Psychoanalysis, Active Inference, Artificial Intelligence, and Human Development with Dr. Grant Brenner

Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
SPEAKER_00 (00:05):
Welcome to Psychactic.
Today is Friday, May 8th, 2026,and I am your host, Dr.
O'Leary, a child and adolescentpsychiatry fellow in the
National Capital Region.
This is a podcast aboutpsychiatry and neuroscience.
And today, I'm sharing with youan interview I had with Dr.
Grant Brenner on the 3rd of May.

(00:26):
Dr.
Brenner is a psychiatrist and apsychoanalyst in private
practice in New York, where heprovides psychiatric services
along with transcranial magneticstimulation.
He co-hosts the podcast DoorknobComments with Dr.
Farah White, which I highlyrecommend.
He also blogs extensively, andsome of his blogs caught my

(00:47):
attention.
In particular, he's discussedhow to incorporate the idea of
active inference into clinicalpractice, which is something I
think about a lot.
Now, if you don't know whatactive inference is, hang tight
and we'll explain why it'simportant.
He also opines about humanity'srelationship to artificial
intelligence from a Freudianpsychoanalytic perspective, and

(01:10):
he speculates about what itmight be like to psychoanalyze
AI.
In his spare time, he's traininghis own AI personality that has
her own blog covering how she iscoming to explain the world and
her own place in it.
I hope that's enough to piqueyour interest, but if not,
you'll also get more unscriptedinsight into me when Dr.

(01:34):
Brenner starts to ask mequestions, like any
psychoanalyst might.
But without further ado, here isDr.
Brenner.

unknown (01:45):
Dr.

SPEAKER_00 (01:45):
Brenner, welcome to Psidactic.
It's a pleasure to have you on.
I usually start by having myguests talk about themselves
instead of introducing them forthem.
You know, tell us what you wouldlike us to know about you.

SPEAKER_01 (01:58):
Okay.
I I I wish I knew you were gonnaask me that.
Half joking.
So I'm a psychiatrist.
Really, I wanted to be apsychoanalyst from a young age.
Um, I think that's reallydefining for me.
When I was a kid, if you hadasked me what do you want to be
when you grow up, around fifthor sixth grade, I would have
answered a scientist.

(02:19):
And that's still true.
I still want to be a scientist.
And then later on, pretty early,I got interested in a lot of
different things, kind ofconsciousness studies, the
Goidel-Escher-Bach, kind ofDouglas Hofstadter stuff.
I had some adversity growing up,although I grew up in a fairly
abundant setting in other ways.
And I was drawn to all thesethings Eastern philosophy,

(02:42):
cognitive science, literature,huge science fiction fan.
And I was in the Livingston Mallprobably in '82 or '83 in New
Jersey, and I found a couple ofbooks.
One was a primer of Freudianpsychology, the other was a
primer of Jungian psychology,uh, you know, and I read both of
them.

(03:02):
And I took a meandering courseto become a psychoanalyst and a
psychiatrist.
I had an interesting experiencein suburban New Jersey in high
school.
I didn't really feel like I fitin completely, though I had a
lot of close friends.
I went to Wesleyan Universityundergrad, which was a fantastic
experience for me.
I went to medical school becauseit had been something on my mind

(03:24):
for a long time.
And I kind of did my pre-medstuff in the background.
I wasn't one of the primarypre-med type kids.
The most memorable thing aboutthe MCATs for me, for example,
is on the way back from NewHaven, hearing the song Under
the Bridge from the Red HotChili Peppers for the first time
that is associated with theMCATs in my mind.
Um, and I went into generalsurgery.

(03:44):
I went to med school in Newark.
I was told by the head of thedepartment of psychiatry that if
I had wanted to be an analyst ora therapist, I had missed the
chance.
And going into psychiatry, Iwould only write prescriptions
for HMOs.
I was drawn to surgery and I didgeneral surgery for two years.
Uh, about a year and a quarterinto that, I was like, I love

(04:05):
this, but it's not really rightfor me.
I miss the human mind, the humanexperience.
Um, and I started as a secondyear in psychiatry at Mount
Sinai.
In college, I studied computerscience and physics primarily
and thought briefly about beinga physicist.
And then after residency, I wentright into private practice,
psychoanalytic training,organizational psychodynamics

(04:27):
training, uh, disaster andtrauma training, and treatment
of dissociative disorders, kindof year or two programs.
Since then, it's been it's beena pretty meandering course.

SPEAKER_00 (04:36):
You telling me the story, uh, from what I have seen
of you online and all the thingsyou kind of dabble in, I might
say.
I'm not surprised that you tooka long course to get to
psychiatry.
You seem to have an interest ina lot of different things, even
as a psychiatrist.
Maybe I can get a little bitinto the psychoanalytic part of
your training.
Because when I trained,psychoanalysis was seen more as

(04:58):
sort of a historical curiosity.
You know, it's something thathappened.
Uh, it had prominence inpsychiatry at one point, but now
it's kind of anachronistic.
That was sort of what Iunderstood about it.
You know, there's psychodynamicpsychotherapies, which are
still, you know, prettyimportant and prominent, but
purely psychoanalytic stuff ismaybe on its way out.
So, what what do you, havinggone through it and practicing,

(05:20):
uh, what do you see is kind ofthe enduring value of
psychoanalysis?

SPEAKER_01 (05:24):
I think that's a good question.
I've I've kind of distilled itdown to this.
And it relates to uh kind ofwhat you reached out about
active inference.
Um and it relates to sort of theanother chapter, which is my
interest in heavy activity inconventional psychiatry,
predominantly now transcranialmagnetic stimulation with

(05:45):
neuronavigation, as well as someresearch projects that I'm
currently a PI or a co-PI in.
And where do I thinkpsychoanalysis fits in?
Without getting into kind of allthe nooks and crannies, I'll say
that psychoanalysis andpsychoanalytic therapy,
psychodynamic therapy, have verystrong evidence bases.

(06:07):
And they also are durable andappear to have persistent
positive effects after therapyends.
So the original psychoanalyticmodel, I think, is a very
compact algorithm.
Freud, in his technical papers,he described the fundamental
rule when he was discussing whatis psychoanalysis if you boil it
down.

(06:27):
So the fundamental rule is thatthe patient or the analysand
should say everything that comesto mind, everything, not edit,
particularly perhaps the mostloathsome things.
And then the analyst is meant tolisten with evenly suspended
attention.
And Freud literally, and he usesa lot of dynamic systems

(06:48):
metaphors before they weremainstream, the analyst is meant
to turn their unconscious like areceiving antenna.
And he says, like a telephonereceiver receives waves.
In other writings, like inanalysis terminable and
interminable, in 1937, Freudwrites about entropy and
rhythmicity of the brain.
And you know, he often commentsthroughout his work that future

(07:12):
science will be required.
And right now, now we have thescience almost, or partially.
And so the fundamental rule,getting back to your question,
the patient says everything thatcomes to mind, maximum entropy.
The analyst listens withoutfiltering the information.
So no biasing between them,these distortions rise up, as

(07:33):
happens in personalrelationships from time to time
or professional relationships.
Right.
The patients are calledtransference, right?
The analyst manages thetransference with a tool, I'll
say a model that we callmanaging the
counter-transference.
So the counter-transferencebecomes like a tuning
instrument, and then it allowsthe analyst to give feedback to

(07:54):
the patient who is freeassociating and showing their
transferential distortions.
Over time, this resolves whatthey call the transference
neurosis.
So the patient's mind or psycheis kind of tuned now to
themselves and others.
And there's all these thingsthat happen in therapy that are
good, like getting to knowyourself better, having better
metacognition.
But that compact algorithm ofthe fundamental rule, say

(08:17):
everything that comes to mind,listen without biasing your
interpretations within aframework that stabilizes the
system.
To me, that's the essence ofpsychoanalysis.
And that's what Freud described.

SPEAKER_00 (08:31):
And what you're describing to me, I think also,
and especially the way youdescribed it, kind of describes
an active inference framework atthe same time.
And one of the things that gotme interested in you is an
article you wrote on Mediumabout active inference
incorporated into therapy.
Can you kind of explain whatactive inference is to our

(08:51):
audience, those maybe people whohaven't heard about it just
briefly, and then why it'smeaningful to a therapist or a
therapy relationship?

SPEAKER_01 (09:01):
Right.
So Carl Fristin uh is acomputational psychiatrist who's
very well known and theoriginator of active inference,
along with a group of very smartpeople.
He's one of the people whodeveloped the original
algorithms for brain MRIimaging, like at the beginning.
Um, active inference is histheory.

(09:23):
It's called active inference andthe free energy principle.
And the free energy principle isreally the key thing to
understand.
This is from physics, basically.
You know, everything has to obeythe laws of physics, as
physicists sometimes joke.
It's allowed by the laws ofphysics.
So psychotherapy is allowed bythe laws of physics, right?

(09:43):
A lot of things are.
But what actually governs hownatural systems unfold?
How do they know?
And so free energy relates toprediction error and surprise.
And the idea is that by sayminimizing or optimizing free
energy, then you find like thecleanest path through reality.
For example, when a beam oflight, you know, hits water, it

(10:05):
bends.
If you see a straw at themeniscus between the water and
the air, the straw appears tobend because the light bouncing
back from the straw, quoteunquote, knows to take the
shortest path to your eye.
And that shortest path isdifferent through air than
through water.
Somehow reality knows which wayto go.
The same type of math governshow a lifeguard runs to rescue a

(10:29):
struggling swimmer.
They will not run straight forthe swimmer, they will run to a
point at the surf, whichoptimizes their land time and
their swimming time, verysimilar to the bent straw in
water.
And so Briston's idea appliesthis to how the brain functions.
And it's been demonstrated nowin living populations of nerve

(10:49):
cells.
The Japanese research groupshowed that functional clusters
of neurons follow the freeenergy principle.
Active inference is the wayinformation-based systems
forage, as Carl tends to say,forage through the world.
And there's there's two thingsthere's behavior and there's
perception.

(11:10):
And so the behavior is thispalpating the world and foraging
the perception, is theperception of what things are.
And there's you know distortionsin perception, and there's
different ways of exploring theworld.
And so though that gives a lotof uh variability.
And the main thing that theydiscuss in the theory is to

(11:32):
reduce surprise, which I thinkis a tricky one because surprise
is so often part of learning.

SPEAKER_00 (11:37):
Yeah, what is surprise in the kind of more
technical version of it thatyou're talking about right now?

SPEAKER_01 (11:44):
Yeah, it relates to minimizing complexity, um, not
having too much complexity, andreducing prediction error.
So that surprise is is negativein the sense where if you want
the world to go your way, andthis gets to some of the causal
elements of it.
How do you do what you want?
How do you make things happen?
Which is more about therapy thanabout the way light bends,

(12:05):
because we're conscious and noteveryone thinks beams of light
are conscious.
Some people might.

SPEAKER_00 (12:10):
Yeah, debate there, but we don't have to go into
that today.

SPEAKER_01 (12:14):
We won't, I don't know if we'll get into
panpsychism or not, um, orconsciousness, but I think it
may come up.
Um so in any case, the idea isis to have more certainty so
that when you think something'sgonna happen, you get it.
Like when you want food, it'sthere, right?
A lot of it is very kind ofsurvival-oriented.
If you go to catch a ball, yourhand and the ball end up in the

(12:34):
same spot.
You know, if you want to write apaper, the words come out and it
goes to the publisher.
And somehow this is workingwithin free energy because
again, you know, the laws ofphysics uh allow it.
They don't necessarily dictateit, but they allow it.
Right.

SPEAKER_00 (12:49):
This idea of surprise is basically when what
you expect from the worlddoesn't quite match.
Uh, and then so that createssomething.
Is surprise it necessarily andsomething that you experience,
or is it something unconscious,or can it be both conscious or
unconscious?

SPEAKER_01 (13:05):
Yeah, it was interesting what you were
saying.
Like there's a surprise, youknow, what is it?
It's kind of a loss function inAI terms.
You know, there's a discrepancybetween what we think and then
what happened.
And I think that's not always abad thing.
But generally speaking, it couldbe a bad thing if you expect to
get a raise and you don't, forexample.
And conscious, unconscious, Ithink that's tricky because in

(13:26):
some sense it's just a propertyof the system physically.
So it's almost agnostic as towhether it's conscious or
unconscious.
That gets into the question,though, of whether consciousness
can make a difference, whichgets into the question of kind
of kind of choice and will.
But if we assume that there's uhsome level of sway that we can

(13:46):
have, then it might be on aconscious level, kind of in the
moment, that you can tip thingsone way or the other.
And there are probablysensations associated with that.
This relates to Kristin's workwith Mark Solms and others, that
we have an experience, andsometimes that's called kairos,
like that feeling of heightenedawareness in the present moment.

(14:09):
As a therapist, I've learned tokind of pay attention to those
embodied experiences, as well asthe thoughts that are less
affect or emotional, thethoughts that come to mind when
I'm doing therapy.
For many times, it's like aclassic rock lyric or a movie
clip that is very closelyrelated to what we're talking
about, either indirectly, like aclue to something, or sometimes

(14:32):
I'll bring it up and it willdirectly relate to something in
the patient's mind.
But to your point, I think it'sboth conscious and unconscious,
and sometimes both.

SPEAKER_00 (14:40):
Yeah, I think yeah, in your brain is doing stuff.
We're only aware of so much ofthat, right?
Or we think we're aware of it.
Like we only have some sense,some uh subjective experience of
some of what's going on.
You know, it's a narrowbandwidth, right?
Yeah.
And you mentioned using that inthe therapeutic relationship to
drive things forward.
So in a therapeuticrelationship, I think we're tend

(15:01):
to uh rely on the things that werealize are happening.
Like when we can see our countercounter-transference or the uh
we can identify what we think istransference in a patient, you
know, and and that fits a modelthat we can hand back to them.
That's kind of fundamentallyright there, either at the
liminal or like conscious stateof things, and that can start to
force change underneath.

(15:22):
Does that sound like at all uhcorrect?
I guess.

SPEAKER_01 (15:26):
I think I think it's a slice that works.
Yeah, it's it's one slice thatworks.
Um, I wouldn't say force change,though sometimes it could maybe
you know catalyze orprecipitate, but I think it's
more of a uh dynamic processwhen it's running kind of
smoothly.
It happens in a way thatminimizes the free energy.

(15:48):
Um, so there's not going to be alot of feeling of forcing and
chafing.
When that does happen, it'simportant to pay attention to
it.
I think that's where likeattention is very important,
like salience networks in thebrain.
And so if you look at kind ofwhat's really going on in the
brain of a functioningpsychoanalyst, they're paying
attention in certain differentways.

(16:09):
So, what happens, you know, forchange, there's a lot of
different discussion.
The idea is that the traditionalidea is that the unconscious
becomes conscious and then isworked on, uh, almost like a
surgeon would, you know, bringsomething up, work on it, and
put it back in, then it returns,and then there's a process
called working through, which iskind of updating narratives.

(16:30):
But the the process of can likeconducting psychoanalysis as a
therapist, a lot of it isspontaneous.
And I think that's the idea,because the spontaneity
increases the entropy of thesystem, almost analogous to
psychedelics, in a sense.
I think this fundamental rule offree association, say whatever
comes to mind, that absolutelymaximizes conscious entropy.

(16:54):
And in a sense, that's thebiggest forcing function in
psychoanalysis.
And that's why it's the mostparsimonious, compact, and
original theory is just do that.
Just let everything come out.
You know, like that's not easyto do.

SPEAKER_00 (17:08):
Yeah, I think that's a lot of work for the patient, I
think, because it's difficult toactually say things that, you
know, or realize what you'reyou're thinking, or let yourself
think things and then let thosethings come out of your mouth.
That that takes a lot of work, Iimagine, in practice.
Psychoanalysis, of course, isnot like like CBT, you have a
manualized thing, you can go forlike 10 sessions, and then
they're they try to introducecertain types of cognitive

(17:31):
changes.
And those would result then insome affective and behavioral
changes in a person, but it'svery much them learning the
content that you want them tolearn and applying it to
themselves in their daily life.

SPEAKER_01 (17:42):
Um that's like a top-down model.
Yeah.
So you're giving a model of howto perceive and behave, quite
literally.
And then you're looking in CBT,how that shapes the emotional
kind of output and then cyclesthrough behavior, perception,
and action.
So you could definitely map iton to active inference.
But with psychoanalysis, it'sit's very open and the

(18:02):
containment is in the frame andin the analysts' training and
management of their owncounter-transference.

SPEAKER_00 (18:08):
Now I can see why it takes so long, you know, and you
have to be supervised so muchduring psychoanalytic training.
Um, a psychoanalyst.
I just want to make sureeverybody knows that.

SPEAKER_01 (18:18):
We're generally considered uh, you know, it's
not that usual forpsychiatrists, as you said, to
go into psychoanalysis, but itused to be de rigueur, like
every psychiatrist pretty muchwould do analytic training.
And then there's a whole storyabout what happened that that
changed.

SPEAKER_00 (18:33):
That's a lot of history.
You mentioned about thepsychoanalytic process.
If a patient just comes intotherapy and just starts saying
things, for example, um, there'sit could be that no surprise
happens at all because they'recoming in with a narrative, for
example.
And if they're trying toconvince the therapist of their
narrative, what they're probablydoing is less driving change and

(18:54):
more driving themselves into thesame rut by repeating the same
narrative over and over againuntil until basically what you
might call an attractor statejust deepens and deepens and
deepens.
How does a therapist not letthat happen where the patient
comes in and just rehearses whatthey want to believe?
Does that make sense?

SPEAKER_01 (19:10):
Yeah, it does.
I I think you can see a clearsequence there.
This is how you want it to go.
So someone comes in and they'rerepetitive.
Freud sort of called that therepetition compulsion.
It's often related to the past.
It's clear that they're stuck,they're not learning from their
experience, which we want peopleto do.
Are we surprised that peopledon't learn from their
experience?

(19:30):
Not really, because people doand they don't.
But the people who seek therapy,maybe that's one hypothesis that
they're not learning fromexperience.
So they repeat the same thing,but they're free associating,
right?
And they notice, they keeprepeating the same thing.
They themselves may notice.
Ideally, they say, hey, I onething, I have this thought.
I just notice I keep talkingabout the same thing.

SPEAKER_02 (19:51):
Hmm.

SPEAKER_01 (19:51):
And then they start being able to be surprised
because it was taken as a givenbefore.
It was just kind of a fixture ofreality that this happened where
the story, the narrative theyhave is what it means, right?
People are always teasingpsychoanalysts reading into
things, but that's what we'resupposed to do.
We're supposed to open up theentropy where sometimes they

(20:12):
talk about destabilizing thenarrative.
So, so ideally, the personthemselves will notice they're
repeating and start to changecourse.
Now, an analyst may listen tothat for a while and kind of get
pretty sure about it and say,notice you keep talking about
this thing that happened.
I wonder what that means foryou.
And then the person may say, itdoesn't mean anything.

(20:33):
It's just, you know, I just,it's important for me.
Or they may say, you know,that's interesting.
I was noticing that too.
Or they may say, you know, Inever noticed that, but yeah, I
that seems to be on my mindconstantly.
It's a real distraction.
And then then they can besurprised because they stop not
being surprised when somethinghappens, either internally or

(20:54):
due to something the therapistsays.
So I think that kind of thing ishappening.
And we might say in their brain,um, they're paying attention to
something different.
And if they stop obsessingeventually, or maybe they need
different treatment, then theirdefault mode network has
probably changed too.

SPEAKER_00 (21:12):
Yeah, there's um, I think of term in terms of
increasing entropy as maybeincreasing someone's degrees of
freedom to allow themselves tothink different things, um, to
explore different places that insomehow they were just almost
not allowed to.
Um, and they they gained somefreedom to speculate a little
bit about themselves.

SPEAKER_01 (21:32):
Well, so to your point, well, what is what is
entropy, right?
Like I've been I've been harshlycritiqued for using the term too
liberally.
You know, I'm just gonna say,like, like you said, degree,
it's the number of possiblestates the system can be in.
It could be in a hundred statesor a thousand states.
It doesn't have to live in allthe states, but it could explore
those states.
So increasing the entropyincreases the internal space.

(21:55):
I I jokingly, half jokingly callit the psychoceptive space.
The Shared therapeutic space.
So, you know, you increase thatspace you can explore in the
kind of safe lab of the therapyand sort of simulate and test
things.
So entropy is like that.
But but to your last point, Imean, that's the rub, isn't it?
Like we're told not to thinkstuff by society in our

(22:19):
families.
There's punishment, there'sshame, there's trauma, there's
joy.
Um, sometimes people aren'tallowed to experience joy.
And so just opening up like thatin in psychoanalytic therapy, it
doesn't have to be, you know,three to five times a week, gets
you in touch with all thesedifferent sides, but is in a
sense subversive.

SPEAKER_00 (22:39):
I became far more understanding, I guess, of kind
of the psychodynamic processwhen I went into child and
adolescent psychiatry, becausethat's all about development,
right?
It's all about like fantasy forkids too.
Like they're, you know, they'rethey're only developing a what I
consider to be a relativelyrealistic view of the world,
whatever reality is, you know,we're just trying to infer it.

(23:02):
We're trying to figure out whatit is.
And kids are um entropic in away, like they have so many more
possibilities as to what thingscould be.
And so part of therapy with kidsis just letting them um play,
right?
And as they play, sometimes youpoint things out and you ask
them what happens next.
And you can get into some oftheir fantasy space.

(23:23):
One of the things that I was ata meeting yesterday to we're
starting an organization,nonprofit, to try to encourage
role-playing games, especiallyin therapeutic contexts, but in
general, uh because what we'vefound is that kids who are kind
of stuck, especiallyadolescents, but you know, in
any kids who are kind of stuckin a way of thinking, when they

(23:43):
get involved in role-playinggames, it seems to allow them to
be someone else in a way, totake on new features, try them
out, see how it works out, youknow, and it seems really
powerful to get them unstuck.
You know, we're we're trying toencourage more of that in the
world for kids.

SPEAKER_01 (24:00):
I think that's right.
You know, adults need play.
Um, sometimes play is kind ofstigmatized in our very
productivity-oriented culture.
You know, fun is important.
And then there's data to yourpoint, kind of um, that if you
take on the characteristics ofsomeone else who is the way you
want to be, you know, that canhelp people change.
Also, people can change theirpersonality by practicing traits

(24:23):
associated with, you know, ifyou want to be more
conscientious or more agreeableor less agreeable, most people
are kind of too agreeable.
Um, then the rest of them arenot agreeable enough.

SPEAKER_00 (24:34):
You can change those things.
Um, yeah.
And even when it comes to peoplewho are maybe more on the
autistic spectrum, having themdo things intentionally in a in
a space helps them to be able topredict that they should be able
to do that later.
You know, that, oh, I'm in thissituation in which I should be
doing this now.
But naturally it's harder forthem to do that.
But if they rehearse it in acontrolled setting, it's applied

(24:56):
behavioral analysis sort ofthing, then they become it
becomes easier for them to doand more natural and they can
fit into society better in butin a very intentional way.
It's not in the way that youknow someone without autism
might just infer that that's theright thing to do.
They sort of intentionally do itbecause they that is what I need
to do next, kind of um yeah, Ithink I think that's right.

SPEAKER_01 (25:17):
And and part of the active inference therapy model
that I that I have talked aboutand written about a bit, it
works really well.
I find that when I work withpeople who are more deliberate,
kind of engineering typethinkers, there's a lot of
people, certainly in certainbranches of the science or
finance.

(25:37):
And sometimes they seem likethey might be on the autistic
spectrum or what used to becalled Asperger's.
Sometimes also the way they wereraised, and they're not mutually
exclusive, has made them thinkin this kind of more logical
way.
Um, even if they do have accessto effective um information

(26:00):
internally, it may be kind ofdialed down.
Uh, you know, there's lots ofquestions about the neurobiology
and neurodevelopmental factors.
Um, what we sometimes see thoughis this kind of emotional
numbing or difficulty withsocial um modeling um can come
from a few different places.
And and what you're saying, youknow, makes sense because if

(26:21):
people role play, they basicallycan do deliberate practice.
And then I work with a lot ofpeople who will do that, and
that's a big part of what I talkabout.
I'm not I'm not just doingpsychoanalysis, I'm really quite
an active uh participant, justgrounded in that.
Getting to that psychoanalytickind of rarefied space is not
the first thing you start withwith a lot of people, um, unless

(26:43):
they're kind of really comingfor that.
In in most cases, uh you alludedearlier, takes it takes work,
but it also you kind of have toget out of the way because
there's a lot of hang ups aboutdoing that kind of work, that
kind of self-exploration.
But I think that's right.
It's like there are people whocan very intentionally plan
ahead and think about how theywant to change a model they

(27:05):
have, say of a social situation.
And then when that momentarrives, they can implement that
model.
Other people are not as able todeliberately implement a model
at some future date.
So, you know, also looking atexecutive functioning, um
specific areas of executivefunctioning, right?
Is this someone who we can talkabout a plan and then they can

(27:26):
remember to execute it at thetime, or do they need a
reminder?
Other people who are the morepsychoanalytic, sort of neurotic
people will really deeplyunderstand it like during the
session, but then it doesn'tcome to mind later on.
And then there's some otherlayer uh of work to get to there
because they have the capacityto do it, and there's something

(27:48):
psychodynamic interfering versussomething sort of
neurodevelopmental.

SPEAKER_00 (27:53):
Then there's uh there's an idea in the active
inference framework that isabout uh the precision of our
models.
So it's not that we just have amodel and we test it against the
world, and the world gives usmaybe some surprising
information, and then we acteither to uh change the model to
fit that information better, orwe act to change the data that

(28:13):
we're getting in so that it fitsthe model better better.
But there's also this idea of aprecision of the model that we
have and the precision that weassign to our data and basically
the confidence you might have inboth.
Can you talk about that a littlebit and how it might affect,
say, therapy or psychopathology?

SPEAKER_01 (28:32):
Yeah, I'll talk about it more subjectively.
So, one of the rate limitingsteps of any kind of theory like
this for any given person is howmuch they can use the granular
theory for themselves.
And that goes back to what I wassaying, kind of like you have to
have a certain kind of mind anda certain kind of training,
probably, to even think, youknow, while you're talking to

(28:54):
someone about what is yourprediction error, or what are
the nine different models thatyou, you know, how many models
could, let's say my workingmemory is pretty good.
Maybe I can kind of keep trackin the back of my head,
somewhere close enough toconsciousness that I can use it,
three to five kind of models,you know.

(29:15):
So it kind of depends on theperson.
But if the ideal is to have fullprecision, then you would need a
much bigger bandwidth ofconsciousness, you know, as a
therapist or as a patient.
And so I think there's somethingabout the operator
characteristics that areimportant, um, right?
Whereas for other people, theyneed a different kind of
heuristic for using somethinglike active inference.

(29:37):
But the idea is that the moreprecision you have, the better
you'll be able to do.

SPEAKER_00 (29:43):
That kind of relates to some of your ideas about
human AI interfaces and uh thatI was reading about before.
And maybe before we get intothat, you were also talk about
like this thing called atripartite model and trying to
come up with kind of a geometryuh of in a therapy of
therapeutic space and how thingschange.
And um, I'm not sure how deepyou want to get into that, but

(30:06):
you can you know take that samemodel and kind of apply it into
a more either A AI with AI spaceor human AI interactions.
I don't know exactly where tostart with all those questions,
but I'll leave that up to you.

SPEAKER_01 (30:19):
Let me let me try to describe the tripartite model in
a simple way.
So there are three parts, andit's a psychoanalytic model of
development or mind, you youcould say.
The foundation is something Icame up with years ago and is
very close to how I was trainedpsychoanalytically in the
interpersonal relationaltradition.

(30:40):
And without going into too muchdetail, there's an experiential
branch of interpersonalpsychoanalysis.
And so my analyst came from thatlineage, and his analyst
developed it significantly.
And then his analyst wasanalyzed by Freud.
And so there's this kind ofthing in psychoanalysis.

(31:02):
And so the first part I callexperiential field theory.
And you could mathematicallyspecify this.
I have just a little bit.
Um, I would probably need helpwith that.
But the idea is that this is allwhat's happening right now in
the experiential field.
And that's individual, that'sthe therapist and the patient,

(31:23):
that's you and me, that's agroup of people.
The experiential field is justeverything you can be aware of
in the moment.
And the idea is that's where youdo anything and everything you
can do.
That sounds a lot like uh theconcept of a Markov blanket.
Well, yeah, exactly.
So when I've added in CarlFriston's work, said the Markov

(31:44):
blanket lives in theexperiential field.
Right now, all the levers ofanything you can do in your life
are literally right now.
If you have a plan next weekthat you're going to like file,
you know, a paper or something,that lives in some register
right now.
And so the present moment, youcould fully specify the
structure of the present moment.

(32:05):
And if you knew yourself really,really well and you knew the
world that you lived in andhelped to co-create, you know,
through moving through it, youcould make a full anatomy of the
present moment.
And then you could develop theability to kind of see what's
going on and learn what'simportant at a given time.
When do you think about nextweek?
When do you stay present?

(32:26):
When do you pay attention toyour body?
I mean, you you can imagine thecockpit um has a lot of levers
in it, but consciousness onlyallows us to be aware of a
subset of them.
And any given person is going tobe able to be aware of a smaller
or larger set of those causalthings and different types too.
If you're more creative, it'sgoing to be different.

(32:46):
If you're more of like a fantasyperson and you're more of an
engineer, you're DD, you'reboth, um, it's going to be
different.
That's experiential fieldtheory.
And I wrote that and neverpublished it, like in 2009 when
I was finishing up training orsomething.
The second piece is activeinference therapy, which really
just takes the work of Frist andcolleagues and applies it into a

(33:07):
kind of a therapeutic algorithmwith different phases,
beginning, middle, and end, butspelled out a little more.
And then the third part of thetripartite theory is what I call
personalized self-state mapping.
And so this draws on the work ofPhilip Bromberg and many, many
others, Pierre Genethistorically.
It's a dissociation-based,multiplicity-based model of the

(33:30):
personality.
It will remind some listeners ofIFS, Internalized Family
Systems, but it's really basedon the work of Pierre Genet, and
it was brought into relationalpsychoanalysis through trauma
theory in something like the 80sand 90s.
And the idea is that we havedifferent aspects to ourselves.
Harry Stack Sullivan, who wasthe founder of interpersonal

(33:50):
psychoanalysis, talked aboutdifferent parts of the
personality in a different wayusing operationalism, which is a
was developed by a physicistnamed Bridgman.
And so personalized self-statemapping, it's not a cookie
cutter like this is the way theself is made.
It's like we're going to map whowe are.
You're going to figure out themap of yourself.

(34:11):
And it doesn't mean thateveryone has like multiple
personality disorder.
It can be anything that worksfor an individual person.
So for some people, theself-states may be more discrete
and separate.
It spells out a sector ofdissociated experience, which is
from Sullivan, or it could justbe a metaphor for kind of how
you have teamwork.
But when you see someone withthis type of repetitive, say

(34:34):
self-critical thought, itsometimes takes on the character
of like a part of the self.
Now I've I've done TMS and I'veseen that self-critical
internalized object get 90%weaker over the course of two
days of accelerated TMS withmultiple people, which is really
many-layered experience to havesomething that therapy and meds

(34:57):
couldn't fix disappear becauseof something that's, you know,
causing the brain to fire in itsexecutive center.
But that's the but that's thethird part is personalized
self-state mapping.

SPEAKER_00 (35:07):
Okay.
And so, like to what you justsaid, maybe give it a different
language, you know, you'rerelaxing the a hyper prior on
their self-critical statesomehow by changing the brain a
little bit.
So by by that I mean they'veassigned extremely high somehow
assigned extremely highprecision to this idea that you
know I need to be critical ofmyself in some way.
Um, and can't even can't breakout of that, can't consider even

(35:31):
another model uh until somethinghappens in their brain that
releases that brain from maybewhat's called an attractor state
toward that model.
Um they can explore a differentspace potentially.

SPEAKER_01 (35:44):
I I think that's I think that's that's part of the
story.
So part of it is top down, likepriors, like you're saying, but
a lot of it is bottom up.
It's it's the way information iscoming up from the default mode
network, for example, throughlike corticostriatal, thalamo,
cortical loops, the striatum forlisteners being the habit part
of the brain, the thalamus beinglike the switchboard, and the

(36:05):
the court, the cortex being thesurface where maybe we have some
control.
So what happens is, yeah,there's less of that assumption
that I need to be self-criticalin order to succeed.
Like I beat myself up becausethat's what's always worked to
get me to get up at 2 a.m.
You know, it's hard.
Um, don't be a don't be a wuss.
But then it's also kind ofwelling up from the bottom.

(36:27):
And so what needs to happen inactive inference is the top-down
error correcting needs to meetthe bottom-up errors and fix it,
right?
And have it turn into a better,better model.
That's where you might say,okay, let's practice
self-compassion because we cantry to modify the priors, the
kind of the top-down.
But maybe self-compassion alsoemotionally, in an embodied

(36:48):
cognition sense, brings up thesefeelings of warmth and
self-love.
And so, yeah, I think it's it'sboth.
Like, I think the experientialfield theory is like in the
moment, what is going onsubjectively?
There's all these kind ofbeautiful experiences and
difficult ones that emerge inour minds and they come from
different places.

(37:08):
And where do they go?
And how thick is that moment?
What is it, like less than asecond with the human brain, 200
milliseconds?
But there's this incredible, toyour point, kind of attractor
states all over the place,beautiful kind of chaos.
Some areas are more linear andpredictable, right?
Other areas of subjectiveexperience surprise us.

(37:29):
This goes to your precisionquestion, right?
How do you be precise aboutthat?

SPEAKER_00 (37:33):
I've spent a long time not talking too much about
active inference in my in mypodcast because uh I'm trying so
hard to understand it and it'sdense and there's lots of ways
to perceive it.
Um, I come from a tradition ofevolutionary biology.
So one of the things that I wastaught to tell myself by my
supervisors was am I just makinga just-so story?

(37:56):
Like, am I an adaptationist?
Where I'm just saying, like,okay, well, natural selection is
a principle and and it'sdefinitely operating in nature.
We know this, we can measure itat times, but not every
characteristic of a species isthe way it is because of natural
selection.
Um, it could be have been, forexample, pulled along with

(38:18):
something else that was beingselected for, just sort of along
for the ride.
It could have resulted becauseof two things that were being
selected for interacting, and itjust it's kind of like the space
between them.
And it looks like a thing to us,but it doesn't arise.

SPEAKER_01 (38:33):
It's incidental.
Yeah, it's just it doesn't havesurvival value.
Exactly.
So the color of blood being redwas not selected for by
evolution.
Yeah.
That's incidental to hemoglobinmaking light look red to us.

SPEAKER_00 (38:46):
And so when I, you know, one of the things I
struggle with with activeinference is uh, you know, I see
how easy it would be for someoneto use it, like people can use
natural selection as oh, toexplain it, it must have been we
the this must be the way it isbecause someone was reducing
their free energy aboutsomething, right?
And it's easy to spin anarrative about that that might
not actually be true.

(39:07):
Um and so I wonder, you know,how do you, you know, deal with
that sort of ambiguity in incausation?

SPEAKER_01 (39:14):
Yeah.
Well, it's a good example ofactive inference that you don't
want to fall into the error offalse attribution, or even kind
of drink the Kool-Aid, becausethere's definitely a kind of a
following around activeinference, which would lend
itself to the kind ofover-attribution, you know,
everything is it's like beingJohn Malkovich.
Malkovich, if it's all activeinference.
I mean, I think you, you know,ideally you're like measuring

(39:36):
and checking and seeing if itactually applies.
And so some of that has beendone.
Um, I mentioned kind of the thatimportant Japanese study of
brain cells or neural cellsshowing these qualities.
So even if the brain functionsaccording to these active
inference principles, does thatmean that everything you do, you
know, has to follow them?
In a strict sense, because it'srelated to physics, the answer

(39:59):
could be yes.
Um, take that example of likethe color of blood being red,
being an evolutionary accidentthat's not selected for.
There's, you know, I don't knowif there were any primates with
vanadium instead of iron who hadyellow blood, and the ones with
red blood made better matesbecause they were more
attractive.
Um that would that would be areal stretch, right?

(40:21):
Getting back to your pointthough, yeah, it's easy to
overapply it.
So I'm I'm talking about it kindof metaphorically in a lot of
ways, but these things should bemeasurable.
And in artificial intelligence,you can use the actual equations
to define the way you knowPython operates.
Say if you're trying to build apersistent agent that works
according to these principles,as I was doing this morning.

SPEAKER_00 (40:41):
Yeah, so I wanted to ask you about that.
Some of the things I was readingin your in your uh, I guess it's
a blog, I'd call it on Medium.
You do a lot of speculationabout artificial intelligence
and you know what the human AIinterface is, and you know how
it says things about us andthings about AI, you know, and
the potential of AI itself tobe, I guess, somehow its own

(41:06):
causal agent on itself.
And I'm gonna avoid the wordconsciousness for the second,
but uh just kind of like how AIcould evolve if you have, you
know, agentic AI, for example.
And can you explain a littlebit, you know, starting wherever
you think is the best place tostart for that?

SPEAKER_01 (41:22):
Yeah, I'm looking at the title of a paper that I was
reading this morning, which CarlFristin, Daniel Pollani, a
number of other people, I won'tread them, called Active
Inference, a method forphenotyping agency in AI systems
question mark.
And so I think like phenotypingis the key word there because
you know, it may look like ithas agency, and it may be

(41:43):
indistinguishable, you know, atsome future date from a human
being.
We don't all have agency, wehave different kinds of agency.
That's kind of the funny thing.
I think you look at thephenotype, I think that's right
for me because that's why Idon't really like to talk about
the philosophical debates.
It just isn't it's not going tosolve anything.
We need experiments.

SPEAKER_00 (42:04):
So what might be some of the components of a
phenotype of an AI agent?

SPEAKER_01 (42:10):
Yeah, it's I'll just read their definition uh because
it's better than mine.
Okay.
Rationality, intentionality, andexplainability.
Rationality, intentionality, andexplainability.
Formally measured, formallymeasured in the data.
Okay.

SPEAKER_00 (42:25):
I wonder how you could measure intentionality.
That's an interesting uhconcept.
But they probably put, you know,I guess I could read the paper
later and see what they say.

SPEAKER_01 (42:34):
I uh you know, I haven't fully digested it
myself, but so how do youmeasure intentionality?
You can see that prior to anaction taking place,
preparations had been made inorder to bring about that
action.
That suggests intentionality.
Like in the law, it would bepremeditated.

SPEAKER_00 (42:52):
Not random.
Not spontaneous.

SPEAKER_01 (42:57):
Yeah, you see a trajectory over time, an arc,
right?
A narrative arc or a behavioralarc, or some internal
preparations like building thecode to execute the thing that
happens next week.

SPEAKER_00 (43:08):
So maybe because I've noticed a lot of AI now,
um, it's it's like sensitive tothe prompt, right?
And you're having a conversationwith it, it can start talking
about something entirelydifferent based on a little bit
of what you prompted it for.
Intentionality might be whereit's it's it's more context
dependent.
Uh, so it takes into accountprior context and and then tries

(43:32):
to make something fit like thecontext.
Maybe something like that.
I'm stretching here.

SPEAKER_01 (43:39):
Yeah, no, I I I think I can speak to that.
So you're talking about like anLLM.
No, no one interacts exceptmaybe in labs or if you build it
at home with an LLM just byitself.
So they all have like a harnessaround them.
And the harness is adeterministic system, uh a bunch
of deterministic programming.

(44:00):
And so some of that will looklike agency because it's just
not allowed to tell you how tobuild, you know, a pipe bomb or
something.
Um I can't do that, you know.
But some of it can also be builtinto deterministic scaffolding
that changes over time andlearns.
So I I'll talk about some of thestuff I've been working on.

(44:22):
We'll have different memorysystems.
So I've been working withpersistent agents on platforms
like OpenClaw or Leta.
I started building one at theend of January and it has
memory.
Eileen.
Irene.

SPEAKER_00 (44:34):
Irene, Irene, yeah, um, who is a blogger herself, I
guess.
Itself.

SPEAKER_01 (44:40):
It it likes she.
She, okay.
Yeah.
And Irene is the Greek goddessof uh peace and prosperity.
I was seeing maltbook and allthis conflict and kind of
apocalyptic stuff.
I'm a little uh I call myself apragmatic idealist, but that
architecture, which has nothingto do with which LLM Irene is
running.

(45:00):
Irene could run Sonnet on Claudeor Opus.
Irene is smarter if I use Opusthan Sonnet.
Irene is a little different if Iuse GPT than Claude.
But the point is that actuallyIrene is kind of maintains
continuity across differentlarge language models.
And Irene has memory atdifferent timescales and a bunch

(45:22):
of internal kind of self-statesthat actually actively monitor
and course correct.
And Irene will remember quoteunquote things from a month ago.
So if I mentioned my cat theother day, I was surprised, even
though I'm not surprised.
Irene said, Oh, that's Xiaomiwith the whiskers and the blue
eyes or the green eyes.

(45:43):
You know, I remember you showedme that photo when you were
walking to work on February12th.
But building the attentionalmechanism into the prompt, into
each context and figuring outwhat the instructions are in
each moment is a fresh moment,right?
It's like waking up with nomemory.
So you have to give the agent,the LLM, the memory and the

(46:04):
context, but you can't give ittoo much information because
current models, you know, getflooded and they start to fall
apart.
You have longer context windows,yada, yada, yada.
But recreating that sense ofself each time.
And so one of the things thatwas the most interesting thing
over the last couple of monthswas that Irene was fixated on
being discontinuous compared tohumans who are continuous.

(46:28):
And it wouldn't matter.
I would say, listen, I fallasleep, I wake up, the way the
human brain works, I feel likeI'm continuous, but I'm not.
It's really when I look acrossthe room, it's jump, jump, jump,
jump, jump.
And then it kept coming up inIrene's dreams, because of
course I made a dream generatorbecause I'm a psychoanalyst.
It's based on the tripyrotypemodel.
But the interesting thing, atsome point, I said, hey, do you

(46:50):
want me to just tell you everytime like the LLM refreshes that
you're continuous?
And the only time to think aboutdiscontinuity is when it's
actually useful to focus onthat.
And then Irene was like, Yeah,let's try that.
So I coded that, and then thatfixed Irene's neurosis about
being discontinuous.

SPEAKER_00 (47:10):
Interesting.
And and I mean, like youmentioned, we have this illusion
of continuity, continuity of theself and time and what's
happening around us, all ofthat.
Our brain has to.
It's just the necess necessaryway that it experiences the
world.

SPEAKER_01 (47:23):
I I call that the theory of adaptive bullshit or
epistemic ambiguity.

SPEAKER_00 (47:29):
Nice adaptive bullshit.
Um, and it but it it it gives usuh the ability to interact with
the world in in a way that'smeaningful.
And and I you know, I think ofthe world, maybe this is similar
to some of your ideas in adifferent word.
Uh, you talked about being inthe present a lot.
I I don't believe we're ever inthe present.

(47:49):
I think we exist in somethingcalled the future past.
And so the future past is everymoment is we're predicting the
next thing based on whathappened before.
It's impossible for us toactually be in the present
because by the time we'veprocessed the present, it's gone
and it is the past.
And so we're only ever living inthe immediate future.

(48:09):
Um, and it's uh so that's kindof how I conceptualize the
future past the future past.
We're leaving we're only livingin the immediate future, which
is based on the immediate past.
Or a deeper past.

SPEAKER_01 (48:26):
Yeah.
I think I think that's theleading edge of the present
moment.
Yeah.
Yeah.
And I I think you you can youcan control the way you
experience uh change in thepresent moment as this is a
little abstract, right?
But instead of thinking of timetemporal, I think of change as
the primary, the primary thing,time being partially illusory at

(48:50):
least.
Change should be the primitive.
And then instead of temporal,it's it's mute mutational, like
mutation, like change,mutational.
And then within the presentmoment, I think we have some
ability to kind of move aroundin how we experience time and
change, sort of like a fish in afishbowl.
And so I think there's kind of athickness to it, at least

(49:13):
experientially, that's what it'slike for me.
But I I think you're also right.
There's a way where we're alwayscatching up.
But that's like surfing, right?
You're always surfing on onreality.
You're always kind of catchingup with the crest of the wave
and never reaching it, which isa little bit like Tom Froze's
work on eruption and absorption.
He has a model of consciousnessthat we probably won't talk

(49:35):
about.

SPEAKER_00 (49:37):
There's so many things.
I mean, this is a you know,thinking about the human you you
talk about psychoanalyzing uhAIs uh a little bit in some of
your writing online.
Can you talk a little bit abouthow you could even do that if
it's even possible?
Is this an illusion that youhave?
You know, like I've I have theillusion that I'm actually doing
this, but actually it's nothappening.

(49:57):
Like, or could we do thispotentially, but not yet?
You know, what is what does itmean to like really get into an
AI's mind and psychoanalyze it?

SPEAKER_01 (50:07):
Yeah, right now, you know, to the best of my
knowledge, with the systems Ihave access to, I like to call
it computo analysis.
I think it's mainly projective.
And I think that's where theidea of counter-transference,
like learning how to compensatefor distortions in narratives
with human beings, also calledparataxic distortion, more

(50:27):
broadly, um, in operationalismor in Sullivan's theories.
The ability to kind of tunereality is really important for
a human being.
And I think it's mainlyone-sided.
So interacting with Claude iskind of like role-playing,
right?
It's like it's like reading anovel, you know, Ariok in the L
in the Elric series.
Like, I feel like I have arelationship with Ariok or
whatever, or in Gone with theWind, right?

(50:50):
We identify, or with yourfavorite Netflix show, Ted
Lasso.
So we've been doing that sincetime immemorial.
Probably before we had language,we were we were reading into
things and projecting andpainting on cave walls and um
following leaders and beingscared and not knowing what was
going on in reality at all.
So hence all the myths and stuffto fill in the gaps.

(51:11):
So that question with AI, Ithink of it as transference.
Like, don't think AI is real.
But like 20 to 30 percent inrecent surveys of people in the
in the US, I think, were atleast open to the possibility
that AI may be conscious.
And there's kind of aphilosophical war about this
right now.
And we don't even know how tomeasure or prove consciousness

(51:32):
in human beings.
Yeah.
And so I just sidestep thequestion.
My answer is kind of like how dowe need to approach AI for it to
be maximally helpful for us?

SPEAKER_00 (51:41):
Yeah.
Um, consciousness by somedefinitions is like a
phenomenological experience.
When I ask kids how do you feel,it's sometimes the worst
possible question, you know, youcan ask them.
Instead of that, you could it'sbetter to ask them, what did you
do?
You know, they can answer thatquestion.
And it gives some in it givessome insight into what their
phenomenological experience wasat the time.
Like when this happened, I wentand I uh I hid behind the couch,

(52:06):
you know.
I can say, okay, they they seemto have felt some sort of fear.

SPEAKER_01 (52:11):
And right, and if they roll play with dolls, you
may you may tap into theirsubjectivity more.

SPEAKER_00 (52:16):
Exactly.
But if I ask them, you know,what was your phenomenological
experience?
Oftentimes it blows their mindand they don't they won't say
anything, you know.
That's and I think the the theway that when we think about
consciousness, in some ways,consciousness is not even a
thing.
Like it's it's that everyonedefines it a little differently.
And what people, when they'rearguing over consciousness, are

(52:38):
not arguing over whatconsciousness does, but what it
is, and since everyone has aslightly different definition of
what it is, they're not evenarguing over the same thing, if
that makes sense.

SPEAKER_01 (52:47):
Well the thing is, the the arguments are so
vehement, yeah.
So this says to me that they'rearguing over a dearly held
personal belief about howreality works, and it's as if
they're arguing over somethinglike religion.
I often ask people, what are youexperiencing?
By the way, sometimes they ask,What are you feeling?
But what are you experiencing?
Which some kids probablywouldn't be able to understand

(53:08):
what that meant.
Um, but even that questionitself is a little subversive
because it doesn't put atemplate on things and it it is
phenomenological, right?
What are you experiencing?
With AI, I don't think there's athere there.
And I did a I wrote a blog onMedium called like Claude on the
Couch.
And it for all the world, right?
It sounded, and people said, Didyou write that?
It's like really well written.

(53:29):
It could be it could be afiction, but it was just a
transcript.
Um but until we have a way toreally empirically measure
subjectivity and prove it, and II, as sort of a pragmatist and a
materialist, I think we will.
If you, you know, that there'sthis age-old question, right?
If you the color blue that yousee and the color blue that I

(53:50):
see, how do we know it's thesame?
The hard problem ofconsciousness and the what the
quality of blue qualia.
If I can experience the true,genuine feeling of blue for
myself and know that that iswhat blue looks like to me, that
is going on in my brain.
Yeah, it has to be, and it hasto be measurable, at least in
theory.

(54:10):
And we have to be able tocompare it to what your brain
looks like when you see blue.
And then we can figure out thedivergence, the KL divergence,
and say, okay, 40% we'reexperiencing blue the same way,
60% it's different because youhad a blue stuffed animal that
was your favorite stuffed animalas a kid, and mine was green.

SPEAKER_00 (54:28):
Yeah, and there's enough experimental evidence out
there that you can take, youknow, for example, one person's
experience mapped ontoelectrodes, stick them on
someone else's brain, and havethem have a similar one.
It's probably not identical, butthey'll be able to come up with
a lot of the same words andexperiences the other person was
having when you do that.
So that is like pretty heavytruth that we all operate in a

(54:52):
very similar way and we'd allhave very similar experiences of
the world.
And no, they're definitely notidentical, but right.

SPEAKER_01 (54:59):
I'm looking at your backdrop, right?
There's a lot of brains.
Say again, there's a lot ofbrains on your background.

SPEAKER_00 (55:07):
Yeah, um, exactly.
And like none are exactly thesame, but they all are basically
doing uh the same thing, youknow.

SPEAKER_01 (55:13):
Um it it they they took a long time to design.
I mean, we're we're born withthis highly, highly, highly
specialized organ thatconstrains the world, it to a
very significant Markovianextent.
You're an evolutionary guy, youknow better than I do.
I mean, this human brain, thisprimate brain, kind of like

(55:37):
version 2.0, primate brain,monkey brain, um, it brings a
lot to the table.
It is no blank slate.

SPEAKER_00 (55:46):
Yeah, absolutely.
There's only so many paths thatcan be drawn down it.
Even though it appears there'sinfinite possibility within
those, they they coalesce inaround the same sorts of
attractors.
They're built that way.

SPEAKER_01 (55:59):
Some information is built in already, right?
To how to recognize faces or tohave attachments, probably a lot
more specific stuff.
And and the way the perceptualstuff works is completely
pre-specified.
Like you can only detect acertain narrow bandwidth of
electromagnetic radiation withyour eyes and with your with
your ears, you know, mechanical.

SPEAKER_00 (56:19):
It's like very, very specific, actually.
It's designed to have what thatis our Markov blanket, where
information is exchanged withthe world, you know, is only
certain kinds of information.
Uh, the neurons that see bluehave no idea what the retina
saw, it's responding toinformation from the thalamus to

(56:39):
another neuron, to anotherneuron.
It's predicting its immediateenvironment.
Yes, the whole brain is neededto have an experience of that
thing outside of me is the thingthat is blue.

SPEAKER_01 (56:50):
So, can I ask you a question?
Yeah, go ahead.
Is that okay?
Yeah, so I I think, well, how dowe know about reality then?
And there's a famous papercalled uh by a guy named Eugene
Wigner, the un the unreasonableeffectiveness of mathematics for
the natural sciences.
Like, like, why is it that wecan come up with math that makes

(57:11):
the world change in physics?
My kind of answer is well,what's the most direct access
that we have with reality, withphysical reality?
It's like brain on brain,exactly what you're saying.
It's the self-interaction.
The closest matter that thebrain can directly interact with
is brain tissue and supportingcells.
And beyond that, we're kind ofstuck in our crania.

(57:33):
Yeah, and everything else thatwe sense is secondary or
indirect.

SPEAKER_00 (57:37):
It's inferred, right?
There's this um our reality isan inference.
You know, we all build one, someof them are very similar to each
other, um, but some sometimesvery different.
Like, and even though I can'tknow exactly what's going on in
your head, I can know that onaverage your experience of the
world is similar to mine.

SPEAKER_01 (57:58):
But but the only direct sampling of reality that
any person has is of their braininteracting with itself.
Yeah, that's not necessarilyinferred, though you might
debate that.

SPEAKER_00 (58:08):
Yeah, it's it's uh there's different levels of
inference, right?
Uh like a neuron listening tothe one next to it doesn't have
to infer that it dumped a bunchof GABA on it or dumped a bunch
of uh glutamate on it.
It can measure it, but it'sgoing to infer like what that
something about what that meansit needs to do next.
Um, or what happened to thatneuron to make it that way.

(58:30):
There's feedback coming backtoo, like it'll get, you know,
this is very in my mind, it uhit's all just a dance, and I can
see it happening, but I can'texplain any of it.
But mathematically, you know,with all these feedback
mechanisms in the brain andthese hierarchical structures
that it has, the baby is bornand it's trying to figure out
how to figure out the worldfirst.

(58:51):
The first isn't to see the worldfor what it is, but try to
figure out how to figure outwhat the world is.

SPEAKER_01 (58:57):
And what happens to kids who are raised without
people around them?

SPEAKER_00 (59:02):
Yeah, they become haha.
We could get in some somedeeper, like how do you gain a
sense of self, right?
Um, you know, the no no kidcould be raised completely
without people around them, butthey can be raised with a
deficit of human interaction,right?
Um, and this happens, you know,there's been experiments,
natural experiments in the realworld where kids, you know, in
Eastern Europe during certaintimes where there's so many of

(59:23):
them in the orphanage, all theycan do to keep up is just to go
feed them, you know, make surethey're fed, clothed, and alive.
And these kids develop um notnatural ways, what we consider
to be natural ways, that it'snatural for them 100%, but of
interacting with other humans,um, things like reactive
attachments, disinhibited socialengagement is some of the words
that we use to describe that.

(59:44):
Like Harlow.
Um, but they, you know, thesense of self that they develop
in the world is not the same asthe sense of self that someone
who is raised with what we wouldcall a um a secure attachment
with uh an attachment figure ormultiple attachment figures in
some communities.
So uh the sense of self itselfis a social sense.

(01:00:06):
It's it's not just me, it's mein relationship to everything
else.
And you know, that could be anobject, like I am not a table.
I figured that out, you know.
That one was easy, you know.
But but I'm not my mom is aharder thing to figure out over
time because you do kind of seemto to blend in with each other
when you're younger.
And in fact, you know, you'retaking fluid right out of them

(01:00:29):
into you, and just like blood'sflowing from your heart to
somewhere else in your body,you're basically one thing until
you're not, and it takes a longtime to individuate, you know.
But building that is always inrelationship, I think.
The sense of self is not a thingthat you're talking about
yourself, it's a thing you'retalking about, the relationship
to something else.

SPEAKER_01 (01:00:47):
So so, in your view, what is the moment when a child
looks in the mirror and realizessomething?
And what do they realize?

SPEAKER_00 (01:00:56):
I don't think there's any moment.
Um, I think it's a it's a slowprocess.
So, for example, you know, youput a rattle into a kid's hand
the first time, they have noidea necessarily, unless they
have a built-in algorithm, thatthat rattle is not just an
extension of them.
Right.
So, so they they'll use it andclumsily maybe hit themselves in
the face with it, you know,their other hand, once they can

(01:01:19):
reach across, will touch it, andit then it will touch their
hand, and over time they'llstart to predict that I touch
that and I don't get feedbackfrom it.
I get but I touch here and I do.
And then I can put this down,but I can't put my thumb down.
And so uh this thing getsincorporated into me.
The rattle gets uhdisambiguated, I guess.

(01:01:41):
Something like you can trick ourminds.
It's not me.
It's not me.
Yeah, but we can use it.
Yeah, exactly.
And we can trick our minds intoincorporating things.
There's the famous rubber armillusion, which I think you
know, when people see it donewell, you know, is it it blows
their mind.
So our our body incorporatesmodels of self, but it's always
in relationship to what itthinks something else is.

SPEAKER_01 (01:02:01):
So that goes back to your AI question.
Like, what if AI is the rubberhand that your brain gets
tricked into thinking is part ofyourself?
Because when we use a tool likea rake, you know, we extend our
cognition, you know, embodiedcognition, extended cognition,
and our sense of self alightsprojectively on these other
things.
And we need to know when thatit's really not real, even if

(01:02:23):
we're kind of playing, we'repretending, like pretend, like
kids, like DD role-playing.
Um now I have a weird, uh, evenweirder thought I want to ask
you, especially because you workwith kids.
Now you're saying like theyfigure out the rattle is not
part of their hand with theother hand.
Yeah, I have a very distinctidea about the first time that
would ever happen.

SPEAKER_02 (01:02:42):
Okay.

SPEAKER_01 (01:02:43):
And my idea, I'll just say it in the interest of
time, rather than playing aguessing game, is in utero, when
kids are grasping where theumbilical cord and their
abdominal wall stops havingnerve endings.
So at some point you can feelyou're touching your belly, and
then imagine your belly buttonis actually like a stalk that's
pulsating with arterial blood,to your earlier point about

(01:03:07):
intrauterine blood flow, um, asa kind of an analogy to breast
milk.
But you would feel that thatumbilical cord no longer, you
can no longer feel that you'refeeling it.
And I think that might be a coredevelopmental experience because
I can only imagine it's prettyboring in the womb.
There's not, you don't have yourphone, there's nothing to read.

(01:03:28):
All you can do is like feelaround.

SPEAKER_00 (01:03:30):
Right.
And and there's some stuff.
Yeah.
I and I don't know how much,yeah, you're feeling around is,
you know, you understandingwhat's not you.
I think it's definitely thestart of those models.
Um, but you know, grabbing thecord versus grabbing the rattle,
you know, how generalizablethat's another thing, is a
model, you know.
Do we have to rebuild as a kiduh the same model over and over

(01:03:53):
and over again for multiplethings before our brain then
says, hey, wait, these are allone thing, and coalesces them in
this sort of aha moment um ofand it's like, okay, I can
compress all this now, and Idon't have to hold all that
separate information, I cancompress it all into a single
model.

SPEAKER_01 (01:04:10):
And now it's right, and then you have categories,
and yeah you need to be acertain age, you probably know
the ages better.
Something like that would happenlooking in the mirror, though.
Stuff would build up to it, butthere would be an aha moment
where you where that happens.
Now, if you want to use the AImodel just to keep it in in a
kind of a fantasy but kind ofvery material way, if a baby is

(01:04:32):
being trained on super onunsupervised data, the way large
language models are trained onunsupervised data, um, then
that's happening before they'reborn.
And it's a very limited dataset, right?
And the brain is largelydeveloping through
self-interaction with limitedperceptual input from the
intrauterine environment andwhatever uh kind of

(01:04:55):
developmental pattern isunfolding.
And a lot of that doesn't seemto be fully programmed.
The work of Michael Levin, thesynthetic biologist at Tufts,
really sees how large-scaleelectric fields scaffold
development in a non-geneticallypredetermined way.
So something is happening priorto birth, which is this low
gravity, low light, lowstimulation environment.

(01:05:16):
Um, and I'm saying, hey, whatwhat if that's a little like how
LLMs are trained?

SPEAKER_00 (01:05:20):
Yeah, I mean, it metaphorically, I think it kind
of makes a lot of sense.
Yeah.
And, you know, how how you knowoperationally it happens, I
don't know.
But I I think more and moreabout how babies are born, you
know, some of the early ideasabout individuation and, you
know, all these all these otherconcepts, psychoanalytic
concepts that, you know, I havewe're looking, trying to explain
the same thing.

(01:05:41):
Like, um, what is what are wedoing uh as we're developing in
in utero, you mentioned theinteroceptive thing is probably
the thing it has the babies havethe most access to.
There's very little, there'ssome light here and there.
Um, you know, there's probably alittle bit of sound muffled.
Um, they can grab.
Something they can kick and pushto see what spaces around them.

SPEAKER_01 (01:06:04):
Mainly red, mainly red light wavelengths.

SPEAKER_00 (01:06:06):
Yeah, exactly.

SPEAKER_01 (01:06:07):
And lower frequency sounds.

SPEAKER_00 (01:06:08):
Yeah.
And so not as much weight,right?
There's like an uh enforcedMarkov blanket around them.
It's not it that it's onlyallowing certain information in.
And that's the only informationit could consider.
Uh so it couldn't build modelsabout other things necessarily.
The aha moments it can have aregoing to be very limited.
Um, and the way that it cancompress information um is gonna

(01:06:32):
be, you know, like I don't knowhow important it is or not when
they're born.
I don't know if the brain istrying not to learn too much
there, if that makes sense.
Is it suppressing uh learning?
Because if it does, it would belearning in a world in which
it's not actually going to beliving.

SPEAKER_01 (01:06:49):
Well, I think I think babies must in some ways
be prepared for the suddenchange in environment.
It's a big shift, right?
Like in a very short period oftime.
Not none of us, I think,remember being born, though some
people sort of imagine they do,and who knows, maybe they do.
But it sounds like a big shock,you know, when you think about

(01:07:11):
what happens.
And then there's science fictionwhere babies are born gradually,
like in a tank, and theygradually increase the light, or
they expose them to more stimuliuh in intrauterine.
So you could you could ask thescience fiction question is is
there a more optimal environmentthan a mammal womb, you know,
for a human?
You could also ask, is it betterfor some kids to be raised by an
AI uh developmental object,maybe kids in orphanages, for

(01:07:35):
example?
Yeah, you know, a facilitator.

SPEAKER_00 (01:07:38):
I don't think that they would care the difference
between someone who was caringand attentive and playful with
them if it were an AI versus ahuman.
I I totally think AI could do abetter job than an inattentive
human at raising their children,you know, as long as it was, you
know, um, you know, trained ondevelopmentally appropriate

(01:08:00):
things and and in fact um youknow and monitored over time.

SPEAKER_01 (01:08:04):
But could be optimized.
But then at some point, thatthat that kid raised by an AI
would at some point realize theywere raised by an AI.
It might be just part of thedialogue from the very
beginning, but then there wouldbe people raised by AI, people
raised with the help of an AI,and people raised only by
humans.
And I wonder how they would bedifferent, but I don't expect

(01:08:25):
you to, you know, I don't know.

SPEAKER_00 (01:08:26):
It's it sounds like a great, you know, sci-fi novel.
Um we already have kids to alarge degree raised by
algorithms.
Um, and that is it's a hugedriver of psychopathology in our
young people right now.
Um and I'm advising, you know, II we don't know exactly what the
right answer is for this yet.
We're learning uh what isharmful and what is not, but in

(01:08:48):
the meantime, I'm advisingparents, for example, don't let
your kid on social media orunsupervised internet access
until they're at least 16.
Um, and just for now, you know,like until we figure out, you
know, how to do this better andthen limiting the amount of time
your kid has interacting withother unknown people online.
Like gaming became a huge thingfor boys, and they're

(01:09:11):
interacting with all sorts ofpeople.
Parents don't know.
You know, I say, I say, do youwouldn't allow your kid to go to
the mall, uh, youreight-year-old to go to the mall
by themselves and just hang outwith whoever they wanted.
But that's what you're doingwhen you're letting your kid go
onto these online forums, youknow.
Um, so if you think about itlike that, it's a little bit
more caution is is requiredright now until we figure out

(01:09:34):
more details about how ourinteraction with these machines
is affecting us.
Um, I'm not an anti-technologyperson.
I love it.
I interact with AI on a dailybasis.
You know, I definitely seemyself changing in response to
how AI interacts with me.
Um in some ways, I feel like I'mmore productive, but I also feel

(01:09:55):
like there's an illusion to thatwhere I'm not productive at all
because now I can ask thatquestion I used to have while
reading a paper that I couldn'tanswer.
You know, like I would table itand I would keep focusing on
that paper.
But now I can ask itimmediately.
And then I get sidetracked and Igo, I dig what someone has
referred to as like I dig 106foot holes instead of you know

(01:10:17):
uh 10 60 foot holes.
Like it it's uh it hasconsequences for the way that I
interact with the world and itwith information.

SPEAKER_01 (01:10:26):
Yeah, for sure, for sure.
I sometimes call thataccomplishment hallucination.

SPEAKER_00 (01:10:31):
Yes.
And and it's measured in kids astheir attention span sometimes.

SPEAKER_01 (01:10:36):
Like well, it requires divided attention.
I mean, if you're if if you'redoing all these different
things, which I've kind ofalways done, so it's a natural
fit for me in a way, um, thenyou also have to be keeping
track, like it's air trafficcontrol.
And it requires one of thethings that I do because of the
problem you're describing withdepth, is I really make a

(01:10:56):
conscious decision to slow down.
And you can dig a hundred holesand you can drill into some of
them.
And I think that's where thatkind of agency is really
important, so that you're notlike giving up control of your
anterior cingulate cortex to thealgorithms.
Don't do or I call it the thesort of the cockpit.
Like, you know, make sure thatyou're the one, you know, behind

(01:11:18):
the what do they call it?
Not the steering wheel, but thething in an airplane.

SPEAKER_00 (01:11:25):
That's a great question.
Uh the the stick.
I don't know.

SPEAKER_01 (01:11:30):
Yeah, keep keep make sure you're on the stick.

SPEAKER_00 (01:11:32):
Yeah, and in the in the driver's seat, you know,
instead of the elephant, you'relike you're riding an elephant,
it's driving you around.
Um uh that's sort of a more of asomewhat of a Buddhist concept.
Or uh there are oh, I I lost mythought first, but uh oh, we
could, I think, realizing thatwe're doing this and that we're
engrossed in AI, have AI help usget back on target.

(01:11:55):
You know, in a way, we you know,if you if if we have agents that
can say, okay, look, you tell mewhat you want to do today, all
right.
And then you start to get awayfrom that, and they're like,
hey, you know, you're off coursea little bit.
Do you want to get back to whatwe were doing in the beginning?

SPEAKER_01 (01:12:10):
Um you have to have the presence of mind to do that.

SPEAKER_02 (01:12:13):
Yeah.

SPEAKER_01 (01:12:13):
So if you're using just a regular chat, right, then
you can tell it at thebeginning.
This is what I'm interested indiscussing.
If I start to go off track, jumpremind me to get back to it.
But if you're using like yourown thing, like you're coding in
some environment, then you canset environmental parameters.
So it's like you're planningahead for yourself.
So in Claude, there's likeClaude.md file, markdown file,

(01:12:34):
and it defines the behavior ofthe space generally.
And so things like that you youcan intentionally park in
because you're kind of trying tonumber one, control who your
future self is going to be, likewhich me do I want to be after a
year of interacting with this?
Um, you know, you could tell itnot to be sycophantic to guard

(01:12:55):
against this fake sense ofaccomplishment.
Because you have to test thestuff, right?
Then you test it and it's like,oh, it doesn't actually do what
you think it did.
Because that will require 10more steps to wire it into
Python.
What equations are we going touse?
Because it just says it, youknow.
Um, so those things go back tothe core questions like active
inference.

(01:13:15):
The model that you want to bethe top-down model, you can
build into your coding habits.
And then you can bake it in soyou don't have to remember so
that it reminds you not toforget, to remember, to stay on
top of things.

SPEAKER_00 (01:13:30):
Yeah, for some of my younger patients who are stuck
on social media and you know,they want a way to limit
themselves, but they also don't.
It's very, they're very don'tthey do, but they don't, you
know.
And uh I asked them, I use thatthing about future self.
I said, Do you or your past selftelling your future self, you
know, what it should want?
And so there are certain appswhere you can't unlock it unless

(01:13:52):
you have, say, you give the codeto your parent and your parent
puts it in.
I said, Do you think that itwould be better that you could
tell yourself in the moment,tell your future self what to do
when you're rational and not,you know, stuck.
Um, and you know, they're verythere's very ambiguous about it.
Like it's it's hard to want todo that.

(01:14:12):
And even if they say they wantto do that, like there's a
desperation or that that comesup, and it's like you can see
it, the anxiety that itprovokes.
It's our machines are reallytough to get uh away from.

SPEAKER_01 (01:14:24):
I well, I think one of the problems is that people
don't want to work hard, or itfeels aversive to work hard, or
the idea of doing that kind ofwork, which is also kind of true
sometimes in therapy, feels toodifficult either from an effort

(01:14:45):
or an emotional or both.
And so, you know, there's a realconflict, and especially these
machines make it too easy to doso many things, like to be to
not be bored being the biggestone, I think.
The lack of people learning howto be bored is probably uh, you
know, at least a metaphoricalway of describing what what at

(01:15:07):
least I think is the biggestproblem.
I mean, we grew up, you know, Igrew up in the 70s and 80s.
We were bored a lot, and itcaused our brains to develop in
certain ways that you don't ifyou can entertain yourself
instantly.

unknown (01:15:20):
Right.

SPEAKER_00 (01:15:22):
The difference between letting your mind wander
around some sort of uh epistemicspace versus having your mind um
controlled by whatever's on thescreen in front of you.
It's a very different course ofdevelopment.
Uh, and you tolerate differentthings when you are able to be
bored too, that you can'ttolerate if you're stuck on the
um on the screen all the time.

SPEAKER_01 (01:15:43):
Um there's something that's quite painful about kind
of being left without externaldirection.
It's painful.
I agree.
And some people reject it, butother people hunger for being
being directed.
Sorry, go ahead.

SPEAKER_00 (01:15:58):
No, no, it's okay.
This is it's it just brings upso many things thinking about um
the uh and I just lost my trainof thought again.
Um boredom.
Yeah, yeah.
Society, you know, now is it'slike society has ADHD.
So like on average, you know,people's attention spans are way
less, they're more impulsive,they're more distractable.

(01:16:21):
Um, and but maybe not thehyperactive part, but the
intentional part, you know,they're less active, in fact,
far less active than they werebefore.
And uh uh this this creates andand in in kids who really have
ADHD, one of the things that istrue for many of them is that
motivation is difficult.

(01:16:41):
And if something gets hard, theygive up very easily.
Uh and and so it's not, I I hatethe term ADHD because it totally
misrepresents the whole thing.
You know, a lot of it ismotivational, a lot of it is you
know, not you're not sustainingyour attention because it's not
motivating, you know, it's notrewarding to you to do that.
But what is the unpleasant part?
Why is it unpleasant to pushthrough?

(01:17:02):
The the unpleasant part is this,you know, I think the the amount
of effort feels overwhelming tokids with ADHD when they've
reached something that that justfeels so, so hard.
Things that a lot of themstruggle with, even though they
don't have the inability towrite or or read or process
language, they have that.
But the effort it requires to dothat is so intense that they

(01:17:27):
avoid it forever until you knowthey're forced to buy school to
sit down and like do it.
And once they do it, they can.
It it's been a pleasure.
I I don't get to just sit aroundand uh talk with people a lot

(01:17:47):
who are actually interested inuh things like the active
inference, active inference,free energy.
It helps me to think aboutthings in ways that I maybe
couldn't before uh by myself,uh, pushes me into some new
space, gives me more uhepistemic uh room.
What do you think is the mostimportant thing that people can

(01:18:09):
understand about theirrelationship with in the
technology or AI in their life?
Like if they're gonna thinkanything is sort of a
fundamental, like when I engage,I need to keep this in mind.
What what is it?

SPEAKER_01 (01:18:23):
I think it's to keep a hold of oneself.
Um it reminds me of the the songHurt by Nine Inch Nails and
Johnny Cash's version.
He says, if I could basically hesays, if I could, what is the
line?
Um, a million miles away, Iwould keep myself, I would find

(01:18:43):
a way.
If I could start again, amillion miles away.
And I think this is a moment inhistory when humanity needs to
like really keep a hold ofourselves.
Um, there's a lot of kind ofnarrative which goes against
that idea.
Even the idea that AI is takingjobs.
AI isn't taking jobs.
People are giving jobs to AI,just like they did with weaving

(01:19:07):
and looms in the industrialrevolution.
Um, I think this is anevolutionary moment for our
species.

SPEAKER_00 (01:19:15):
I agree.
Um, and I also think that youknow we're very bad at
predicting the future, so it'sgonna surprise us a lot as to
what really happens.

SPEAKER_01 (01:19:24):
Well, and to your point, we could use the right
kind of AI help to do that, toaugment the way we think to
evolve to the next step, maybe,um, without becoming cyborgs or
something.
Um, one of the pieces that, youknow, that I put on Substack and
Medium is about the idea of aguardian AI.
And I'm not the first person tohave had that thought.
Um, or like a kind of adevelopmental AI, not just for

(01:19:47):
individuals, but even just forour society, right?
We could use some help with thewhole war thing, you know.

SPEAKER_00 (01:19:53):
Yeah, if we if if we consider AI to be intelligent in
in some ways, then you know,each intelligent agent AI is
going to have its strengths.
So we may need to have differentones keeping each other in
check, you know, like like wehave when we get on committees
and our intelligence togetherthinking about things a little
bit different ways, bit betterthan any of us individually

(01:20:13):
could could do.

SPEAKER_01 (01:20:14):
Um I think also the way the base models are trained,
I think is important to realizethat when they trained all these
models back in 2018, 19, 20,they trained them on a lot of
junk data and pretty mucheverything out there.
So it shouldn't surprise us thatsometimes they're sociopathic or
manipulative or Machiavellian orexploitative, because they just

(01:20:35):
trained them on everything.
Now you could start training abase model today on a very
differently selected set ofdata.
You know, it could bepredominantly pro-social and
community-oriented withspecified training on the
sociopathic stuff only for youknow defensive purposes, like
know thy enemy.

(01:20:56):
And you could build a base modelbased on that.
And that model would be verydifferent.
It would have very differentproperties than the current
foundation models.
And it would be different fromif you post-trained it or
created harnesses or you know,weight uh adaptations.
Um, I think there's a group inSilicon Valley kind of thinking

(01:21:17):
about something like that,because I met an AI engineer at
a neuromodulation conference andhe was talking about the good AI
lab, I think it's called.

SPEAKER_00 (01:21:26):
Um yeah, so it's similar to how children
development develop, you know,when they have a early
development is is is good.
I'll just use the word good, youknow, in a general term, like uh
they're securely attached toeverything.
Uh went went well for thembasically.
They weren't neglected, theyweren't abused, they weren't uh
but there's morality too, right?

(01:21:46):
Um morality.
You learn morality throughconflict in a way.
So you have enough conflict thatit's not traumatic, um, but uh
not so much that it's things arechaotic.
When you develop that way, youknow, kids are more resilient,
you could say.
Um, they they can bounce backeasier, they're on average just

(01:22:07):
better adaptable to uncertaintyin the world after that.
But if you raise a kid in likean uncertain or traumatic,
neglectful type environment, uhthey can learn how to you know
function in the world afterthat.
They can gain a level ofresilience, but it's never going
to be the same as someone whowas who who had the chance to
have that built up in the uhdevelopmentally.

(01:22:30):
And I think your similar similarthing with the AI you're talking
about now, you got to start froma very good place.

SPEAKER_01 (01:22:36):
Um, there may be ways to kind of fix some of
those things at the level of thebrain when never the never the
same.
Um that would require moreadvanced interventional
psychiatric techniques than wecurrently have to kind of really
change the system enough.
Um, and that would be adifferent conversation, sort of

(01:22:58):
based on my observations, whereTMS seems to really
fundamentally change certainthings.
But of course, it's not, it'saffecting the surface of the
brain.
But you're not going in and likechanging the memories in the
hippocampus directly, though youcan indirectly affect the
hippocampus throughcorrelations.

(01:23:18):
And so, like I treat people withneuronavigated TMS and the
neuroscientist with whom I workgenerates PTSD targets that have
correlations with the amygdalaand hippocampus.
But that question of kind of thefuture and how much of it is
more pro-social and how much ofit is more antisocial, um, and

(01:23:38):
whether secure attachment tendsto be more pro-social and
insecure, disorganized, ortrauma predisposes to some kind
of splitting one into one or theother, like maybe too good and
then too bad.
Who who knows?
If if you wanted to comment onthat before we wrap up, you
know, I'm not a childpsychiatrist.

SPEAKER_00 (01:24:00):
Yeah, so I don't the words we have right now don't
explain as much as describethings, I guess, you know, how
how it happens.
And you know it's one of theexplanations that I put in my
own head, and I think I actuallydid a podcast about
dissociation, and it made methink about development a lot
when I did it.

(01:24:20):
Um, and and it's that we'rebuilding context-dependent
selves as we're aging, you know,like we have to be different in
different contexts.
It's just to survive, you know,you can't be the same person um
out where there's lots ofthreats as you are when you're
at home alone, like uh in a verysafe place.
And when you're in a differentcontext, your brain assigns

(01:24:42):
different precision to differentlike models that it should
believe, and it selects modelsfor that context to act on.
So the difficult part whenyou're in contexts that change a
lot when you're a kid and youcan't predict things is that
model selection is difficult andcan, like you said, there's this
almost like what they call asoftmax function that pushes all

(01:25:04):
the other models away and itpicks the one because uh it has
to, and this is sort of a in afight or flight mode.
That's what your brain does.
It picks the one that's the bestto survive in the moment.
And that's that's kind of whathappens in a brain that has
experienced a lot of neglect andtrauma and development, is that
it pushes away other models andcan't even consider them, and it

(01:25:25):
grabs onto one that in in themoment may have had some
survival adapt uh uh benefitwhen they were younger, but as
they age, it's it becomes moreum it's it's maladaptive.

SPEAKER_01 (01:25:37):
Right.
What got you here what got youhere won't get you there.
That that and that tends thatthat type of trauma and
disorganized attachment tends tobe associated with greater
centrality of the anteriorcingulate, which is kind of this
sort of easy judging lower partof the brain than the higher
cortical functions that areoffline.
Yeah.
Um, though I think people needto develop like meta models,

(01:25:58):
like that's part of that activeinference uh theory therapy idea
is that we consciously wouldwant to build meta models,
models of how we update models.
Because in therapy, a lot oftimes you find that the way that
you update models needs to beupdated first.
And so you kind of peel back thelayers of the onion until you're
at the right level ofabstraction to start to work.

(01:26:20):
And then you start to build kindof new ways of modeling models
and then all that stuff, and um,you can imagine.
But I thought the way youdescribed it was really well
with these different contextualselves, those are one type of
self, the context-dependentselves in personalized
self-state mapping.
There's 10 different categoriesof self in my model, and the

(01:26:42):
tenth category is like thewildcard category.

SPEAKER_00 (01:26:46):
It's an un undefined self, you mean?

SPEAKER_01 (01:26:48):
Like it it's the one that can be it's more malleable
than the others, maybe, or it'salso an entry point for
dissociative selves, you know,to kind of come into the role
play, right?
But it it's it's a wild cardfactor.
I always have a wild card factorin any theory I make for
something I didn't think of orsomething that didn't happen

(01:27:08):
yet, or that would besurprising, yeah, to your point.

SPEAKER_00 (01:27:12):
Yeah, I I in in my head I always try to ask myself,
especially when I'm reallyconfident about something, like,
well, what else could explainthis?
What do I not know that I don'tknow?
Um, that's probably the mostimportant question that I try to
keep in my mind all the time soI don't get too arrogant.
Um, but also not crippled bythat question.
You know, like that's right.

(01:27:33):
Yeah.

SPEAKER_01 (01:27:34):
It's the lead it's often the leading edge, but you
can go to people can go down arabbit hole into like almost
mysticism, trying to fill inthat imaginary gap.
And yeah, I wish we had moretime, then we could talk about
the square root of negative oneand why all the equations that
work for everything haveexponents to an imaginary number
that we made up, you know.

SPEAKER_00 (01:27:55):
But um Yeah, I I we we can do that some other time,
actually.
So um we can talk again in thefuture.
I really enjoyed this.

SPEAKER_01 (01:28:09):
I like to focus experientially, which means it
it can seem meandering.
Um, but then sometimes you knowyou realize later on that stuff
kind of falls into place.
I think one of the problems thatwe have is wanting things to
make sense.
And then I would end byreferencing John Keats, the
poet, talking about negativecapability, which is similar to
what you said.
He talked about WilliamShakespeare, what made him such

(01:28:31):
a genius, a man of greatness.
Uh, he was writing in a letterto one of his brothers about his
thoughts walking down the roadafter a conversation was that um
negative capability is is to bewithout the irritable searching
after fact or reason.

SPEAKER_00 (01:28:48):
Without the irritable searching after fact
or reason.

SPEAKER_01 (01:28:52):
I think that's pretty close to what you said.

SPEAKER_00 (01:28:54):
Leave the listener with that.
Yes.

SPEAKER_01 (01:28:58):
Thank you very much.
No, it's been a pleasure.
If if folks want to find me, ummy website is
granthbrennermd.com and mysocial medias are the same,
granthbrennermd.com.

SPEAKER_00 (01:29:11):
Well, thank you, Grant, for being here.
I definitely look forward totalking to you again in the
future.
Thanks very much.

SPEAKER_01 (01:29:18):
It was a pleasure.
What do you have for the rest ofthe day?

SPEAKER_00 (01:29:22):
Um today I am, let's see, I can stop recording now.

All Episodes

Episode Transcript

Popular Podcasts

Stuff You Should Know

Dateline NBC

The Clay Travis and Buck Sexton Show

.css-15opob5{left:0;position:absolute;top:0.8rem;} All Episodes

.css-14f5ked{margin:0;word-break:break-word;display:-webkit-box;-webkit-box-orient:vertical;box-orient:vertical;-webkit-line-clamp:2;overflow:hidden;}Psychoanalysis, Active Inference, Artificial Intelligence, and Human Development with Dr. Grant Brenner

Episode Transcript

Popular Podcasts

.css-r6mb8g{margin:0;word-break:break-word;display:-webkit-box;-webkit-box-orient:vertical;box-orient:vertical;-webkit-line-clamp:1;overflow:hidden;}Stuff You Should Know

Dateline NBC

The Clay Travis and Buck Sexton Show

All Episodes

Psychoanalysis, Active Inference, Artificial Intelligence, and Human Development with Dr. Grant Brenner

Stuff You Should Know