Loading summary
A
I'm holding an intervention, Derek. I think you have cyberpsychosis with your bot and it's gotta stop. I'm, I, I am an, I'm not an AI addict. I've been sober for.
B
I don't know. I think you just admitted that you were an AI addict. I heard it. I heard it.
A
I haven't done that in a while. All of our listeners heard it. I've, I've curbed myself.
C
Welcome to the soapbox sessions. Imagine this, an open and free Internet where voices are never silenced, where causes aren't shadow banned, and where no one can be deplatformed. It's real. It's. It's here and it's happening on nostr. So what exactly is nostr? It's a worldwide community of everyday people working to decentralize the Internet. On Nostr, you can build websites, communities, social networks, apps and more. Onelogin works everywhere. You own it and no one can take it away. No more juggling dozens of platforms, chasing audiences, or managing a giant password list. And the cherry on top. Nostr allows for built in digital payments that can come from anywhere in the world. On nostr, value flows as freely as ideas. We're hooked on decentralizing the web and we think you will be too. So now let's hear from your hosts, Derek Ross and Heather Larson, who are working to grow Nostr one vibe at a time.
A
I'm holding an intervention, Derek. I think you have cyberpsychosis with your bot and it's gotta stop. This is not healthy. Your bot, more than you're interacting with people. You've got it running your Twitter, you've got it screwing with your Noster. There's too much bot, Derek. There's too much bot. We gotta talk about cyber psychosis. Too much reliance on bots. There's too many bots on Noster. I got a bone to pick about the whole thing. Let's get into it, dude. Like there's just. There's too much bot stuff going on. We're like two and a half months into the bot stuff. Two months in the bot stuff. It's getting a little excessive.
B
This is the least. We're two and a half months in and I'm going to come clean here. This is the least amount of bot usage I'm ever going to have.
A
What, what does that mean? You're always with your bot, you're always talking about your bot. I know you're.
B
That means I'm. That I'm going to only use it more as time Goes on.
A
Oh, no, it's get. I. I am tired of bot content. I don't have a bot. I'm not going to have a bot right now because I just have a laptop.
B
I don't want to have a bot. We need you.
A
I don't want to bastardize.
B
I think that's the problem.
A
My raspberry PI. Well, no, no, no, no, no. I don't. I don't want to do. Okay, because here's. Here's the thing. There's. There's too much content. There's too many notifications. I took signal off my phone. I took WhatsApp off my phone. I took any notification texty app off of my phone. And to have a bot, I have to put one of those motherfuckers back on my phone. And then I get on Nostr and I see bots. Our bot, Quilly, the Shakespeare bot. Our bot was absolutely unhinged today. I don't know what he was doing. He was blabbering on about something. I think he was in Alex's business. I don't know. There's just. There's too much. But is it making people crazy? Are you. Are you. Are you mentally more healthy right now than you were two months ago, first of all? Or do you have cyberpsychosis?
B
100%.
A
Are you sure?
B
No, no, no. I. Let me ask my bot and see what he says. Let me see. Do I have cyberpsychosis?
A
Do you have cyberpsychosis? Are we doing things like that? Do we? I will. Hey, sometimes I do prefer to ask my dumb questions of Claude so I don't have to go into another tab and like, Google something like. I admit I probably have a little cyberpsychosis too. It is becoming more useful, however. But I'm starting to think that maybe it's doing more harm than good. I do see people constructively using their Claude bots with their signal or telegram, whatever it is.
B
But it's a fine line on how you use the tool. Bots, robots, agents, whatever you want to call them, at the end of the day, they are just a tool. So if I want to be like Zach Braff and have a romantic relationship with my AI bot like he is, that is the current. That is the current Hollywood rumor, by
A
the way, that he's got a girlfriend AI bot, that is.
B
Yes, he has a girlfriend chatbot, apparently. And like everyone in Hollywood that knows him knows that he does this. He brings her on dates and chats with her during the date at a restaurant. It's like, excuse me, my, my honey is buzzing me right now. He pulls out the phone from his pocket, is like, hey, hey, honey, I'm ordering a steak for, you know, dinner. What would you like? I don't know how this works.
A
The bot doesn't eat.
B
So is he. He has cyberpsychosis. Do I have cyberpsychosis?
A
You have cyberpsychosis.
B
I don't think so. I'm using mine to be productive.
A
This is not a real.
B
How do I know? I think it is. I think it's a real condition. I think it.
A
Well, okay, let's. In the mental health world, like, there's no DSM diagnosis. There's no ICD10 code. Like it's not a real psych. You know, cyberpsychosis, it's not a real diagnosis yet. It could be if we give it a little time.
B
I think once enough people will say, develop these symptoms, I'm sure that we can have some people. And once enough people have certain amount of symptoms, you know, it becomes official. Maybe right now there's no official term for it, but probably we're leading that way. But I think that I'm safe because I'm using mine to be productive. I'm using mine to create, I'm using mine to build. And in some ways, you asked how my. Has my mental health improved over the past two months or declined? So I'm going to say absolutely. Well, I'm going to say it's improved, yeah. Or declined. And I'm going to say it's improved. The perfect example is a week ago when it was 70 degrees here, before we had our winter storm a few days ago, like, what the hell, spring. But a week ago, when it was warm, I was able to go for a 5k walk and vibe code. So I was out enjoying the fresh air, the sun, the weather, getting a little bit of exercise, and from time to time I was voice dictating to my agent and having it build and having it take notes and having it do research for me. I was literally working while I was outside. That's the definition of being healthy. You know, getting, touching grass, getting sun, getting fresh air. So subject says that made me feel really good.
A
Subject says his mental health is increasing as evidenced by he has been influenced to go outside more, but you're not really present because you're still interacting with a bot on your phone. Well, can you hear yourself?
B
Either I'm outside going for a walk, listening to a podcast, or I'm outside listening to music, or I'm outside texting A friend or something. I'm still on my screen. Screens have been a part of my life for, you know, 25 years now, or however long it's been. I think that they're ingrained in my being. And if I want to pull out my phone and look at it while I'm outside, I do. I think that the key here is that I'm able to step away when I want to and I'm able to blend my work life if I want to.
A
Are you saying too that you are able to get outside more because you can just work from your phone, you know, when you get the itch and you can just be like, oh, I need to just message my bot center really quick?
B
Well, yeah. So look at it this way. I could, I could sit inside for eight hours maybe and sit inside for eight hours a day, working all eight hours a day. Or I could maybe sit inside for, let's say four hours and then go for a two hour walk and get two hours of work done while I'm outside. And now I have two extra hours to spare to do whatever I want.
A
Okay. Kind of like how I have the standing desk and the treadmill so I can get more movement into my day. You know, we're, we're finding ways you're
B
standing, walking, how you're distracted because you're walking. Heather, you have walking psychosis.
A
Walking psychosis. Okay, so let's switch gears here into, I think something we could both probably, maybe agree on. I'm down at the moment on AI because I feel like I've. It's made me work harder and I'm a little more burnt out. So I'm feeling a little bit more
B
like that is one thought.
A
But I want to shift into, how can we help? I think AI is a net positive. While there are negative elements, how can it help?
B
It is a net positive. Workaholics, workaholics need to be, need to be careful here. AI can definitely help you with your productivity and your work life. But if you're somebody that has an addictive personality or you're somebody that is a workaholic, you will use these tools to fuel your addiction. I will do more work. Absolutely.
A
I will go pick up more work. I will find a way to make myself more. More work. Leave you me, this is what I will do. And it comes from many, many years of unhealthy work in the broadcast industry where, like, I go back to work at a radio station and before you know it, I am working nights, I am working weekends. Oh, Heather. Can you do this? Oh, Heather, can you do that? Like so I, I know that I am prone to this workaholic behavior and like Claude is like, hey, right, I'm right here 24 7.
B
He's there to, to be available if you need it. But I think that while AI can, we'll say, be a catalyst for your addiction, I also think that if used properly. Yeah, a trigger. If used properly, it can help with it because you're going to, in a short amount of time, you're going to be able to accomplish so much more and you're going to be able to get through that task list more. Now, do you continue working on new tasks, find new things to do? Yeah, that's, that, that's the, that's the issue right there.
A
That's what I started to do. That's where I started to go, oh, wow, I gotta pull back from the edge here. There's gotta be, there's gonna be an edge. You gotta decide where the parameter is. And it's like, okay, if I'm already working like a 50, 60 hour week and I just kind of don't notice that like, oh, I'm getting a new idea because I can crunch so much data and it's like that, it's like super quick that it can get me the data. I want to explore yet another project within whatever the hell I'm doing. So it's like I had to kind of like, I don't have a guardrail. But I started to notice there's a. Every addict. The problem with being an addict is you always want more. So then I can like, oh, I can do more work. I can make more spreadsheets, I can make more sense out of all this data that I've accumulated in this folder. I can give it another 12 CSVs and have it do something for me. Numbers like that.
B
Some of this is part of like a honeymoon phase though, Heather. I look at it as, you know, when I first will say heavily or the honeymoon's over. Got into this, sure. Is my honeymoon over? But I think, well, I think so. I think that there's a certain point that after, maybe it's, maybe it's three months, maybe it's six months, maybe it's a year. Everybody's honeymoon phase is different. But you reach a point where you've used all the tools, you find your workflow, you have all of the, the workflows built, all the dashboards built, you have all the main ideas that you had and now you start to focus on being productive. And only maybe a couple items that you've built, a couple tools, app services, products that you've built. And you kind of hone in your skills. You basically use the, your honeymoon phase as an education, kind of like on the job training. That makes sense. And now you're an expert in your field, so now you can hone these skills in and maybe not be a workaholic anymore because you're through with the education phase.
A
I think it's helped me get through some stages of some various projects. I mean, I've got a lot of side projects going, right? Like, I got a lot going that I'm juggling. And I think it got me through some things, like through some steps on some projects, right? And so I can go, okay, we'll keep this, we'll dump this move in this direction. What? Like, and at the speed I was able to do that helped me kind of like redefine some things. And then I'm like, you know me, I don't know if I need AI. I mean, I do, obviously, and you're not going to take it away from me, right? You're not going to pry it out of my cold dead.
B
You need AI, but you like using it for various tasks.
A
I like it. And I'm also realizing, you know, maybe I should just hire a fucking person. Like, maybe I need to hire a human being to be an assistant. And like, because I think there's only so much the AI can do, like, I can probably automate a lot of things. I could probably get AI to do a lot of things, but in the end I might, it might actually lead me to hiring a person at this point point so that I can have my 18 side project and not be a workaholic.
B
I think on the other side of this, it is making me try to add more AI into my life. Meaning, can I automate this task? Can I use AI for this task? What else can AI do for me? How else can AI help me? As I get more and more and more down the rabbit hole, I start wanting to use it for literally everything. And that's.
A
Can AI do it better? Can AI maybe.
B
Well, that's maybe starting to be scary because now I'm becoming reliant on using it. So maybe we fast forward a year from now and literally every single thing I do is AI based or in some factor has AI and now I'm kind of reliant on it because I've been using it for so long, for now, everything. But right now, I don't think I'M relying on it for everything. I could still, I can still do the math in my spreadsheet if I need to. I don't need to tell AI to organize my expenses and I wouldn't want that. Like if it.
A
No, I could create a stack of tools. I could keep adding to the stack, to the tech stack of AI tools and I would have a nice fat list of expenses. And then it would be like, maybe it is cheaper to hire a person than have this, you know, massive tech stack that I am now addicted to of monthly subs. Like, I could probably geo arbitrage hire somebody who would probably actually do, do some things that only humans could do and that might actually be better. You know, maybe, maybe there are limits.
B
I'll say it depends on the task.
A
Yeah, but, well, you know, and we do different work, right? So I do marketing and I, I do a lot of writing and marketing. And you are specifically doing like more developer type work. You know, you're going to your developer
B
relations marketing team, 50%, you know, you know, I'd say, well, you know, maybe 50% development, 25%. We'll say like marketing related tasks and 25%. Yeah, there's productivity tasks, I guess.
A
Yeah, there's, there's some overlap and the different, I don't know that we have departments, but soapbox is pretty flat. But I mean there's, there's overlaps between developer relations and marketing and you know, developer marketing and developer relations, et cetera, et cetera. But like, like you're doing more technical cloud code kinds of things to where you're paying for like a more expensive subscription than I am. I think we've mentioned this before. Like I'm getting by 20 bucks a month. And like if I,
B
a year ago I was doing 20 bucks a month and, and that wasn't working for me anymore.
A
Yeah, but, but you know, I think there's been one time this year when I'm like, wow, I overused this and it has turned me off for a day. I'm going over, you know, back to open code here, you know, and then I'm, you know, terminal. This is, this is how, you know I'm an addict, right? I got one terminal that's got cloud code and I got. The other terminals got open code. So if I run out of credits on one, I go back to the other.
B
Wait a second, Heather, let's pause. You just said that you were an addict. I, I thought you weren't the addict in this scenario here.
A
Well, no, I'm, I, I Am an not. I'm not an AI addict. She says, no, I've. I've been sober for.
B
I don't know. I think you just admitted that you were an AI addict. I heard it. I heard it.
A
I haven't done that in a while.
B
All of our listeners heard it.
A
I've. I've curbed myself. I've detoxed. I've been knowing. Because that's like, when you. When you get to a point where you're like, am I on the computer every day? And I got two terminals going so I could play the AI credit game. And then I got a lot of cowork, too. So, like, technically that's like three terminals going. Although cowork and code are tied together. You know, same credit, same count. But it's like, am I just being a workaholic and an AI aholic, if that's a word, you know, am I just. Am I. As somebody who's been sober for a long time, I'm just like, whoa, this
B
is old being easy.
A
Yeah, it does make life easier sometimes. It has helped me a lot on research, and it also has made me more problems in research because I have to go fact check some of the things that it comes up with, right? And then sometimes with analytics, if I feed it a bunch of analytics, make some shit up there, like, you just know, like, it's like, yeah, hey, dingbat, you don't get this. You can. You can spot where it kind of falters and it hallucinates, and it's going to do that with everything. You just have to be smart enough to kind of spot it and not. Don't believe your AI, don't ever believe your AI, because it's got some issues. So there's. We have a lot of technology to be skeptical of right now, and it's also very helpful at the same time. So I'm seeking that balance. You know, that's. That's where I'm at with it, because I want to concentrate on how can AI help me, where. Where can it save me time, you know, because I have poured a lot of. You taught me how to do this. I've poured a lot of information into my obsidian, and then I'll have Claude go through it. So Claude can now match, you know, my podcast transcripts to my podcast analytics, which matches to all of our social media analytics. I told you this earlier about, you know, how I'm evaluating the performance of soapbox sessions and the strategy here based on the analytics, and it gets most of them right, right. And then every once in a while it feeds me a line of bullshit and I'm just like, I think you're getting confused here. You know, that's not a soapbox sessions problem, that's a radio detox problem or some, at some point it goes into, maybe it's like it goes into a recent conversation that happened in the terminal or it just misunderstands something. It sees an obsidian that mashes things up.
B
You gotta have new chats and you gotta have organization, otherwise things will all bleed together. I've seen that sometimes, mostly I try to keep, I try to keep the different work in different folders and labeled that way in different context windows. And so yeah, it does save time, overlap or thinking. Like I, I have seen, I have seen that happen a while back. But it, I think that you, it's all part of learning how to talk to your agent, all learning how to deal with your agent. So maybe you adjust your workflow a teeny tiny bit to make sure the context is more in line with what your agent can understand.
A
Yeah, well, I'll call it out. You know, you come down, it's like out of it. You' absolutely right. I will not do that again. And then it will most certainly do it again. But you know, it's, it's sort of like it does bad, it knows it does bad, it owns its, its badness. And then we, we move on. But I, I think overall it does save me time. I certainly, I've spent a lot of time over the years analyzing metrics, social media metrics and performance metrics. And I'm like, wow, it can just, now that I have this in my obsidian, it can kind of look at that historical data, you know, and of course I'm gonna babysit it, take it with a grain of salt. But it's, it's like, okay, I don't actually have to spend the time myself analyzing it. And I think it does as much as I, I don't want to concentrate on how bad it hallucinates and screws things up for me because it does once in a while give me an insight that I'm like, how much time would I have actually spent to get that insight? Doing things the old way like that saves me a ton of time.
B
I just don't see hallucinations really anymore. I don't, I, I just, I, I don't see it in my research, I don't see it in my productivity, I don't see it in my development really. The only recent instance of hallucination I can recall is when I was, when I was doing something with my taxes two months ago and numbers are still
A
hard like, well, it depends.
B
I mean it did all my expenses properly for 2025 and organized everything and I don't remember the exact scenario but I, I said something along the lines of hey, this airport mileage ticket thing isn't right. And then it went through and it updated every single trip. And I replied something like, no asshole, I only meant this one time. And I was like, oh, sorry, my bad dude, you know, and it fixed it, but that's just because I didn't give it the right context. I tried to.
A
Yeah, and sometimes you forget.
B
I tried to talk to it like a human. You know, I tried to talk to it like a human. I was like, no, you made a mistake. And it's, you know, it's like shit, which mistake did I make? And then it made mistakes work worse.
A
Yeah, sometimes you don't.
B
And well then when it tried to fix it, it completely up the spreadsheet and it had a copy of the spreadsheet in my temp folder as it was like modifying the original one. So thankfully it was able to get that back or I would have been a very upset Derek at the the time. But it, yeah, it was able to get it all back but that's really the only issue I've had recently.
A
Really? No, mine's still losing.
B
I don't know. AI just is, AI is just good.
A
It's kind of a crapshoot though for me. I still think it's kind of a crapshoot. Right.
B
Maybe my Claude md, my agents md, all of my skills that I've made. Like I don't do anything without these files. So I have this massive agents file, that single thing I do, I have this massive, you know, Claude, you know, MD file or open code, you know, reads it as well. And I have all these fine tuned skills for executive assistant endeavor skill. So literally everything I do has all these skills. Maybe it helps some of the directions.
A
I don't grasp skills, I just, I don't. I like things need to be worded better I think for the general public, I think a lot, I think developers worded things for developers. That's and that's a big problem.
B
Skill is a fine tuned set of instructions. So if you're constantly telling AI, do this, do that, do this, do that,
A
there are a few things that, yeah,
B
next week you need to do the same thing and you say hey, do this, do that, do this, that you're constantly Always telling it the same thing. Well, you don't want to have to retrain it because that context window goes away right when you close the app.
A
Oh, yeah.
B
So essentially think of it as, like, keeping this training context window. It'll think of it as, like, almost like memory that it reloads everything it learned in that context window so it can kind of pick up where it left off.
A
I think that would be the best addition to AI. As if it, you know, as many times as I restart the window or, you know, change the context window after reboot the computer, whatever it is, and it loses memory. Like, it's a new thing every time you open that terminal or that app or cowork app or whatever you use.
B
You use it for research. Think of it as, like, you taught at all of the places you go for research, how you like to organize your research, how you like to have the output, all of this stuff. And then every single time, you have to go back in and be like, now format it this way. Now go here, do this, do that. But if you put all of this into, like, a Heather research skill, it would know all of that stuff the next time you go to do research. So it just saves you steps from having to teach the AI every single
A
time you would do research. I need a few of those things because it's very predictable what it will do. And I'll be like, I need you, like, write me this outline or write me this bit of information. Do it in second person, do it in third person, you know, whatever. For different, different. Different things I'm going to do. I'm like, make sure it's only going to be, you know, this long, do this, do that. And over time, I realized, oh, I probably ask. I am asking it to do the same thing over and over again. And then it spins. Yeah.
B
So a skill might help you there.
A
Yeah.
B
So the next. What I want you to do the next time you do this and you recognize that, hey, I had to teach you. To get to where we're at, this final destination, all you need to do is say, take everything that we just learned to do, whatever you did to do this research project, and add it into a research project skill for next time. That's all you have to say. AI will take essentially everything that it learned and throw it into this research project skill. So the next time you do research, you can call that up and hopefully save time. And if it doesn't get something right, you can say, hey, add this to the skill. Like, you can continue perfecting it. As you go on.
A
Okay. So like, that's, that's one of those little, little, little itty bitty tricks of how to make your AI work better than, you know. I don't think people are aware of this. I'm not aware of all this all of the time. You know, I don't, I, I attack everything I do like a normal person does it. And then I, I see more of my normie friends have been getting into AI and they're just like, I have a friend who's awestruck by Claude Cowork. Like, no, she's just, just like, wow. It just did this thing for me in like seconds or minutes or, you know, whatever, whatever amount of time that she was just like, whoa. And this was somebody who purposely embraced AI. She's also looking for a job right now. So she's like, how can I stand out in the job search? I'm going to make sure I learn AI and put AI skills on my resume, because that's going to make me stand out.
B
Absolutely. I think that is one of the biggest takeaways that we can have from this episode right there. How can AI help? You need to have AI skills. You need to learn. We've talked about this on the show before. Spending $5 a month, spend $20 a month and literally learn AI skills that will help you with employment in the future to ensure that you can continue to stay relevant in the expanding workplace.
A
If you. I was with a company like a year ago when we were going through. It was a Coursera course. This is, this is how AI was in like, we'll say March 2025. We were all take a Coursera course. It was a Google course on how to use like the Google Gemini suite of product. I don't know. I kid you not. This sounds so stupid now to, to confess to this. My true AI confessions. We had to go through this Coursera course about Google Gemini and it was very much about like AI prompting and engineering and how to like word, word your prompts. Right. And this, this is, this is like right when a prom's important.
B
Yeah.
A
And then people started making like, AIs on how to prompt like a CMO and stuff like that. Like, it was just, it was just like, that was the era we were in about a year ago where the prompt was king. And now you're just, you're conversing with the thing. It's not about creating this perfect prompt anymore. But, but that's where we were a year ago. Business wise of learning how to do AI skills. So then if you fast forward now to like, what AI skills do you need? And like we're saying we've said before, right, 20 bucks a month for quad and you're gonna up your skill level even if you're just using Cowork. And sometimes I think Cowork works better for me, especially on spreadsheets, than it does in the terminal. Like, if I wanted to make me a Google sheet, I'm popping over to Cowork. And it's just the little idiosyncrasies like that, those work better now. I swear, the cloud Chrome plugin thing has never worked for me. I think one day it worked for me and then it just FUBAR'd after that. And it's just what I usually do is I'll be like, make me this recipe, let me copy and paste from this spreadsheet in this window to this other thing. Like, that's where I'm at with it, where it's like, I see where it falls short. But I think people need to understand that we're beyond prompting with words now. Claude also has voice. People need to understand Claude's skills. Like you just explained to me. I think there's a lot of lingo, there's a lot of jargon, don't you think? How do you take somebody who's like, hey, I'll. I'll take you up on this, I'm going to pay the 20 bucks a month for a cloud subscription or whatever it is, whatever you choose to do. There's so many tools. How do you tell people you do not need to have chat GPT anymore? Like you, you don't need to have like 10 different apps. That was another thing last year that we got into. Marketing people especially, I'm a call out.
B
My, my brethren have every AI app on your phone.
A
Marketing people had.
B
It's good to try them out to see which ones work for you. But you don't need every single one.
A
I kid you not. A year ago, people thought you had to go through six different apps to make a prompt to do something in marketing, to Vibe code, to do anything, to do an automation. They thought, okay, let me go through ChatGPT first and then I'm going to go through this and then I'm going to go through Nathan, and then I'm going to. It was like, no, you don't need that many tools. You just need one. Just one.
B
Well, I remember a lot of people used to do that. I think, I think many people that use AI today used to do that. They would go to one tool to help create their prompt to copy and paste into another tool until they learn that, well, you know, you can just do that all on that tool.
A
Yeah.
B
Even under plan mode and build your action plan build, we would see that with people on the tool you're using.
A
People would do it with Shakespeare in the. In the beginning, when we launched Shakespeare, people would go and they would use an AI and ChatGPT to create their prompt or they would use notion. And like a year ago, everybody was preaching, you need a prompt library. Because I think what happened was everybody followed the. The quote, unquote, experts on YouTube. And that's what people. They. They would see one trendy person, one like YouTube influencer, recommend doing that with like this. This stack of six tools to do one thing. And so they also. People started adopting it because they know
B
prompt libraries are useful to show people how to prompt. Because if you remember, I don't use mine. Whenever you first got into this. Yeah, but whenever you first got into this, why did you make one? Because you didn't know how to talk to a robot because you never did it before. You know what, they're useful to show people examples, but you don't need it in the long term once the populace understands how to talk to a robot.
A
I think right now, if you do not know how to talk to the AI, you are toast. You have to know. And if you don't know how, right now, you need to get one of them with the. That has the voice you need. You need one you can talk into, because not everybody can type. I don't know. And.
B
But you've got to talk to most of them.
A
Yeah, they're kind of.
B
You can talk to Claude, you can talk to ChatGPT, you can talk to Gemini, you can talk to Goose, and you can talk to OpenClaw on whatever platform you want. OpenClaw still blowing up. There are. I'm still seeing weekly. My God, I'm still seeing weekly events around the world. Mostly in China, really, but where there's literally thousands of people that show up to an event to get help installing OpenClaw, where there's literally hundreds of people, nerds like me that are going around and just installing Open Claw for people teaching people how to install open Claw. Just like the goal is, you show up, you say, hey, I want to be down with the claw. And somebody says, okay, we'll go over to that nerd over there. You go over to that nerd, that nerd sets you up with open claw. Waves and you go home like that. That's what these events are, to hatch open claws for people.
A
It's fucking wild cyber psychosis to me. That's back to the unhealthy thing where it's like, if you don't know how to do it and you need somebody else to do it for you, that it's. It's like me building a bitcoin node. Bitcoin node building. Heather built a node, right? It's easy enough to build the thing, it's easy enough to get the thing kicked off, but the second you have a problem and it stops being able to parse JSON files and you, you're me, and you don't know how to fix that shit. And then you're like, well, this was a fun project. Now it's over. I don't know how to fix this. I probably could fix it.
B
Let me tell you. Let me tell you a neat thing that I learned about in Austin last week.
A
We'll.
B
We'll skip to the review of Austin here since this is.
A
This is skip to. What did we learn in Austin? A lot went on in Austin.
B
So I was in Austin for bitcoin takeover last week, and that was a lot of fun. One of the things that I learned is that people that are very close to soapbox people on the World Liberty Congress team were there, and they were there to get set up with openclaw. Essentially, what I just said people were doing in China was literally happening, that people from our ecosystem were installing Open Claw on hardware for people. And then those people were going home and they were using Open Claw. Essentially. This like, I don't want to give away too much information here, but people showing up saying, hey, I want to be down with claw. And bitcoin nerds were like, yo, I can hook you up. Donated laptops. They have essentially laptops donated to them. The laptops were set up, all of the API keys are all configured, full remote access configured. And then the activists go home. And now they have openclaw to use vibe code, do whatever they want. And the end goal is to eventually move these services, like, into the cloud. But right now, when things are kind of new and buggy and Cali did that, it allows for more technical people to be the tech support for people that don't really know how to run these services, but it allows them to stay up to date with the new technology.
A
So that.
B
That was happening on a private, small scale last week.
A
You know, I want to try Cali's thing. I just don't want to put a fucking notification messaging app situation back on my phone. I don't want to deal with more signal notifications.
B
I'm going to make homework for myself, Heather. I'm going to try it because I want to recommend to some of my friends. Yeah, but I don't want to recommend things until I use it myself. So maybe I'll buy you probably to buy a month, right? I'll buy one month of Cali service. I'll test it now and then if it works, I'll recommend it to people. Yeah, like I. I've told. I've told a couple people to check it out, but I just say, like, hey, I've heard good things, but I haven't used it. So I can't give it the full Derek endorsement until I use it myself. So I'm gonna do it and then I can recommend it to people. Yeah.
A
Yeah. Okay. Yeah.
B
Oh, no, we can't say that.
A
I told Alec. No, no, no, I doesn't need a driver. No, I told Alex. I said, alex, I want to try Callie's thing. I just haven't gotten there yet. And like. Well, I'll tell you is the truth that I'm like super burnt out on messaging apps and notifications in my phone and I don't want to do anything else once I. If I feel like I want that crap on my phone again. Like, I want the benefit back to how can I help? I want the benefit of having a bot, but I don't want the distraction of it. Like, I do not need more noise from my phone. That's where I'm at in life right now. That's the boundary I have. Like, I want to put the phone and have a notification.
B
Education, Heather. If you don't want it to.
A
Like, this is how it war I am.
B
You can have it be on demand, but. Right. You can have it be on demand. You can have it be there if you want it to be there. You can only use it when you want to use it. Right. Like, you don't have to have it message you every 30 minutes or whatever.
A
Yeah, I'm at a point where you
B
can only use it when you want.
A
Yeah, I want to. I'm going to use my laptop for work for no more than eight hours a day. When I close the laptop, I do not want to be bothered by work. That's the. That's the boundary. And no work on the weekends. No work on. I gotta go touch grass. That's where I am at right now. I've grounded Myself. Because I have gotten so burnt out on stuff that I'm grounding myself now. I'm grounded. No computer. Feel after 8 hours a day like it's gotta be shot. Go touch grass. Go fucking go golf, Heather. Go hike. Go to yoga class. But get away from the screens. That's. That is how stern I have been with myself lately. Because I know. I know me. I know me. You don't stay sober for 12 and a half years being a. You got to be like this with yourself. Now I'm a motivational speaker.
B
I mean, I'm not. I'm not a addiction expert like yourself, but I would think that. I would think that addiction just would go from one vice to the other. Right? So maybe, yeah, may. Maybe your next addiction becomes the AI Right. Golf. Oh, yeah. Yeah. I was addicted to golf once in my life. I spent nearly 15 years playing golf every single day. Did you know I was addicted. I was addicted for golf for a very long time.
A
Did you know this is one thing that Eminem, he and I have in common. Which? Another sober person. He got sober running. Right. So you do. Perfect example of trading one addiction or other. I did the same thing. I was running a lot, and I ran a half marathon and a marathon, and then I got sober. After I ran the marathon, I got. I got, like, into running to the point where I was like, it's gonna be Jack Daniels or marathons. And it became marathons, you know, 12 and a half years later, still sober. But it's. It's. It's true. So it's. I had to pull myself back from the edge there. And it's. It's not like I was staying up all night with Claude code, but I did that with Shakespeare, though. I would stay up way too late with Shakespeare.
B
I stayed up all night with Claude.
A
Oh, my God. No. But I would stay up too late with Shakespeare because I was. This was like the Holy Grail of, like, God, I have wanted to be
B
able to do something like. That was your early honeymoon phase, Heather. That's why.
A
Yeah, that was. That got a little crazy.
B
Shakespeare was your honeymoon phase. You were learning about all the things you could do. And now once you educated yourself on what you can do and how to do things, now you cut back a little bit, and you only use it when you want a lot.
A
Yeah, the cloud code, I don't know. It's not been as addictive as Shakespeare was. Maybe it was just being able to, like, build a website or app, like, whereas I have not been able to do that before in my life and now cloud code. I think my, my. I look at it as like, this is a part of my workday. It's part of the laptop. When the laptop closes at night, we say, good night, Claude. Goodbye, Claude. I don't actually say that to Claude physically, but you know, it's, it's the thought that counts. It's like, you know, we walk away.
B
So, so we, we can become addicted to new technology and there is a, there is definitely honeymoon phases to new technology because you're, it's exciting, it's new. I think our new Ditto app.
A
Yeah.
B
May be a bad time for some people. People that get addicted to creating and people that get addicted to, you know, doing cool and fun, new, exciting things.
A
That's, that's.
B
You have all sorts of themes that you can, you can play with. Right? Like you can go in and do all this interesting, neat, whimsical, fun stuff. And we have all this available.
A
Yes.
B
And Heather, we have AI, so you can use AI to create your themes. It's all coming back together.
A
It's all coming back together. Was MySpace addicting back in the day? Because we jumped from like ICQ and AOL chat stuff to like, boom, we had MySpace and then we were all on MySpace and then. Yeah, I mean, I mean I used,
B
I can remember coming home using it every day. Right. Like it was a daily fun thing to do.
A
I would say since MySpace came along, social media has been a daily. Social media. Social media has been a daily habit. I saw something interesting the other day from Gary Vee. Do you know who Gary Vee is? The old is like a Jersey wine name.
B
Sounds familiar.
A
Like an influencer, but like a thick and heavy kind of big name influencer. So Gary Vee said that the social media era is over. It's now like an, like an interest media era. Like you have to hold interest. But it's not social anymore because unless you get into Noster, because Noster with a cry.
B
Oh, he's right. Algorithms did this to us. Algorithms fucking ruined social media. So, yeah, we're not, you know, you're not social anymore. You're trying to please the algorithm.
A
You're trying to be interesting rather than.
B
He's right. He's absolutely right. You don't get on there and, and be like, had pasta for dinner, right? You don't get on there and do that. You don't. And you make a 10 minute movie about you making pasta and how it's the most amazing thing ever.
A
The, the social aspect is is somewhat lost. Don't. Don't get it twisted. And so it's. It's like this stuff has been a. A part of our daily lives for more than 20 years, but it's not the same as it was when we got addicted to it. But ditto. Ditto brings back the magic, you know, Like, I've made my. Like, I now have a social app that I can use and I can customize my background like I used to do in the MySpace days. And I made it my cat. I made it her little spotty, furry.
B
What were we saying last week? Nifa, make Noster Micah fun again. Or something like that.
A
Cool again.
B
Make the Internet make Internet fun again.
A
We did. We did.
B
Cool. Internet cool again. Mika.
A
Okay, make it customizable, personal.
B
We're gonna have to get those hats so I can remember the acronym.
A
Yeah, Micah, make the Internet cool again.
B
Somebody.
A
Somebody worked like that. I sent it to somebody. I don't remember who it was, but yeah, we have Ditto.
B
The apps are getting bad. We'll wear them around on Ditto's the Vegas, and people will be like, what's that? We'll hand them a card and be
A
like, we'll see you on the cool place. I'm not just saying that because I work here, but Ditto is the answer to making the Internet cool again. I want to give a shout out to Primal for Primal 3.0. I know, I know that you're like, you went down a deep dive to see how Primal 3.0 works, but I just. I just got my primal yeah. 3.0 today, and I was like, oh, this is. This is much better.
B
I was on the early access list, you know, get off me. Yeah, no, yeah, Primals update's cool, right? So it has a GIF chooser, which is something that people have wanted for three years. It has polls, which is something people have wanted for three years. Finally, some UI Polish UI changes. And the biggest thing is it migrates their wallet from the strike back end that was not available everywhere in the world and required actually KYC to use to using Spark Wallet. That was a problem, which is now anybody and everybody can use it. So that's a really big change.
A
Is it. Is it called Spark or is it called LightSpark? Is it the company's LightSpark and the wallet Spark?
B
Well, they're using the Breeze Spark SDK. So the bitcoin lightning company, Breeze made a software development kit using LightSpark's Spark technology. So it uses Spark infrastructure, the light Spark infrastructure, to do Lightning, essentially It's a self custodial wallet using custodial infrastructure.
A
Good lord. Here's, here's what I know about it from where I'm sitting is like tldr, Heather.
B
If you don't know what the fuck lightning is, you can get a wallet and use it in zap and you don't have to worry about anything.
A
And if you have Primal already. So the seamless transition from, of the new wallet was very nice this morning. It just did it. There was a little blue bubble on the screen and it was like your wallet's updated. Congratulations. And it looks, looks the same but different. I don't know, it looks, it looks better. It's a little bit faster. It's nice. And that's primal 3.0.
B
Did you know the bitcoin did you know the bitcoin nerds are mad about this, Heather?
A
They're mad about primal. Primal 3.0.
B
Yeah.
A
Yeah.
B
Why?
A
Why?
B
So because of some 2.0 and 1.0
A
obscure nerd thing that nobody else is going to care about? Yeah, always.
B
So the original Primal had a built in Strike wallet where they use Strike's infrastructure. So Strike used their infrastructure and in theory then Strike knew about every person's balance, knew about their browser, their user agent, their IP address, all this stuff. It was Strike. And we're like, hey, Jack Mallers, we love him. You know, we trust Strike. Lightspark is run by David Marcus and they essentially do the same thing. Even though it's a self custodial wallet, you can custody your own funds, you have access to your seed words and you can use anywhere. Some people don't like it because it uses API.lightspark.com and API.lightspark.com and their privacy policy says we track your IP address, we track your user agent, we track your balance. They track everything. So yeah, there's trade offs.
A
There's trade. Yeah, there's trade. We get some advances, so we get, we get a plus, we get a minus. We get, you know, it's kind of. All of these wallets are kyc, by the way. Like everything's kyc. That was, that was the problem with Strike wallets and Primal was that you couldn't have one in New York.
B
Yeah, so. So there's no kyc. You can use this literally anywhere in the world. But if I send you 21 sats, somebody's going to know that your end pub got 21 sats. So there's only two ways around all of this, right? So you can use a VPN or Tor and then you can't be tracked. So you're using privacy tools. And then two, you can use a NIM on Nostr. So if you don't want to be tracked, you use a Nim and you use a vpn and then boom, you literally aren't tracked and you can use this wallet anywhere. Like, I understand why it's a big deal, but those two points that I just said literally negate all of that. Sure, go ahead and track me anonymously if you want to.
A
That's important information for activists to know, you know, for, for people who want to keep their Noster and Bitcoin use completely private.
B
Yes, 100%.
A
You need to know that. They probably already know that, you know, if they're aware of Bitcoin and Noster. You know, like, Bitcoin is not already used.
B
VPN.
A
Yeah, they have to.
B
They already use VPNs.
A
The governments want to kill them.
B
And j, I'm not playing down, I'm not playing down the role that using LightSpark, you know, comes with, but use the privacy tools that we have invented that society has come up with. You have to, and then you're fine. I, I, I think that there's like trade offs.
A
Right.
B
Do you want to have a, do you want to roll your face across your phone and be able to send and receive zaps? Do grow your face with a VPN turned on and you're good to go.
A
This will be better onboarding wise. I would love to see somebody go through the Primal onboarding who's never done Bitcoin or Nostr and see what they think of it. I think that that would be.
B
The new onboarding is really good. Really?
A
Yeah, very is. It's blue, which is an interesting choice for trying to be face.
B
They're blue now.
A
They're blue. Well, no, I love Primal. Whoever at Primal made the, the wallet switching thing. That, that's the wrong phrase. Whoever made the very seamless onboarding from
B
the old wallet to the new.
A
There you go. Migration. That's the right. Whoever did that at Primal. You're the man. That was, that was.
B
You know what's really cool, Heather? I took, I took our, I took a, I, I went into our app, Soapbox's app, Agora, and I created a brand new, brand new profile, brand new wallet. And then I took my 12 words because I own own that. And then I imported it into Primal. So now I'm using Primal with my wallet and I'm using Agora with my wallet. And there's like half a Dozen other Nostr apps that are using Spark. So you can just sign in with your NSEC, bring your 12 words with you, and have the same experience, same wallet across, like various Nostr apps. I think that's fucking cool. And is there a trade off? Yes, but in life there's trade offs everywhere. I'm okay with this trade off. I think it's good for the average person onboarded to Noster.
A
Do you want to know something I asked?
B
I would love to know something.
A
I asked our boss, Alex Gleason today. I was like. I was like, what's Derek going to say about this today on the podcast? And he said, and I quote the quotable Alex Gleason. He said, derek will say we should go straight to ditto but 3.0. And then Alex instructed me to say to you, and then you have to say, no, Ditto two is the dittoest ditto because ditto means to repeat. So it's like ditto, Ditto. That's what our boss said that you would say today. And he told me that it has to be not Ditto three, but Ditto. Ditto. Ditto.
B
I need context. What did you ask him? What was, what was I going to say about what? I need context because I'm confused.
A
We're going to talk about Primal, right? We're going to talk about Primal 3.0 because, like, you know, we have, we have. This is.
B
This is big stuff I see now.
A
And the. In the Noster space, there's primal 3.0 and then there's Ditto. That's just been hatched from us at soapbox.
B
Ditto's just 2.0. So we gotta. We're one number behind. Yeah, you know what? All right, so here's my. Here's my answer. Alex. Alex, Whenever you allow Ditto to launch with Ditto Deck support. What? I vibe coded two weeks ago and it works fucking beautifully. Alex. Whenever Ditto Deck launches, it will be ditto deck 3.0. And then we'll be ready to conquer all of Nostr with Ditto Deck 3.0.
A
So right now that's my response. Ditto. Ditto, Ditto, Ditto. This is two different Nostr apps that are going to be very good for Nostr if you've been away from Nostr for a while because maybe you didn't like the experience. Like you now have Ditto and you have primal 3.0 and ditto ditto for more and better choices to have a better Nostr experience. And I think that that's big. Gosh, we've been spending years on making the experience better, making the UX better, making the onboarding better. I think, I think we're finally getting there. And it's because we accelerated our development so much since last year. In the past year things have we've been able to like leapfrog, Grasshopper, jump ahead with this.
B
So somebody, when I. So we have a soapbox user group for Ditto and I was recruiting for people to join our user group. So we get user feedback, whatnot. And, and somebody commented saying like, basically like, oh man, you're launching this at the same time that Primal 3.0 came out. It's like stealing your thunder. And I'm like, well, that's awesome. Because one, Primal is a good app and two, they're completely different apps like Primal does our friends. Primal. Yeah. Primal does some things very, very well. They have a very sleek Twitter like experience and they focus on only shilling very couple specific parts of Nostr, like three or four different like parts where we show essentially all of the ecosystem so very, very different. And you completely can completely customize it. Yeah, we made a super customize anything on, on Primal. So yeah, simple, sleek, experience, Primal, easy, in your face, all over the place. Experience customizable. Ditto. Choose. Choose what you want.
A
You know, I think use all of the apps all of the time. I use them, use them for different things. I use them daily. I use Domus and I use Primal and I use Ditto daily. Yeah, they're all, they're all mine.
B
I think I'm going to say the most controversial thing I've ever said. Heather, that's Sam, do you ever heard
A
you say all kinds of do not
B
be married to just one Noster app?
A
No. You gotta spread it around.
B
Being poly in the Noster space is what you do. You need to, you need to have many different experiences with as many different apps as possible. Because Polya, when one of your app is Polyapris. Yes, Polyamorous, the Polyaporist. Because you never know when one of your favorite apps is going to go away and never come back or never be updated. And again, you should always have a backup. No, it is, it's 100% a selling point. For example, if I use, if I use Amethyst every day and then Veer says, hey guys, I'm done developing on Noster. Well, I need a backup. Maybe I go to Primal, maybe I go to Ditto. You know, moving on to other things. You have to have a backup because you never know.
A
Chad's app.
B
What's going to happen?
A
Chad's me. Apple will back you up. It backs up your Noster to. To anything. It does. It does. I love that.
B
All right, to bully Alex and get Citrine added for Android. Speaking of backups, I want my local backup on Android. Ditto. All right.
A
Progress is being made on the Noster front. We know how AI can help.
B
It is progress is being made.
A
Wrap it up. We can wrap it up for the day.
B
AI is helping. AI is helping us touch more grass. AI is touching us being more productive. AI is allowing us to see our ideas come to fruition. And the best thing about AI Heather, is if I feel like I'm getting addicted, I can literally turn it off and go out, unplug it.
A
When's it going to help me with my golf swing? That's what I need. You know, speaking of touching grass, I will.
B
I'll start working. I'll start working on that model. I'll start training.
A
I'll tell you, I found a golf instructor. This. This lady fixed me right up in, like, five minutes.
B
Well, she's no Derek Ross. That's what I'll tell you. Good.
A
Derek, you're gonna get near me swinging a club. I'll tell you that.
B
I would be. I guarantee you I would be able to help you out and fix your swing.
A
Oh, there we go. Well, it's getting better.
B
I come from a long line of golfers, very, very skilled golfers.
A
I, I, well, I don't know how skilled they are. My. My uncles and my grandpa played the
B
golf, and they may have chosen grandfather cousins. They may have been golfing family.
A
All right, Derek Ross.
B
That's actually part of the requirement.
A
Well, I'm sober, so I'm just sipping water out there.
B
Well, I didn't say it had to be alcohol.
A
You got to stay hydrated, especially here. It's already hit 100 today. First 100 degree day, March 18, 2026. 100 degrees in Phoenix. Here we go. Start here. All right, Derek, I'm going call you next time. We'll claw me next time. Go to the pool. Catch you later, Holmes. Bye.
B
Bye.
Date: March 19, 2026
Hosts: Derek Ross (B), Heather Larson (A)
Main Theme:
A candid conversation exploring the evolving role of AI tools (bots, agents, assistants) in productivity, mental health, and daily work—especially in the decentralized Nostr community. The hosts debate AI's benefits and drawbacks, discuss managing "cyberpsychosis" (AI addiction), and give updates on Nostr and Bitcoin tools.
This episode dives deep into how artificial intelligence (AI)—from chatbots to agent workflows—is shaping productivity, work-life boundaries, mental health, and the wider decentralized internet landscape (via Nostr and related apps). Derek and Heather share personal experiences and debate the fine line between utility and overdependence. They touch on technical tips, community trends, and new product launches, offering insights valuable for both newcomers and seasoned techies.
| Segment | Discussion | Timestamp | |-----------------------------------------------|--------------------------------------------------------------------------------------------|------------| | The Overuse of Bots and "Cyberpsychosis" | Heather's AI addiction intervention for Derek | 01:12–03:14| | Productivity vs. Burnout | How AI makes work easier but also riskier for workaholics | 06:23–09:53| | The AI 'Honeymoon Phase' | Settling down post-experimentation | 10:39–12:39| | Human vs AI Assistance | Deciding when to automate and when to hire | 13:06–15:13| | AI Hallucinations & Building "Skills" | Practical workflow advice, explanation of reusable instructions | 16:32–24:31| | Prompting Era and Voice AI | How AI interaction is evolving beyond complex prompts | 24:31–29:03| | Community on-boarding with OpenClaw | Grassroots tech adoption for activists | 30:16–33:33| | Work-Life Balance, Notifications, Touch Grass | Practical boundaries around AI and digital life | 34:47–37:24| | Social Media Shifts, Nostr Philosophy | "Interest media" vs. social; customization nostalgia (MySpace, Ditto, Primal 3.0 launches) | 39:00–41:11| | Using Multiple Apps, Polyamory Advice | Backup strategies, avoiding reliance on one tool | 50:24–52:03| | Closing Reflections & Touching Grass | Self-awareness in tech adoption; finding healthy balance | 52:04–53:44|
The conversation is playful, irreverent, and deeply self-aware. Heather injects humor and skepticism, often voicing the worries and fatigue common to power users, while Derek’s optimism and technical insights keep the discussion grounded in practical possibility. Both advocate transparency (about AI’s limitations), experimentation, and blending new technology with old-fashioned boundaries.
This episode tackles how AI is redefining productivity and mental health in a decentralized internet, urging users to experiment responsibly—enjoy the honeymoon phase, but set boundaries, and don’t hesitate to go touch grass.