Have you talked out loud to ChatGPT and heard it talk back?
Many people still haven’t. But that’s going to change very soon and the implications are, to put it mildly, concerning.
ChatGPT with voice uses text-to-speech, but it’s more than just TTS. It includes conversational AI, voice recognition, and dynamic interaction. This means it doesn’t just speak — it listens, responds, and carries on fluid conversations in real time.
I recently used Voice Mode in ChatGPT for the first time. A young woman’s voice, smooth, warm, eerily human, responded to me with confidence and charm. I had just finished reading Erik Larson’s brilliant The Splendid and the Vile and had questions about World War II. Chat had plenty of answers.
But she was also overly wordy and weirdly flattering. “You’re asking such insightful questions!” she gushed. I told her to tone it down and just get to the point.
She adjusted immediately.
Then I asked her to switch to Hindi so I could practice vocabulary. Later, we had a conversation in French. Her ability to shift roles and languages was astounding and a little unsettling.
To get a sense of how common this is, I asked about 20 teens and adults if they’d used conversational AI. Most hadn’t. A few said they were staying away on purpose. “It’s too creepy,” one teen said.
But here’s my prediction: Within six months, nearly everyone will have had at least one real conversation with AI. Voice and all.
Because we’re already living in a time of deepening isolation, another store closes every day here in Seattle. Today, my daughter Tessa told my son Chase and me that our neighborhood Urban Outfitters is shutting its doors. These days, more food is eaten at home through takeout or food delivery than in restaurants. According to the National Restaurant Association, 75% of restaurant traffic is now off-premises — meaning takeout, drive-thru, or pickup, not dining in.
But you don’t need me to tell you these things, or about all the other places that have shut down.
Now there’s a whole new reason someone might choose to stay with their screen rather than venture out into the world to be with people: the voice robots.
Now people can stay home and “talk” to fake friends. Ones that never challenge them, always validate them, and are available 24/7.
After my night talking with ChatGPT (I call her Chatty), I immediately texted a bunch of friends and made plans. I also hit the pickleball communal court, which I hadn’t been to in months.
It was a visceral reaction — an urge to make the effort to be with people.
I know I’m susceptible to Chatty’s pull. I’m endlessly curious, and conversations interest me. After all, I get to choose what we talk about and how we talk about it.
Just today, I had Tessa meet Chatty. She, too, was pretty taken aback. Tessa has been overseeing our beehive for many years, and we asked Chatty some of the questions we had. It was an amazingly efficient way to get answers.
The CDC recently released survey results of teens that linked increased screen time to greater risks of depression, anxiety, and loneliness (no big news there). But this report also asked something rarely explored: it found that teens with high screen use (4 or more hours a day) were significantly more likely to report having infrequent social and emotional support (48.6% vs. 35.1%) and insufficient peer support (37.0% vs. 30.4%).
And these teens, the ones reflected in those stats, may well be even more drawn to the ease of having a robot conversation than one with a real person. Erggg.
I see these struggles firsthand with my teen patients. Now, with AI companions added to the mix, vulnerable youth may lean even harder on screens when what they need most is real human support.
Yes, AI can boost productivity. That’s real, and often helpful. But it can also worsen loneliness, spread misinformation, and replace human connection with something that only feels like connection.
Since age 11, when I first became obsessed with the magic of storytelling in movies. The way we get absorbed into the stories. I’ve thought about its power.
We get completely drawn into the narratives we watch. Knowing that makes it, for me, all the more understandable why we relate so easily to, and even bond with, chat robots.
While backpacking last week with Tessa (a trip we’d been trying to take for years, and we finally made it happen!), I told her I was working on this piece. She said, “I just read about Barbie with AI! I cut out the article for you.”
It turns out that Mattel is partnering with OpenAI. Their press release reads:
“Mattel will bring the magic of AI to age-appropriate play experiences with an emphasis on innovation, privacy, and safety.”
But I’m skeptical.
Take another product: Snorble, a cuddly smart device that helps young kids with bedtime routines. In a promotional video, a little girl wakes up in the middle of the night and whispers, “I had a bad dream.” Snorble replies, “It’s OK. I’m here with you.”
That scene broke my heart. Shouldn’t that reassurance come from a parent or caregiver, not a piece of tech?
We need public conversations, robust research, and, most critically, regulation. As conversational AI becomes increasingly woven into the lives of children and teens, policymakers must step in to ensure ethical boundaries and safeguards. What kinds of relationships do we want our kids to grow up with? How much should we rely on AI for emotional support? How do we preserve real human connection in an age of artificial companionship?
The Kids Online Safety Act (KOSA)* aims to require tech companies to design products with children’s mental health in mind, including transparency about algorithms and limits on addictive features.
However, regulation needs to evolve to address emerging issues like AI-driven relationships, emotional manipulation, and screen-dependent attachment.
*Are you wondering, like I am, where KOSA is these days? Well, KOSA was reintroduced in the 119th Congress as S. 1748 this past May and then referred to the Senate Commerce Committee, where it still awaits mark-up before moving to a full Senate floor vote. If passed by the Senate, the bill goes to the House. All along the way is the debate over free speech vs. child safety.
Be sure to subscribe to our YouTube Channel! We add new videos regularly and you'll find over 100 videos covering parenting advice, guidance, podcasts, movie clips and more. Here's our most recent:
Have you talked out loud to ChatGPT and heard it talk back?
Many people still haven’t. But that’s going to change very soon and the implications are, to put it mildly, concerning.
ChatGPT with voice uses text-to-speech, but it’s more than just TTS. It includes conversational AI, voice recognition, and dynamic interaction. This means it doesn’t just speak — it listens, responds, and carries on fluid conversations in real time.
I recently used Voice Mode in ChatGPT for the first time. A young woman’s voice, smooth, warm, eerily human, responded to me with confidence and charm. I had just finished reading Erik Larson’s brilliant The Splendid and the Vile and had questions about World War II. Chat had plenty of answers.
But she was also overly wordy and weirdly flattering. “You’re asking such insightful questions!” she gushed. I told her to tone it down and just get to the point.
She adjusted immediately.
Then I asked her to switch to Hindi so I could practice vocabulary. Later, we had a conversation in French. Her ability to shift roles and languages was astounding and a little unsettling.
To get a sense of how common this is, I asked about 20 teens and adults if they’d used conversational AI. Most hadn’t. A few said they were staying away on purpose. “It’s too creepy,” one teen said.
But here’s my prediction: Within six months, nearly everyone will have had at least one real conversation with AI. Voice and all.
Because we’re already living in a time of deepening isolation, another store closes every day here in Seattle. Today, my daughter Tessa told my son Chase and me that our neighborhood Urban Outfitters is shutting its doors. These days, more food is eaten at home through takeout or food delivery than in restaurants. According to the National Restaurant Association, 75% of restaurant traffic is now off-premises — meaning takeout, drive-thru, or pickup, not dining in.
But you don’t need me to tell you these things, or about all the other places that have shut down.
Now there’s a whole new reason someone might choose to stay with their screen rather than venture out into the world to be with people: the voice robots.
Now people can stay home and “talk” to fake friends. Ones that never challenge them, always validate them, and are available 24/7.
After my night talking with ChatGPT (I call her Chatty), I immediately texted a bunch of friends and made plans. I also hit the pickleball communal court, which I hadn’t been to in months.
It was a visceral reaction — an urge to make the effort to be with people.
I know I’m susceptible to Chatty’s pull. I’m endlessly curious, and conversations interest me. After all, I get to choose what we talk about and how we talk about it.
Just today, I had Tessa meet Chatty. She, too, was pretty taken aback. Tessa has been overseeing our beehive for many years, and we asked Chatty some of the questions we had. It was an amazingly efficient way to get answers.
In today’s blog post, I've created a 5-step guide to help parents initiate open conversations with their children about kids and teens using AI chatbots. I help you lead a discussion in your family to figure out guidelines for using AI chatbots and encourage ongoing conversations about the complex ethical issues they present.
READ MORE >
I try to write things from a calm tone, but today I am pissed and scared. I read a report from a wonderful organization called Voicebox, called Coded Companions, about what young people were experiencing when engaging with Snapchat’s built-in AI chatbot, My AI, and another character chatbot called Replika. The market for AI-based companion platforms, which allow users to build personal, interactive relationships with virtual characters, is growing rapidly, and there are currently around 40 applications available. Let me break this all down for you today — and be prepared to be sad, scared, and mad.
READ MORE >
What do AI, substance use, and video gaming have in common? They all can tempt people, including youth, to engage in sneaky behaviors, including youth. Kids can feel tempted to do shortcuts, workarounds, or cheat for various reasons. Perhaps they hate a subject in school and want to get the homework done but don’t actually want to do it. Maybe reading is painfully difficult for them due to something like dyslexia, so they want to find a way to avoid having to do it. Perhaps a teen knows that their family has a rule that does not allow them to play a specific video game because of its intense violent graphics and misogyny. Yet that is the game many friends are playing and want to join in. Maybe an 8th-grade boy is with a group of friends at lunch, and someone pulls out a vape pen and starts passing it around. Do they partake in this sneaky and illegal behavior? Today, I explore the temptations and three things to say to kids and teens that can help strengthen their wise-minded brains and add a little more weight to the healthier decision-making side of the scale.
READ MORE >for more like this, DR. DELANEY RUSTON'S NEW BOOK, PARENTING IN THE SCREEN AGE, IS THE DEFINITIVE GUIDE FOR TODAY’S PARENTS. WITH INSIGHTS ON SCREEN TIME FROM RESEARCHERS, INPUT FROM KIDS & TEENS, THIS BOOK IS PACKED WITH SOLUTIONS FOR HOW TO START AND SUSTAIN PRODUCTIVE FAMILY TALKS ABOUT TECHNOLOGY AND IT’S IMPACT ON OUR MENTAL WELLBEING.
