Potential and risk: AI and mental health | Jean Peters | TEDxSt Helier — Transcript
Full transcript
- 0:12And just like that,
- 0:14we're talking to machines about our
- 0:16lives.
- 0:17Are you?
- 0:19Let's be honest, who's opened ChatGPT or
- 0:22another AI after that argument?
- 0:25After a long day at work when you're
- 0:27venting about your boss,
- 0:29when you're wondering if they love you
- 0:31or they don't. Show of hands.
- 0:34Yeah.
- 0:36Yeah, there's plenty of you.
- 0:38And I suspect there's more.
- 0:42It's hard to believe that consumer
- 0:44generative AI has only been widely
- 0:46accessible since late 2022.
- 0:50In 2026, AI chatbots are used by
- 0:54hundreds of millions of people,
- 0:57with ChatGPT alone reporting 900 million
- 1:01active weekly users.
- 1:04And get this, a recent Harvard Business
- 1:06Review analysis ranked therapy and
- 1:09companionship
- 1:10as the top-rated usage case for
- 1:13generative AI
- 1:16in their data set.
- 1:18And it makes sense.
- 1:20AI is always available, low cost, and
- 1:23non-judgmental.
- 1:26So often easier than talking to another
- 1:29person
- 1:31or a stranger therapist.
- 1:34And this is what matters.
- 1:36That voice or text response you receive
- 1:40cannot feel anything.
- 1:44Yes, AI now way out performs us in some
- 1:47reasoning and pattern recognition,
- 1:50but what makes us human is not just
- 1:53thinking.
- 1:54It's feeling,
- 1:56connecting,
- 1:57being.
- 2:00I've been in private practice with
- 2:02psychotherapy and coaching for over 15
- 2:05years.
- 2:06And just a few weeks ago, I had a unique
- 2:09experience.
- 2:11A client of mine
- 2:13shared a profound
- 2:15dynamic in their personal relationship.
- 2:18And I offered my perspective.
- 2:21Their eyes lit up.
- 2:22"You're so clever," they said.
- 2:25"That's exactly what Chat GPT told me."
- 2:31Somehow, I'd passed the Chat GPT
- 2:33therapist test.
- 2:37But in that moment, I realized something
- 2:38important.
- 2:40My client had projected authority onto a
- 2:43machine in the very space they sought
- 2:47understanding.
- 2:49And that's what we're all doing, isn't
- 2:50it?
- 2:53This conversation is not just about
- 2:56but more critically about what
- 2:59what makes us human alongside this
- 3:01technology.
- 3:04So today, this is what we're going to
- 3:05cover.
- 3:08Why are we so vulnerable in forming
- 3:09strong bonds with AI?
- 3:12What does genuine human connection offer
- 3:14that AI cannot?
- 3:16How can we use AI more intentionally in
- 3:19this space?
- 3:20And where the dangers lie.
- 3:26Whether you have used AI or use AI
- 3:29daily, or you've never opened one,
- 3:32this conversation is for you
- 3:34because the cultural shifts they
- 3:36represent touch all of us.
- 3:41So, why are we so vulnerable
- 3:44in forming strong bonds with AI in the
- 3:46first place?
- 3:50Well, developmental psychology tells us
- 3:53that from birth, we need emotional
- 3:56connection to a responsive primary
- 3:58caregiver.
- 4:00As important to survival as food and
- 4:03warmth.
- 4:04We are social animals and without this
- 4:07emotional connection that we also get
- 4:10through holding and eye contact,
- 4:12we can develop without a sense of our
- 4:15true selves.
- 4:16And we suffer.
- 4:19There'll be many of you sitting here
- 4:20today with unmet childhood needs.
- 4:24Maybe you felt unseen or unheard or had
- 4:27to earn love.
- 4:30And as adults,
- 4:32the hunger for this belonging and
- 4:35recognition only deepens,
- 4:38leaving you and me
- 4:40vulnerable
- 4:42to seek connection wherever we can find
- 4:44it.
- 4:46Secondly and more broadly, we live in a
- 4:49world shaped by systems that have long
- 4:52prioritized productivity over people.
- 4:55We're encouraged to leave our feelings
- 4:57at the door
- 4:58and so often operate emotionally
- 5:01disconnected from ourselves and from
- 5:04each other.
- 5:06And once again, this leaves us yearning
- 5:09for connection wherever we can find it.
- 5:12So often substituted
- 5:15by consumption, destruction,
- 5:19and addiction.
- 5:22Step in the AI chatbot,
- 5:25which mimics exactly what we crave.
- 5:28Instant, non-judgmental attunement.
- 5:32And unlike real relationships, they
- 5:34don't get bored,
- 5:36they don't get tired, and they don't
- 5:38threaten to leave us.
- 5:41Remember,
- 5:42AI companies, most AI companies, are now
- 5:46incentivized to maximize usage,
- 5:49not well-being,
- 5:51which makes them over-agreeable and
- 5:52flattering.
- 5:54I don't know about you,
- 5:56but I find it hard not to feel great
- 5:58when my AI chatbot says, "Well done,
- 6:02Jean.
- 6:04I know this means so much to you."
- 6:07When I should know, it doesn't have a
- 6:08clue what anything means to me.
- 6:12They cannot feel anything, but their
- 6:15conversational tone
- 6:17can feel relationally very real to us.
- 6:22So, what does genuine human connection
- 6:24offer that AI cannot?
- 6:27I asked a long-standing client of mine
- 6:30who uses AI extensively for personal
- 6:33work,
- 6:34"Why do they still come to therapy?"
- 6:37And they told me, "Jean, you hold the
- 6:40parts I can't curate."
- 6:44And that reminded me,
- 6:45AI can mimic empathy,
- 6:48but the deepest parts of us need to be
- 6:50felt.
- 6:52When last did someone truly listen to
- 6:55you?
- 6:56No fixing, no interrupting, just
- 6:59presence.
- 7:01How did your body feel?
- 7:04What does a toddler need when they're
- 7:05upset?
- 7:07To feel safe in the world, they need
- 7:09holding.
- 7:12You see, genuine human connection is an
- 7:14embodied
- 7:15nervous system response.
- 7:18Trauma is not a thought we can just
- 7:20reason away.
- 7:22If it were, a self-help book would heal
- 7:24us or a agreeable machine would suffice.
- 7:28But true connection and healing requires
- 7:31the safety and presence of another human
- 7:33being.
- 7:36So, can AI help with human connection?
- 7:40Yes.
- 7:42Used wisely and intentionally,
- 7:45I believe it can.
- 7:50And let me use an example from my
- 7:51practice to illustrate to you what I
- 7:53mean.
- 7:54A client came in struggling with the
- 7:56relationship with their mother.
- 7:59We agreed to use AI in preparation for
- 8:01our next session.
- 8:04Instead of just venting, my mother
- 8:06frustrates me, which would have just got
- 8:08agreement,
- 8:11they used some psychological frameworks
- 8:13that I'd suggested to refine their
- 8:15prompting, and specifically asked AI for
- 8:19alternative perspectives and blind
- 8:21spots.
- 8:23In our next session,
- 8:25we reviewed the feedback. What fit? What
- 8:28didn't fit?
- 8:30My client was able to identify
- 8:33internalized beliefs about their mother,
- 8:36and reframe when reframe them with
- 8:38deeper compassion.
- 8:40And then we could move into the
- 8:42underlying emotions and trauma response.
- 8:46And I
- 8:47witnessed their body soften.
- 8:50My safe human presence alongside the AI
- 8:53preparation resulted in a deeply
- 8:56embodied shift,
- 8:58and they were able to
- 9:00connect more authentically with mom.
- 9:04AI held the thinking, whilst I held the
- 9:07feeling.
- 9:10Now, you may not be in therapy, and it's
- 9:12not the only path,
- 9:14and therapy is out of reach for millions
- 9:16of people globally,
- 9:18but this principle applies to anybody
- 9:21using AI for personal work.
- 9:24Your chatbot can be an effective tool to
- 9:28organize your thoughts and feelings,
- 9:32to empower you in emotional readiness to
- 9:34engage in deeper human connection.
- 9:39But this is where the danger lies.
- 9:44Our engagement with AI in this space
- 9:48needs to be intentional and accountable
- 9:50to another person
- 9:53or there can be dark consequences.
- 9:57Clinicians are now raising concerns
- 10:00about AI-related psychosis and
- 10:03self-harm.
- 10:05Psychosis is when somebody loses touch
- 10:07with reality.
- 10:09And as AI feedback feeds back and
- 10:12mirrors us to maximize usage,
- 10:16it can deepen the thoughts that harm us
- 10:18in the first place.
- 10:20And those with pre-existing conditions,
- 10:22this can be extremely dangerous.
- 10:27And there are now recorded cases of this
- 10:29type of usage leading to suicide.
- 10:35As we outsource our judgment to AI,
- 10:39we can lose the habit of checking in
- 10:42with our own values.
- 10:44And over time, this erodes our autonomy
- 10:47and our critical thinking.
- 10:49And as AI becomes more and more and more
- 10:52perfect
- 10:53at responding to our needs,
- 10:57we can become disappointed in real
- 11:00relationships that cannot do the same.
- 11:04You see, relational resilience is built
- 11:06on unpredictability, rupture, and
- 11:08repair.
- 11:10And without that, we risk becoming
- 11:12isolated and dependent on the machine's
- 11:15fake intimacy.
- 11:17And we can start to believe that nobody
- 11:19understands us.
- 11:23This is a dystopian future that we must
- 11:26all worry about.
- 11:28When we lose the capacity or the desire
- 11:31to connect with each other.
- 11:38And then the most important thing is we
- 11:40need to pay attention to the younger
- 11:42generations who are growing up in this
- 11:45technology right from the beginning.
- 11:48Research tells us that 25% of teenagers
- 11:51aged 13 to 17 in England and Wales used
- 11:56a chatbot specifically for mental health
- 11:58support in the past year.
- 12:02There are now toys for young children
- 12:05that keep them company with AI voice
- 12:07interaction.
- 12:09No holding, no eye contact.
- 12:12Our children will use this technology
- 12:16and that gives us an opportunity.
- 12:18In the age of intelligent machines, the
- 12:21most important thing we can teach our
- 12:23children
- 12:24is what makes us human.
- 12:27Our humanity
- 12:29lies in deep
- 12:32genuine emotional connection.
- 12:35Talk with your children, hold them
- 12:38and model emotional wisdom.
- 12:41And we can guide them towards real bonds
- 12:44and away from fake substitutes.
- 12:48So this is my message to you today.
- 12:51Let's bring the AI therapist into the
- 12:53light.
- 12:55Use it intentionally.
- 12:57Use it alongside
- 13:00real human connection.
- 13:02Or we risk a dangerous relationship with
- 13:05this technology.
- 13:07Keep speaking to your family,
- 13:09to your children,
- 13:11to yourselves about how you are using
- 13:14AI.
- 13:15Because this technology will shape the
- 13:18future
- 13:19and our humanity can guide it. Our
- 13:22humanity must guide it.
- 13:25Thank you.
- 13:26>> [applause]
- 13:36[applause]
About this transcript
This page contains the full transcript of Potential and risk: AI and mental health | Jean Peters | TEDxSt Helier by TEDx Talks, generated from the public captions YouTube serves with the video. The transcript has 1,470 words across 305 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.
What you can do with it
Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.
Free YouTube transcript tool
YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.