---
title: "Pathologically Positive: Children, AI, and the Internet"
date: "2025-03-29"
updated: "2026-11-10"
duration: "about 52 min"
format: "Podcast (audio)"
host: "How to Ruin Dinner podcast, University of North Florida"
source: "https://marytreyz.domains.unf.edu/podcast/episode-83-pathologically-positive-children-ai-and-the-internet/"
canonical_url: "https://zoltner.com/speaking/transcripts/how-to-ruin-dinner-pathologically-positive.html"
speakers:
  - "John Zoltner, Guest"
  - "Mary McGovern Treyz, Religious Studies Department, University of North Florida (host)"
  - "Hayden Heller, Religious Studies Department, University of North Florida (co-host)"
  - "Rose Donna, Chief Operating Officer, AIChildSafety.org"
topics:
  - "AI and child safety"
  - "Parents and families"
  - "Regulation"
  - "USAID funding cuts"
---

# Pathologically Positive: Children, AI, and the Internet

*How to Ruin Dinner podcast, University of North Florida · Episode 83*

> Machine-generated transcript; it may contain mistakes. See “How this transcript was made” at the end of this page.

## Summary

Mary McGovern Treyz and Hayden Heller of the University of North Florida’s Religious Studies Department host this episode of How to Ruin Dinner.

I talk about how AI is changing childhood, including recommendation feeds, misinformation, sexual exploitation and over-reliance on AI, and about what parents, technology companies and regulators can do. I also talk about how the sudden pullback of US development funding hit child protection and other work. My message throughout is that we can’t do with AI what we did with social media.

## Key points

- Children meet AI most often in social media feeds and recommendation systems, where it reinforces what they already engage with and makes misinformation more persuasive.
- I don’t favor banning AI in schools or spying on children with spyware. I favor teaching pro-social, effective use, and building an open, trusting relationship with your child.
- Governments need to set basic responsibilities for technology companies, such as age verification for content that harms children, and regulators need to be active from the beginning.

## Questions asked

The questions as put in the conversation, with the time each was asked.

- [00:00:00] **Mary McGovern Treyz:** Do you want to tell us about your background and what you do?
- [00:03:06] **Mary McGovern Treyz:** Like how long have you been working with AI technologies?
- [00:06:20] **Mary McGovern Treyz:** But do you, do you two both think of it as a democratizer?
- [00:11:12] **Mary McGovern Treyz:** I can’t control myself, how is a young child supposed to find the discipline to control that want?
- [00:27:04] **Hayden Heller:** Do you like have any thoughts or opinions on AI being like the next evolution of these responses in this technology?
- [00:32:24] **Mary McGovern Treyz:** What, what are some of the regulations that can be put into place working with children specifically?
- [00:35:54] **Mary McGovern Treyz:** How much do you think sort of the corporate and the technology producers are considering their responsibility towards protecting children and protecting old people like me from scams?
- [00:37:58] **Mary McGovern Treyz:** What would you, when you talk to these trust departments, what are you suggesting for them or how are you working with them?
- [00:39:31] **Mary McGovern Treyz:** What would you recommend parents do to protect their children?
- [00:42:04] **Mary McGovern Treyz:** And is it really that different from, you know, watching too much TV?
- [00:47:25] **Mary McGovern Treyz:** Can I get you to tell us a time when you’ve ruined dinner with technology?

## Transcript

The wording is as spoken, with punctuation and obvious errors corrected. See “How this transcript was made” below.

**[00:00:00] Mary McGovern Treyz:** Hi, and welcome to How to Ruin Dinner, conversations from the university. I’m here with my regular co-host, Hayden Heller, and then my lovely sister-in-law, to whom this season is dedicated, Rose D’Hanna. And our guest comes from Rose. She roped him in, shall we say. And our guest today is John Zoltner, who has been in the technology field for a long time and worked with Save the Children. And he’s particularly interested in the concerns around child safety and the internet, largely. And today we’re going to be talking about how AI is going to impact children for good and for bad. So we’re really glad you’re listening and get ready for some dinner-ruining conversation, we hope. John, welcome to How to Ruin Dinner. It’s a pleasure to have you here. Do you want to tell us about your background and what you do? And I guess I should have said your last name too, John Zoltner with a Z.

**[00:01:25] John Zoltner:** Sure, sure. Thanks, Mary. Thanks for having me on. First of all, like you say, I’m John Zoltner. I have been focused on the intersection of technology and social impact for my whole career. And in the past several years, I started focusing on online safety in children, and most recently on AI and online safety.

**[00:01:52] Mary McGovern Treyz:** And you studied, you went to Georgetown. I saw that on your LinkedIn page. And what did you study in college?

**[00:02:00] John Zoltner:** I was an English literature major and philosophy minor.

**[00:02:05] Mary McGovern Treyz:** Okay.

**[00:02:06] John Zoltner:** Later on, I did an MBA. So nothing related to technology whatsoever, which always, I lived in Latin America for a while. And it always causes issues because they assume I’m a computer scientist. And that’s the farthest away from what I am.

**[00:02:24] Mary McGovern Treyz:** Well, how did you get into technology?

**[00:02:28] John Zoltner:** I, you know, right after school at Georgetown, I was working with kids on juvenile probation in D.C. And one day someone came in from Howard University’s computer science department and said they wanted to tell the stories of the kids we were working with using multimedia. And so I said, great, what’s multimedia? This was at the beginning of being able to just do video and images and sound and all together.

**[00:03:06] Mary McGovern Treyz:** So we’re right back to a brand new thing with AI, right? Like how long have you been working with AI technologies?

**[00:03:16] John Zoltner:** The past few years, I’d say about three or four years. And yeah, my whole career has been kind of surfing to the next thing.

**[00:03:27] Mary McGovern Treyz:** Yeah.

**[00:03:28] John Zoltner:** Arriving just barely before the rest of the field. I’m like, I know what I’m talking about. And I’m doing the same thing in AI. But I had a lot of opportunity while I was at Save the Children. At Save, I started what we called the Safe Digital Childhood Initiative. And that gave me a lot of space, a lot of entree into both the companies and other global nonprofits or local nonprofits that were working in the issue. And I had a chance to experiment with some of the technologies also.

**[00:04:04] Mary McGovern Treyz:** Wow. Well, I know you worked with Rose in technology as well.

**[00:04:09] John Zoltner:** Right. Yeah. When I was down in Chile, Rose was running Datadyne, it was called at that time. And so I was a Latin American representative and we started up a project called Datagro, which was at the beginning of when you could start to use cell phones for development work. We were sending market prices to smallholder farmers in Latin America.

**[00:04:38] Mary McGovern Treyz:** Well, that was one of the things I’m most interested in, I have to say, the way in which technology can open up, you know, fields and access to information for people who really would never have access before. And that seems to be happening again. And Rose, I don’t know, is it similar in that way in that AI is going to give people a tool that really only experts have had access to or maybe experts is not the right word, but.

**[00:05:13] Rose Donna:** No, that’s I think that’s a very valid point. I mean, again, you know, what John was talking about, you know, we’re working with rural farmers who maybe probably only had a cell phone maybe a year or less than that. And then they were able to jump right in into using their cell phones to collect data and get information that they may not have access to because they were living out in these rural communities. Without internet or, you know, just general knowledge of any kind of technology. So, yeah, I mean, I think that’s exactly where AI is now and the importance of what John is doing because it is so unknown and there’s, you know, it’s still the Wild West and there’s a lot of things, you know, there’s positives and negatives to it. And so it’s important to, you know, kind of put some, put things into context and start thinking about regulation.

**[00:06:20] Mary McGovern Treyz:** But do you, do you two both think of it as a democratizer? I mean, this.

**[00:06:26] John Zoltner:** I do. Yeah, I do in a way. I was part of the community technology movement, which grew up in through the 90s, mainly in small, basically computer labs or they call telecenters in most of the world that provided access to low income neighborhoods, the kids or adults there. To teach how to use technology, to provide access to the internet and to teach how to use technology tools so that they’re competent in their use. And yeah, like Rose was saying, in that project with Datagro, we had to teach people the basic skills for how to send and receive a text message. And it’s the same thing now with AI, we need to teach basic skills and it provides access to information that so it provides inclusion in that way and democratizes that access to information. But we talked a lot about that in the community technology movement and the dream was that, and it’s true in a way, but the dream was that people would now be able to publish to the entire world and make their ideas heard. And it’s true, but commercial interests still drown them out, so, you know, an individual can’t publish in the same way the New York Times can, so it provides some of that possibility and it’s, we want to see them have access to information in these kinds of ways. But at the same time, it doesn’t do what we had dreamed about in the 1990s.

**[00:08:31] Mary McGovern Treyz:** Yeah, well, that’s kind of everything, isn’t it? But in full disclosure mode, you sent me some great questions and now I’m throwing out all my questions and I’m going to go to your first question because it sort of fits in here. And that is your, how is AI changing childhood?

**[00:08:54] John Zoltner:** Well, I think when people talk about AI now, they’re generally talking about and thinking about generative AI. So, about using things like ChatGPT and Anthropic’s Claude, but really, I think where kids are most interacting with AI are in their social media feeds and in recommendation systems that they may not be aware of. Or they’re at least not, not necessarily aware that AI is controlling them. So, these feeds that continuously reinforce information you’re interested in or information you come across while browsing and it’s creating, just as it is with adults, it creates these echo chambers that reinforces messages that are sometimes misinformation. And kids are more vulnerable towards believing misinformation and kind of going down that rabbit hole to, I don’t know, think the earth is flat or believe conspiracy theories.

**[00:10:19] Mary McGovern Treyz:** Yeah, that seems the big danger, along with the advertising, I would say. I mean, the culture of envy and greed and what other vice can I throw in there? But that culture of insufficiency, not enough, is one that is, I think, really problematic, where we, in America, I think, certainly in the United States, that feeling of it, I’m not enough. I don’t have enough, I need newer, better, bigger, whiter, everything, right? We’re exporting that, it feels like, on these social platforms where we’re allowing advertising, especially if it’s going to be directed at kids.

**[00:11:12] Mary McGovern Treyz:** I can’t control myself, how is a young child supposed to find the discipline to control that want?

**[00:11:25] John Zoltner:** Sure, and I’d say it’s not just advertising because a lot of what we know is harmful about social media is comparing yourself to other people. So, you read a lot about people only posting positive information about themselves or, you know, their vacations or, you know, things that they have, fashion, makeup. And especially with the work that I do in the global south around the world, you know, those kids in Bangladesh are watching kids in the U.S. who are unboxing these incredible technology goods. And so you can imagine what they feel as they’re watching that, in addition to, you know, girls watching and wanting to be skinnier girls or watching other girls with filters applied so they look more beautiful. And so it’s very easy to fall into this negative comparison frame of mind.

**[00:12:32] Hayden Heller:** I noticed, Mr. Zoltner, a lot of what you’re saying here is that AI can be a powerful tool for globalism, for kind of broader capital L liberalism. I noticed you had a paper recently or like a conversation talking about the pausing of USAID funding. How is AI kind of a part of all of this and how does this kind of affect kind of the U.S. soft power?

**[00:13:01] John Zoltner:** Or like going forward, using AI, so yeah, I mean, I’d say AI is one line of work in the sort of development and innovation work that I do. So my work is at Save the Children and in general, before that I worked at a health organization called the FHI 360 that also did a lot of work with USAID. And half of that work is technology for development and innovation, which is basically making development projects more effective or more innovative. That’s like that work with the dot agro program I did with Rose in Latin America. And the other half is online child safety, which at this point is more than half. And a lot of that work, especially in the developing world has been done with USAID funds, either directly. Or with USAID as a partner that was really leading that work. And so especially the sudden pulling out of aid and humanitarian dollars has kind of dropped the floor out of not just that work, but global health and global education work and humanitarian response. And so it’s really an emergency moment in the field where we’re trying to figure out what we can keep going and what we need to stop doing. And in that post, I talked about these awful choices that we’re having to make between, you know, do we respond in a big way to an earthquake in one country? Or do we do the kind of services we always have to improve maternal and child health so that babies can live, a higher percentage of babies can live.

**[00:15:20] Mary McGovern Treyz:** I think it’s worth going back for just a second and spelling out what USAID stands for, the United States Agency for International Development. And really what a small part of our federal budget it is, but it seems to me what important work it does for the United States, community building outside of our country. And I mean, I think for most Americans, maybe that work has gone unnoticed and all of a sudden we’re hearing about it in the news without understanding its full impact. And why it benefits America to have a relationship on such important level in people’s lives in foreign countries. And meaning their basic health, their basic access, that kind of work, we just don’t hear about it enough, I think, so that we don’t appreciate when the funding’s gone, what that will mean to us. We talk a lot about pulling military aid, that kind of stuff, or tariffs, but this kind of work often seems to go unnoticed. I don’t know. You two are the experts. Do you think that’s true or?

**[00:16:45] Mary McGovern Treyz:** I mean, it’s very true. And John, you know this because, you know, even though it’s called the U.S. Agency for International Development, the staffing, let’s just say an example of Ethiopia. So there’s a U.S. aid office in Ethiopia, and sure, there’s what, maybe, I don’t know, maybe four or five U.S. expats there, but they’re also the people that they’re hiring to implement these projects, such as if they’re doing a food distribution, they’re hiring local drivers, they’re hiring local staff, secretaries, project managers. Those are all, like, the local, building the local economy and making that image of, okay, well, yes, this is United States funding, but I’m Ethiopian, and I’m working for this agency, and I’m helping my fellow Ethiopians. And so, I mean, again, it’s not just, the problem isn’t just not having access, you know, people having access to the services. It’s all these people who were working for them are also out of jobs, and also that image of, you know, the United States, they’re taking away my jobs, I was, you know, going to send my kids to college, et cetera, et cetera. So, the implications are so far-reaching, and I think that’s something, to your point, that a lot of people here in the United States don’t get. It isn’t just Americans over there working, and now they have to come home.

**[00:18:23] Mary McGovern Treyz:** Right, right. Or what the mission of USAID is. I don’t know that people know what that mission is, that you’re talking about ending hunger and educating and protecting children, that that is what makes America good. And in its goodness, maybe there’s where the greatness lies. But we don’t have that conversation at home much, it doesn’t seem to me. And so, I’m happy to talk about it today, because it’s so finely in the news, as it’s having its funds cut. It’s just, it’s ironic, or maybe ironic’s not that’s, disturbing is maybe a better word.

**[00:19:09] John Zoltner:** Yeah, and, I mean, USAID, the most area of criticism I’ve always heard is people saying that USAID is a tool of the government, of the State Department. That it’s more conservative, not more liberal. So, I mean, two things. People think that a lot more money goes out in humanitarian assistance when it’s about 1% of the budget, I think.

**[00:19:37] Mary McGovern Treyz:** That’s what I read, too.

**[00:19:39] Mary McGovern Treyz:** Between 1 and 5%, 0.5%. Right, and that people think it’s about 20% of the budget. So, there’s this huge chasm between the reality and what people think.

**[00:19:52] Mary McGovern Treyz:** Well, when you say $63 billion, it sounds like a huge number. And you need to know the percentage, because these numbers are, you know, outrageous. And who can keep trillions of dollars in mind?

**[00:20:10] John Zoltner:** Sure. Yeah, and, I mean, the rice that goes out to feed people or grain through USAID, everything is printed, help from the American people. So, it’s really low cost comparatively to the Defense Department.

**[00:20:30] Mary McGovern Treyz:** It’s marketing and, like you say, the soft power from the U.S. So, when we invade a country in order to accomplish some sort of military objective, like in Afghanistan, the armed forces only carry it so far. And then USAID and nonprofit organizations, international NGOs, come in to provide that support so that the country doesn’t fall apart after the military pulls out. So, there’s a pretty important global role that USAID has always supported.

**[00:21:17] Mary McGovern Treyz:** Yeah. Well, one of your other questions that I’m very interested in was this idea of what are the biggest risks that AI poses to children. Could you talk about that a little bit? Because I think the educational risks are sort of outweighed by the educational potentials. But I wonder what you’re seeing as the most dangerous part or riskiest parts of AI.

**[00:21:51] Rose Donna:** Right.

**[00:21:52] John Zoltner:** I would say, and this isn’t in order, but I’ve done a lot of work in online child sexual exploitation and abuse. So, one thing, of course, we’re seeing is deep fakes or AI-generated content that’s explicit of children or youth and or just access to even if it’s not meant for children, that children have easy access to those kind of materials now. And we’ve done some research I just finished with Western Sydney University and saved the children at the end of last year, a study on online grooming. We know that children are much more liable to believe people who say they are somebody online and whereas with some adults, at least, it’s more obvious when people are posing as someone who they’re not. So, it’s much easier for children to get access to that kind of explicit material and even to create it. I mean, we know that a lot of what we call online grooming is being created by children now. You could imagine children in high school everywhere, if they have access to these Nudify apps, are going to use them because it’s fun, it’s a joke, it’s they don’t have the mature enough decision-making capability not to understand the true effects and understand the potential dangers of them, of being arrested or encountering law enforcement and then misinformation, as I talked about before. It’s AI is a way of just really cranking up the volume of misinformation and also the effectiveness. So, I think it can make much stronger arguments and in many cases you can even ask AI models for how can you talk to a child of X age about whatever it is you want to convince them about. Scams, we know that adults fall for financial scams all the time and of course children are, it’s much easier to fool children about those things and what we’ve seen, I guess it was two years ago now, but there was a sudden spike in what we call online sextortion financial sextortion for using sexual content or tricking kids into sharing sexualized content of themselves. And we know that happens through organized groups. So, they were sharing scripts and ways of kind of techniques for entrapping children and then getting them to send money however they could access it. But AI makes it much easier to do. The information bubbles I talked about, just creating an information environment where the children are steeped in misinformation, lies, either concrete stuff like flat earthers or information that, you know, can draw them into extremism. Like people who are recruited to go fight with Al-Qaeda even though they had no connection to that before. And then I think also just in over-reliance on AI.

**[00:26:16] Mary McGovern Treyz:** Yeah, I wanted to ask you about it. Because that’s so easy to use now.

**[00:26:23] John Zoltner:** And I think, look, I know a lot of schools have bans on the use of AI. I don’t believe in doing that. I think you need to teach pro social effective use of any technology tools, including social media and AI. But I think you, if the first step is to go to an AI tool in order to complete your homework or come up with arguments for a debate or an essay, you’re going to lose that ability to make those decisions yourself.

**[00:27:04] Hayden Heller:** Yeah, I wanted to talk a little bit about like a personal anecdote here. I’ve kind of avoided the AI bug and I’m generally very against, you know, its usage. I think the history department, which I’m in, kind of all has that same opinion. But Dr. Treyz convinced me to kind of try it one time, you know, see what happens. I was grooming. So I apologize to Dr. Bevel, but I tried to use it for a historiography paper. It’s basically like a long lit review in the history department. And the first, like I basically asked it for some historiography sources on like immigration in this time period. And I kind of used, like, I talked to it like it was a person. And the first source that it gave me was a made up book by an author that’s in a completely different field. And me, I think having a decent basis of understanding for what the paper was already, I knew to question that. I said, hey, I don’t recognize that name or that source or the context that you’re putting it in. And it just replied, oh, sorry, you were right to be suspicious of me and my involvement in, you know, giving you sources. And it feels like if students are kind of coming into school with this, combating that over-reliance so that the misinformation is going to be that much harder because it’s very convincing in its responses. But I also wanted to connect this to a previous episode that we had with Dr. Scherer. I mean, she, her research is on nonprofit usage of social media, especially during crises. And she basically said that social media has been used to communicate compassion and communicate benevolence in times of need for people. And I feel like AI is like another step in this direction as well. I know, Mr. Zoltner, you had kind of comments on California wildfire policy for AI or something like that. Do you like have any thoughts or opinions on AI being like the next evolution of these responses in this technology? Like, do you have anything around that?

**[00:29:13] John Zoltner:** Well, I think it can be, and I think it obviously provides tremendous access to information. And to synthesizing ideas. But you’re right. And you talked about that first resource that the, whatever the model showed you. If that were the 37th resource, you might’ve missed that no matter how technically knowledgeable you are. And what’s happening now is these deep research capabilities of, I know, Google had the first one and then ChatGPT does, Grok does, Perplexity does now. So, and they’re really good. Some of them are great. And you could generate a long, well-resourced paper, but sprinkled within the, that paper might be some untrue facts. And a lot of times maybe they’ll be inconsequential, but sometimes they could be hugely consequential and they could change the whole argument you’re making. They can make you or your readers make assumptions that aren’t true. So, yeah, I think, I do think it’s a huge step forward. But I think we need to solve some problems like the hallucination problem. And there are some ways of doing that, that are emerging, like using AI agents that challenge each other. So that one, you know, maybe provides the first paper argument, but the other one is basically fact checking it. So you have more than one agent talking amongst themselves.

**[00:31:08] Mary McGovern Treyz:** I hate to say, I don’t know what the hallucination problem is.

**[00:31:12] John Zoltner:** Oh yeah. It’s hallucination within the AI concept is a term for AI making things up basically. So you get these information presented as facts, like Craig was talking about. And they’re a lot of times made up and the way these AI models work is that they’re looking at the preponderance of information in their, not just on the internet, but in their training data. And they’re predicting what either the next token, the next word should be. And they’re making arguments based on that. So they tend to be kind of general and based upon what people say. And so they’ll give you this return of information based on what people talk about on the internet.

**[00:32:21] Rose Donna:** It’s very human.

**[00:32:22] Mary McGovern Treyz:** Yeah. Right.

**[00:32:24] Mary McGovern Treyz:** I think I have lots of hallucination problems in my conversations. Um, John, one thing, I mean, so the, of course, burning question is what, I mean, what can be done? What, what are some of the regulations that can be put into place working with children specifically? I mean, of course that’s on the tech level, um, and, uh, you know, government regulation level.

**[00:32:55] Mary McGovern Treyz:** I mean, what can be done?

**[00:32:57] John Zoltner:** Sure. Yeah. First of all, there’s a concept. We talk about safety by design. There are a set of safety by design principles developed by the eSafety Commission in Australia, that’s one of the leading regulators, and youth participation. So making sure that youth are consulted in the development of these, we call it frontier AI models. The large language models or systems, base systems, groups like OpenAI or Anthropic develop. And the systems or tools are tools based upon those models. And so both for the tools and for those frontier AI models, children or youth need to be consulted. Or people from other vulnerable groups, so that they can use their input to do some kind of risk analysis to understand how it might affect those children. And basically I think we can’t do with AI what we did with social media, which is not anticipate negative effects and think that, you know, ‘Oh, these are tremendous new tools. We’re going to connect the world.’ And because we found out with social media that maybe part of the world shouldn’t be connected. And, you know, there are people we don’t want talking to each other. So we can’t be pathologically positive as you called yourself.

**[00:34:46] Mary McGovern Treyz:** Right. Right. I love that. It’s eternally hopeful, but I’m using your own words against you, but you are kind of warning us about being pathologically positive.

**[00:35:02] John Zoltner:** Right. I think the next few years will define how AI is used in the coming decades and whether you believe in AGI or artificial general intelligence, which is basically machines, computers, reaching the level of intelligence of humans. And once they do that, they’ll be able to recursively improve themselves, though, they’ll be designing software and hardware that can make themselves smarter and more generally intelligent. So a lot of people believe that will happen within the, really within the next five years. And if we’re not ready for that, we’re in big trouble.

**[00:35:54] Mary McGovern Treyz:** How much do you think sort of the corporate and the technology producers are considering their responsibility towards protecting children and protecting old people like me from scams? Scams, and how concerned do you find this larger corporate world is?

**[00:36:26] John Zoltner:** Yeah. Well, I have a lot of contact with the people from the, what they call trust and safety teams at these companies, which are the teams responsible for protecting children or protecting other groups. And so they’re great and they’re really committed. They’re doing interesting, innovative work to try to make their platforms safer for kids. But the company has to invest the appropriate resources. So, and that’s rarely happening. So I think what happened, the base model for a lot of internet commerce is getting people to spend time using those tools, whether it’s social media or now AI. They’re not AI yet because it doesn’t have an advertisement model. But if the model is to get kids on these tools as much as possible, you know, that’s what’s going to replace the critical thinking. And I think companies are thinking about it more than they were before and taking some steps to prevent harm against children, but not as much as they should be.

**[00:37:58] Mary McGovern Treyz:** What would you, when you talk to these trust departments, what are you suggesting for them or how are you working with them?

**[00:38:10] John Zoltner:** One thing I’m doing now, or I was doing towards the end of my tenure at Save the Children was working with Western Sydney University with a group called the Young and Resilient Research Centre led by Amanda Third there, to do a big study in Australia on children’s attitudes towards AI, positive and negative. And the protections they’d like to see built into AI models. So a big thing is consulting children, giving them a voice, but also it’s making sure that children in the low and middle income countries in the Global South are protected. So you can’t just protect people in your major markets where you’re making money and not in other countries because they’re not important markets to you. You have to be looking at the effects of your technology on everyone. Like I talked about before about the children in Bangladesh or Yemen who are looking at US students or children doing unboxing videos.

**[00:39:31] Mary McGovern Treyz:** What would you recommend that? Well, let me ask it this way. Let’s get this way. What would you recommend parents do to protect their children?

**[00:39:41] John Zoltner:** I think, you know, first of all, I’m not one of those people who believe in spying on your kids using spyware. So I think it’s more having a healthy discussion about their use of technology and getting even with my six year old or when she was five, maybe even younger. Or I did, worked with ChatGPT at that time, doing stories, you know, creating images connected to those stories using DALL-E, one of the OpenAI tools. And you know, talking with her about what she’d like to hear in a story, what should happen and generating that using the model. And I think, and also with my other daughter, with my older daughter planning a trip to California together and figuring out, we were speaking with also ChatGPT in this case, and asking questions and having it give responses. So I think that shows positive use of these models, but also people ask me, how can I protect my children from technology? What can I do? What can I do? And they’re thinking about technology fixes, I think, or they’re thinking about, you know, what software could I put on their phone?

**[00:41:13] Mary McGovern Treyz:** Right.

**[00:41:14] John Zoltner:** How many times a day should I be checking their text messages? And I’m not saying that shouldn’t happen at all, but I really think the best way to protect your child is to have a good, open relationship with them and talk to them about their lives and how they’re using technology, right? They need to trust you. And what happens is people, parents will, the most obvious response to any kind of harm that you suspect technology is doing to your kid is to take away that technology.

**[00:41:53] Mary McGovern Treyz:** Yeah. The banning, the lazy banning, right?

**[00:41:57] John Zoltner:** Right.

**[00:41:58] Mary McGovern Treyz:** Isn’t that?

**[00:41:59] John Zoltner:** Yeah. So what kid is going to tell you that? They’re going to do anything. That is happening.

**[00:42:04] Mary McGovern Treyz:** Yeah, for sure. And is it really that different from, you know, watching too much TV?

**[00:42:12] John Zoltner:** Right. Right. Yeah.

**[00:42:15] Mary McGovern Treyz:** I mean, when we use technology to be the babysitter, we can’t really complain about what happens when we’re not watching. Right. The responsibility does fall on the parents to be involved. And I think your point about trust and building the trust is tremendous. I think it’s easier said than done like everything, but building that trust and showing them how to use it to be curious. I mean, that’s always what I’m coming back to is it’s a great tool to be creative and curious, but it does take a lot of guidance.

**[00:42:58] John Zoltner:** Yeah, it’s difficult. And even with my own daughter, we just applied new rules to how long she could watch Netflix or PBS Kids. And yeah, I mean, that’s all now that’s old stuff, but that’s, we’re all used to those kinds of restrictions. And it’s been my contention for a long time that you don’t learn self-discipline without being taught.

**[00:43:31] Mary McGovern Treyz:** No. Right. As a child that no in so many ways teaches you the self-discipline, whether it’s no, we don’t drink soda in our house or no, we’re not going to a fast food restaurant.

**[00:43:45] John Zoltner:** Those no’s lead to a muscle or develop a muscle of self-control that if you don’t learn it when you’re young, it’s really hard to learn as you get older. I think.

**[00:44:00] Mary McGovern Treyz:** Yeah.

**[00:44:01] John Zoltner:** Especially I think with these new technologies and how addictive they are as well. I think, like I’ve thought about teaching in the future and I always think, am I going to be the enemy for, and I’ve thought about high schoolers, but for high schoolers and be the one that tells them, you know, you can’t use your phone in the class. Because I feel like, you know, if they’ve had years and years of formulative years with these, you know, with technologies, if you are the one to say no, though, you also kind of, I think become the enemy as well. And I fear that for parents too, they fear that too, their kid’s going to go to school. He’s going to, he or she’s going to go and see all their other friends, you know, and socialize with them and see them using technology and at home they can’t use it. I don’t know. It seems like our expectations for parents are so much higher now to kind of navigate these tough environments. I don’t know. John, do you think that, in terms of regulation, like what is the role of any type of government or government body on this? I mean, I know I saw Instagram now has something like, you have to be 16 in order to buy, to purchase apps and get your parents permission. I mean, what role do you see government playing in any of these regulations?

**[00:45:23] John Zoltner:** I think it has to set the responsibilities that technology companies have, basic responsibilities. So the self policing doesn’t always work or doesn’t work about every aspect of the technology that they create. So I think they should mandate some certain ways of protecting children, like a big area right now is age verification. How, because we know there are no negative effects on growing children of access to pornography, even legal pornography. So I don’t take a stance in general against pornography. But I do take a stance against open access by children to pornography. So there needs to be some way of making sure that children don’t have access to that content. And it’s hard right now. There are a lot of experiments going on, on how we can do age verification by, I mean, some systems map out the face and use that.

**[00:46:52] Mary McGovern Treyz:** Oh, that’s creepy.

**[00:46:54] John Zoltner:** Right.

**[00:46:55] Mary McGovern Treyz:** So creepy. Oh, but I mean, it’s already happening anyway with airports and what have you, you know, you probably really don’t even need a passport.

**[00:47:06] Mary McGovern Treyz:** Will they be able to see my guilty conscience?

**[00:47:09] John Zoltner:** Yeah.

**[00:47:10] Mary McGovern Treyz:** Oh, will it be able to map that out? Oh God, I can’t, I’m going to have to hole up in my closet.

**[00:47:16] Mary McGovern Treyz:** Oh, that’s it.

**[00:47:17] Mary McGovern Treyz:** With that baby face. Yeah.

**[00:47:19] Mary McGovern Treyz:** Oh, that’s it. You won’t get through this. I won’t pass any tests.

**[00:47:25] Mary McGovern Treyz:** John, I know we’ve gone past my usual time. Can I get you to tell us a time when you’ve ruined dinner with technology?

**[00:47:35] John Zoltner:** With technology or without it?

**[00:47:39] John Zoltner:** Yeah.

**[00:47:40] John Zoltner:** Well, like a lot of people I’ve had to control myself with using devices at the table.

**[00:47:47] Mary McGovern Treyz:** Yeah.

**[00:47:48] John Zoltner:** And we have ruined a few dinners or reduce the quality of family time during those dinners. But now we keep our devices away from ourselves. Well, it’s sad too, but one Thanksgiving dinner a few years ago, I’m usually very careful about what I say. And, you know, I’ve, I used to give speeches at Thanksgiving, motivational speeches for our friend group. But so I was at a small Thanksgiving dinner with friends a few years ago and we went around and did that thing where everyone talked about what they’re grateful for. And I still don’t know exactly what I said, but I offended everyone at the dinners. Somehow. And.

**[00:48:47] Mary McGovern Treyz:** Offending with gratitude. That’s quite a specialized trait.

**[00:48:52] John Zoltner:** I know. I know. And so it really affected our friendship and, you know, took a lot of kind of making up afterwards. But I was sort of trying to be funny and it really didn’t work. So I not only ruined dinner, I ruined Thanksgiving dinner.

**[00:49:12] Mary McGovern Treyz:** Oh my gosh. Well, I’m glad you were able to patch things back up. It sounds like.

**[00:49:18] John Zoltner:** Yeah. Yeah.

**[00:49:20] Mary McGovern Treyz:** John, thank you so much for coming on. It was kind of frightening actually talking to you, but I do think the takeaway is that we need to be vigilant.

**[00:49:34] John Zoltner:** Right. Right.

**[00:49:36] John Zoltner:** I agree. Again, I mean, one of my points that I constantly repeat is we can’t do with AI what we did with social media and we need to be thinking about it and we need to be making decisions and regulate regulators need to be more active from the beginning. So that we can’t pretend we don’t know what the unintended consequences will be.

**[00:50:06] Mary McGovern Treyz:** It’s kind of what you’re saying, right? Like we can look to the past and have a pretty good idea of what will corrupt our use of this technology.

**[00:50:20] John Zoltner:** Right. I think that’s true for the most part, but also we can’t pretend that there won’t be negative consequences that we can’t anticipate.

**[00:50:30] Mary McGovern Treyz:** Yeah.

**[00:50:31] John Zoltner:** So we have to be ready to deal with those. And as they’re emerging, we have to be taking action to stop them somehow. There was just at the Paris peace summit in February, there was an AI action summit, or they focused on AI during part of it and especially AI and youth. And I think it’s great that on the highest levels of, well, what are purported to be the highest levels of society. There’s people are considering how to protect children and potential negative effects on children and starting to contemplate what to do about it.

**[00:51:19] Mary McGovern Treyz:** Yeah.

**[00:51:20] Mary McGovern Treyz:** Well, pathologically positive, John Zoltner. Thank you so much. We really appreciate it.

**[00:51:27] John Zoltner:** Thanks for having me.

**[00:51:28] Mary McGovern Treyz:** Yeah. We’ll talk soon. I hope. Okay.

## How this transcript was made

This transcript was produced automatically in October 2026. Speech-to-text software converted the recording’s audio to text, and an AI language model then added punctuation, divided the text into turns, and proposed who was speaking from the event description and the conversation itself. It has not been checked line by line against the recording, so it may contain mistakes: misheard words, wrong names or numbers, and lines credited to the wrong speaker. The summary and key points were drafted with AI assistance from the transcript. John Zoltner assigned the speakers for the turns the software could not attribute. Square brackets mark an editor’s addition or an omission. The recording is the authoritative version.

To report a mistake, use the [contact page](https://zoltner.com/connect) on zoltner.com.

## More transcripts

- [What Parents Must Know About AI](https://zoltner.com/speaking/transcripts/mobicip-1-what-parents-must-know-about-ai.html): Video interview, February 2026; also covers Parents and families
- [When Kids Turn to AI for Help](https://zoltner.com/speaking/transcripts/mobicip-4-when-kids-turn-to-ai-for-help.html): Video interview, February 2026; also covers Parents and families
- [Online Grooming in the AI Age](https://zoltner.com/speaking/transcripts/mobicip-5-online-grooming-in-the-ai-age.html): Video interview, February 2026; also covers Regulation

John Zoltner is the founder and CEO of [AIChildSafety.org](https://aichildsafety.org). [About John Zoltner](https://zoltner.com/about). [All speaking transcripts](https://zoltner.com/speaking/transcripts/index.html).

