybe we can talk about—I feel like you’ve been vibe coding the last year since the term was coined. Vibe coding was coined in February 2025. A few days ago. An exact one year anniversary.
08:30Melody: And so Henry, since then, I’ve been hearing him talk about vibe coding. He’s been telling me about Claude Code, and it always still felt a little daunting to me. Just the fact that you’re interfacing with it in the terminal. But for some reason it just started to really gain steam in November, December.
08:55Melody: At least for me, this is the way I felt. I didn’t even know what was happening from a technical perspective, but just my feeds were starting to be full of Claude this, Claude that. And I was getting so annoyed. I was like, I’m gonna mute Claude Code. And then two days later I finally sat down to play around with it.
09:13Melody: I think a good starter thing was, people were talking about organizing their downloads folder. That’s why I had everyone do that for the workshop. And that was a magic moment where I was like, whoa, this is so crazy. Because Claude went in and was organizing all my files.
09:32Melody: And then from there I just kept building and building and all of a sudden I built an app in a week. Idea to production. I had authentication. I had a database. I had all these things I’ve never been able to do before. I’m pushing to git, I’m using git—I’ve never used—
09:58Melody: You can tell them.
09:59Henry: I forgot what month. It doesn’t matter what month it was.
10:02Melody: It was August of last year because it was my sister’s wedding.
10:06Henry: I just remember Melody was like, “You’ve been working with printers.” And I’ve been working with printers for, maybe it’s two years. I don’t even know anymore. But that was when AI was still more like autocomplete. So I did use it, but ultimately I still coded everything and it was so hard to get anything working. And then also I think printers are cool now for some reason, receipt printers. So I made one where you can message me like a fax machine. And I had a lot of people ask me to help them make one. I guess I said yes for you.
10:48Henry: It was a custom website for the wedding where anyone can use the laptop to do a selfie.
10:56Melody: That’s when I was trying to vibe code. And then I knew I couldn’t—
11:00Henry: And I gave you access to my GitHub. And then you can fork it or whatever. I gave you access. I didn’t think Claude would automatically push. So basically you ended up force pushing and it just broke everything. So I was like, at least I’ll make my own branch or do something else.
11:32Henry: It was just your first, not your first experience with git and all that. But just in a real project. And then the first thing that happened was force pushing broken code.
11:45Melody: And it still didn’t really make sense to me. I technically was using the terminal at the time. So that was August of last year. I wasn’t bit by the bug then, even though I was dabbling in it. And then it was in January of this year.
12:03Melody: It’s early February, I have to say. I got so Claude-pilled. If you talk to anyone that has started to play around with Claude, it just kind of takes over your life.
Why It Feels Like Gambling
12:03Melody: Someone mentioned that it’s essentially like gambling because it’s variable rewards. It was actually Omar that said this. There’s something about, I guess you are awaiting the output.
12:39Henry: What are you gambling? That’s a good question. You wanna go deeper into that?
12:44Melody: Sure. But I think what I’m trying to say is, it’s just very addictive.
12:49Henry: It definitely is. We talked about staying up late. Normally you stay up late ‘cause you have to finish a line or all these things. You’re not even in the flow. It’s coding for you, but you’re in the flow of thinking about what you wanna make.
13:06Henry: And the problem is for me, and maybe for a lot of people, you can get distracted. You want to start with something and then it ends up like you want to polish this random button. ‘Cause you can. And it can do it. And maybe the gambling part—I’ve had experiences where there’s a bug and then you’re just prompting and not looking at the code.
13:28Henry: And then the hours go by and you’re just stuck. You’re just waiting for it to finish and you could’ve just, well, if you know how to program, you can just fix it.
13:39Melody: That’s not me. I’m wasting tokens on the stupidest things. Like cd, I could just cd—
13:47Henry: The git thing.
13:48Melody: Yeah. Do the git thingy. So this is how I’m truly a vibe coder. Karpathy coined a new term, which is agentic engineering.
How the Design Process Has Changed
14:03Henry: If you’re still thinking about what’s happening—in a way, there’s no such thing as actual vibe coding other than, you just say, “Make the thing.” And then you don’t describe anything about it. I don’t know why you would expect it to be good, because you yourself didn’t give the AI any context. It has to assume so much. That’s why planning is good. I don’t even know what I want it to be.
14:33Henry: And so maybe that’s what’s fun, is that you’re using it and building on the fly. I see the output and then I continue—
14:42Melody: That is one thing. There’s so many things to talk about, but one of the ways my design process has changed is that the output is so easy to get to that it actually shapes the idea.
14:57Melody: Normally it’s a much more linear process to go from idea to wireframe to prototype to end. But now that it’s so easy to get to the polished end state, it’s actually faster. It crystallizes your idea faster for you to go to the end, recognize where your mental model or data model was wrong, iterate. You kind of go from the front and back and then you whittle. That’s kind of how I’ve been thinking about the design process for me.
Abstraction All the Way Down
15:26Melody: The other thing is, we’re now operating at a higher level of abstraction. One way to think about what’s happening is, back in the day, there was binary. We were working with hardware and zeros and ones. That’s a signal that’s coming out. Then you have another layer of abstraction where you’re designing languages to interpret these things and imbue meaning.
16:01Melody: Software programming languages have evolved and different technologies have evolved over time. We’re kind of just moving further and further away from the metal. And now it’s really just at this really high level mental model where you’re doing the work that software engineers have to do at the beginning of a project—systems design, data architecture, the data model—doing things at that higher level. Which obviously really have to map to the user experience.
16:46Henry: Yeah, you could do all of it. That’s why I like having a sense of design or caring about the product, if it is a product, the users, all that. You can do all that at any moment.
”The Hottest Programming Language Is English”
16:56Henry: I think I probably had another tweet that went viral: “The hottest programming language is English.” Or whatever language you use. But it’s all language all the way down. Whether it’s zeros and ones or assembly or a programming language—English is still a language. It’s just that you don’t have to know the specific version of it.
17:09Henry: And it’s funny because you don’t have to care about the binary, low level hardware. You might not have to, but all of this technology is based on needing to know hardware at the fundamental level to make it efficient and cheap and easy to use.
17:35Henry: It only works because GPUs are so good. We shouldn’t forget that. At the end of it, the cloud is hardware.
17:50Melody: And then on the flip side, for someone who has lived at this more abstracted level, who thinks at this abstracted level—to be able to then execute from that level has just been so fun.
The Faucet Turned On
18:14Melody: This is just a longer way of saying I’m an “ideas guy.” I’ve always just had so many ideas. And one of the things that I remember texting our vibe coding group chat was: “Guys, this is seriously a problem. I can’t go to sleep. I feel like a faucet has been turned on and I cannot turn it off.”
18:31Melody: And so it’s 2:00 AM. I’ve tried to end my session multiple times. And then I’m sitting in bed and my mind is just swirling with—
18:49Henry: What’s possible.
18:50Melody: It’s kind of overwhelming because the capability expansion is so wide and it’s no longer unavailable to you. You can actually do it. And so now it’s more like, what is the path you focus on?
19:12Henry: If you know that it’s possible—
19:15Melody: And back in the day it’s like, could you do it? No.
19:20Henry: Well, I would say everyone could have done it. It’s just the amount of time. It’s kind of like anyone could do whatever they want. You have agency to do it. But realistically speaking—
19:31Henry: And then now, even as a programmer, things that I wouldn’t have done, now I can do. Or at least I feel confidence. That’s why for me, a lot of it’s just confidence. We could have done all of it. It’s just that you didn’t think it was worth the time. And maybe now you think it is because you can offset that time by AI doing it for you or exploring the idea for you.
19:53Henry: And even for low level stuff. Say if I only have experience doing front end. Now I can make backend, now I can do an iOS app without needing to know the language, but I can still use the principles of programming or vibe coding to make that stuff. This is gonna change everything because even open source—I don’t even know how that’s gonna be the same anymore because you could theoretically make everything from scratch now through the AI and just using other people as reference.
20:25Henry: It’ll be interesting to see what dependencies will still be necessary. The problem with the XKCD comic was that this one person maintains this thing forever—that was Babel for me and other projects. And there’s no time. There’s different problems.
20:47Henry: It just changes the shape of it. There’s still gonna be the same amount of dependencies on one person. ‘Cause it’s more like, does anyone care enough to run it? Is there enough money to do that? And then also, if it’s that important, we can’t just run AI to make things work. It’s not gonna be secure. There’s gonna be people trying to attack it on purpose.
21:05Melody: Well, Anthropic is using Workday. The best place where they can probably create their own solution for something. They’re still buying SaaS software.
21:24Henry: That’s a good point. And the whole joke around that—they’re still hiring people for a lot of money. If it was that good, they wouldn’t need to hire people anymore. It should be obvious, but those are good examples. They’re still using software ‘cause most of the time all the problems are people problems and you’re not gonna solve that by just—
21:48Melody: Yeah, on an aside, the way I think about it with the Workday thing too is you also have to think about ruthless prioritization. Same thing for Anthropic. It’s a core capabilities thing. You have to just focus on what your core thing is, and that’s why they’re gonna pay for top engineers to still work there.
22:22Melody: But I’ll just say, I’ve been jokingly calling it Claude-cosis or whatever, being super Claude-pilled. I definitely felt a sense of urgency. I wanted to talk about the workshop. Very quickly I was like, we need to host a workshop. Tomorrow. Eventually I ended up posting it the Wednesday afterwards.
22:51Henry: It was crazy both to prepare and then to get that many people to want to come.
Will AI Replace Us?
22:55Melody: I posted this to Instagram and I was just so surprised at the number of close friends that were interested in the workshop. It was almost a hundred people signed up. It wasn’t even very formal. I just posted a few stories. The stories got circulated. So that was awesome.
23:09Melody: But I think from there there was this deep sense of—I shared how excited I was about the technology to my friend Hajin, and her first reaction was just like, “This sounds scary. If you’re so excited about it and it can just do everything, what does that mean?” I was so caught up in how excited I was about the technology I wasn’t even thinking about how most people would be feeling about it.
23:51Melody: But that’s been a thread recently. I wrote a little Substack post about this too, but I feel more strongly than ever that all these tools will help expand our capabilities. We shouldn’t think of it as, “AI is coming to replace us.” I think that’s a very negative way of viewing it.
24:18Melody: And 100% it is going to replace some things. Literally this morning—Goldman, did you hear about this? Has had Anthropic engineers come in and replace all of their compliance and accounting.
24:36Henry: Well, they actually went in. That’s funny.
24:37Melody: Yeah, they were embedded. There are like four deployed Anthropic engineers. Today’s headline—my sister works in consulting, I send it to her. It feels like if Goldman’s doing this, they’re kind of the standard. Other companies are gonna follow suit pretty soon. And also, just personally, I know so many people around me that have been getting laid off. And I think these are AI efficiency things. So there’s definitely going to be some short term pain, where a lot of these jobs are gonna get automated away.
25:18Melody: But for those that are dealing with that or are nervous about it—one, I think executives should not just be cutting people. It shouldn’t be a cost savings thing.
25:33Henry: Yeah. DOGE and things.
25:36Melody: Yeah. So DoorDash right now is being super AI forward. They’re having their entire company—they want everyone to just be using AI.
25:49Henry: Even look at OpenAI themselves. I think Greg posted something.
25:54Melody: That every employee should—whether you work in HR or IT or whatever—you have the domain knowledge and you will know how to use AI better in that specific domain than a random person.
26:16Melody: That’s like designing AI applications for efficiency. No, you are the person that will know how to do that. All to say, I’m just excited about how AI can expand our creativity. And I think there’s so many interesting applications in the home, in our community.
26:35Henry: Yeah. If we speak to local knowledge—or even, we were talking about this concept earlier, legibility and illegibility. I’m referring more to domain knowledge. You know what the problems are in your organization or your life or your family, all this stuff. And you’re the only one that can, depending on how local and focused it is, do something about that or understand the problem space. Another person just doesn’t understand that.
McLuhan: Extension and Amputation
27:15Henry: And so that’s your unique capability. And it’s just augmenting that. And of course if we bring in McLuhan again—the laws of media that he came up with, the four of them. The one that we’re always focused on is extension. It augments our ability. Better, stronger, faster. And that’s very clear, very obvious.
27:52Henry: And then the other obvious thing is that there’s plenty of bad things too, but it’s really hard to think about what that really means. In his book Understanding Media—understanding not social media, but media in general. When he means media, he means technology. And it also means anything we make. He says it’s “the extensions of man.” But it’s also the opposite. The opposite of extension is amputation.
28:12Henry: So it also can amputate us. Whether literally—if I use a car, then it kind of amputates my leg. Not literally, but I end up taking the subway or taking a car instead of walking. Or instead of biking. And then with AI, the parallel would be: is it thinking for me? Is it amputating my brain’s ability to—
28:40Henry: And how do you balance that with, maybe you don’t need to know how to code, or maybe you do need a little bit, or you need to somehow catch up to it. People are talking about debt. One way to put it is, when you start a project with AI, you have an understanding debt that you need to catch up to.
28:59Melody: Of how software works, or—
29:00Henry: Or how this project works. I’m already assuming you’re an engineer, but if I vibe code this whole thing, I have no idea. I’ve never looked at the code. You can’t just assume that AI is gonna keep getting better over time. It will, but either the problem gets too complicated or whatever—
29:20Melody: It’s like editing in post. You don’t wanna assume that the AI’s just gonna clean up your mess.
29:27Henry: Right. Philosophically you wouldn’t wanna do that either. But maybe the ultimate sin is laziness. We’re assuming that it’s gonna be good. Why not put in the effort upfront? For ourselves, for the sake of understanding. Which is something we value and want.
29:44Henry: I was talking with Alexey Guzey a while ago, and he was saying about math and science and understanding how things work. He’s like, “I wanna know it for myself."
"The Seed Needs to Be High Quality”
30:05Henry: The whole point of all this is not so the AI would know. It’s that I would know. If it doesn’t help me understand, then what’s the point?
30:17Melody: Yeah. That’s why, especially if you think of AI as being autocomplete on steroids—it’s predictive. The seed needs to be extremely high quality. The seed of the idea, and then it’ll produce the fruit of your project. And that kind of planning upfront just takes a lot of effort. My brain hurts after a day of doing that kind of work.
30:54Melody: Even today as I’m planning my AI projects, you’re definitely still thinking. And that’s where I’m like, when I’m designing stuff using AI, there’s no way someone’s just gonna replace this whole process that I just went through. They’re definitely still going to have to pay someone to do this because the amount of brain cells I used to orchestrate this thing was just so intense.
31:21Henry: I think it’s more like the 90/10 or 80/20 rule. Anyone can make something in one minute or one second now. And that kind of looks like it works. But the things that I think are worth making are still gonna take time.
31:42Melody: Exactly. I genuinely, from our design studio’s perspective, it truly has just expanded the capability. For any engineer, the first time you encounter it, it does kind of freak you out ‘cause it is doing what your job looks like as it currently stands. But I think it’ll just expand what you’re able to do. And it’s been really fun to explore the boundaries of all the various fun ideas. We get to focus on more fun, interesting ideas rather than spending our time on the minutiae.
32:25Melody: And I hope people can welcome that change, even though it’s going to be painful in the short term.
32:36Henry: You don’t have to use it. And also you don’t have to use it in the way people are telling you on Twitter or something.
32:42Melody: I know the Claude bot and Bolt hype—people are so confused by that. A lot of the Bolt posts are fake, right?
32:51Henry: That’s just funny because to me it just shows that the media of Twitter is still the same. People are just trying to go viral, whatever. You don’t have to do any of that. Even what I say, you don’t have to listen to any of it. It’s just one perspective.
33:08Henry: What I would try to say is you don’t have to listen to people and you can learn on your own and use it the way you want. Some people were like, “No, I think my own coding ability—I don’t want it to atrophy.” Which it will, if you just vibe code.
33:30Henry: Maybe they use the agent differently. Geoffrey Litt was talking about how he’ll use the AI to create this basically foolproof spec. As much plan as possible, down to almost how you should do it, especially if he doesn’t know how. And he’ll just type it out manually.
33:43Melody: He’ll type out the spec?
33:44Henry: Or the code. He’ll actually write the code, but he uses AI to just help him plan it. And so you’re still using AI, but you are doing the work. And then obviously in the process of typing, you’re probably like, “Actually that probably shouldn’t—”
33:57Henry: So you have full understanding.
34:01Melody: Slower, but—
34:02Henry: Again, just ‘cause it’s fast doesn’t mean it’s right or it’s good or it makes sense for what you’re doing. And maybe for the things he’s working on—in this current moment, if you want to make your own UI, a new UI interface or HTMX, that kind of thing—it cannot do that. You really need to come up with something new. It’s just gonna copy whatever was there before and you have to think through what’s worth spending your time on.
Orchestration and Human Judgment
34:30Melody: 100%. The things that I’ve been thinking about lately—I’m creating systems for myself too. Even my design process. If everything’s really about orchestration and workflows now—I don’t love the term “human in the loop,” but what are the most valuable moments for the human to be steering the project in the right way?
35:07Melody: There are very certain crucial steps where my input at that moment is super valuable just to direct the AI in the right way. And then you can let it go forth and then there’s a human judgment moment again.
35:34Melody: But it just helps us be more efficient, and also allows us to explore a much larger territory of ideas. And that’s just been really fun.
Building for Community, Not Companies
35:46Henry: I think that’s the thing that is fun. Like you said about the business way or trying to remove employees—if you’re a company, the only way you can see the world is by getting rid of people, which is funny. Ultimately it’s just gonna be an automated company or whatever. When I’m talking about building for community, a person has to be at the center. People have to be at the center.
36:09Henry: I can’t help but see from a church point of view—what is the church? It’s not a building, it’s not the institution, it’s the people, the body of Christ, the people of God. If that’s not at the core, then we’re losing something. I just don’t agree with that attitude. I’m sure that’s gonna happen, but that’s not what I wanna see in the world.
36:31Henry: So how do I think about using AI in a different way? Like you said, it’s not about making me efficient. It’s about giving me confidence that things I could have done, now I’m willing to do anyway. And then, expand my imagination. It’s funny—ironically, AI can remove your creativity too, if you let it think for you. But then also, because it can do all this work, it might make me dream bigger. To do things I never would’ve thought was worth the time or even possible. And that’s the fun part. I think that’s the part that’s exciting to get anyone to build things.
37:06Henry: Jasmine had that post a while ago about—I forgot the term she used, but I think it was like “software vision” or something. Lucas had the metaphor of “parkour vision.”
37:30Henry: When you’re doing parkour, it’s all about looking at this landscape and then knowing where to jump and go around and do all that stuff. So the same with software: when you look at your life, how can I use software to change things?
37:44Henry: And it doesn’t have to be in a small sense of, “What app can I make?” Maybe you can make your own hardware now. Maybe you use a 3D printer and then combine that with the AI. It’s not really about code at all. It’s just making.
38:01Melody: Exactly. I think the big shift for me has been: for so long, we’ve always known how transformative software can be. Literally, our lives are transformed by this technology over the last however many years. And we just kind of had to wait until a software engineer decided this problem was worth solving. And now that bottleneck has been essentially opened way up.
38:43Melody: And so I’m really excited about—I’m hearing about moms using Claude Code and how they’re using it to build tools for education. Or like you were saying, tools for the community, how to use it to get your neighborhood together. Or niche things. Our friend did something where it’s skiing and cheap flights. That one still feels a little in the tech realm. But my friend’s building a garden planner so she can plan different plots.
39:06Melody: A lot of people think of AI and it’s like, “How am I gonna use it in the workplace?” But right now with Claude Code and all of these things, every single person in their domain knows it better than any software engineer. You know your problems really well and you now have the tools to build.
39:26Melody: I think that’s so amazing. My personal goal is I wanna help more people learn about this, not be scared of it. The headlines are really nerve-wracking. It can feel like you’re getting left behind. There’s obviously layoffs and all these sorts of things.
39:53Melody: I just wonder how this can help nurture and recover people’s creativity.
Culture, Belief, and What Makes Us Human
39:58Henry: It’s kind of similar to the dinner that we had.
40:04Henry: Religion and culture. I just think everything comes down to your beliefs. It doesn’t mean belief in God, I just mean belief in life, in society and what’s gonna happen.
40:15Henry: And you read the news about how everyone’s gonna get replaced, all these things. Why do we believe that being human is not worth it or has no meaning? Or our meaning is only in our work? It comes down to these fundamental questions about who we are.
40:32Henry: And then AI is challenging those things and that’s, I don’t think that’s bad. Hopefully it gets closer to what that means. And of course, if you don’t have a worldview that has that kind of view of people—that we are inherently valuable, created in the image of God—
40:51Henry: Then you might lead to, if you’re at OpenAI, one of these companies: “We have to create meaning. We’re trying to discover what meaning humans have.” And that’s because you already have the implication that humans don’t mean anything. Because once we create a robot that’s better than us, transhumanism, all this stuff—there’s no value. We have no utility as long as there’s something that can do better. And that has nothing to do with life and love and family, community, suffering. What does that even mean for people?
41:10Henry: I’m starting to read about the dignity of dependence. This idea that if you are dependent, that means that you can’t do everything yourself. You’re not self-sufficient. You’re not God.
41:37Melody: Does that make you not viable?
41:38Henry: Yeah. So what are we supposed to do with these people? Whether you’re an older person, you’re disabled, you have health problems. Or you’re a child. If we really were that utilitarian, we wouldn’t need anything. But that has nothing to do with life.
41:58Henry: I think people, including myself—we all have to be rethinking, which is painful. It is painful when all the things that we thought were true are changing. Whether it’s how I work—we have to always adapt.
42:18Melody: In general.
42:19Henry: In general. I don’t like the language of, “It’s inevitable. AI is here, you gotta just change your life.”
42:27Melody: “You gotta escape the underclass.” I hate all—
“There Is No Inevitability”
42:32Henry: Because it’s good to take account of what people think, but no one knows what’s gonna happen. No one knows the future, other than the fact that people think that we can make it in a certain way, but that way is still through a person. Whoever has money or power or whatever, they can at least try to convince everyone that they can do something about it. And it’s like, where’s our agency?
42:59Henry: Another McLuhan quote: “There is no inevitability as long as there’s a willingness to understand what is happening.” Isn’t that funny? It comes down to understanding. If we’re not willing to understand, then we’re just gonna let people that don’t care about this make it happen.
43:16Melody: Or let other people decide what the narrative is. The narratives are so powerful because that’s what you eventually believe. People who are leading the AI narrative in the media—that’s why people outside of tech who don’t understand the thing feel so nervous about it.
43:36Melody: Just go in and actually understand the thing. Don’t just believe what people are telling you about it.
43:45Henry: Including the people that are working on it. Not that people working on anything are inherently bad. It’s just that it is literally self-motivating. When you work at a company, it doesn’t matter how good of a person you are, you’ll have incentive to care about the company. ‘Cause you’re employed. Or if you’re making an AI model, of course you’re gonna say it is good. But sometimes you can say it’s bad, and then you’re still working on it. It gets confusing—are you supposed to listen to the experts? Or random people?
44:19Henry: I get that. But we need to be willing to try to understand it. And if it’s too overwhelming for one person, that’s why we have community.
44:34Melody: Right. 100%. Really quickly, the thing I wanted to say about the dinner the other day—we probably have to go. But the last thought is, the talk was on “culture is religion externalized.” And it’s really just about how our beliefs shape culture and it just makes so much sense.
45:06Melody: And then I looked up the root of what culture is, and it comes from cultivating. And it’s really interesting to consider. That’s what he was alluding to. He was like, there’s tomatoes, but ketchup is culture. Because ketchup imbues a lot of meaning into the thing. You decided to puree it, you decided to add sugar to it, you decided to consume it in this particular way. And these are your beliefs, passed down.
45:39Henry: You’re saying the raw materials into the way that people decide to use them.
45:45Melody: Exactly.
45:46Henry: Culture is literally just how people live. All the ritual around life and life with other people. What we do when people are born and when they die, when they transition. That’s culture, and it’s unique to each location, people, group.
46:08Henry: And we could talk about Babel and Pentecost and all these things. One language, one way versus the diversity. Unity in diversity, right? Versus, “It’s all AI.”
46:28Melody: Yeah, exactly. Like this is all gonna squash us down into one.
46:35Melody: Well, I was just thinking about how we’re talking about all these different kinds of people who can now be using these tools, and the positive view of this is that now all these people have the tools to cultivate either their homes or their communities. They’re just more empowered to cultivate and bring forth culture using the tools that they now have access to.
47:04Henry: There’s an infinite amount of things to talk about, and I know that I tend to think of things from a very local point of view. We have to talk more, or with other people, about how that affects the whole country or the world or policy. Those are bigger questions. I’m just trying to figure out what I can do here.
47:26Henry: And maybe I need more imagination myself around how that affects the greater whole. ‘Cause it’s scoped out and maybe I also have a loss of agency around wanting to impact the larger thing.
47:41Melody: But often we think too much about the macro, which removes our agency locally. That could be the thing that brings us hope and joy day to day.
47:58Henry: And that’s just its own level of abstraction. We’re talking about abstraction. Do I care about people at a global level? Yes, we’re all human. But ultimately we affect specific people who are materially around us.
48:15Henry: Our neighbor. I don’t wanna say that everyone is our neighbor. It’s not true. Yes, everyone can be our friend, but doesn’t mean literally they will be. You don’t have enough time. You are one person in one place specifically.
48:30Henry: And there’s this weird dynamic about abstraction and concreteness. We live in this world. We’re not just a disembodied thing. But then we’re using these disembodied things to help us understand this world, but ultimately we have to live it.
The Checkpoint
48:52Melody: Exactly. On that note—end of our internet checkpoint there. There were probably more things we could have covered.
49:00Henry: Maybe the checkpoint is great because we didn’t plan to do this. It’s just, what are we thinking at this moment? And it needs to be raw.
49:10Melody: Exactly.
49:11Henry: We’re all over the place. That’s just where we are.
49:16Melody: If we later with Josh and Parth or something, we do. That’s fine.
49:21Henry: That’s it. I think of it as a snapshot.
49:23Melody: Exactly. A snapshot.
49:25Henry: Our brain at this moment.
49:27Melody: Yeah. I think that’s why I feel motivated to capture this right now. We’ll see how we’re feeling.
49:36Henry: Yeah. Maybe feeling horrible.
49:37Melody: Everything goes to crap and we’re looking back on this time. Who knows? Okay, well this is the checkpoint for today, February fifth, I think.
49:52Henry: Okay. We’ll see.
49:53Melody: We’ll be back another time. Who knows when. Bye bye.