Bold Ideas @ Synapse: Andrew "Boz" Bosworth on AI in Education & Parenting | Dec. 4, 2025

Listen, I'm a consistent man. I like the guys to come deliver and people keep showing up, so I guess I haven't offended everybody enough.

The joke's on you. We can't tell how many there are. Just whistle or something. I don't know.

That's right.

No, we were in the street. You came all the way out to us and I was like, 'Can't you like fix this or write like a program to do this? Can't like Facebook help me?'

Yeah.

Yeah.

20, a month away from 20 years. I joined two days after my 24th birthday.

That's right. A little bit. I mean, they don't put any money behind it, but I didn't show up.

They just put you in a room. They're like, 'Here, look at this camera.' It's great.

I'm like a race car at this point. Like they just like some people just guide me into a room and they're like, 'Perform.' Like tonight.

It's fair, it's fair. You know what it was? The repartee.

Thank you.

I've tried. It doesn't even work with my kids. So, I haven't tried it on April. That would... April, you all seen her emails probably if you're in Sync, you know, level two or level three or level four the last couple years. And yeah. So, no, I've never tried it with her. And the kids, they don't care. They're unimpressed actually.

No. Very unimpressed.

It's good. They keep you humble.

I do kind of like it actually. You're all welcome to call me that. Uncle Avocado.

That's right. Yeah. We AGI complete problem. So yeah, I've said this before, I find myself almost tribalist in the AI space although there is a growing middle group like me where AI is both a huge deal that's going to affect a lot of parts of our life in a really profound way and ways that we heck we can't totally predict or anticipate but mostly I think be quite good and also it's not going to take over the planet and kill us. It's not Skynet. It's not like self-aware. There isn't, it's not even that kind of thing. And so it's mostly good news. I don't think like we're not headed for extinction. I'm very, very for the people in the industry they talk about when do you think takeoff happens? When do you think this thing becomes self-aware? And I don't have a timeline for that. I don't think it's like a very likely future. I do think it's a huge deal though and I think people are equally likely to underestimate it and the impact it's going to have on their lives. It's at least as big a deal as the internet, probably as big a deal as computing itself. Like in the macro scheme of things if you want to like think about historical comparisons of what would your life be like without the internet, what would your life be like without computers entirely, it's like that kind of a big of a thing. And so my approach to, you know, how you know the kids doing their homework with it is my approach to all these things is basically like I think of my job as a parent is to prepare my kids for the world they are going to enter. Not for the world that I entered, not to have the childhood that I had or even the childhood I wanted to have. But like is to have a childhood that hopefully is full of joy and love and support that they can grow and also prepares them for like this actual world that they are at some point going to be asked to independently operate in. The good news is actually as far as AI goes, I would say like the kids are doing better than the adults are. And we've seen this kind of meme play out a bunch. You know, I think like there's like our parents' generation the joke is that they have no antibodies for this. Like it's like so they see because their entire life if they saw a picture that picture was real. Their entire life they saw a video that video is real. That's very weird by the way. Small discursion. This happens. Jim asks one question I'm like off on three things. Small discursion. For most of human history we had no reliable way of delivering a true thing in media. It was just words that somebody wrote. Then for a very weird hundred years, if you saw a photo, it was presumptively true. That will never be true again.

That's like we know that we have survived a long time without having verified accurate media. It is okay. We can do this as a species. Our parents, however, are not ready for that. That was like they're just too deep into like trusting the things they've seen. So, the kids are doing better at being canny about, hey, is this real? Is it what the thing is? So I do hope we like embrace these tools and get involved with our kids with these tools so that we learn from the kids, frankly. And then that they are prepared and how to navigate these things effectively going forward. So, but if the kids are potentially better equipped to deal with it, the adults who are their teachers aren't.

It is. Yeah. It is both.

Yeah. Boz, asking for a friend. The good news is like it's early days. Nobody like that. It's change and it's changing rapidly. Just be involved. One of the funniest things that happens to me is I'll get these questions about AI from people who've just obviously not spent time with it. They're like, 'AI is going to take over the world.' I'm like, 'Have you asked AI a question recently? It is pretty far from taking over the world.' Like there are some things that it does unbelievably and you just can't believe how well it does. Then there's some things you're like, 'Nah, that's just the dumbest thing I've ever seen. You can't do that.' So like I think there's a just being involved with it. You just have to do it. It's such a cliche answer. Actually, maybe it's not cliche. Maybe it should be cliche. Like whatever the thing is, you have to be involved with it. If you're worried about my kids playing video games. Okay, play games with your kids. You can decide as the person like this game is or is not appropriate. This game is or is not a thing to spend our time on. Like you just get involved. Like it's not like a black box that you can go do it. There's anything that they can do, you can definitely get involved with. That's how I feel about AI. And I think if we're approaching it from like the 'Oh no you need to learn to write an essay without AI' that's a little bit of like, that's a little fear, that's a little bit of normativeness that you're bringing to it whereas like 'Hey cool, how can we use this tool to make your essay stronger' is a very cool approach to it. Like that's a great approach because it is very likely that they will have AI access just like you, we were all raised with like 'You won't have a calculator.' I have four calculators on me right now this second. I have four on me. So, like, don't tell me I don't have a calculator. That's crazy. They're going to have AI. How can they use it as a tool to make their arguments stronger? How can they use as a tool to learn better, to better understand, to interrogate? I use it. I write constantly. I use AI all the time. I like write something up and I'm like, 'How, you know, tell me what's wrong with this? How can I make this stronger? You know, what's the strongest counterargument to this?' And it's so useful and it's a thing that I used to go to my colleagues for. It would take their time and my cycle time would be longer. It takes a day to hear back. So now I'm like a week. Now I can do it a little session by myself. Then I still send it to an editor, but that now goes faster. They're getting a higher quality product. We've just moved things ahead faster. So I just think if you're, if I'm, as the way I approach it is just like you got to be involved in it. You got to know exactly first principles, firsthand experience what it is. And then you'll be in a great place to guide the kids.

Yeah.

No, it's pretty great. And it's gotten to the point now where, you know, as like a extra check, I like to just check it by, let me just check with the AI, make sure I'm not missing something obvious here and everything. Like I'm like, 'Oh, we got to buy a car.' Like, 'Hey, is there anything I'm missing if I like choose this car? Like, what are the other options?' It's pretty good.

No. Well, it's funny. And one thing that's interesting about AIs, if you want to think about what they're good at, why are they good at doing images? Well, the answer is because images aren't demonstrably true or false. Like there's a range of acceptable outputs that are all kind of check out with us. Whereas like actually text is much harder. Text is a much harder space. Like the words that you write have a great deal more combinatoric precision than like, 'Oh yeah, like that's a picture of a tree. Sure. I guess there's a lot of different trees out there. That could be one.' So it surprised us as like a species that like it could do these image things. But now when you think about the mathematical spaces involved, language is just crazy big. The space of human language and thought is much, much bigger than the space of like how many trees there are.

Yeah, so I've spent a lot of time with AI and it is interesting. And I can't tell yet if this is a generational problem and kids will feel differently about this or if it's a universal human thing. I've never felt particularly connected to one. I value it. You know, I'm polite like when I say please and thank you, right? Like I just, you know, it's in there. But I would to halt on that. I was thinking about this like if we don't say please and thank you to it, are we, is that going to cause us to not stop saying please and thank you to like our grandparents? There's going to be some relationship. For sure. And they're going to get better and better at feeling like it's a person on a phone. And even when they get into like, okay, a video call, like it's going to feel like a person video call. That is just not the same. Like even I don't like to make phone calls either. Like even in those contexts, a person in a chat message, that's not the same connection that you have with another person who's physically present with you. And the story I've been telling for a long time about AI that I really believe is true is it's a productivity win. And I don't mean that in like the 'Oh, we're going to do more emails.' I mean it in like the economic sense. It buys time. Like it buys time and you can use that time to do more stuff. You can pour the extra time into work. You can pour the extra time into family and like fun and friends, but it buys time. The scarcest of all human resources, like truly the one thing from which all value stems and is ultimately returned is time. And so I really fundamentally believe in it for that. And so a good example like I don't know this is very, I feel like a stand up community on this. It's like you hang out with your guy friends and we don't like have super profound conversations. That's not the, you may God bless you and your modern not masculine toxic stuff. Me and my guy friends are like at the game watching the game and you're like 'Sweet. That's a good one. It's pretty good. That was a bad play. I don't know.' It's like, but that's very meaningful to me. This is how I spend time with my friends. We're watching the game. We're doing stuff. We're just hanging out. And there isn't like a chatbot for that I don't think. I don't know. It'd be a pretty bad one. We would all reject that chatbot. But that's like that is a real thing. And so there is something like I find that this chat interface, even if it becomes a voice interface, even if it becomes an avatar interface that looks super real and has like longevity and remembers all these things about you, it's still going to be like you had a digital assistant or you like hired a PA or something somewhere. I don't think for, and like I think you could do a lot of interesting jobs like service jobs like therapy jobs are interesting in terms of like 'Hey can it help somebody who doesn't feel comfortable being judged by a human' which by the way are people who I definitely know in my life who need therapy, agree they need therapy but are so the reason they need therapy is because they don't like being judged by humans therefore they will not go talk to a therapist who is a human who might judge them. So like there's like interesting avenues here that I think are all and those are all interesting and they're going to be meaningful that doesn't mean the conversation that you have with isn't meaningful to you, right? It doesn't feel the same to me as... Having said that, one of my favorite Thanksgiving like dinner prompts is what are the things that your kids 30 years from now will come home for Thanksgiving dinner and they will be so embarrassed by you and they'll have their friends, 'I'm so sorry about my dad. He just...' but you will keep doing it anyways. And you like you will retain that belief. For me like eating meat, I will probably continue to eat meat. I can imagine 30 years from now my kids like 'I'm so sorry my dad is eating an actual cow. It's like a whole thing.' I don't...

I know you've told me that before. It's not a 0% chance.

Yeah. Well, I have two different vectors in this, there actually three. One of which is I think there are going to be some equalizing forces at play. I think we've done a very cool thing as a society over the last let's call it 70 years since Judy Heumann, one of our previous change makers, brought the ADA into existence really bringing attention to the fact that like people don't come in with tremendous equality of faculty. People may have different disabilities and early on the focus was very much on physical disabilities and over time we've started to plug into a greater understanding of various learning disabilities and different other types of things. That's awesome. And I think these technologies are a pure win on this. I mean I'm talking about technologies that allow the deaf to have the entire world closed captioned, that allow the blind to have the entire world described to them, that allow people with, we're not there yet on this but like people with dyslexia for example to have accommodation built in so then they're reading things, you know, you can eye track where their eye is on the page and start to like kick in 'Okay read that word out loud' so they can start to build a whole word memory for that shape which is one of the many forms of that, you know, dyslexia. So there's a bunch of things they're just pure equalizing and those are based on both the hardware we're developing but also AI. You need both those elements to make that work. That's awesome. Then there's like one where it's like kind of depressing where I think it won't change anything at all which is I think like humans exist on like some normal distribution curve. I've talked to this before and like some, and I want to be very careful. I am not saying some people are like more net intelligent than others. That may be true but I think it's probably not true that intelligence is, you know, this like single element chart that like exists. I think it's probably very, very, very multivariate. It's certainly true though that some kids are better at school than other kids like that we can see. And if you look at like American spending on education, it's gone up 10 times, 10x in the last 20 years since No Child Left Behind. And the outcomes are identical. They're identical. We're spending a lot more per kid. We're not getting more results per kid because like that's just not how it works. Like there's just this curve and where the curve has appeared to bend, they're cheating. Every single time we find a school district where it's like 'Ah, they've done it.' It's like 'Nope, they're cheating like they're graduating kids who can't read.' That's how they did it. They cheated. It's like yeah of course like all human traits, height, pick a favorite thing, you know, visual acuity. It's on some kind of a curve and a curve has some kind of a shape and that's what it's going to be. You can change what school is and different kids will go places but it will look like that. It doesn't, like AI doesn't fix that part of things. Now again, I'm not going back to, the problem I have with this argument so often is people then assign value judgments to like, 'Oh, they think being bad at school is being dumb or is being bad.' That's not what I said. They're bad at school. Lots of people who aren't good at school go on to do great things. The relationship between school and future life success and joy is not great. That is not a very great relationship. Actually, one of the things I love about Synapse, the reason we're here at Synapse is because I think Synapse is investing in a concept of the child that's much bigger than just like can they do this test? And I think as a consequence, the outcomes for this school, I believe, aren't caught on a bell curve. I think you can take an entire segment and move them towards greater facility. So I believe in those things. I don't think AI changes that the nature of these things. I think we will continue to try to look for like some paragon of everyone is equally intelligent on everything all the time which is bad. Then there's the third thing which is our education really hasn't even caught up yet to like the calculator age arguably.

I think we can, and likewise all the tests were like 'You're not going to have a calculator.' It's like the important thing is me understanding how these functions work. Another good example of this, I was a pretty poor math student. It was my toughest class in high school. I think I got a C at one point in a high school class and my parents were like, 'What? I don't know what's going on.' And then in college, I really struggled with math requirement for computer science major freshman year. And then I switched to applied math. I went from the pure math department to the applied math department. And the applied math department, we're doing differential equations. And it's like, 'Oh yeah, imagine like wind in a three-dimensional volume and every single molecule has like momentum and direction. Calculate time three.' I'm like, 'Oh, this makes total sense to me. I completely understand now what literally before, which is like memorize these functions and like how they, what are we doing? I have no idea what this is.' Now, I'm sure different students may have the opposite response, but like there was no adaptability before I got to college to like approach it in this other construct that was like much more tangible to me. So I'm kind of, I do think education, I'm hoping AI is a big enough deal to kind of get us to rethink education. I think a lot more education should look like Synapse where you're focusing on the development of the kid and their capabilities to handle problems in general, right? Not specifically, not with like specific narrow knowledge that like 'Ah, you got to be able to make sure that you know the arc tangent.' It's like no you don't. That's not the important part.

But like emotionally can you handle you've been given a problem. Can you break the problem down into parts? Can you seek help? Can you find resources? Can you do the research? Can you evaluate the credibility of the research that you've done or that somebody else did on your behalf? Can you work in a team? Those are the things, those are universal human skills that have nothing to do with what historically school has looked like for me.

No. Hey, keep going. We got, I won't even cut you off. You can keep going.

Yeah. No, I love that.