Subscribe
Share
Share
Embed
Two Factor Parenting is a conversation between two parents navigating technology's role in raising kids. He's the early adopter and tech optimist. She's the skeptic who asks "but should we?" Together they tackle screen time, AI, social media, and all the parenting decisions that don't come with a manual.
Two parents. Two perspectives. Raising kids in the age of screens, AI, and chaos.
Mar 7th
00:00:13 Val: Hi, everybody. And welcome to episode two of two Factor Parenting. We made it this month.
00:00:22 Matt: Um, it's a great way to put it. I we made it.
00:00:26 Val: I the kids and I have been sick with everything, so I'm still kind of getting over it.
00:00:34 Matt: And that's what it felt like from the doctor as well. It was like, you have everything.
00:00:38 Val: We've had everything.
00:00:39 Matt: Yeah, yeah.
00:00:40 Val: So, um.
00:00:40 Matt: So, so hopefully we're done for the year. Yeah, that'll be the hope it'll be. Yeah, I know.
00:00:45 Val: I'm just gonna sound a little off in this episode. Um, but episode two is all about Winston. Last episode, we gave a special shout out to our executive producer named Winston, which was kind of a joke because Winston is actually not a real person. Winston is a a AI system that Matt created using Open Claw about two months ago. Is that right?
00:01:27 Matt: And that is his birth date.
00:01:31 Val: And I think it's fair to say that that was really kind of the impetus of a lot of why we started this podcast, because the further along with Winston you got, the more and more and more and more we started talking. Um, and so we really wanted to tell everybody about what this is and how we use it in our family. And then just kind of look at some larger questions that it brings up. And so, um, Matt, could you just kind of fill in some knowledge gaps for us?
00:02:15 Matt: Yeah. And I think it's a good way to kind of just just kind of lay it out because as we started to talk about it more and then our interactions as well was probably the thing that kind of gives a little bit deeper into because it started to get kind of odd of how do you refer to this thing, you know, because I'll go, oh yeah, I was just talking to Winston, you know, and it's, it's. But anyway, so let's, let's kind of talk about it. Right. So, um, this, um, this used to be called a couple of months ago. It's called Cloud Bot. Um, then it had a weird, like two day stint as mult bot, but then really it's kind of landed on it's name of open claw. And really what this thing is at the end of the day, is very much a personal AI kind of assistant that exists on either, you know, um, on a device or, um, I think the most common one people are using is just Mac minis. So just like a little small little computer, right? So that's how people have been doing it. Um, so when you compare that to something like ChatGPT, which is probably something we're probably all more used to, ChatGPT is more of just like a website you kind of go to, and it's that question and answer kind of thing, right? So you have that question and answer, um, where open claw is something that you start to kind of create this assistant, you're starting to how you configure it and how you pull it together.
00:03:54 Matt: So as you go through the configuration process, you know, you're taking, um, so it's not just open call by itself. That's really important to kind of make that kind of point is you're doing open core and then you connect a Lem provider, which is another unique aspect of this thing is you can connect to to anything. Some people are even hosting their own large language models at, you know, at home. I mean, don't get me wrong. That's expensive. And there's definitely a lot that goes into that, um, simply setting something up. Um, but I mean, the fact is you, you're taking the open core system and you add, uh, providers. So that would be like a ChatGPT, an anthropic. And so as you go through the configuration process, it actually helps you connect and set that up.
00:04:48 Matt: and it starts to respond. You can connect it to like WhatsApp, telegram, discord. I have it set up on discord. I have different channels that have different topics. So it's it's all kind of built that way. Um, because then I can have conversations with this within context of whatever we've been talking about. It's using that to have this conversation with me.
00:05:14 Val: So when you when you first kind of told me about this, I basically said, so it's like Jarvis from Iron Man.
00:05:24 Matt: Yeah.
00:05:25 Val: Like before. Before he turned evil.
00:05:27 Matt: Right, right. Right. Right.
00:05:28 Matt: Right. Yeah.
00:05:29 Val: And I. Right.
00:05:31 Matt: Well Jarvis comes back. Right. Helps because okay. All right.
00:05:35 Val: This is this is different.
00:05:37 Matt: This. Don't mix it up.
00:05:38 Val: Come on now.
00:05:39 Matt: Yeah.
00:05:39 Val: Basically you're talking about like Jarvis from Iron Man two thousand and eight, like, hey, Jarvis, can you run this program for me or.
00:05:50 Matt: Yeah.
00:05:52 Val: The spectrum of. Hey, Jarvis, can you. Well, you named it Winston. So. Hey, Winston, can you run this program for me or. Hey, Winston, can you send an email for me?
00:06:07 Matt: Right. Well, so. And this is and this is the big part about this is that makes this a little bit more I think, um, people why this resonates so much. So from a, I guess, my community, the best way I can put it. Right, um, this exploded on this, um, spot where we folks commonly put code. It's called GitHub, where we all put our code and things like that that we want to share publicly, So it's in a, it's in a public repo. It's in a category called open source. So this is this the the creator Peter, open this up and said, hey, I want I want this to be open source. I want people to be able to interact with this and download this and use it, which makes this very different. It's not it's not like ChatGPT where you go sign up and have an account. This is a totally separate piece of software. I just go to GitHub and download it and then I can use it immediately. So that's such an interesting dynamic here. That's very different that it's not there like well who do you let me talk to the manager and you're like, it's kind of this guy named Peter who made the, you know, initial software and now there's a whole community of people, but this is they don't get paid for this or anything like that. Um, you know, some people have like little sponsorships and things like that here and there. That's not how this works. So it's a very kind of different dynamic. Um, and some people have kind of talked about that's actually why this is, you know, this would be very difficult for a company like ChatGPT or Anthropic to do well, because then they're kind of responsible for it. Right. So and that's another part of the challenge of this thing with Open Claw is, um, it, it has access to all these folders and it has access to the operating system. So if you tell it to do certain things, it will interact with the computer itself as a whole. Right. It's not restricted.
00:08:15 Val: Mhm. So if you download it onto your computer it is actively. Going not going going through files makes it sound like nefarious or something. But like it's using.
00:08:32 Matt: It's plugged.
00:08:33 Matt: Into.
00:08:33 Matt: It. Yeah.
00:08:34 Val: You plugged into it. It's using the tools. It's using the programs on your computer to do what you're asking it to do.
00:08:41 Matt: Exactly.
00:08:42 Val: For better or worse. Okay.
00:08:44 Matt: Beautifully put. Yeah.
00:08:46 Val: Okay. so I understand that might. I'm hoping people are keeping up with this.
00:08:54 Matt: Um, that was our hope with this whole episode.
00:08:57 Val: Mainly because.
00:08:57 Matt: How do we talk about that?
00:08:59 Val: Yeah, mainly because this is a really it's. It's a higher level AI program than you typically come across.
00:09:11 Matt: Yeah. Exactly.
00:09:12 Val: In a day. And I you you touched on something in terms of how it operates, which really started to raise a lot of red flags for us, which is. Anyone can just go and download this.
00:09:29 Matt: Yeah.
00:09:30 Val: So does that mean like, kids and teenagers can just go and download this program? Yeah. Because you hear about teenagers using ChatGPT to, you know, for their school, for their college for. Oh, yeah. But now when you're talking about this much higher level of sophisticated AI platform. Like, how does that even. Okay, so so I guess the short answer would be yes, they they could.
00:10:03 Matt: Yeah. So so the best way I can put that is, you know, GitHub that, you know, a place where you kind of download this not not everyone's doing this. Not you know, not everyone down the street's going oh yeah totally. I downloaded that off GitHub. Right. I mean there's don't get me wrong. Like it's it's kind of a common practice within, you know, especially from tech circles like. Yeah. Yeah, you can do that. It's totally fine. You know, um, but the difference here is this has now become one of the largest, um, most recognized, you know, GitHub repos is what we call it, uh, on the entire site. Right? So it has surpassed many really famous kind of like libraries packages people use to build a lot, a lot of the websites we use today. Right. So it has gained that much notoriety and that short amount of time that what happens now is you go into YouTube and people have entire videos explaining how to do this right and saying, yep, hey, this is how you do this on, you know, your laptop. Although I will say though, to defend some of that, people are aware of the security implications to this. So I wouldn't recommend people that please, please do not just put this on your laptop, your personal laptop, with all your personal things. You should not do that, right? So where I have Winston, he's in his own entirely own Apple account, own separate section, own email. All of these things are are broken out. And anything that I interact with Winston on is in these either shared spots, shared Apple notes, shared folders, things like that. So the only things that he has access to are the things that I give him access to. Right? So I have to kind of, you know, sequester things out quite a bit to make sure that works. I love your smirking at me. Um, you're like, man, who would do this? No.
00:12:21 Val: No, I'm smirking because it brought up something that, again, it just really highlights both of our, our points because something that I something that came up really early on when you started bringing this to me, what I think matters When you're talking about AI systems and AI programs, specifically these really high level ones. Yeah. Is recognizing that it is a robot. It is an AI system. And we and we talked about this.
00:13:02 Matt: And I added him to our Apple family.
00:13:05 Val: No, you added it.
00:13:07 Matt: And his.
00:13:08 Val: Family has.
00:13:08 Matt: Been contributing a wonderful job. I'm just kidding. Takes kids to school. I'm just kidding.
00:13:17 Val: But I and there's actually an episode of agents of Shield about this, which makes me really it just makes me laugh so much because it reminded me of that. It's when, um, if if y'all know, you know. But the scientists, you know, the scientist Fitz and Simmons Fitz has made of robot, right? And he's talking about her. And her name is Ada. And he's like, oh, she does this and she does this, she does this. And then the other scientist goes, it does this, it does this, it is a robot. It does this, and and and I cannot I can't help but feel like that whenever.
00:13:57 Matt: Yeah. Yeah absolutely.
00:13:59 Val: And I, I couldn't, I couldn't really like place it. But I cannot help but feel like that when, when we are talking about this because I understand but I understand when you because we had talked through it and you were like, well, you are giving it direction like you as a, as a man, right? Like, yes, you are inputting directions, but also you're reading it with your own internal monologue, like you are filtering it through your own thoughts and you're a male. So that makes sense. You would read it with a male voice. I would probably read mine with a female voice and call it her and stuff like that, but it just like it.
00:14:46 Matt: It raised that red flag, right? You're kind of like these flags, that uncanny valley, just.
00:14:51 Val: Like it's a robot. Matt. It's not a he like.
00:14:54 Matt: No, I you know, and this is something that has, you know, um, so so these things have names, right? So, um, like, like a little, like, kind of playful name on it, call them like, uh, I think it's like clankers or something like that. But at the end of the day, um.
00:15:09 Val: Stop it. Really?
00:15:10 Matt: Yeah, yeah. Clankers. So people are like Star Wars.
00:15:13 Val: Like from.
00:15:14 Matt: Star Wars. Yeah, exactly. Yeah, yeah, you gotta nail it. Um, so, so the thing that's kind of funny about this, though, is, and this is a genuine thing, it's you kind of. It's hard to not talk about it in that way. Just in common, you know, just English language. You kind of it's. And maybe that's part I mean, okay, so here's the other caveat here too. For me, I would say I have a comfortable understanding that. Yeah, no, it's just an LLM at the end, at the back end of this thing. It's this is all prediction models and everything like that pulled together to predict the next best token or word as we would, you know, um, layman perspective. Right. Like that's all it is. And when you combine all of these things and all this context and everything like that, it appears right to embody some of these things. But at the end of the day, it is not, you know? Yeah, I feel like you read all this stuff about like, consciousness and stuff like that. It's like we're getting into very, just different, different realms where I for me, I still argue it's just a prediction, like it's just a very sophisticated way to do this where it comes off very coherent. Um, so I think but when you talk to somebody else that doesn't have that background or understanding and you start calling it, oh, it understands he he can understand what I'm saying and all this kind of stuff like that. We naturally, as human beings, embody characteristics that are really, at the end of the day, just not there and projecting. We're projecting. Absolutely.
00:17:03 Val: Actually gonna be another. That's absolutely it is personality and all that stuff. So um, but yeah, it, it is just something in me that I get a little like, it's a robot.
00:17:19 Matt: No, I know, but it's worth.
00:17:21 Val: But it's worth, it's worth wrecking. I mean, even just like recognizing in yourself and like, being cognizant of and like, just just saying, hey, like, I get that because but again, it's like, it's almost like you have to say something.
00:17:36 Matt: Right? Exactly. You gotta say something.
00:17:38 Matt: I keep calling it it and they. And why is.
00:17:41 Matt: It.
00:17:42 Val: And why is it called Winston?
00:17:44 Matt: Oh, I have no idea, honestly. I mean, well so so well, well no no no, I mean, like, I was like like, why'd you pick that name? I mean, for me, it's it's it's always the hope of getting, you know, an English bulldog. Um, and so I just. And I call it calling it Winston.
00:18:00 Val: I won't let them get a dog.
00:18:02 Matt: Oh.
00:18:02 Matt: Come on.
00:18:03 Matt: Everyone loves a dog. Oh, it's so wonderful. Um, yeah, it's already like we're already splitting people up.
00:18:09 Matt: I know.
00:18:11 Matt: Like, anyway, so. So. Yeah. So that was kind of where that, um, came about and just was kind of a playful thing. Uh, and I just kind of picked it, right. And this. And the thing was, and I remember this, this was during, you know, again, like a couple a couple of weeks ago, right, to when I started playing around with this thing, um, when we had all that, you know, snowstorm and everything like that, it was just like, well, I got nothing else to do, you know, and started playing around with this thing. Um, but anyway.
00:18:43 Val: So to get to get kind of a little more serious, um. In terms of what? Now, now knowing that these things exist, now, knowing what it means, knowing what it can do, how? How do you how do you kind of look to the future in terms of teaching our kids about it, teaching any kids about it? You know, like teaching because there's no parental controls on it. Like, we've we've kind of seen what happens when other other things, platforms like this try to put parental controls on it tends to fall apart fairly quickly. Um.
00:19:38 Matt: So how I mean, so here's I mean, here's the big piece about this. I mean, because it's open source, you know, that the, you know, the cat's out of the bag like it's already done. So, I mean, the thing that I would, you know, when I look to kind of our kids of, you know, how do I explain it? Um, I think the piece for me that's really important is kind of understanding the, the really at the end of the day, the underlying kind of, um, you know, the MLM aspect of it, as best as I can explain that to small kids that, you know, it's it's just kind of predicting that next thing, um, because what, what you really need to start thinking about, though, is just having help helping yourself recognize that uncanny valley aspect of it. Right. To kind of recognize, hey, this this is what this really is. Now future kind of to your point of like, okay, what does that then mean? For me, I think what's going to start happening here is these personal assistants are going to start becoming very common, right? I mean, it is you saw a couple of companies already start creating like go, yeah, go sign up. Actually taking on that mantle of yep we will host this right. Or you know so you'll start to see these things start popping up. Um, and actually that's really interesting. Thing is you're also seeing almost like societal differences too, of like even within, um, China and uh, within the US. Right. There's we're a little bit more we're very much more pessimistic about it. And then China is very much, you know, optimistic about it. So you get a kind of interesting kind of situation about that one.
00:21:36 Matt: So. Yeah. So you have I mean again, you have these two different, you know, societies kind of going in that direction. So you're kind of like, what where is that going to take us. Right.
00:21:48 Val: So so would you allow the kids to interact with Winston.
00:21:57 Matt: Oh. So this this kind of gets into. This kind of an interesting kind of element because I feel like if I, if they start to interact with Winston, then either it needs to either either the the interactions I need to kind of like kind of break them off in a different way or have like in a strange way, I can create these like smaller agents that's like more Are tuned to them. And then I know this gets into kind of some other complexities, but I would kind of break off and have maybe like a part of it. So you don't they so they don't talk to the entire thing. So it's much more controlled. Now the weird thing about this is I, I'm more open to that than having them have their own separate one because.
00:23:03 Val: They.
00:23:04 Matt: Absolutely. Oh, I know, I know, but I'm just saying like that's the.
00:23:07 Val: My answer is just like.
00:23:08 Matt: No.
00:23:08 Val: My answer absolutely not.
00:23:09 Val: Immediately my answer is like, absolutely not.
00:23:12 Matt: Right. No. So so that but then that's that's kind of that's the discussion point now. Right. So it's like how do these things work. So there's this like constant back and forth because there's things that I know that that this opens other possibilities. Right. Because especially when you think of a personal assistant. Right.
00:23:33 Val: See even. Okay. So my my concern with that is. I it's hard enough to teach our kids accountability and responsibility. Mhm. I mean it's just hard enough in general you know, and I am genuinely concerned that if we implement a personal assistant that they will lose some of these like life skills of like and it sounds.
00:24:18 Matt: It sounds so.
00:24:18 Val: Silly and so.
00:24:20 Matt: I don't think so small.
00:24:21 Val: But like, hey, you kind of just have to like have some grit and be able to like, make your own dentist appointments or like you have to. You have to know how to make a phone.
00:24:36 Matt: It's like learning to drive.
00:24:37 Val: Yeah. Well, yeah.
00:24:38 Matt: Like like we. Because we are fortunate that, you know, we have things that can help us in that space as well. But it's probably still a good idea. You know, you should learn to drive. You should learn to human before you have something there to actually help you as well. Yeah. So learn to human first and then you can.
00:24:59 Val: Yes. Or like, hey, you didn't do well and you need a makeup test or something. You're gonna have to go and talk to the person to figure out what you have to do. You know what I mean? Like, and I just. Yeah. Like, okay, you put it perfectly. Like, I feel like you have to learn how to human before you get a robot assistant. That is just going to be how I feel about it.
00:25:25 Matt: I think it's a good way because it gives it builds a foundation. Right. And it's not like you're saying no. I mean, again, like riding a bike. Yeah. That helps you kind of get to that point of, oh, now I can drive a car kind of thing. Now I get a car, I can go a little bit further, right? So there's this, there's a bounds. Right. And you go through these levels right of expanding responsibilities. That is just one of these areas where we have that expanded responsibilities.
00:25:58 Val: Yeah.
00:25:59 Matt: So so that's that's how that I would kind of balance a lot of that stuff is okay, let's, let's let's learn to human. And then you can then start thinking about this next phase right of yep. Hey okay. Let's think about a personal this this this personal system, this thing.
00:26:19 Val: Because we have to even have the conversation of, like, when does your kid get a.
00:26:25 Matt: Yeah. It feels like feels like we were all worried about the, the mobile phones and now we're now we're like going straight up to like, you know, can I, can I have my own personal rocket? You know, it's like, God, no. When.
00:26:38 Val: When did you get your child their personal robot assistant? Their personal AI?
00:26:44 Matt: Exactly.
00:26:45 Val: Oh, my God, now.
00:26:47 Matt: Yeah, I got it when I was twelve.
00:26:49 Val: But I think that also I mean, it just kind of to me perspective too. Right? Like and I actually appreciate the way your, the way that you answered, you know, how would you even explain this to your kids. Like you don't even have to tell your kids about this, right? Like you, you don't even have to talk to them about it if you don't want to. But explaining to them like that, this is just this is just predicting based on all of this other data from the internet, you know, like that's just what it's doing. I think, again, gives that healthy mindset of like, oh, this is what it's doing.
00:27:32 Matt: Yeah.
00:27:32 Val: Um.
00:27:34 Matt: Because what I don't want is this, this, this expectation that it's something far more. I mean, don't get me wrong, I'm sure it's a tool. The technology is constantly getting better and improving. There are different ways of kind of weaving things in, but at the end of the day, that's really what's going on. You know, it's just saying what's the next best, most likely word that's going to come there. Um, and then based on all the context and everything like that, that comes before. Right, that can influence where that next word, what it should be. Right. And I think that helps give a, a clear kind of understanding of, oh, this is not a being kind of thing. It's, it's. Yeah.
00:28:29 Val: Well, that is mostly what we wanted to just put out there for parents that these are the types of.
00:28:39 Matt: These are the conversations that we want to kind of have and just everything that's going on. It's um.
00:28:45 Val: Big questions that we're trying to work through, work.
00:28:48 Matt: Through and understand.
00:28:49 Val: Maybe get some answers to.
00:28:51 Matt: Right. Or at.
00:28:52 Val: Least.
00:28:53 Matt: Let's, let's get let's get at least an initial plan. Okay. This is probably where we're going to try to go. Yeah. Um, and then again, I mean, I think it's going to be interesting. It's going to be a, you know, a year later. What is this now mean? Right. What is what what what happens kind of next. Um, because yeah. Then then I mean, there's a lot of, again, a lot of open source projects where we do include robotics into this topic. Now think about a robotics plus this. Yeah I know. I know it gets into the whole. And that's a whole.
00:29:31 Val: That's a whole other episode to.
00:29:33 Matt: Itself.
00:29:34 Val: Ready for that.
00:29:34 Matt: one. But again, it builds on this. Right. And that's what we're trying to do. Right is let's let's kind of get, you know, let's let's balance this within our day and age and, and the delightful chaos that we have.
00:29:47 Val: So just real quick as a little fun exercise, what did Winston help us do with this podcast?
00:29:54 Matt: Oh, well, so so.
00:29:57 Val: How many different ways did AI help us make this podcast?
00:30:00 Matt: Well funny enough. Okay, so, um, he connects a lot to the platform that we actually use to distribute. So he helps fill in kind of the details. He'll, uh, pull the transcript of the episode, put in the appropriate keywords. Um, I am going to try to work start connecting him to some social media pieces as well, to kind of keep an eye on stuff or when people, you know, bring things up. Um, so that's kind of where he's starting to plug into things. You know, I will pass on like articles or ideas, um, things that we've talked about, like.
00:30:40 Val: Put it in a shared folder.
00:30:42 Matt: Shared folder, and then then, you know, he reviews and looks at it and then, um, pulls, pulls that stuff together again. What you would hope from an executive producer, right. So again, it's, it's that's that's how we're kind of using it to kind of help pull things together.
00:31:02 Val: So we can actually do this podcast and still raise kids and have a day job and.
00:31:09 Matt: Live.
00:31:10 Val: You know.
00:31:11 Matt: Get sick for weeks. Come out. Exactly.
00:31:16 Val: Anyway, thank you all so much for listening.
00:31:18 Matt: Thank you everybody.