Latest / Play Therapy Podcast: A Master-Class in Child-Centered Play Therapy / 440 | Do AI-Generated Notes Have a Place in Child-Centered Play Therapy?

Transcript

440 | Do AI-Generated Notes Have a Place in Child-Centered Play Therapy? · Play Therapy Podcast: A Master-Class in Child-Centered Play Therapy

• 0:00
  You're listening to the Play Therapy Podcast with Dr.
• 0:03
  Brenna Hicks,
• 0:04
  your source for centered and focused play therapy coaching.
• 0:08
  Hi,
• 0:08
  I'm Dr.
• 0:09
  Brenna Hicks,
• 0:09
  The Kid Counselor.
• 0:10
  This is the Play Therapy Podcast where you get
• 0:12
  a master class in child-centered play therapy
• 0:15
  and practical support and application for your
• 0:17
  work with children and their families.
• 0:19
  In today's episode,
• 0:20
  I'm answering a question from Leanna in Virginia.
• 0:24
  And this is about using AI
• 0:26
  in the playroom.
• 0:28
  So,
• 0:29
  this is gonna be an interesting question to tackle together.
• 0:33
  Very excited,
• 0:34
  Lena.
• 0:34
  Thank you for
• 0:35
  sending it in.
• 0:36
  Before I dive in,
• 0:37
  I just wanna give you all a couple of reminders.
• 0:38
  I know I keep preaching to the choir here,
• 0:40
  but
• 0:41
  we only have a few weeks until the New Jersey
• 0:44
  event.
• 0:44
  So,
• 0:45
  if you are interested in the VIP event on Friday,
• 0:48
  if you are interested in the conference on Saturday,
• 0:50
  if you are interested in the podcast meetup on Sunday.
• 0:54
  There's lots of stuff going on October 16th,
• 0:56
  17th,
• 0:56
  17th,
• 0:56
  18th.
• 0:58
  So,
• 0:58
  iamccpt.live is where you can find info about it.
• 1:01
  playtherapynow.com is also kind of a one-stop shop for everything.
• 1:05
  Make sure you are signed up for the newsletter so that you're in
• 1:09
  communication with me.
• 1:10
  And you're running out of time if you want to register for the VIP,
• 1:13
  the conference,
• 1:13
  or the VIP,
• 1:13
  the conference,
• 1:14
  or the meetup.
• 1:14
  So,
• 1:15
  please make sure
• 1:16
  that you are
• 1:18
  aware of that.
• 1:19
  For those of you that are not able to attend on
• 1:22
  Saturday,
• 1:23
  specifically because of Sabbath observance,
• 1:25
  I'm specifically doing
• 1:27
  a meetup on Sunday
• 1:29
  so that I'm still able to hang out with you when I'm up in Jersey.
• 1:32
  So,
• 1:33
  if you are unable to attend the Friday or Saturday because of Sabbath and or
• 1:37
  you're just not able to make it work.
• 1:40
  I will be available on Sunday for a meetup.
• 1:42
  So please make sure that you RSVP for that if you're interested.
• 1:45
  I'm really looking forward to hanging out with y'all.
• 1:47
  So,
• 1:47
  let me read parts of this question and then we'll dive in together.
• 1:51
  I'm actually really thrilled to tackle this.
• 1:53
  So,
• 1:54
  the question is,
• 1:55
  can you speak to the idea of using AI
• 1:57
  for notes in the playroom?
• 1:59
  This is being pushed hard at my employer and
• 2:01
  I've not implemented it because of ethical concerns.
• 2:04
  I understand that if the parent consents to the
• 2:06
  use of an AI recording or note generator,
• 2:08
  it is technically considered legal.
• 2:10
  However,
• 2:11
  I have ethical concerns about the use of AI generally,
• 2:14
  particularly for children who,
• 2:15
  in my view,
• 2:16
  have no meaningful ability to even offer assent for the use of AI
• 2:20
  the way they might offer assent for dumb device video recordings of sessions,
• 2:24
  giving children's developing cognitive abilities
• 2:27
  and the ever-changing nature of what
• 2:29
  AI can do and what guard rails are in place.
• 2:32
  In theory,
• 2:33
  The value of the AI use is only for me as a clinician,
• 2:37
  and using it with children has only possible harms for them.
• 2:41
  Given this,
• 2:41
  it seems to me that using AI for session recording and note writing
• 2:45
  would amount to taking advantage of my client's
• 2:47
  limited cognitive ability for my own gain,
• 2:50
  and that violates my professional and personal ethics.
• 2:53
  I always appreciate how you gently invite question askers to examine our own stuff
• 2:58
  and how it might be affecting our practice.
• 3:00
  Is there anything I'm missing here in my thinking?
• 3:02
  Do you have a framework or suggestions
• 3:05
  on how to think about the potential use of AI tools in the playroom?
• 3:08
  And then there's a little asterisk here
• 3:11
  with the possible harms that were referenced before
• 3:14
  when it said using it with children has only possible harms for them.
• 3:18
  Here are some that
• 3:19
  were cited.
• 3:20
  Clients' data has not been handled as promised if
• 3:22
  the system goes rogue as has recently happened.
• 3:25
  Their data being hacked and info about them and
• 3:27
  their families or family members being sold to advertisers,
• 3:31
  their voice recordings being used or sold on the dark
• 3:34
  web for all other sorts of bad and worse reasons,
• 3:37
  and the general lack of accountability for tech companies and big business
• 3:41
  and the gutting of various regulatory bodies
• 3:43
  should any of the previously listed things happen.
• 3:47
  OK.
• 3:48
  So,
• 3:49
  let me begin with,
• 3:50
  y'all know my anti-technology bent and my anti
• 3:54
  anti-AI bent.
• 3:55
  So,
• 3:56
  let's be real clear.
• 3:59
  I'm not necessarily saying that I'm unbiased in this
• 4:02
  discussion here.
• 4:03
  However,
• 4:04
  I'm gonna try to be really neutral and factual in my response because I
• 4:07
  think that neutrality is important when we're talking about a question like this.
• 4:13
  I absolutely agree with you,
• 4:14
  Leanna,
• 4:15
  all of your assessments.
• 4:17
  A child
• 4:19
  gets no benefit from it.
• 4:22
  We are using it for our benefit.
• 4:26
  Even though we don't need to.
• 4:29
  We can
• 4:30
  write in session notes.
• 4:32
  We can write after session notes.
• 4:34
  We can do our documentation
• 4:36
  in other ways that do not
• 4:38
  use AI.
• 4:41
  And yes,
• 4:41
  technically it is considered legal if parents give consent.
• 4:45
  But a child has no idea,
• 4:47
  even if we were to present
• 4:50
  to a child in the playroom.
• 4:53
  You can choose to
• 4:55
  let
• 4:55
  AI,
• 4:56
  do they know what that means?
• 4:58
  You can choose to let this computer system,
• 5:00
  do they know what that means?
• 5:01
  You can choose to let this computer system,
• 5:01
  do they know what that means?
• 5:02
  You can choose to let this video recording or audio recording device,
• 5:05
  do they know what that means?
• 5:06
  Probably
• 5:07
  no to all.
• 5:09
  Do you choose to let me use this
• 5:11
  so that I have a recording of what happened in the session
• 5:15
  for my notes,
• 5:16
  or do you choose for me not to use it?
• 5:20
  Because of the cognitive capacity of children
• 5:23
  and the lack of intellectual
• 5:25
  and abstract reasoning
• 5:27
  and logical
• 5:29
  capacities.
• 5:31
  There's no way that they're able to make
• 5:34
  an informed decision
• 5:36
  one way or the other.
• 5:37
  And even if that was all we were considering,
• 5:40
  can the child actually give consent to this?
• 5:43
  From a cognitive development standpoint.
• 5:47
  A child certainly has no idea.
• 5:50
  What it means.
• 5:52
  In other words,
• 5:54
  We allow AI to record.
• 5:57
  Where does that data go?
• 6:00
  Where is it stored?
• 6:01
  When is it stored?
• 6:02
  When is it deleted?
• 6:05
  How accessible is it?
• 6:07
  What's the security protocol?
• 6:10
  And nothing is unhackable,
• 6:11
  y'all.
• 6:12
  Nothing.
• 6:14
  So I don't care if they say it's HIPAA compliant.
• 6:15
  I don't care if they say it's HIPAA compliant.
• 6:16
  I don't care if they say that it's protected.
• 6:18
  I don't care if they say that it's protected.
• 6:18
  I don't care if they say that no one's going to
• 6:21
  access it and they have all this stuff in place.
• 6:23
  It's not.
• 6:25
  And
• 6:26
  I absolutely agree with you,
• 6:28
  there are more and more and more studies coming out.
• 6:32
  About how people have consented.
• 6:35
  To AI use of recording
• 6:37
  and it has been used in ways that they did not agree to.
• 6:41
  They are training AI to do therapy
• 6:45
  through these
• 6:46
  apps
• 6:46
  and these programs.
• 6:48
  They're training AI
• 6:50
  to
• 6:51
  interact in a more human way
• 6:54
  through these apps.
• 6:55
  They're training AI how to have conversations
• 6:58
  through these apps.
• 7:00
  And I agree,
• 7:01
  we have no idea what happens
• 7:03
  once it's recorded
• 7:05
  because they can say they have all these systems and protocols and whatever,
• 7:08
  we will never know.
• 7:10
  And I have a very
• 7:11
  technologically savvy
• 7:14
  computer guy husband
• 7:17
  who actually kind of has like a hacker brain.
• 7:20
  And the first time I brought up to him that one of
• 7:23
  the therapists that I was working with said that they were using
• 7:26
  AI for notes,
• 7:28
  he about flipped his lid.
• 7:31
  Like
• 7:32
  was
• 7:33
  You gotta be kidding me.
• 7:34
  That is so unethical and that is so inappropriate.
• 7:36
  I mean,
• 7:37
  and that is not secure,
• 7:38
  and you have no idea where that's gonna go.
• 7:40
  I mean,
• 7:40
  he had his hacker brain saw all of the possibilities
• 7:45
  of how badly that could go.
• 7:49
  And at the minimum,
• 7:50
  it's a breach of confidentiality.
• 7:55
  Any client that comes in our room,
• 7:57
  no matter the age,
• 7:59
  whether they're 3 or they're 93.
• 8:02
  We are
• 8:03
  committed and obligated
• 8:05
  to provide them with confidentiality.
• 8:09
  And letting a third party recording device managed by who knows who,
• 8:14
  who knows where,
• 8:15
  who knows what.
• 8:17
  That is a breach of confidentiality at minimum.
• 8:21
  And I think there are all kinds of other ethical considerations
• 8:24
  at worst.
• 8:26
  So,
• 8:27
  I think you kind of have your finger on the pulse,
• 8:30
  Leanna,
• 8:30
  of,
• 8:30
  of what I would also say.
• 8:33
  Regarding this.
• 8:34
  And let's go back to
• 8:37
  Why do we need this in the first place?
• 8:41
  We were trained
• 8:43
  Even if we weren't trained in CCPT,
• 8:45
  most of us weren't,
• 8:46
  as a matter of fact,
• 8:47
  so we were trained.
• 8:49
  As therapists,
• 8:50
  counselors,
• 8:51
  social workers,
• 8:52
  clinicians.
• 8:53
  To take notes
• 8:55
  We,
• 8:56
  that was part of our job,
• 8:57
  that was part of our training,
• 8:58
  that was part of our learning,
• 9:00
  how to take psychotherapy notes.
• 9:04
  Why do we all of a sudden feel like we can't take psychotherapy notes anymore?
• 9:08
  We absolutely can
• 9:11
  So,
• 9:11
  this is
• 9:13
  maybe
• 9:14
  in some way,
• 9:15
  shape or form to,
• 9:16
  to,
• 9:17
  I,
• 9:17
  I guess maybe powers that be might argue that this is making things easier.
• 9:21
  It's not difficult to take in session notes.
• 9:25
  Hey,
• 9:25
  I'm adding this in
• 9:27
  post original recording,
• 9:28
  which I never do.
• 9:29
  I just do my
• 9:31
  podcast episodes in one take,
• 9:32
  but
• 9:33
  I'm splicing this little part in because I
• 9:35
  wanted to add one more element to this discussion
• 9:39
  after the fact.
• 9:40
  Eric and I were talking about it,
• 9:41
  and
• 9:42
  I think one of the things that I need to address that I didn't.
• 9:46
  Is
• 9:48
  In adult therapy.
• 9:50
  The conversation is the therapy.
• 9:54
  So an adult
• 9:55
  and an adult therapist are talking.
• 9:58
  That's the entirety of the therapy.
• 10:00
  So if you were to receive a transcript
• 10:02
  of
• 10:03
  the adult therapy session,
• 10:05
  it would be the entirety of the therapy session.
• 10:09
  children
• 10:10
  The conversation is not the therapy.
• 10:15
  The play
• 10:16
  is how they communicate.
• 10:19
  And a transcript is not going to be complete.
• 10:24
  Nor understand
• 10:27
  why the play was significant,
• 10:29
  what the play was,
• 10:30
  why it mattered,
• 10:31
  what symbolic play emerged,
• 10:33
  what thematic play emerged.
• 10:36
  It's not going to know any of that.
• 10:39
  And let's run a scenario.
• 10:41
  So we don't label anything,
• 10:43
  we don't identify anything.
• 10:45
  We don't add words that the child hasn't provided.
• 10:49
  So a child is quietly
• 10:52
  Throwing
• 10:53
  pom poms up into the air and letting them rain down.
• 10:58
  And they're saying nothing.
• 11:00
  And we say,
• 11:02
  You're throwing those up.
• 11:07
  We can't say
• 11:09
  you're throwing the pom-poms and letting them rain down on you.
• 11:14
  Because that would change the model.
• 11:18
  So we have to be neutral and intentional in our response to be adherent.
• 11:22
  So we would say you're throwing those up.
• 11:26
  And if the child says,
• 11:28
  this is fun.
• 11:31
  We would say,
• 11:32
  you're really enjoying that.
• 11:36
  AI has no clue.
• 11:40
  What is going on,
• 11:41
  symbolically,
• 11:42
  thematically or significantly?
• 11:45
  When all they hear is,
• 11:46
  you're throwing that up,
• 11:47
  this is fun,
• 11:48
  you're enjoying that.
• 11:50
  So what's the summary going to be that's gonna be any,
• 11:53
  in any way meaningful or helpful to you
• 11:55
  as a CCPT
• 11:57
  Child had fun throwing something.
• 12:00
  What are we getting out of that?
• 12:01
  The play
• 12:03
  is the language,
• 12:04
  and y'all know I don't like that phrase.
• 12:05
  So,
• 12:06
  but you need to hear me in this context.
• 12:09
  The child is communicating through their play.
• 12:12
  They're not communicating through their conversation.
• 12:16
  And
• 12:17
  we obviously are verbally engaged,
• 12:19
  we're going to talk.
• 12:22
  Using skills and engage verbally,
• 12:25
  but we can't change the way that we communicate to the child to give AI clues.
• 12:30
  That's what would end up happening.
• 12:32
  Child is playing quietly
• 12:34
  and all they say is this is fun.
• 12:37
  And in our heads we're like,
• 12:39
  oh,
• 12:39
  AI is gonna have no idea what's going to have no idea what's going on,
• 12:41
  so I'm gonna have to say,
• 12:42
  you're really having fun throwing the pom-poms up
• 12:45
  and letting them rain down on your head.
• 12:49
  Now you're giving context clues to the robot,
• 12:52
  so that your notes are going to be more helpful to you.
• 12:56
  In the transcript
• 12:59
  We're absolutely changing the model
• 13:01
  and we're drifting specifically to accommodate
• 13:05
  the incapacity of AI.
• 13:07
  And I think that is just as much of a danger
• 13:10
  as all the other things that we've talked about.
• 13:13
  So,
• 13:14
  I'm concerned and confused as to why even though the technology exists,
• 13:19
  we would jump on a bandwagon
• 13:21
  that we have no idea.
• 13:24
  How secure and safe it would be.
• 13:27
  And
• 13:27
  we don't need it.
• 13:30
  So,
• 13:30
  I think that your concerns are very valid.
• 13:33
  I'm really very similarly positioned on this.
• 13:37
  And
• 13:38
  at the end of the day,
• 13:41
  A child comes in and plays with us.
• 13:44
  We are invited in when we're invited in,
• 13:46
  otherwise we're verbally engaged.
• 13:49
  We can
• 13:50
  be aware of what's happening in session and we can write stuff down
• 13:53
  and
• 13:54
  that's all we need to do presence in the playroom.
• 13:58
  is
• 13:59
  we're just right there with the kid.
• 14:01
  We know what's happening.
• 14:02
  We don't need an audio recording afterwards.
• 14:05
  To,
• 14:05
  to tell us what happened.
• 14:07
  We don't need any kind of device
• 14:09
  to fill us in on what happened,
• 14:10
  to summarize it for us to,
• 14:13
  these are not
• 14:14
  people
• 14:15
  that are trained clinically in child-centered play therapy.
• 14:19
  These are robots and computers,
• 14:21
  and we know AI lies all the time anyway.
• 14:25
  Eric
• 14:26
  was asking
• 14:27
  chat GPT I affectionately,
• 14:29
  uh,
• 14:30
  sometimes affectionately,
• 14:30
  sometimes affectionately,
• 14:31
  sometimes not referred to her as Chatty Cathy.
• 14:33
  So if I say chatty Cathy,
• 14:34
  that's who I mean.
• 14:35
  He was typing in Chatty Cathy the other day
• 14:37
  looking for an art of the session,
• 14:39
  sorry,
• 14:40
  Art of the relationship
• 14:41
  reference.
• 14:43
  So,
• 14:44
  he was like,
• 14:45
  hey,
• 14:45
  where in Art of the Relationship does it talk about whatever.
• 14:50
  And Cathy's response was,
• 14:52
  Oh,
• 14:53
  it's on this page in this chapter.
• 14:55
  Here's the quote.
• 14:57
  Totally made it up.
• 14:58
  Like
• 14:59
  didn't even exist,
• 15:00
  was not a valid reference at all.
• 15:03
  And Eric typed back in,
• 15:04
  did you just make that up?
• 15:07
  And it admitted that it created a reference
• 15:09
  out of the oblivion
• 15:11
  and passed it off as fact
• 15:13
  as an answer.
• 15:14
  Y'all,
• 15:15
  AI is not trustworthy.
• 15:18
  So
• 15:19
  let's ignore all the other things we've talked
• 15:21
  about as if those aren't big enough considerations,
• 15:24
  then we're trusting a robot that lies
• 15:27
  to give us summaries
• 15:28
  and analyze what happened and come up with themes and write our notes for us.
• 15:32
  This is so scary.
• 15:35
  We have a place in this relationship.
• 15:36
  We have a place in this relationship.
• 15:38
  We have a place in this role,
• 15:40
  and we're not gonna give that away to anyone else,
• 15:43
  let alone a computer.
• 15:45
  So please be very careful.
• 15:46
  I,
• 15:47
  I'm really cautious and concerned about this.
• 15:49
  And so
• 15:50
  be very thoughtful.
• 15:52
  Just because there is technology available to do something
• 15:55
  does not mean that we should be doing it.
• 15:57
  We need to preserve
• 15:58
  classic
• 15:59
  CCPT.
• 16:01
  As it was intended,
• 16:03
  as it was written
• 16:04
  in its original form.
• 16:07
  The further we get away from that,
• 16:09
  the more we lose the model entirely.
• 16:12
  So,
• 16:12
  be very careful.
• 16:13
  That's my caution and my plea.
• 16:16
  I am really against
• 16:19
  AI in the playroom of any kind
• 16:22
  for a lot of reasons.
• 16:23
  So,
• 16:23
  Leanna,
• 16:24
  thank you so much for the question.
• 16:25
  I think that that was really helpful for us to dive into.
• 16:27
  If y'all wanna reach out,
• 16:28
  please do.
• 16:28
  brenna@thekidcounselor.com.
• 16:30
  I love y'all.
• 16:31
  We'll talk again soon.
• 16:32
  Bye.
• 16:33
  Thank you for listening to the Play Therapy Podcast with Dr.
• 16:36
  Brenna Hicks.
• 16:37
  For more episodes and resources,
• 16:39
  please go to www.playtherapypodcast.com.