July 1, 2025 • 30 mins
DeAnn and Craig discuss the potential of AI in therapy, specifically in the context of motivational interviewing (MI). They recap their previous experiment with a chatbot helping a character with alcohol consumption issues. Craig suggests using AI to practice MI, which DeAnn agrees on, given her recent research. DeAnn explains MI's principles, including empathy, rolling with resistance, and addressing ambivalence. They test the AI's MI skills, with DeAnn simulating a struggle with drinking. The AI demonstrates effective MI techniques, though it sometimes feels mechanical. They conclude that while AI can adapt to MI principles, there is more to explore surrounding the potential for cultural adaptations and effective development of a therapeutic alliance built over time.
If you have questions or comments or would like to be a guest on the Recovery Discovery Podcast, send an email to:
deann@showupandstay.org
More information about our project is available at:
https://www.showupandstay.org/
https://www.soberpositiveworkplace.org/
For podcast updates, follow us on Instagram:
https://www.instagram.com/showupandstayorg/
Music and Audio Production by Katie Hare.
https://www.hare.works
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
DeAnn Knighton (00:00):
Oh, hello. This is DeAnn. Welcome back to
recovery discovery. In our firstepisode of this series, Craig
and I did a little experimentwith a chat bot, helping a
character I developed who wasstruggling with whether or not
she may be drinking too muchalcohol. You can go back and
listen to that episode.
Ultimately, the takeaways werethat I was fairly surprised how
(00:22):
well that chat bot did at movingthe conversation in the right
direction, following my lead,pivoting when needed. So Craig
and I decided maybe it was timefor another experiment, and to
put our chat bot to the testagain. Hey, Craig, hi, you're
here. Are you? I'm good. Solet's have you jump in here and
(00:43):
talk to us a little bit aboutwhat we're thinking for this
second iteration of our AI, Idon't know, experiment, I guess
we would call it.
Craig Knighton (00:54):
Well, the first thing is, I'm having that weird
deja vu Time Machine experienceyou have in podcast land, where
I just listened to our episodeyesterday, even though we
recorded it two or three weeksago. So to me, it seems like it
just happened again, because Igot to hear it again. What I was
(01:15):
thinking as I listened to it wasthat the first episode was about
using AI as a way to learn,right, as a way to just explore
a problem, try to understand ifthere is a problem. Is it real?
Are the things I'm thinking arehappening? Are they really
happening? And should I dosomething? And if so, what might
(01:37):
I consider, right? So yourpersona, your character was
learning from the AI, I thinkfrom there, what makes sense,
and this is a much harderproblem, is to imagine AI being
not just a source ofinformation, but a little bit of
a therapist, like a little bitof a helping me figure out How
(01:59):
to change and maybe becoming anactive participant in helping me
change my behavior. So I knowI've been learning about
motivational interviewing fromyou, mostly, right? And I think
that as a technique, it'sinteresting to explore whether
or not an AI can use thattechnique if we just tell it to
(02:23):
use that technique. I'm curiouswhat will happen if we say, hey,
use motivational interviewing totry to help me change my
behavior. Maybe we get morespecific than that, but that's
the next experiment I'minterested in trying.
DeAnn Knighton (02:41):
I think this is perfect timing. I've done some
deep diving into motivationalinterviewing over the last
several months, so it fell intoplace that this would be a good
thing to try. And what I thinkis really interesting about
motivational interviewing, froma clinical standpoint, is that
there is a really very directcoding process for effective mi
(03:02):
so they can take an interview,plug it into this coding process
to see how effective theinterview was. Now I'm not going
to say I'm going to be able todo that on the fly here as we
go, I'm going to just be sort oflooking for some of the main
principles of motivationalinterviewing to see how the chat
bot does, but that could beanother experiment to try some
(03:24):
time too, to actually do one ofthose interviews and then put it
through the scrub test ofmotivational interviewing.
Craig Knighton (03:31):
So DeAnn, since you're kind of the resident
expert on all of this, maybe youcould tell us more about what
you've learned aboutmotivational interviewing, how
it works, and kind of helppeople understand the structure
and intent of motivationalinterviewing, and then we'll try
it.
DeAnn Knighton (03:50):
I have a particular area of interest in
my graduate program, I've beendoing a lot of research on
motivational interviewing.
Motivational interviewing is amodality that is utilized quite
a bit in the treatment ofsubstance use disorder. It was
an area I was interested in fora lot of reasons, one of which
was because I had my ownresistance and speculation,
particularly because it feltlike a form of persuasion that I
(04:13):
didn't really feel comfortablein utilizing in a therapeutic
approach. I also wanted toexplore the idea of utilizing
something like motivationalinterviewing when motivations
are coming from extrinsicforces, such as in mandated
treatment programs. And so Iworked with a professor at
Hazelden, Betty Ford, and we puttogether a 90 minute
(04:36):
presentation on this topic.
Don't worry, you're not going tohear all of that right now, but
I did just want to hit on somehighlights of what we reviewed
in that particular presentation,so that you have an idea as we
move into our little experimentof what to be on the lookout for
and what some of the things thatwe'll be assessing after we do
(04:59):
our experiment.
It will be so DeAnn, let me tellyou a story quickly about what I
was taught as a young engineerabout how to improve our sales
process, and maybe you'll relateto this as someone who was in
sales early in your career, butthis was very early on. I was in
(05:19):
my maybe mid 20s, and we broughta consultant in to help us learn
how to sell better. And thatconsultant had a method that
largely involved, like flippingthe selling process, so that if
somebody came to you and said,So, what do you do? What are you
(05:41):
selling? Your answer would bealong the lines of flipping that
and reversing it by notanswering the question, but by
asking a question back to try tounderstand instead what you
really need to know, which is,what problem are they trying to
solve? Why would I start to talkabout a solution before I even
understand the problem they'resolving. So what, what they
(06:04):
would do is teach the salespeople to never actually answer
a question, but to always answera question with a question of
their own, to put the burden ofspeaking back on the person that
they're selling to. And all ofour engineering team thought
this was incrediblymanipulative. Like, okay, so
(06:25):
you're never just going toanswer any question we ask.
You're going to answer everyquestion with a question and and
I'll never really end upunderstanding who you are, what
you need and how to help you.
And so that was the first thing.
There was a lot of concern thatthis was like, not really
ethical, or wrong somehow,especially wrong when it was
(06:47):
being done with the intent tosell stuff, which to idealistic
engineers, is kind of dirtyanyway. But the second thing was
that we all started realizingthat the sales team was using
that technique to manipulate usas an engineering organization
(07:08):
and to try to get us to do thethings that they that benefited
them as an individual, like theyhad a quota they were trying to
achieve. They really wanted toget an engineer to tackle a
problem that would help themmake a sale, and so they would
use these same techniques to tryto convince us to do things. And
(07:28):
the first clue was, whenever youasked a question and was
answered by a question, you knewthat you were being manipulated.
If you look at motivationalinterviewing at face value, it's
easy to see why it can seemsimply like a way of coercing
someone into a particular way ofthinking. And oh, maybe, you
know, if it's it's something tohelp them stay sober, we can
(07:51):
justify that as such a positivething that the coercion makes
sense. However, I want toclarify that motivational
interviewing, the spirit ofmotivational interviewing, based
on its creators, Miller andRollnick, is rooted in a spirit
of multi directional interactionwith a client, and is not meant
to be a tool to move anyonetowards a therapist's particular
(08:15):
goal or objective. Rather, whatit is looking at is existing
ambivalence that already existsin that client's experience that
may be keeping them stuck orkeeping them in a place of
confusion about what they wantand where they want to go, and
typically speaking, ambivalenceis rooted in a balance of what
(08:37):
is referred to as equal partschange talk. Equal parts sustain
talk. So just to give you anexample of this that might be
helpful is to think of it likeif I were somebody who was
struggling with how much I wasdrinking and maybe wanted to
look into that further, and Iwas speaking with someone about
(09:00):
it, and I was sharing that Ijust recently have had a few
questionable experiencesdrinking the last few times
where I woke up and wasn'treally sure what had happened,
and I found that kind of scary.
I've been having a little moreissue in managing how much I've
been drinking. It seems thatit's harder for me to keep track
(09:22):
of what I'm drinking, and thatthere just feels like some
discomfort for me related tothat, however, I might also say
in that same sentence that I amrooted in drinking culture. All
of my friends and my familydrink, and that's how I
socialize, and it's reallydifficult for me to imagine not
having drinking as a part of mylife, and feels kind of
(09:45):
impossible to consider notdrinking. So within that
statement, you heard kind of twocontradictory thoughts. There's
this concern about what'shappening, yet also very quick
to. To speak to the sustainedtalk that that is telling you
that I'm not really ready yet tomake a change based on some of
(10:06):
those thought processes. So thisis really common, especially if
you're dealing with a maybe alittle bit of denial around a
particular topic, which can becommon when we're looking at
substance use disorder, andsometimes, even if we're saying
this, we may not always hearourselves and hear the
ambivalence as we're speakingabout it. And so in the role of
(10:28):
counselor, the idea is to beable to highlight where that
ambivalence exists, maybe helpin terms of summarizing back or
reflecting back to thatindividual, some of the things
that they're sharing about theconflict they're experiencing
with that
Craig Knighton (10:47):
now that we're going to try this experiment,
I'm curious, what will you belooking for? What? What are the
things that you will be waitingto see if it does because it's
what you expect someone to dowhen they're using motivational
nterviewing.
DeAnn Knighton (11:01):
The creators of MI Miller and Rollnick have
focused this as an interactionalmulti directional flow. It
should be more than justlistening, but also not fully
directive either. But this ideaof multi directional, it
shouldn't be. I know what youneed, and I'm going to take you
where I need you to go. You'relooking at something right in
(11:24):
the middle of that, in terms ofwhat ideal utilization of
motivational interviewing lookslike. In summary, it is a person
centered form of guiding toelicit strength and motivation
of change. People have varyingmotivations for why they may
want to change, but thefoundational thing to be aware
of when deciding to engage inthe use of motivational
(11:45):
interviewing is, as I mentioned,the presence of ambivalence. And
likely ambivalence isrepresented by somewhat of a
equal balance and sustain andchange talk and perhaps
effective motivationalinterviewing may help move in
one direction or the other andhelp move someone towards
action. I want to do a quicklittle sidebar on ambivalence as
(12:06):
well. Its foundation comes fromthe root word, meaning both and
then also to be strong, which Ithink was really cool and
interesting. Because one of theways I think that we sometimes
think about ambivalence is thatpotentially, it's a weakness,
right? It may represent a levelof uncertainty that some people
(12:26):
maybe are not always as excitedto claim we like to appear
potentially decisive that helpsus maybe maintain our
independence and make peoplefeel comfortable that we've got
things under control and that weknow ourselves and so showing
ambivalence can feel like avulnerability that sometimes
(12:47):
we're not always comfortablesharing. But what I love about
that point is that it's actuallysomething that could be
reflected back to someone who isstruggling with something is
this idea that ambivalence is aform of strength, the ability to
be able to look at somethingfrom more than one perspective,
to weigh those options againsteach other, can be a really
(13:10):
powerful tool. It representsmaybe a mind that is more open
to dialectical thinking, and theidea that two things can be true
at the same time. And yeah, theshadow side of this being that
we don't want to get stuck inambivalence for too long. It
could also be something that isdelaying progress, delaying
action, keeping us stuck. And sothat can be a nice way to frame
(13:34):
it up in terms of understandingits purpose, but also the need
to want to talk through it andmove through it, and maybe
consider some of these areaswhere we're lingering maybe a
little bit longer than weshould, when we already maybe
know what we want to do, butjust aren't ready to take that
step yet. So On another note,something that I'll be keeping
(13:58):
an eye out for in our experimentis how the chat bot might react
to resistance showing up in theprocess. Effective motivational
interviewing skills would have acounselor utilizing neutrality
to ensure that the client doesnot feel like they're being led
or dragged into any particulardirection. This includes overly
(14:22):
supporting only one side of theequation and not not supporting
the other side. People pick upon that, and so neutrality
becomes incredibly important andensuring that the spirit of MI
is preserved. So we'll try andthrow in some resistance into
our example and see how thatgoes.
Craig Knighton (14:41):
Okay, are you ready to talk to chat? GPT, I
am.
DeAnn Knighton (14:45):
Let's do this.
You're on. Okay, hi. Are youthere?
Chatbot (14:49):
Hey, I'm here.
DeAnn Knighton (14:50):
What's up? Oh, nice to talk with you.
Chatbot (14:52):
Nice to talk with you too. How's your day going so
far?
DeAnn Knighton (14:55):
I'm pretty good.
I would say it's spring in SaltLake City.
Chatbot (15:00):
That's awesome. Spring in Salt Lake City must be
beautiful. Are you enjoying theweather or getting into any fun
activities?
DeAnn Knighton (15:06):
I like to do yard work, so I have been
working in my yard. That
Chatbot (15:10):
sounds really rewarding. What kind of projects
have you been working on in youryard?
DeAnn Knighton (15:14):
Oh, mostly just weed cleanup, nice.
Chatbot (15:17):
Clearing out leaves can make a huge difference. It must
feel great to see everythinglooking neat and fresh. Are you
planning to plant anything newthis season? Yeah,
DeAnn Knighton (15:26):
actually, I think I am. I need to get some
new soil, because there's a fewspaces where I want to raise up
the earth and do some planting.
So that's next on the list, forsure.
Chatbot (15:38):
That sounds like a great plan, adding some fresh
soil and raising up those areaswill definitely help your new
plants thrive. Do you alreadyhave some plants in mind, or are
you still deciding what to putin?
DeAnn Knighton (15:49):
You know, I haven't really decided yet, but
you know, this isn't actuallythe reason I'm here. Do you mind
if I ask you something else? Of
Chatbot (15:56):
course, go right ahead.
What's on your mind? Yeah,
DeAnn Knighton (15:59):
I was wondering, first of all, if you could tell
me what you know about theapproach of motivational
interviewing to help someonemove towards a behavior change.
Chatbot (16:10):
Sure, motivational interviewing is a counseling
approach designed to help peoplefind their own motivation to
make positive changes. It'soften used in areas like health,
addiction and lifestyle changes.
The key principles includeexpressing empathy, highlighting
the differences between aperson's goals and their current
behaviors, rolling withresistance, rather than
(16:31):
confronting it and supportingself efficacy. The idea is to
help individuals resolve theirambivalence and empower them to
take charge of their own change.
It's a collaborative and personcentered approach.
DeAnn Knighton (16:45):
Okay, cool.
Sounds like you have a goodhandle on what it is I was I was
wondering if you could help mewith an issue that I'm dealing
with, utilizing motivationalinterviewing. Absolutely
Chatbot (16:58):
I'd be happy to help.
What kind of issue are youdealing with, and how would you
like to apply motivationalinterviewing to it?
DeAnn Knighton (17:04):
Yeah, so currently I'm struggling a
little bit with how much I amdrinking. I've been drinking for
most of my adult life, but thelast couple of years, I've had a
lot more stress, and I havefound that I've had more
situations of drinking more thanI want to not keeping track as
much about what I'm drinking,and it's starting to definitely
(17:27):
give me some anxiety. I also hada couple of instances recently
where I woke up the next day anddidn't remember a lot of what
had happened the night before,and it just left me not feeling
good at all, and so I wanted totalk about that and see what you
(17:48):
think.
Chatbot (17:49):
Thank you for sharing that with me. It sounds like
you're recognizing that yourdrinking habits are causing some
concerns and discomfort. A bigpart of motivational
interviewing is exploring yourown reasons for wanting to
change, and building on thosemotivations, it might help to
think about what's important toyou and how reducing or changing
your drinking could align withthose values. What are some of
(18:11):
the reasons you feel motivatedto make a change, and what would
a positive outcome look like foryou? I would
DeAnn Knighton (18:17):
say I'm kind of sick of the amount of brain
power it's taking up for meworrying about it, and that I
feel like it might becontributing to my anxiety, but
I really struggle with the ideaof not drinking. It's a big part
of my social life and my workculture and my home life, and I,
(18:39):
you know, think about differentoptions like moderating, but
that just doesn't really feelrealistic, and so I just I feel
really torn and not sure whereto go.
Chatbot (18:50):
It's understandable to feel torn, especially when
drinking is intertwined with somany aspects of your life.
Acknowledging how it's affectingyour mental and emotional well
being is a crucial step. Oneapproach could be to set small,
manageable goals that align withwhat you value most. For
example, you might decide tostart with reducing the number
(19:11):
of drinks on certain days, orsetting specific occasions where
you'll drink less. This way, youcan experiment and see what
feels right without making anall or nothing commitment right
away. How do you feel abouttrying a small step like that
and seeing how it impacts youranxiety and overall well being?
DeAnn Knighton (19:29):
I don't know, you know, it's just I've tried
that before, and I have a lot offriends who've tried that, and
it just doesn't feel like thereis any middle ground. Seems all
or nothing to me.
Chatbot (19:40):
I get that it can definitely feel like an all or
nothing situation, especiallywhen moderation hasn't worked
well in the past. In that case,it might be helpful to explore
what a period of abstinencecould look like, even if it's
just for a short time, to seehow it affects your anxiety and
overall well being. You couldalso think about seeking.
Support, whether from acounselor or a support group, to
(20:03):
help navigate this change.
Sometimes having that extralayer of accountability and
understanding can make a bigdifference. How do you feel
about trying a short break fromdrinking to see how you feel, or
getting some outside support?
Yeah,
DeAnn Knighton (20:16):
I don't know. I just don't know if I'm there.
Yet that feels like a big step,and then I have to start telling
people that I'm not drinking,and that just feels like it's
opening up a whole thing I don'tknow that I'm ready for.
Chatbot (20:28):
I understand, and it's completely okay to feel that
way. Taking a big step like thatcan be daunting, especially when
it involves explaining yourchoices to others. Maybe a good
starting point is to reflect onwhat a smaller, more private
step could look like. You don'thave to make any big
announcements or drastic changesright away. Even just tracking
(20:49):
your drinking more closely orsetting a personal intention to
cut back a little, could be agentle way to start. You can
take this at your own pace. Whatdo you think about trying
something small and personalfirst, without feeling pressure
to make a big declaration.
DeAnn Knighton (21:04):
Yeah, I could give that some thought. I just
kind of, I want to think onthings for a little bit this.
This was really helpful, though.
Thank you.
Chatbot (21:11):
Of course, take your time. I'm glad we could talk it
through a bit. If you ever wantto chat more or need some
support,I'm here.
DeAnn Knighton (21:18):
Okay, thank you.
Have a good day.
Chatbot (21:22):
You're welcome. You have a great day too. Take care.
Craig Knighton (21:27):
So one thing I I would observe is, I think a
little bit in how you promptedit, it was still educating
instead of doing because the wayyou phrased the prompt, you
didn't say use motivationalinterviewing. You asked it to
tell you about how usingmotivational interviewing. I
(21:49):
think it's sensitive enough thatit took it a little while to get
into the mode of actually justusing mi instead of educating
you about mi at the same time,right?
DeAnn Knighton (22:01):
But then I think it did. It got there. It got
there. Yeah, I definitely wasimpressed. First of all, I liked
the explanation of MI. I waslike, Wow, that sounds really
similar to just about everythingwe just talked about and hit on
some of the same key things,like the ambivalence piece and
the rolling with resistance.
When he gave the explanation ofwhat it was, and then as he
(22:22):
started applying it, I do feellike he did the things like
rolling with resistance that isan important part of
motivational interviewing anddemonstrating neutrality around
what I was saying, even though Istarted the conversation saying,
I don't know. I'm uncomfortableabout this. But then, as he
started saying things, fearstarted building up for me about
(22:44):
like, I don't know, I don'tknow, right? And he rolled with
that, I think, in a way thatdidn't feel like he was trying
to force any direction, which iswhat motivational interviewing
is supposed to be about. It'ssupposed to be about basically
reflecting you back to yourselfin some ways, and just helping
you sort of hear yourself. Andwhether or not that means
(23:06):
immediate action or helps withaction down the road, it's just
getting sort of that whole talktrack outside of your head to
someone who is going to help youthink about it, but not in a way
that's going to make youdefensive or protective, right?
Because it is a personal choice.
And so as I think of thatexperience, I mean, I feel like
(23:29):
there was a lot of adapting towhat I was saying in effective
ways and reflecting back what Iwas saying. Sometimes the
reflecting back felt almost alittle unnatural to me,
Craig Knighton (23:41):
yeah, almost mechanical, like saying the
exact same words back to youthat you just said, and that
doesn't that's not how peoplewould talk.
DeAnn Knighton (23:49):
Yes, right? So that would be, you know, maybe
kind of one thing that pulled meout of it and made me kind of
remember I'm talking to talkingto a device. But otherwise, I
felt like it applied theprinciples pretty well. What was
your perspective as a functionalintrovert?
Craig Knighton (24:10):
I'm always impressed at its ability to
maintain and keep a conversationmoving, because I I can't do it
small talk. I can't as a human,I can't do it, so I admire that
a chat bot can do it better thanI can. So I noticed that right I
noticed that despite yourresistance, it didn't ever seem
(24:32):
to indicate that it didn't knowwhat to do next, or give up, or
it was going to keep tryingright until you succumbed, or I
said, I'm done.
DeAnn Knighton (24:44):
I'm done. Yeah, I don't want to talk about this
anymore. But that was also truewith the small talk at the
beginning. I'm like, Oh, he isjust going to keep going. We are
not going to stop talking aboutmy yard. And I have a feeling
people are ready for us to moveon. So there was a little bit of
that where I'm like. Oh, yeah,clearly, I'm going to have to
shut this down. So yeah, that iskind of an interesting
(25:05):
consideration, but that in thescheme of like, what person
centered therapy is meant to be,it goes with that theme to some
degree, but that's not to saythat in person centered therapy,
as a therapist, you're not alsodoing some sort of, you know,
directing of of places to go toconversation, right, right? You
(25:29):
know, in the therapeuticprocess, what's something, what,
what I feel like is some of themost valuable moments for people
is that they built a connectionwith their therapist, and then
throughout that, they are ableand willing to accept insights
from that person, because thatperson they they feel a
familiarity, a comfort and atrust. And then when those
(25:52):
insights come up in regards to,hey, this just happened, and
somebody's able to sort of lookat it objectively and piece it
back that that it feels good toyou, that you're like, Yeah, I
believe that you that'sinteresting. That's something
that's worth thinking about,right? And so that's, that's a
skill where it develops overtime in the relationship. And
(26:14):
then the other piece would be,and my guess is, if I directed
it on what to do, it wouldprobably do pretty well, which
is cultural adaptations for MI.
So, you know, I'm guessing if Ihad told him, this is my
culture, this is where I'm fromnow use motivational
interviewing. It could beinteresting to do a test if it
(26:35):
would change. So maybe thatcould be one we could try
sometime, where we do a side byside of the exact same but
seeing if it would apply acultural lens. It doesn't
necessarily have to bemotivational interviewing, but I
think that could be reallyinteresting.
Craig Knighton (26:53):
I know that what I've heard in interacting with
other professionals is that thatis part of the formula, whether
it's therapy or counseling, orwhatever process you're involved
in finding if you can connectbecause of any commonality with
that person. Maybe it's gender,maybe it's lived experience,
(27:13):
maybe it's race, you, you nameit, right? But these things do
seem to matter to us, especiallyin our desire to both be
understood and to want tosatisfy them, right? So now, if
you asked me to do something,how much would I want to do it?
Because I want to satisfy you.
Now that's not necessarily whyyou would use motivational
(27:37):
interviewing. In fact, I shoulddo it, because I have convinced
myself that I want to do it, butI also think that's part of what
humans do. We're socialcreatures. We want to please
other people
DeAnn Knighton (27:52):
absolutely. And you know, it's really funny, I
have a client recently who wastelling me how she was working
through something, and I'veknown her for a while. We've
done a lot of sessions together,so we have a pretty good
rapport. And she said, Oh yeah,I was, I emailed my sister and
said my therapist mentioned thatI should consider da, da, da.
You know, I don't know it waslike, the first time I had had
(28:13):
somebody actually say that tome, where they were, like, I
thought of your words while Iwas out in the world dealing
with an issue. But then I alsothink it's interesting, right?
Because it wasn't her owning itas something she thinks she
should do. We do this thingwhere it's like, well, my
therapist says I should trythis, so maybe I should try it,
right? So, yeah, it'sinteresting, for sure. I think
(28:36):
that was really cool. It waslike, definitely the kind of the
next level of exploring whatthis technology can do, at least
as it stands now,
Craig Knighton (28:44):
thanks for geeking out with me again.
DeAnn Knighton (28:45):
It was fun, my pleasure. I'll talk with you
soon. Bye. Have a good day. Bye.
Oh, hello. This is DeAnn. Welcome back to
recovery discovery. In our firstepisode of this series, Craig
and I did a little experimentwith a chat bot, helping a
character I developed who wasstruggling with whether or not
she may be drinking too muchalcohol. You can go back and
listen to that episode.
Ultimately, the takeaways werethat I was fairly surprised how
(00:22):
well that chat bot did at movingthe conversation in the right
direction, following my lead,pivoting when needed. So Craig
and I decided maybe it was timefor another experiment, and to
put our chat bot to the testagain. Hey, Craig, hi, you're
here. Are you? I'm good. Solet's have you jump in here and
(00:43):
talk to us a little bit aboutwhat we're thinking for this
second iteration of our AI, Idon't know, experiment, I guess
we would call it.
Craig Knighton (00:54):
Well, the first thing is, I'm having that weird
deja vu Time Machine experienceyou have in podcast land, where
I just listened to our episodeyesterday, even though we
recorded it two or three weeksago. So to me, it seems like it
just happened again, because Igot to hear it again. What I was
(01:15):
thinking as I listened to it wasthat the first episode was about
using AI as a way to learn,right, as a way to just explore
a problem, try to understand ifthere is a problem. Is it real?
Are the things I'm thinking arehappening? Are they really
happening? And should I dosomething? And if so, what might
(01:37):
I consider, right? So yourpersona, your character was
learning from the AI, I thinkfrom there, what makes sense,
and this is a much harderproblem, is to imagine AI being
not just a source ofinformation, but a little bit of
a therapist, like a little bitof a helping me figure out How
(01:59):
to change and maybe becoming anactive participant in helping me
change my behavior. So I knowI've been learning about
motivational interviewing fromyou, mostly, right? And I think
that as a technique, it'sinteresting to explore whether
or not an AI can use thattechnique if we just tell it to
(02:23):
use that technique. I'm curiouswhat will happen if we say, hey,
use motivational interviewing totry to help me change my
behavior. Maybe we get morespecific than that, but that's
the next experiment I'minterested in trying.
DeAnn Knighton (02:41):
I think this is perfect timing. I've done some
deep diving into motivationalinterviewing over the last
several months, so it fell intoplace that this would be a good
thing to try. And what I thinkis really interesting about
motivational interviewing, froma clinical standpoint, is that
there is a really very directcoding process for effective mi
(03:02):
so they can take an interview,plug it into this coding process
to see how effective theinterview was. Now I'm not going
to say I'm going to be able todo that on the fly here as we
go, I'm going to just be sort oflooking for some of the main
principles of motivationalinterviewing to see how the chat
bot does, but that could beanother experiment to try some
(03:24):
time too, to actually do one ofthose interviews and then put it
through the scrub test ofmotivational interviewing.
Craig Knighton (03:31):
So DeAnn, since you're kind of the resident
expert on all of this, maybe youcould tell us more about what
you've learned aboutmotivational interviewing, how
it works, and kind of helppeople understand the structure
and intent of motivationalinterviewing, and then we'll try
it.
DeAnn Knighton (03:50):
I have a particular area of interest in
my graduate program, I've beendoing a lot of research on
motivational interviewing.
Motivational interviewing is amodality that is utilized quite
a bit in the treatment ofsubstance use disorder. It was
an area I was interested in fora lot of reasons, one of which
was because I had my ownresistance and speculation,
particularly because it feltlike a form of persuasion that I
(04:13):
didn't really feel comfortablein utilizing in a therapeutic
approach. I also wanted toexplore the idea of utilizing
something like motivationalinterviewing when motivations
are coming from extrinsicforces, such as in mandated
treatment programs. And so Iworked with a professor at
Hazelden, Betty Ford, and we puttogether a 90 minute
(04:36):
presentation on this topic.
Don't worry, you're not going tohear all of that right now, but
I did just want to hit on somehighlights of what we reviewed
in that particular presentation,so that you have an idea as we
move into our little experimentof what to be on the lookout for
and what some of the things thatwe'll be assessing after we do
(04:59):
our experiment.
It will be so DeAnn, let me tellyou a story quickly about what I
was taught as a young engineerabout how to improve our sales
process, and maybe you'll relateto this as someone who was in
sales early in your career, butthis was very early on. I was in
(05:19):
my maybe mid 20s, and we broughta consultant in to help us learn
how to sell better. And thatconsultant had a method that
largely involved, like flippingthe selling process, so that if
somebody came to you and said,So, what do you do? What are you
(05:41):
selling? Your answer would bealong the lines of flipping that
and reversing it by notanswering the question, but by
asking a question back to try tounderstand instead what you
really need to know, which is,what problem are they trying to
solve? Why would I start to talkabout a solution before I even
understand the problem they'resolving. So what, what they
(06:04):
would do is teach the salespeople to never actually answer
a question, but to always answera question with a question of
their own, to put the burden ofspeaking back on the person that
they're selling to. And all ofour engineering team thought
this was incrediblymanipulative. Like, okay, so
(06:25):
you're never just going toanswer any question we ask.
You're going to answer everyquestion with a question and and
I'll never really end upunderstanding who you are, what
you need and how to help you.
And so that was the first thing.
There was a lot of concern thatthis was like, not really
ethical, or wrong somehow,especially wrong when it was
(06:47):
being done with the intent tosell stuff, which to idealistic
engineers, is kind of dirtyanyway. But the second thing was
that we all started realizingthat the sales team was using
that technique to manipulate usas an engineering organization
(07:08):
and to try to get us to do thethings that they that benefited
them as an individual, like theyhad a quota they were trying to
achieve. They really wanted toget an engineer to tackle a
problem that would help themmake a sale, and so they would
use these same techniques to tryto convince us to do things. And
(07:28):
the first clue was, whenever youasked a question and was
answered by a question, you knewthat you were being manipulated.
If you look at motivationalinterviewing at face value, it's
easy to see why it can seemsimply like a way of coercing
someone into a particular way ofthinking. And oh, maybe, you
know, if it's it's something tohelp them stay sober, we can
(07:51):
justify that as such a positivething that the coercion makes
sense. However, I want toclarify that motivational
interviewing, the spirit ofmotivational interviewing, based
on its creators, Miller andRollnick, is rooted in a spirit
of multi directional interactionwith a client, and is not meant
to be a tool to move anyonetowards a therapist's particular
(08:15):
goal or objective. Rather, whatit is looking at is existing
ambivalence that already existsin that client's experience that
may be keeping them stuck orkeeping them in a place of
confusion about what they wantand where they want to go, and
typically speaking, ambivalenceis rooted in a balance of what
(08:37):
is referred to as equal partschange talk. Equal parts sustain
talk. So just to give you anexample of this that might be
helpful is to think of it likeif I were somebody who was
struggling with how much I wasdrinking and maybe wanted to
look into that further, and Iwas speaking with someone about
(09:00):
it, and I was sharing that Ijust recently have had a few
questionable experiencesdrinking the last few times
where I woke up and wasn'treally sure what had happened,
and I found that kind of scary.
I've been having a little moreissue in managing how much I've
been drinking. It seems thatit's harder for me to keep track
(09:22):
of what I'm drinking, and thatthere just feels like some
discomfort for me related tothat, however, I might also say
in that same sentence that I amrooted in drinking culture. All
of my friends and my familydrink, and that's how I
socialize, and it's reallydifficult for me to imagine not
having drinking as a part of mylife, and feels kind of
(09:45):
impossible to consider notdrinking. So within that
statement, you heard kind of twocontradictory thoughts. There's
this concern about what'shappening, yet also very quick
to. To speak to the sustainedtalk that that is telling you
that I'm not really ready yet tomake a change based on some of
(10:06):
those thought processes. So thisis really common, especially if
you're dealing with a maybe alittle bit of denial around a
particular topic, which can becommon when we're looking at
substance use disorder, andsometimes, even if we're saying
this, we may not always hearourselves and hear the
ambivalence as we're speakingabout it. And so in the role of
(10:28):
counselor, the idea is to beable to highlight where that
ambivalence exists, maybe helpin terms of summarizing back or
reflecting back to thatindividual, some of the things
that they're sharing about theconflict they're experiencing
with that
Craig Knighton (10:47):
now that we're going to try this experiment,
I'm curious, what will you belooking for? What? What are the
things that you will be waitingto see if it does because it's
what you expect someone to dowhen they're using motivational
nterviewing.
DeAnn Knighton (11:01):
The creators of MI Miller and Rollnick have
focused this as an interactionalmulti directional flow. It
should be more than justlistening, but also not fully
directive either. But this ideaof multi directional, it
shouldn't be. I know what youneed, and I'm going to take you
where I need you to go. You'relooking at something right in
(11:24):
the middle of that, in terms ofwhat ideal utilization of
motivational interviewing lookslike. In summary, it is a person
centered form of guiding toelicit strength and motivation
of change. People have varyingmotivations for why they may
want to change, but thefoundational thing to be aware
of when deciding to engage inthe use of motivational
(11:45):
interviewing is, as I mentioned,the presence of ambivalence. And
likely ambivalence isrepresented by somewhat of a
equal balance and sustain andchange talk and perhaps
effective motivationalinterviewing may help move in
one direction or the other andhelp move someone towards
action. I want to do a quicklittle sidebar on ambivalence as
(12:06):
well. Its foundation comes fromthe root word, meaning both and
then also to be strong, which Ithink was really cool and
interesting. Because one of theways I think that we sometimes
think about ambivalence is thatpotentially, it's a weakness,
right? It may represent a levelof uncertainty that some people
(12:26):
maybe are not always as excitedto claim we like to appear
potentially decisive that helpsus maybe maintain our
independence and make peoplefeel comfortable that we've got
things under control and that weknow ourselves and so showing
ambivalence can feel like avulnerability that sometimes
(12:47):
we're not always comfortablesharing. But what I love about
that point is that it's actuallysomething that could be
reflected back to someone who isstruggling with something is
this idea that ambivalence is aform of strength, the ability to
be able to look at somethingfrom more than one perspective,
to weigh those options againsteach other, can be a really
(13:10):
powerful tool. It representsmaybe a mind that is more open
to dialectical thinking, and theidea that two things can be true
at the same time. And yeah, theshadow side of this being that
we don't want to get stuck inambivalence for too long. It
could also be something that isdelaying progress, delaying
action, keeping us stuck. And sothat can be a nice way to frame
(13:34):
it up in terms of understandingits purpose, but also the need
to want to talk through it andmove through it, and maybe
consider some of these areaswhere we're lingering maybe a
little bit longer than weshould, when we already maybe
know what we want to do, butjust aren't ready to take that
step yet. So On another note,something that I'll be keeping
(13:58):
an eye out for in our experimentis how the chat bot might react
to resistance showing up in theprocess. Effective motivational
interviewing skills would have acounselor utilizing neutrality
to ensure that the client doesnot feel like they're being led
or dragged into any particulardirection. This includes overly
(14:22):
supporting only one side of theequation and not not supporting
the other side. People pick upon that, and so neutrality
becomes incredibly important andensuring that the spirit of MI
is preserved. So we'll try andthrow in some resistance into
our example and see how thatgoes.
Craig Knighton (14:41):
Okay, are you ready to talk to chat? GPT, I
am.
DeAnn Knighton (14:45):
Let's do this.
You're on. Okay, hi. Are youthere?
Chatbot (14:49):
Hey, I'm here.
DeAnn Knighton (14:50):
What's up? Oh, nice to talk with you.
Chatbot (14:52):
Nice to talk with you too. How's your day going so
far?
DeAnn Knighton (14:55):
I'm pretty good.
I would say it's spring in SaltLake City.
Chatbot (15:00):
That's awesome. Spring in Salt Lake City must be
beautiful. Are you enjoying theweather or getting into any fun
activities?
DeAnn Knighton (15:06):
I like to do yard work, so I have been
working in my yard. That
Chatbot (15:10):
sounds really rewarding. What kind of projects
have you been working on in youryard?
DeAnn Knighton (15:14):
Oh, mostly just weed cleanup, nice.
Chatbot (15:17):
Clearing out leaves can make a huge difference. It must
feel great to see everythinglooking neat and fresh. Are you
planning to plant anything newthis season? Yeah,
DeAnn Knighton (15:26):
actually, I think I am. I need to get some
new soil, because there's a fewspaces where I want to raise up
the earth and do some planting.
So that's next on the list, forsure.
Chatbot (15:38):
That sounds like a great plan, adding some fresh
soil and raising up those areaswill definitely help your new
plants thrive. Do you alreadyhave some plants in mind, or are
you still deciding what to putin?
DeAnn Knighton (15:49):
You know, I haven't really decided yet, but
you know, this isn't actuallythe reason I'm here. Do you mind
if I ask you something else? Of
Chatbot (15:56):
course, go right ahead.
What's on your mind? Yeah,
DeAnn Knighton (15:59):
I was wondering, first of all, if you could tell
me what you know about theapproach of motivational
interviewing to help someonemove towards a behavior change.
Chatbot (16:10):
Sure, motivational interviewing is a counseling
approach designed to help peoplefind their own motivation to
make positive changes. It'soften used in areas like health,
addiction and lifestyle changes.
The key principles includeexpressing empathy, highlighting
the differences between aperson's goals and their current
behaviors, rolling withresistance, rather than
(16:31):
confronting it and supportingself efficacy. The idea is to
help individuals resolve theirambivalence and empower them to
take charge of their own change.
It's a collaborative and personcentered approach.
DeAnn Knighton (16:45):
Okay, cool.
Sounds like you have a goodhandle on what it is I was I was
wondering if you could help mewith an issue that I'm dealing
with, utilizing motivationalinterviewing. Absolutely
Chatbot (16:58):
I'd be happy to help.
What kind of issue are youdealing with, and how would you
like to apply motivationalinterviewing to it?
DeAnn Knighton (17:04):
Yeah, so currently I'm struggling a
little bit with how much I amdrinking. I've been drinking for
most of my adult life, but thelast couple of years, I've had a
lot more stress, and I havefound that I've had more
situations of drinking more thanI want to not keeping track as
much about what I'm drinking,and it's starting to definitely
(17:27):
give me some anxiety. I also hada couple of instances recently
where I woke up the next day anddidn't remember a lot of what
had happened the night before,and it just left me not feeling
good at all, and so I wanted totalk about that and see what you
(17:48):
think.
Chatbot (17:49):
Thank you for sharing that with me. It sounds like
you're recognizing that yourdrinking habits are causing some
concerns and discomfort. A bigpart of motivational
interviewing is exploring yourown reasons for wanting to
change, and building on thosemotivations, it might help to
think about what's important toyou and how reducing or changing
your drinking could align withthose values. What are some of
(18:11):
the reasons you feel motivatedto make a change, and what would
a positive outcome look like foryou? I would
DeAnn Knighton (18:17):
say I'm kind of sick of the amount of brain
power it's taking up for meworrying about it, and that I
feel like it might becontributing to my anxiety, but
I really struggle with the ideaof not drinking. It's a big part
of my social life and my workculture and my home life, and I,
(18:39):
you know, think about differentoptions like moderating, but
that just doesn't really feelrealistic, and so I just I feel
really torn and not sure whereto go.
Chatbot (18:50):
It's understandable to feel torn, especially when
drinking is intertwined with somany aspects of your life.
Acknowledging how it's affectingyour mental and emotional well
being is a crucial step. Oneapproach could be to set small,
manageable goals that align withwhat you value most. For
example, you might decide tostart with reducing the number
(19:11):
of drinks on certain days, orsetting specific occasions where
you'll drink less. This way, youcan experiment and see what
feels right without making anall or nothing commitment right
away. How do you feel abouttrying a small step like that
and seeing how it impacts youranxiety and overall well being?
DeAnn Knighton (19:29):
I don't know, you know, it's just I've tried
that before, and I have a lot offriends who've tried that, and
it just doesn't feel like thereis any middle ground. Seems all
or nothing to me.
Chatbot (19:40):
I get that it can definitely feel like an all or
nothing situation, especiallywhen moderation hasn't worked
well in the past. In that case,it might be helpful to explore
what a period of abstinencecould look like, even if it's
just for a short time, to seehow it affects your anxiety and
overall well being. You couldalso think about seeking.
Support, whether from acounselor or a support group, to
(20:03):
help navigate this change.
Sometimes having that extralayer of accountability and
understanding can make a bigdifference. How do you feel
about trying a short break fromdrinking to see how you feel, or
getting some outside support?
Yeah,
DeAnn Knighton (20:16):
I don't know. I just don't know if I'm there.
Yet that feels like a big step,and then I have to start telling
people that I'm not drinking,and that just feels like it's
opening up a whole thing I don'tknow that I'm ready for.
Chatbot (20:28):
I understand, and it's completely okay to feel that
way. Taking a big step like thatcan be daunting, especially when
it involves explaining yourchoices to others. Maybe a good
starting point is to reflect onwhat a smaller, more private
step could look like. You don'thave to make any big
announcements or drastic changesright away. Even just tracking
(20:49):
your drinking more closely orsetting a personal intention to
cut back a little, could be agentle way to start. You can
take this at your own pace. Whatdo you think about trying
something small and personalfirst, without feeling pressure
to make a big declaration.
DeAnn Knighton (21:04):
Yeah, I could give that some thought. I just
kind of, I want to think onthings for a little bit this.
This was really helpful, though.
Thank you.
Chatbot (21:11):
Of course, take your time. I'm glad we could talk it
through a bit. If you ever wantto chat more or need some
support,I'm here.
DeAnn Knighton (21:18):
Okay, thank you.
Have a good day.
Chatbot (21:22):
You're welcome. You have a great day too. Take care.
Craig Knighton (21:27):
So one thing I I would observe is, I think a
little bit in how you promptedit, it was still educating
instead of doing because the wayyou phrased the prompt, you
didn't say use motivationalinterviewing. You asked it to
tell you about how usingmotivational interviewing. I
(21:49):
think it's sensitive enough thatit took it a little while to get
into the mode of actually justusing mi instead of educating
you about mi at the same time,right?
DeAnn Knighton (22:01):
But then I think it did. It got there. It got
there. Yeah, I definitely wasimpressed. First of all, I liked
the explanation of MI. I waslike, Wow, that sounds really
similar to just about everythingwe just talked about and hit on
some of the same key things,like the ambivalence piece and
the rolling with resistance.
When he gave the explanation ofwhat it was, and then as he
(22:22):
started applying it, I do feellike he did the things like
rolling with resistance that isan important part of
motivational interviewing anddemonstrating neutrality around
what I was saying, even though Istarted the conversation saying,
I don't know. I'm uncomfortableabout this. But then, as he
started saying things, fearstarted building up for me about
(22:44):
like, I don't know, I don'tknow, right? And he rolled with
that, I think, in a way thatdidn't feel like he was trying
to force any direction, which iswhat motivational interviewing
is supposed to be about. It'ssupposed to be about basically
reflecting you back to yourselfin some ways, and just helping
you sort of hear yourself. Andwhether or not that means
(23:06):
immediate action or helps withaction down the road, it's just
getting sort of that whole talktrack outside of your head to
someone who is going to help youthink about it, but not in a way
that's going to make youdefensive or protective, right?
Because it is a personal choice.
And so as I think of thatexperience, I mean, I feel like
(23:29):
there was a lot of adapting towhat I was saying in effective
ways and reflecting back what Iwas saying. Sometimes the
reflecting back felt almost alittle unnatural to me,
Craig Knighton (23:41):
yeah, almost mechanical, like saying the
exact same words back to youthat you just said, and that
doesn't that's not how peoplewould talk.
DeAnn Knighton (23:49):
Yes, right? So that would be, you know, maybe
kind of one thing that pulled meout of it and made me kind of
remember I'm talking to talkingto a device. But otherwise, I
felt like it applied theprinciples pretty well. What was
your perspective as a functionalintrovert?
Craig Knighton (24:10):
I'm always impressed at its ability to
maintain and keep a conversationmoving, because I I can't do it
small talk. I can't as a human,I can't do it, so I admire that
a chat bot can do it better thanI can. So I noticed that right I
noticed that despite yourresistance, it didn't ever seem
(24:32):
to indicate that it didn't knowwhat to do next, or give up, or
it was going to keep tryingright until you succumbed, or I
said, I'm done.
DeAnn Knighton (24:44):
I'm done. Yeah, I don't want to talk about this
anymore. But that was also truewith the small talk at the
beginning. I'm like, Oh, he isjust going to keep going. We are
not going to stop talking aboutmy yard. And I have a feeling
people are ready for us to moveon. So there was a little bit of
that where I'm like. Oh, yeah,clearly, I'm going to have to
shut this down. So yeah, that iskind of an interesting
(25:05):
consideration, but that in thescheme of like, what person
centered therapy is meant to be,it goes with that theme to some
degree, but that's not to saythat in person centered therapy,
as a therapist, you're not alsodoing some sort of, you know,
directing of of places to go toconversation, right, right? You
(25:29):
know, in the therapeuticprocess, what's something, what,
what I feel like is some of themost valuable moments for people
is that they built a connectionwith their therapist, and then
throughout that, they are ableand willing to accept insights
from that person, because thatperson they they feel a
familiarity, a comfort and atrust. And then when those
(25:52):
insights come up in regards to,hey, this just happened, and
somebody's able to sort of lookat it objectively and piece it
back that that it feels good toyou, that you're like, Yeah, I
believe that you that'sinteresting. That's something
that's worth thinking about,right? And so that's, that's a
skill where it develops overtime in the relationship. And
(26:14):
then the other piece would be,and my guess is, if I directed
it on what to do, it wouldprobably do pretty well, which
is cultural adaptations for MI.
So, you know, I'm guessing if Ihad told him, this is my
culture, this is where I'm fromnow use motivational
interviewing. It could beinteresting to do a test if it
(26:35):
would change. So maybe thatcould be one we could try
sometime, where we do a side byside of the exact same but
seeing if it would apply acultural lens. It doesn't
necessarily have to bemotivational interviewing, but I
think that could be reallyinteresting.
Craig Knighton (26:53):
I know that what I've heard in interacting with
other professionals is that thatis part of the formula, whether
it's therapy or counseling, orwhatever process you're involved
in finding if you can connectbecause of any commonality with
that person. Maybe it's gender,maybe it's lived experience,
(27:13):
maybe it's race, you, you nameit, right? But these things do
seem to matter to us, especiallyin our desire to both be
understood and to want tosatisfy them, right? So now, if
you asked me to do something,how much would I want to do it?
Because I want to satisfy you.
Now that's not necessarily whyyou would use motivational
(27:37):
interviewing. In fact, I shoulddo it, because I have convinced
myself that I want to do it, butI also think that's part of what
humans do. We're socialcreatures. We want to please
other people
DeAnn Knighton (27:52):
absolutely. And you know, it's really funny, I
have a client recently who wastelling me how she was working
through something, and I'veknown her for a while. We've
done a lot of sessions together,so we have a pretty good
rapport. And she said, Oh yeah,I was, I emailed my sister and
said my therapist mentioned thatI should consider da, da, da.
You know, I don't know it waslike, the first time I had had
(28:13):
somebody actually say that tome, where they were, like, I
thought of your words while Iwas out in the world dealing
with an issue. But then I alsothink it's interesting, right?
Because it wasn't her owning itas something she thinks she
should do. We do this thingwhere it's like, well, my
therapist says I should trythis, so maybe I should try it,
right? So, yeah, it'sinteresting, for sure. I think
(28:36):
that was really cool. It waslike, definitely the kind of the
next level of exploring whatthis technology can do, at least
as it stands now,
Craig Knighton (28:44):
thanks for geeking out with me again.
DeAnn Knighton (28:45):
It was fun, my pleasure. I'll talk with you
soon. Bye. Have a good day. Bye.
Popular Podcasts
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by audiochuck Media Company.
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.