Larry I’m Larry Magid, and this is Doing Tech Right, a podcast from ConnectSafely where we and our guests speak about parenting, tech, and related topics, along with strategies and policies to make tech safer and more civil. And if you like what you hear, please subscribe and visit us at connectsafely.org.
Hey, it’s Larry Magid from ConnectSafely, and I’m with Kerry Gallagher, our Director of Education, who flew all the way out to California from Massachusetts, and she’s with me in my home studio. This is the first time I’ve had somebody in the studio since before the pandemic. I used to have people here all the time, but for the last few years, it’s all been by Zoom. So thanks for reinitiating my studio and making me set up that second mic. It’s great to have you here in person.
Kerry In some ways, the pandemic feels like a million years ago, and in other ways, it feels like just yesterday, both in terms of remembering what happened, but also how much technology has advanced and how many new and different skills we need to learn together.
Larry In some ways, you’re right. It seems like yesterday. In some ways, it seems like a million years ago. It was before AI, and we know the word before C, we know what that means, BC, but before A is before artificial intelligence became commonplace. AI has been around since the 50s, but not in everybody’s homes and in everybody’s minds. But now it’s omnipresent in our lives.
Kerry It is. We can access it while we’re in our car, and we can access it on our devices. Our children are accessing it as part of their toys. It really is everywhere.
Larry If you go to connectsafely.org, you’ll see an article I wrote about what it’s like talking to AI, usually in my car, sometimes on walks, versus typing on AI. When I type on AI, I always know I’m talking to a machine. I’m not an idiot. I think everybody listening is fully aware that there are not human beings behind AI that are sitting there talking. It’s a machine talking.
But when I’m actually having a conversation with it, it feels like I’m talking to a person. I know it’s not, but it feels like a person. And I say please, I say thank you, I argue with it. And I sometimes think, am I crazy? Why am I talking to a machine this way? But it’s so seductive.
Kerry Yeah, I mean, is it seductive, or is it that it was programmed to respond to human communication patterns and it’s doing exactly what it was programmed to do?
Larry Oh no, it’s doing exactly what it’s programmed to do. It’s also doing what its masters, well, I hope they’re its masters, I hope it’s not their master, but what its creators want it to do, one of the two, which is to get you engaged, make you comfortable, make you feel like it’s empathetic and caring, even though we all know that it’s not. It has no empathy. It doesn’t even have knowledge. It has pattern recognition. It has the ability to figure out patterns of words.
You think it has knowledge. You ask it a question, it gives you an answer, and it might be right, it might be wrong, but it’s not because it has some database of information, other than the entire internet.
Kerry Quite the database.
Larry Quite the database.
But anyway, I think about how if I, as a grown person, hopefully a mature one, can get lulled into the sense that I’m talking to a person, again, my brain understands it, but my feelings go a different way, what does that do to a 10-year-old, an 8-year-old, a 13-year-old? What impact does that have on their brains?
Kerry Yeah, I think that’s what’s both exciting and concerning about AI, right? It can allow us to learn things and interact in a way that feels very natural to us. So we are able to gain information or ideas in a way that’s very human, while the thing we are interacting with has the benefit of giving us information from the entire internet, right?
Larry And you can trust everything on the internet, right?
Kerry Absolutely.
Right, and you’re transitioning for me. But also, the benefit of the entire internet could be used to allow folks who have potentially bad intentions, or greedy intentions, to use that same power to engage us in the way that you just described in a way that could be risky or harmful for us.
Larry Or sloppy intentions. No intentions at all. People make mistakes.
It’s one thing for me to make a mistake. I make mistakes. I’m a human being. But I have to own my mistakes. And if I’m going to write something, whether it’s an article or an email to my friend or a text message, if I say something, it’s on me to be accurate. It doesn’t matter whether I got the information from somebody else. I have to vet the accuracy.
And that’s why it is so important, and I’ve said this a million times, never to act on or share what you learn through AI without going to a primary source to make sure it’s true.
Kerry Yeah, we talk a lot. And our listeners may know, and I know, Larry, you know that I’m an educator too by profession. We talk a lot about researching and verifying information.
Our students learn to take information they get from one source, verify it from another source, and when they’re writing academic essays or producing a video or presentation, they’re sharing information from various sources that corroborate each other.
We need to practice that when we’re using AI. We need to verify or corroborate information from alternative sources to make sure, before we act on anything, that we have the highest quality assurance.
Larry Yeah. Sometimes what I’ll do, if I’m in a meeting and everybody agrees to allow me to record it, I might take a transcript and have AI summarize it and clean up typos and things. But even then, I still vet it, because it can make mistakes even if you feed it content. It can elaborate and make assumptions despite the fact it has a transcript it’s working from.
Kerry And I think that’s really what teachers are struggling with right now, and parents too, when they see their children using AI to help them do homework or accomplish certain tasks.
Do children who don’t have the benefit of experience and years of developing their skills trust themselves enough to know they need to vet something? Or does the output exceed their own skill level, so it’s tempting to just go with the AI output because they think, “That’s better than I could do”?
And therefore they’re not experiencing the same struggle you and I experienced in school, and potentially not learning in the same ways. Will their skills as adults be of lesser quality as a result? It’s something parents and teachers are asking themselves right now.
Larry Well, you may remember, during our Safer Internet Day event, we had a panel you moderated, and there was a young woman on that panel, a high school junior and member of our Youth Advisory Council, who said she doesn’t use AI because, at this stage of her development, she’s afraid of not developing these skills.
I’m not sure I agree with her, but I respect her decision. But I would like to think she could use it and still develop skills so she could function with or without it.
Kerry Yeah, it’s a valid concern.
As a professional educator and a mom of teens, though, that’s not what I practice. I’ve taught my children how to use AI as part of their homework, but we’ve talked about ways of using AI that benefit them and ways that would put their learning at risk long term.
And that’s led to some really great conversations.
Oftentimes now, they use AI to help create practice tests to study for a real test. My daughter just took the SATs and asked AI to help create some practice sections.
I have another daughter who had to read a super challenging book for English. She used AI for chapter summaries. She didn’t use that instead of reading. She used the summary before reading, so she knew what to expect and it helped her interpret challenging literature. It gave her better reading comprehension.
So it isn’t about circumventing learning. It’s about adding layers of support.
Larry Yeah. And speaking about AI, last night, when we’re recording this, in mid-April of 2026, we had a really interesting event in Lafayette, California. Assemblymember Rebecca Bauer-Kahan held a meeting for women and girls. I was one of two men in the room, actually the only man in the audience.
Members of the audience asked some really good questions.
One of them was: How are AI deepfakes specifically impacting young women online?
Kerry One area of great concern for both young men and young women, but especially young women, is the use of AI deepfake tools to nudify images.
And yes, “nudification” is a real word.
It describes using AI tools to take an image in which a person is clothed and make them appear unclothed.
Unfortunately, people use those tools without consent. And women, young women, and girls can become victims.
Larry And this comes up among your students too.
Kerry It does. We’ve had examples of boys being victims, and unfortunately students making these images of peers.
It’s important for all of us who care for adolescents to build awareness so they know what to do if it happens.
Larry What should a young person do?
Kerry The first answer from a girl in our audience was exactly right: tell a trusted adult.
That could be a teacher, coach, parent, neighbor, best friend’s parent, older sibling, cousin.
It’s important kids have trusted adults in their lives.
Larry And in addition?
Kerry If it becomes sextortion, slow down. Breathe.
Know you have done nothing wrong.
You are not in trouble.
That’s the first step.
Then tell a trusted adult and get support.
And adults need enough awareness to know what to do next.
Larry And your life is not over. Even if the worst happens, you’ll get past it.
Kerry Exactly.
And schools need policies ready to go.
Response should be victim-centered. Support the victim first. Help tell parents if needed. Provide counseling support. Help report and remove images.
I recommend the National Center for Missing & Exploited Children, especially the CyberTipline and entity[“organization”, “Take It Down”, “NCMEC initiative”].
Larry And what about the child who created the image?
Kerry There should be accountability, yes.
But it also matters why they did it.
If we don’t understand motivations, we miss an opportunity to prevent it from happening again.
Consequences and support can coexist.
Larry Another question was how schools and parents can prepare young people to deal with deepfakes.
I think you have to have these conversations.
Kerry Yes. And they don’t have to be long conversations.
The car, the couch, a walk, those are often the best places.
Use the actual words. Even if your child rolls their eyes.
It’s only weird if you make it weird.
A quick check-in after a news story can be enough.
Larry Any last thoughts?
Kerry I’d say we should never make parenting decisions based on fear.
Make them based on accurate information.
If fear kicks in, acknowledge it, then flip to curiosity.
Why am I scared?
What more can I learn?
Go to connectsafely.org. Look to trusted organizations.
Make informed decisions.
Larry Talk to people in your life.
Recognize people are resilient.
I raised two digital natives. They turned out great.
Did they make mistakes? Yes.
Did I? Sure.
But people are more resilient than we give them credit for.
Kerry And don’t beat yourself up when you make a mistake.
Every child is different.
Know your child.
Be curious with them.
Help them reach their goals in healthy ways.
We can’t protect our kids from every mistake.
They need to make some in order to learn.
Larry Kerry, it was great seeing you in person. We should do this again.
Kerry Thank you again for hosting.
Larry Take care.