November 12, 2025 • 59 mins
🎬 From film school dreams to the operating room reality, Dr. Colin Son’s journey is anything but ordinary. In this episode of Med School Minutes, he shares how he went from aspiring screenwriter to board-certified neurosurgeon, and how his creative background now shapes his work at the intersection of medicine, AI, and entrepreneurship.
💡 Tune in to learn:
- How nontraditional paths can strengthen a medical school application
- Why AI is transforming healthcare without replacing doctors
- How storytelling skills translate into better patient care
🎧 Watch now and see how following curiosity, not convention, can lead to extraordinary careers in medicine.
#SaintJamesSchoolOfMedicine #SJSM #MedSchoolMinutes #AIinMedicine #Neurosurgery #FutureofMedicine #MedicalInnovation #HealthcareAI #MedicalEducation #DoctorsandAI #NontraditionalPath #MedicinePodcast
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
SPEAKER_00 (00:01):
Hello, and welcome to another episode of the Med
School Minutes Podcast, where wediscuss what it takes to attend
and successfully complete amedical program.
This show is brought to you bySt.
James School of Medicine.
Here is your host, Kashik Gua.
SPEAKER_03 (00:20):
Welcome to another episode of Med School Minutes
where we talk about everythingMD related with a focus on
international students,specifically students from the
Caribbean.
My name is Koshik Guha, andtoday we're joined by a unique
guest, Dr.
Colin Sun.
He's trained as a neurosurgeoncertified by the American Board
of Neurological Surgery,specializing in minimally
(00:40):
invasive brain and spine care.
But Dr.
Sun didn't stop at surgery.
He's now moving intoentrepreneurship, applying
artificial intelligence totransform patient access to
physicians.
In this conversation, we'll diginto his transition from
medicine to business, explorewhat real-world AI means in the
neurospace today, the challengeshe's facing, and his vision for
(01:02):
where things go next.
Let's turn to Dr.
Sun now.
Thank you so much for your time,Dr.
Sun.
Really appreciate uh you makingtime to come on our podcast.
Uh, why don't we start by yougiving us a little bit about
your background and uh yourjourney in the field of
medicine?
SPEAKER_01 (01:20):
Yeah, that thank you for having me.
Um I really, really appreciateit and uh appreciate what y'all
y'all are doing with thispodcast in terms of you know,
uh, you know, I'm sure it's amarketing tool, but it's also
seems super educational formedical students and and
especially pre-med.
Uh so my name's Colin.
(01:41):
I am a neurosurgeon.
I'm up in Texas.
Uh I uh uh am also anentrepreneur, so I've gone
part-time clinical.
I've totally jumped on uh kindof the AI hype train and how I
think it's gonna, you know,really, really change uh how we
(02:06):
train physicians, how we educatemedical students, how we access
care, how we pay for care, howwe regulate care.
Um uh and so I have uh a companytrying to get off the ground in
that space.
Um, but still practicingneurosurgery part-time.
Uh I came to neurosurgery verycircuitously.
(02:27):
I went to film school.
Okay out in out in Californiaand tried to be a screenwriter,
and uh uh it quickly becameclear uh maybe uh maybe some of
those things like selling myselfor or maybe even my talent as a
(02:47):
writer were were gonna leave mewaiting tables.
Um and so I started, I was like,well, I'm pretty good at school.
unknown (02:54):
Right.
SPEAKER_01 (02:55):
Maybe I'll do something, uh go back to
professional school and andlooked into um, you know, uh
business school and law school,applied to both, applied uh to
medical school uh for the samereasons.
And while I was doing all that,I shadowed a neurosurgeon and uh
had never been in the OR before,uh, didn't really watch a lot of
(03:18):
ER as a kid.
Um, but uh that was I was like,oh my goodness, this is um this
is really impressive.
So I came into medical schoolpretty much knowing I wanted to
be a surgeon.
It was the whole reason I wasthere, and uh led led to this.
SPEAKER_03 (03:36):
Wow.
So I I do want to focus a littlebit on on about this uh uh film
writing uh aspect.
That is so interesting.
Um and and and one of thereasons I do want to focus on
that is because this would betruly inspirational.
Because um what I hear,especially in the United States
for more mainstream medicalschools, uh, is that if you have
(04:00):
uh if you vacillated on yourcareer, that is typically not
the ideal candidate for a lot ofUS medical schools.
Um however, it seems like youhave obviously turned that
around and you were in LA atsome point, waiting tables,
trying to get into Hollywood.
But then in my opinion, I wouldthink that neurosurgery is
(04:22):
probably should have been a muchdifferent difficult path than
trying to get break intoHollywood.
But you obviously chose theharder path.
But I do want to focus on alittle uh hear a little bit more
about that.
Can you tell us a little bitmore about the how you you went
from film writing or scriptwriting to um uh neurosurgery?
SPEAKER_01 (04:43):
Yeah, so um so I I went to film school.
Um and uh and it was uh it was afour-year program.
I was I was already getting mybachelor's, um, but I was not
looking at medical school.
I was not pre-med.
Right.
In my junior year, I was like,oh my goodness, um uh this is uh
(05:09):
this is not, you know, I'm notnot sure uh trying to do this
long term is for me.
And so like I said, I startedlooking into all these schools.
When I when I started doingthat, um, you know, the second
semester of my junior year, Iactually took a biology class,
which counted as a prereq justin case.
(05:31):
I was also like, you know, Itook like a pre-law class, like
my electives were uh, and thenum when I decided that year, I I
really wanted to try for medicalschool.
Um I did uh I did summer work ata local community college, um,
which allowed me to get a lot ofcredits and then uh more more
(05:56):
kind of prereqs work.
Um, you know, I I hear whatyou're saying that it is so
competitive nowadays, uheverywhere, and that uh you know
I think I definitely had someknocks against me doing doing
some of my prereqs hurriedly,like at community college.
(06:18):
Um, but uh I also you know,maybe it was just the way I kind
of went into the process, but Ialso kind of viewed it as um,
you know, this kind of atypicalpath as uh as a real selling
point, honestly.
(06:40):
I don't remember getting a lotof questions like in interviews,
like uh, you know, people wantto know, you know, as they
always do, why you want to be aphysician.
But I uh I really, you know,when I went into these
interviews or even my personalstatement, um I really kind of
(07:00):
pitched it as uh as like aunique path, right?
Uh this is part of the reasonI'm sure about medicine.
These are the experiences I haduh before deciding on medical
school.
This is how I think my creativeside um makes me more empathetic
(07:22):
uh and uh and ideal for dealingwith patients.
And so um, you know, I I think Ihave lots of friends who took
very non-traditional paths, youknow, time off, uh worked, uh
had careers before beforemedicine.
And I I think sometimes whenyou're applying to medical
(07:43):
school, that can be, you know,when you're selling yourself,
that can be a real benefit.
SPEAKER_03 (07:48):
Okay, well, I mean, uh that's really interesting.
So it almost sounds likespending and and the film school
was, I'm assuming, full fouryears.
SPEAKER_01 (07:56):
It was, yeah.
SPEAKER_03 (07:57):
Yeah.
So film school gave you thetools to write a strong story
that strengthened yourapplication.
It probably did.
Yeah.
Because I I will say, I mean,uh, I've I've spoken to I speak
and interact with a lot ofphysicians, and obviously the
physicians who make it to uppermanagement are very, very strong
(08:18):
people persons.
Um, uh, I think without thatthey can't really get to
business side.
Yeah.
Yeah.
Um however, also I will caveatthis by saying people who are
hospitalists aren't necessarilythe most uh people oriented, uh
generally speaking.
I mean, and I don't mean anydisrespect, I just think that uh
(08:40):
it's just not a skill that uh isabsolutely essential to be a
good physician.
So it's compared to otherprofessions, I would say that
that tends to be a little moreuh rudimentary for most
physicians.
But it it sounds like um yourexperience in film school really
(09:01):
bridged that gap and made you abetter, more holistic uh
candidate.
Would you agree or disagree tothat?
SPEAKER_01 (09:09):
I think I think I did.
Um I am uh uh I'm a prettyintroverted, kind of nerdy, uh
you know, not I mean I've gottenbetter over time, but like uh if
I had if I had been going intomedical school interviews,
(09:30):
having been a pre-med, havingbeen only around pre-meds, just
with my social skills uh comingout of high school, I bet I
would have been a worseapplicant than having gone
through this uh the field schooland um yeah.
Okay.
SPEAKER_03 (09:49):
So uh going back to today, where as you mentioned,
it's incredibly competitive.
Um medical school, generallyspeaking, is incredibly
competitive.
Um and I feel like a lot of themedical schools generally really
focus on reasons to rejectstudents.
And if you don't have a moretraditional pathway, from what
(10:12):
I've seen, at least nowadays, itseems a harder task.
You don't necessarily see peoplenowadays saying, I went to film
school, or as you said, you hadfriends.
Um and and I feel like thismight have been a paradigm shift
sometime in the mid-2000s.
Um and do you think that todaythe programs are probably doing
(10:35):
a disservice by looking atsomebody who's done pre-med,
then they want to go to medicalschool, and then they're looking
for residents, uh, especially inthe uh uh very elite specialties
like neurosurgery, orthopedicsurgery, ENT, etc.
etc.
Uh, what are your thoughts onthat?
SPEAKER_01 (10:51):
Yeah, I can imagine that really being the case.
Um uh you know, for people whoare non-traditional, um you
know, the path I kind of tookand probably would be my advice
uh is to really lean into it.
(11:13):
Um I d I you don't need to bedefensive about it, but I I
think these non-traditionalpaths can certainly uh make
people better physicians.
And if you're able to articulatethat argument like in your
interviews, like in yourpersonal statements, like when
(11:34):
you ask for letters ofrecommendation, like discussing
that with your you know, uh withyour mentors.
Um you know, I I I don't thinkthese are things that should be
you should have I I agree thatsometimes it probably feels like
you have to, but uh that youshould try to excuse.
(11:55):
Um, and I I think they can be astrength, really.
SPEAKER_03 (11:59):
Okay.
Wow.
So, you know, this is somethingwe tell our students all the
time.
Everything in this world isdetermined by storytelling.
And how do you formulate yourstory, your life?
And it sounds like you're reallysaying the same thing that it it
all depends on how you presentthe things.
(12:19):
Um I I have to ask one morequestion about your background.
You did say you looked at uh MBAand law, but you chose not to do
that.
You chose to go to neurosurgery.
Uh what made you go to medicalschool over MBA or law?
SPEAKER_01 (12:35):
Oh, that's shadowing experience.
It was a single time in the inthe OR.
So I had taken the LSAT, uh, I'dtaken GMAT.
I I had applied along withmedical school to a bunch of a
bunch of those programs.
And um I went out and you know,in terms of strength, you know,
(12:56):
knowing what I wanted to do andof course strengthening my
application, uh, went lookingfor actual shadowing
opportunities.
And the one I came across umhere in town through a family
friend, happened to be aneurosurgeon.
And uh, you know, most ofneurosurgery uh is is spine
(13:16):
surgery, especially in privatepractice.
But this turned out to be abrain tumor case.
And I was like, this is amazing,holy moly.
And uh, and so uh yeah, thatthat singular shadowing
experience while I was deciding,yeah, totally totally.
SPEAKER_03 (13:33):
So it almost sounds so it almost sounds like it was
the aha moment when you werethere.
Definitely what I want to do.
SPEAKER_01 (13:40):
Yep.
I went into I went into applyingjust being like, I am good at
school.
Yeah, uh, I was very good inhigh school.
I don't know what I want to do.
It seems safe to stay in school.
Like I went into it very scared.
Uh, and it was, you know, meapplying to all these schools
(14:00):
was very much just driven kindof by fear.
Like, I I am not gonna make itin Hollywood.
Uh, I have to do something else.
School seems safe, but afterthat aha moment, I was like all
in on medicine.
It it totally changed myperspective, yeah.
SPEAKER_03 (14:15):
Right.
I and I can I can definitelyassert that uh you don't
necessarily become aneurosurgeon by fluke.
There has to be some focusthere.
But um, you know, I mean, uh canwe switching gears?
Do you said that you actuallyare an entrepreneur now?
Can you tell us a little bitabout your business?
SPEAKER_01 (14:34):
Yeah, so this is my first company.
It's a startup.
So from scratch, uh, it is atech company, and I'm not in
Silicon Valley.
So we'll see how this works out.
But um, like I said, I I havemaybe I've been one-shotted, but
um, but I'm all in on the AIhype.
(14:56):
Okay.
I really think it is uh gonna betransformational for healthcare.
It's as big as the internet,it's gonna be as big as mobile
phones, it's uh I think really,really gonna alter society over
the next few decades.
And so um uh my company is uh isa telehealth platform.
(15:18):
There are a million of those,but it's AI native.
So connect your connect yourmedical records, right?
Patients can connect theirmedical records, have that as
context for AI, just forinformational conversations.
Hey, what does this mean?
What were my last labs?
But when you need a physician,get connected seamlessly,
(15:40):
basically in chat.
Um so I'm very much like focusedon kind of urgent care, uh
simple primary care.
When we get off, we're in betaright now.
Um but this is this is my firstuh my first entrepreneurial
steps.
And I know you have a lot ofexperience, but it's uh it's
(16:02):
it's different.
Even though I was in privatepractice as a neurosurgeon, so
basically kind of a smallbusiness owner for for a while
with partners.
Uh this is this is definitelysome different skills, skill
sets.
SPEAKER_03 (16:17):
Yeah, right, right.
Um, so you mentioned that youthink that AI will be
transformative.
Obviously, the jury is out onthat one.
Nobody really knows.
Uh, I mean, everybody has thesame feeling, but there seems to
be a lot more apprehension thanthere there is adoption and uh
people embracing this as a newtechnology.
(16:38):
So, as an example, I'll tell youa lot of my uh our students come
in saying that, oh my god, youknow, we're gonna be out of a
job.
Uh AI is gonna take, you know,like everybody seems to pinpoint
on radiology specifically, thatoh, AI is gonna do so much uh
better than radiologists.
And, you know, I've spoken toDr.
(16:58):
uh Patel, who's the CMO of uh uhPrime Health.
Uh we've spoken to uh thepresident of the uh University
of North Florida, and their viewis pretty diametrically
opposite.
That they're saying that AI isactually going to supplement and
make physicians better.
What's your view?
SPEAKER_01 (17:17):
I think it's gonna be a mix, but um but I think
clearly, right, this kind offear-mongering about job losses
and things, at least in the inthe short term, has has been
doomerism.
There was a there was a recentgreat article um looking, you
(17:38):
know, uh some very famous peopleuh in the AI field, obviously
famously, you know, five plusyears ago were like, oh,
radiology is right for thepickings.
It is true that most uh AIsoftware as a medical device
that have been approved by theFDA are AI uh radiology tools.
(18:02):
But um this great article, it'slike two weeks old, but looking
at the average radiologist's youknow compensation, uh, which is
outstripped the averagephysician compensation.
Right.
Like it doesn't it uh AI rightnow is not taking radiology
jobs.
Um and I I think uh you knowthere's a there's a lot that
(18:26):
radiologists do just besidesjust looking at images beyond
just the the technical aspectsof hey, out of the lab, how do
these radiology models perform?
And obviously, you know, an LLMcan't consult, you know, with
another physician, it can'tperform procedures, it can't so
(18:49):
I um you know I would never Iwould not be telling people who
are considering medical school,hey, don't because AI is is
gonna take physician jobs in 10years.
I I just don't think that's thecase.
SPEAKER_03 (19:05):
Right.
Um and and I think uh, you know,we've spoken to some uh leaders
of some large hospital systems,and uh, they've always obviously
echoed the same thing thatradiologists tend to make a lot
more money.
Uh however, their feeling wasthat they're not gonna hire less
radiologists.
SPEAKER_01 (19:21):
Yeah, I agree.
SPEAKER_03 (19:22):
Their expectation is that they're expecting the
radiologists to do more.
Um, a case in point, uh uh thisis a little uh deviated from uh
what uh medicine, but uh I havea friend who works at uh one of
the largest uh truckmanufacturers in the world.
(19:43):
And they are in the cutting edgeof technology as well, where
they're developing AI trucks, etcetera, et cetera.
So I asked him, does this meanthat truck drivers are going to
become obsolete?
And he said that, well, we'llneed less truck drivers, but
what this also means is that ifyou are a truck driver, you just
cannot really drive a truck.
You have to have a morebroad-based skill where you have
to do a little bit of codingbecause if we are doing uh truck
(20:05):
caravans that are AI poweredwith the first, second, third
trucks all AI driven followingthe first truck.
If something goes wrong, and itdoes always, the person driving
the per uh in the first cab hasto be able to service this.
So just having a CDL is just notgoing to be good enough in down
the road.
Um and and I think you know,when I spoke to these uh
(20:27):
hospital system leaders, a lotof them have suggested the same
thing that radiologists will bethere's not going to be any cut
in compensation, they're notexpecting to hire less, but they
are expecting in due course oftime radiologists to be able to
do more.
Um would you what are yourthoughts on that?
I and and we've seen that inGoogle and some of these IT
(20:48):
software companies too, wherethey're expecting the production
to triple quadruple per softwareengineer.
SPEAKER_01 (20:54):
Yeah.
Well, we'll see, even as even asAI writes what you know,
whatever these these techcompanies claim 90% of of code,
right?
But um yeah, I I will tell youthat that is that is probably my
more realistic fear.
Um that uh you know what youknow, as you have these clinical
(21:23):
decision support tools or thesecomputer assisted assisted
diagnosing tools that likeradiologists can use, what it
what's that gonna do forinsurance reimbursement and um
the expectation on like theradiologist work workflow and
work workload, right?
(21:43):
Like, oh the the computer'sdoing some of this for you, read
three times as many films,right?
Um so uh we'll s yeah, that thatmight be how this plays out.
Um I do think that uh as thesetools proliferate, just having a
(22:04):
basic understanding of them, AIin general, um is is gonna be
important for physicians.
Okay.
SPEAKER_03 (22:13):
Uh so uh in your specific field, uh surgery and
neurosurgery more specifically,do you think AI has impacted
your field directly yet?
SPEAKER_01 (22:26):
I do.
I'll give you I'll give you acore example.
So I I even have a little, I'meven more subspecialized in
neurosurgery.
I I do mostly neurovascularstuff, so blood vessel diseases.
Um and so big strokes, likethrombectomies, brain aneurysms,
things like that.
(22:46):
And this is very, very niche,but I'll give an example where
AI has mattered a lot to mypatients.
So I take a a lot of strokecall.
Um in stroke call over the past10 years, uh for large vessel
occlusions, so when a big bloodvessel gets blocked to the brain
(23:10):
and is causing a stroke, there'sbeen this paradigm shift where
we know that much like whenpeople have a heart attack and
the cardiologists go in andreopen the vessel, that
reopening these large vesselsemergently can uh prevent people
from having a stroke or keeptheir stroke small.
(23:31):
And so that uh that is a veryemergent thing.
So people will come into the ERwith a stroke, they will get
these this imaging, a head CT, aCT angiogram that looks at the
blood vessels of the brain, andthey get that um pretty much a
standard of care urgently.
And if you see a blocked bloodvessel, right, you have to
(23:53):
activate the team, activate me,get them over to the angiosuite
to get the blood vessel open.
And there have been aproliferation of tools, even
though radiologists uh areobviously very good at detecting
these blocked blood vessels.
These um these apps can do thatin 15 seconds, right?
(24:17):
What used to maybe take 10 to 15minutes of kind of radiology
workflow.
Oh, the images get pushed overto a radiologist, the
radiologist looks at it, theradiologist calls the ER doc,
the ER doc calls me.
Now uh you get the scan and theAI like pings immediately.
(24:37):
There's a blocked blood vessel,and it just shows up on your
phone.
And and even those 15 or 20minutes, right?
We know from the literature uhcan affect outcomes in terms of
the chance of people doing well.
So that's one place that I useweekly that AI has really made a
difference.
There's no doubt.
SPEAKER_03 (24:57):
Oh wow.
And um has AI reduced theworkflow though for you?
Do you think so?
SPEAKER_01 (25:06):
I think I think it's in that very, very niche
example, yes, it has made usmore efficient.
There's no doubt.
I use a lot of other AI tools,so other popular ones, right?
Um clinical decision supporttools like open evidence, right?
Oh, this is, you know, put in,you know, ask questions rather
(25:26):
than going just to PubMed andsearching.
Uh I do use an AI scribe.
It's kind of plus or minus,right?
These are tools that listen toyour conversations in clinic
with patients and then spit outa note for you.
Um, I think that's a little bithelpful.
But I I would say that um thatthat probably that stroke
(25:47):
example has been the mostmeaningful in terms of for my
patients.
SPEAKER_03 (25:52):
Awesome.
Um, so one thing I do want totouch upon and switching gears
to like more um uh less, youknow, like internal medicine and
family medicine.
One thing I get a lot ofquestions from from students is
that um, you know, a lot ofpatients are using uh ChatGPT to
(26:12):
really diagnose initial problemsthat aren't major medical
emergencies.
That, oh, I have a skin spot,this is the situation, they take
pictures, chat GPT gives a bunchof diagnoses, uh, et cetera, et
cetera, which has made a lot ofum students who want to do
family medicine or internalmedicine really nervous about uh
their professions.
(26:33):
What are your thoughts on that?
SPEAKER_01 (26:35):
Yeah, I th I think that's uh an increasing use
case.
It's even kind of it's even kindof the principle that that I'm
building a company around,right?
Um, but I don't think we arenear uh automating away
physician knowledge work.
Um uh and so I I do if I wasgonna go into a mostly
(27:00):
knowledge-based field, uh evenyou know, primary care or a
medical subspecialty.
I I think uh the future isbright.
Um even with AI.
Uh I don't think uh these toolsare, even though they're
changing rapidly, who knows whatit looks like in five years, but
(27:22):
I don't think they're they'rethere in terms of uh safety.
Um I think there are gonna behuge non-technical things to
replacing physicians (27:33):
societal trust, reimbursement, liability.
I do think these tools, youknow, 10 years from now might
how we practice medicine look alittle bit different.
Um AI is triage or things likethat.
Uh yeah, of course.
(27:54):
But uh I think there is a longfuture for primary care, uh,
even with AI.
SPEAKER_03 (28:01):
Okay.
And um so so obviously we'vetalked a lot about what AI can
do.
Um but I I think that there aresome downsides, and I'd like to
talk about that.
What you feel like the downsidesof AI is.
One thing that I have seen isthat the proliferate and as this
is not just AI, I think it'sjust generally information.
(28:24):
The proliferation of informationum is good and bad.
We we can see some of the uh badconsequences of democratization
of information in society.
But one thing I have noticed isthat there seems to be a lot
more pushback from the patientside in the sense that a patient
will go in and they have somesort of a diagnosis in their
(28:46):
mind that Chat GPT told them.
And the the the physician willcome and tell them that no, I
don't think this is what is it,what it is.
Uh this is probably more likely,and the patient comes like, oh
yeah, the physician doesn't knowanything.
Uh I trust ChatGPT.
And I think that that is anincredibly dangerous aspect, and
(29:06):
I and I dub it as uh the age ofexpertise is over to a certain
extent, because now people thinkthat I could type in a prompt in
ChatGPT and get all theinformation I need, or I can
watch a uh five-minute YouTubevideo that ChatGPT pulls out for
me, and I'm just as good as umuh any physician or any other
(29:28):
specialty profession.
Uh, what are your what so Ithink I feel like that's a
pretty big downside.
Um, I don't know if you'veexperienced that and what your
views on um that is.
SPEAKER_01 (29:43):
Yeah, I to be honest, even when I heard you
talking, I was about to about tosay we've even before you use
the term, we've we've enteredthe age of of anti expertise,
right?
Yeah.
Um I I don't I don't thinkthat's uh You know, I I think uh
that a little bit uh predates uhAI.
(30:06):
Um I mean, not to get intopolitics right, but you're you
see, you know, RFK yesterdaybasically saying, hey, we don't
we uh we shouldn't trustexperts, that's not science.
Um and so I do think there's uhthere's I do think that's uh
there's a lot of that and thatAI might might make that worse.
(30:30):
I do think it's kind of abroader societal trend though,
uh not just AI.
Um you know, even when it'sfrustrating, even when it uh
seems personal, that I have justcome to be like have grace for
patients.
There is nothing wrong withpatients, even if it's it seems
(30:53):
misdirected to you, like guidedby AI, of them going and getting
a second opinion.
You know, I think we have totrust our our colleagues to to
handle it the same the same way.
Um but it can be a frustratingaspect of practice, there's no
doubt about it.
SPEAKER_03 (31:13):
Well, I mean I will I will go out on a limb and say
this is not just for medicine,it's for every profession.
Um people, you know, um I'm aCPA by training.
I I have a lot of colleagues whoare accountants, and they come
and say that they ran it throughAI.
And in some cases, this is true,this does work, but they're
like, oh, my tax refund could beso much larger because AI told
(31:34):
me so.
Um and there are certain nuancesthat they don't understand,
there are certain physicalitiesthat they don't understand.
So, for example, in certaincases, um, you know, an
accountant might know better thefamily that just had a kid, and
AI missed that because theyeither they didn't put it in the
prompt, or you know, I mean it'sthe AI is as good as the prompt.
(31:55):
And I think that that humantouch is pretty necessary for
almost every profession, and Ithink more so for physicians
than than any other profession.
But uh one thing that I'venoticed is that AI is absolutely
terrible at managing chaos.
Um, it cannot handle chaos, andthat's and uh let's face it,
(32:18):
human beings are chaotic, ourthoughts are chaotic, our
thoughts are not organized,logical, and streamlined.
Um, as you mentioned, RFK Jr., Iwould say that his thoughts are
pretty chaotic too.
Um, and and that's what uh tricktrickles down.
Do you see any majordisadvantages of our adoption of
(32:40):
AI beyond obviously the societalimpact?
unknown (32:44):
Yeah.
SPEAKER_03 (32:45):
From a technical perspective.
SPEAKER_01 (32:46):
Yeah, I mean, uh, technical challenges with AI,
where we stand right now.
I mean, with the caveat thatthings are still continue to
change very rapidly, whateveryou think about the newest
models.
I mean, uh, I completely agree.
Um, AI is pretty poor, even whenyou try to design tools around
these models, at going out andseeking information, okay.
(33:10):
Which physicians are great at,right?
You know, uh patient comes intothe exam room, medical student
is interviewing them, theymention something, oh, this
triggers something there, I'mgonna follow up, right?
Maybe this wasn't the chiefcomplaint.
Or you know, AI is amazing atsynthesizing data, um, but not
(33:31):
seeking it out itself.
Uh it's also not very novel.
Um, you know, that maybe that'sgetting better as as as it tries
to get into science, but it youknow, things it has uh not seen
before and not been trained on,it's not it's still not great
with.
Um and so these are these arethings that I think physicians
(33:56):
are still much, much, muchbetter at.
Yeah.
If you give AI all the context,right, yeah, it's it's great at
synthesizing it as long asyou've given it everything it
needs.
But um, yeah.
SPEAKER_03 (34:10):
Okay.
And um do at some point, I wouldassume AI, AI's been
transitioning in leaps andbounds.
Just two weeks ago, there havebeen updates across all the AI
platforms um that haveessentially, you know, replaced
a lot of uh or will be replacinga lot of consultants because uh
(34:31):
as long as you can communicatewith AI appropriately, it can do
uh Excel sheets, it can do procharts and and and all of that.
Um but do you feel that that'llchange at any time in the near
future, in the distant future,where we don't need to go to a
physical person, but we're justsitting and I don't know,
talking to a chat GPT persona?
SPEAKER_01 (34:54):
Um we may get there, uh especially in terms of like
initial care.
You know, you I can I canenvision a you know, a kind of
medium-term future, maybe wherephysicians do a lot of triaging,
a lot of like edge cases.
(35:16):
Hey, the AI is unsure, hey,review the AI's work, um click
in when when it needs more.
It that might happen.
Um I will say, you know, we justin a broader sense, uh there's
there's a lot of talk about AItaking jobs.
(35:38):
Um I uh not just for physicians,but I just think historically,
right?
Um with uh as technology hasbecome you know, everything of
of the twentieth andtwenty-first century, right?
(35:58):
The driving force.
Even as technology has advancedit has it has really shifted the
type of work that humans do.
Um but it's it's not like youknow the internet killed male
mail carriers or something likethat.
(36:18):
And so um I think AI may reallychange how physicians practice
and maybe how uh patients accesscare.
Probably not in the in the inthe very, very short term, but
longer term.
But I don't think um physiciansare gonna become like dinosaurs
(36:42):
at all.
Right.
SPEAKER_03 (36:43):
Right.
I I w I would definitely agreewith you.
Uh one thing I do want to saywhen you talk about jobs, from
everything that I've spoken to,uh everybody I've spoken to, it
really seems like uh these largehospital chains are not
necessarily interested incutting or looking to cut any
physician positions.
(37:04):
What they are looking to changeis they are looking to change
the headcount and looking to uhuse AI and leverage AI as much
as possible, but they're lookingto eliminate a lot of
lower-level administrativepositions.
And um, but but because of that,do you anticipate the nature of
(37:26):
physicians changing wherethey're doubling up and doing a
lot of the functions that uh notall of it, but a lot of the
functions that potentially uh anurse is doing now?
Hypothetically, I'm you know,because uh because uh they might
have more free time now becauseAI is doing some of the analysis
(37:51):
work or some of that uh busywork, as you said, for
radiologists.
They're going through many, manyreports, and something that used
to take 10 minutes is taking 30seconds now.
Uh and and if there is not anincreased patient volume,
they're looking at uh doing morejust like a truck driver is
(38:12):
being considered to essentiallybe a software engineer as well,
or a software engineer is goingto be driving trucks also
nowadays.
Um, do you anticipate that kindof a world where a physician is
also doubling up as uh doing alot of the basic work just
because they have more time?
SPEAKER_01 (38:30):
Um that's a really good question.
I uh I had never considered likephysicians doing more.
I mean, not that this is whatnurses do, but more more like
scut work.
I I I guess I'm probably alittle bit skeptical of that.
Okay um I'm interested in, youknow, I don't I want to be
(38:55):
careful, but um you know, healthhealthcare has you know, we're
all in this to help people, buthealthcare as a business does a
pretty remarkable job of uhcreating its own demand, right?
(39:19):
So, you know, I think a lot ofpeople um look look at medicine
as some sort of fixed thing.
When you're sick, you go to thedoctor, but I uh you know, I'm
not sure that's a hundredpercent how and you know, and
people only get sick so much,but I'm not sure that's a
hundred percent how it works inthe real world.
And so um I think there willalways, you know, in the in the
(39:42):
medium term, uh even if AI makesus more efficient, that there's
there's gonna be enoughpatience.
Um and so I think it's probablymore realistic that like we
talked about, uh, you know, Ithink it's gonna be important
for physicians to know how touse these tools, troubleshoot
(40:04):
these tools, absolutely.
Um and I think like we talkedabout with the radiologist, that
maybe there's demand to be formore throughput, more be more
productive.
I can I can envision that.
SPEAKER_03 (40:21):
Okay.
Um and and I think this kind ofgoes back to the uh the the
pyramid structure of ourhealthcare system, where at
least in the United States, umthey focus a lot more on primary
care in the beginning, so thatspecialists have less let's not
(40:42):
call it a workload, but uh lessacute patients at the latter
end.
Because if you're dealing withhypertension at a relatively
early age at the PCP level, uhyou're probably less uh likely
to develop complications thatyou would be seeing uh at your
level.
And that's generally thepyramid, but uh I think what
you're suggesting that makes alot of sense where um maybe the
(41:08):
level of care is increased tohave a healthier population.
Um I definitely think thatthat's possible.
That and and I and that makes alot of sense to me.
Um so I want to switch gears alittle bit, and since we are an
educational organization, whatdo you think?
And and I have to caveat this bytelling you that education moves
(41:31):
at a glacial pace.
Um, I can tell you right now, somany conversations I've had with
faculty members, programdirectors, um people who are in
various leadership roles invarious uh residency programs
who absolutely abhor AI.
They have clearly said that ifyou are using AI, you're
probably an idiot.
(41:52):
You don't know what you'redoing.
I want people to do all of thesethings hands-on.
So there's still a lot ofpushback, and I think that
that'll stay for at leastanother half decade or so, if
not longer.
Um and even from our accreditorsand the accreditation
standpoint, um, there have beenno mandates on any particular
(42:15):
course that's tech-driven,AI-driven.
Um they are, in fact, when wesay that we want to introduce
some of uh some courses likethis, the general reaction is
that you know, we have acurriculum, we want to stick to
it, we don't want to take awayfrom what we have.
So from the educationalstandpoint, at least uh general
(42:38):
leadership, we don't necessarilysee a whole lot of movement on
what schools should be doing totrain our physicians um
appropriately.
What are your views on whatschools should be doing to train
our the physicians of tomorrow?
SPEAKER_01 (42:57):
I think training people on tool use is uh
absolutely important even rightnow, right?
And and you know, you'retraining young people who are
generally more technologicallyinclined, right?
They're probably learning thesetools already um by themselves.
(43:19):
I uh, you know, uh it'll bereally interesting how AI comes
into education, even medicalschool education.
I am I am pretty sympathetic,and I've heard similar arguments
that that you have.
I'm pretty sympathetic to theidea of using, you know, of what
(43:40):
is what are the costs of usinggenerative AI.
Uh like you know, I thinkthere's a lot of value in terms
of critical thinking andorganizing your thoughts to uh a
simple task like writing.
And when you offload that, whatare what are kind of the
long-term costs?
And so I'm sympathetic to peoplewho really don't like that.
(44:03):
But um, you know, I I'm I'mreally curious to see how, you
know, holy moly, right?
Uh does every does every lectureneed to be live, right?
My goodness, right?
We we asked AI about this topicand it generated this whole
(44:25):
hour-long audio lecture andslides itself and um or uh you
know using it to to do research,things like that.
So uh yeah.
I think it'll be reallyinteresting.
I mean, you're you're obviouslya lot closer to it than I am um
in the education field.
(44:46):
Uh I do think you have to becareful when you're trying to
offload your critical thinkingto it.
I think it is an amazingresearch tool.
It can be amazing for likehelping you learn, um, for
synthesizing stuff.
But if you're like, well, I'mjust gonna have this AI write
(45:09):
this essay for me, um I I don'tknow.
SPEAKER_03 (45:14):
Right, right.
Well, I mean, uh absolutely soso that's just being lazy.
Like if you're disparating AI tocompletely substitute something,
but um so one of the challengesthat we've had is that right now
AI is pretty disparate.
You have ChatGPT, you haveClaude, you have Perplexity, you
have uh Google Gemini, you haveGrok.
(45:35):
They all are different and theyall have somewhat different
answers.
And and we've tested all ofthese.
All the ones, the five that Imentioned, we've uh I have
personally used all of them.
Um and the results are notreliable in the sense that uh
obviously uh I I'm asking itrelatively subjective questions.
(45:56):
I just basically choose theanswer that I like best, that I
think fits my narrative thebest, which is obviously a
little bit of critical thinking.
Um but when you're saying if youwere to hire somebody or or if
you were to get a neurosurgeonresident, a first-year resident,
uh, for your uh in youruniversity, would you care that
(46:21):
they know what these fivesystems are and how to use them
and how to uh create a username.
Does that matter to you as as oftoday?
SPEAKER_01 (46:33):
Uh no, to be honest.
Uh I uh, you know, there wouldcertainly be expectations.
Hey, these are the tools we use.
Uh this is the ambient scribe weuse in our clinic.
This is the, you know, this isthe stroke alert tool we use in
(46:55):
our clinic or hospital.
Um, you know, learn the basicsof these.
But would I care if they do thenuance of, hey, I've discovered
Claude is better at you know atwriting and ChatGPT codex is
better at coding.
(47:15):
Uh no, I wouldn't.
Okay.
SPEAKER_03 (47:18):
So uh when you're when we're talking about AI
education in general, do youthink uh what what would matter
to you?
Like how how to write a prompt?
They have a general idea of howto write a prompt, um, having
some guidelines or guardrailsthat they personally know that
(47:38):
we shouldn't do these thingswith AI.
What what's the flow that youwould care about as an AI
entrepreneur, as a neurosurgeon,when you're hiring somebody,
what would what would matter toyou?
SPEAKER_01 (47:50):
Uh I want them, yeah.
I want them to have berelatively technologically
savvy, right?
All these, even the tools we'rebuilding on top of models, they
all work generally the same,right?
Hey, you're coming in and youknow, uh I'll give you an
example, right?
I mean, this is not a newchallenge.
(48:11):
Uh AI is not new, but mygoodness, looking at my we, you
know, we we recently, well,about a year and a half ago,
sold ourselves to a bighealthcare system.
We were private practice and webecame employed.
And that uh entailed, becausethis was entirely new employees,
(48:32):
entirely new hospital, uh, a newEMR.
And like watching my you know,60-year-old partners try to get
a handle on that, uh, you know,it's a it's a slow, painful
process.
So, you know, these a lot ofthese tools work very, very
similarly.
Being familiar, being able topick up new tools, I think being
(48:54):
able to prompt um and tocritically assess AI's
responses, right?
Oh, that that's like not right.
Um, and to be able to ask it uhquestions and give it context
where that's a lot less commonand it gives you uh answers that
you're looking for, yes, isabsolutely key uh as well.
unknown (49:18):
Okay.
SPEAKER_03 (49:19):
So so generally the concept of prompting, which I
think is really the basis andbackbone.
I don't know uh I've been usingAI for about two years now, and
I think the sophistication thatI needed to create a prompt to
give me meaningful answers twoyears ago has dramatically
reduced today.
Uh back in the day I felt likeit almost needed to follow a
(49:42):
certain formula.
Yeah, perfectly.
Yes, and whereas today I feellike I can just have a just
talking to a person and I cankeep elaborating until I get to
that answer.
So I do think AI is getting moresophisticated, but at the same
time, having that knowledge ofwhat the structure originally
should be like helps me getbetter answers.
(50:02):
So I think uh having a coursethat kind of talks to students,
not necessarily about uh what uhAI is, but generally having a
general understanding.
And you know, especially onething I've noticed is even today
we have this these token limitson AI, which means that if you
(50:23):
keep the same chat window openand you keep asking questions at
some point only as so muchcontext, yeah.
Yeah, uh and and it's at somepoint it starts spewing
gibberish because it's run outof tokens.
Um, this is something that I'vebeen a little surprised to know
that most people don't know.
So the answers that you get inthe beginning of the
conversation versus the answersthat you will get towards the
(50:46):
end, they're a lot looser.
Um AI hallucination is still afactor, which is again, I'm
surprised that a lot of peopledon't know.
But I think having a course thatreally talks about these
concepts, and again, these arechanging so much uh over time
and may not maybe completelyobsolete at some point for
hallucinations.
(51:07):
Um, and and if they have thatkind of a structure and a
knowledge, would that bebeneficial to a person like
yourself?
SPEAKER_01 (51:14):
Honestly, that'd be super beneficial.
Yes.
Okay.
SPEAKER_03 (51:17):
Uh awesome.
My final question about AI isreally do you see AI ever
leveling the playing field?
Because let's face it, today inmedicine, uh, if you're a
neurosurgeon, if you're an ENT,if you're an orthosurgeon, uh,
for one reason or the other,there is dramatically uh a lot
(51:39):
more prestige, let's say.
And and and not to say that Ithink just generally being the
doctor, uh, you know, I have myview that there are only four
real professions in the world.
Um, and I think uh being adoctor is definitely one of
them, no matter what kind of aphysician you are.
Um but I think uh do you thinkAI will ever level the playing
(52:00):
field between the discrepancybetween all these specialties
and doctors?
Number one, number two,obviously, uh in some of the
more elite fields likeradiology, anesthesiology,
surgery, the different kinds ofsurgery, uh there's still a
little bit of stigma for IMGs.
Um I know that stigma has goneaway quite a bit for medicine in
(52:21):
general.
Um, as I said, our school issince 1999.
We saw a lot of that in theearly years.
Uh, I've personally hadconversations with the program
directors who've said that ifyou're from a Caribbean school,
this is not the program to applyfor.
Um, which is, in my opinion,completely baseless because in
the United States we have astandardized system.
(52:41):
You have everybody takes thesame tests.
If you score well on thesetests, I don't see the problem
with with uh applying.
But um but do you see AI everreally leveling the playing
field and uh decreasing thediscrepancy between the various
medical professions and thenwithin the medical profession,
(53:02):
say the DOs, the M uh theCaribbean MDs and the American
MDs?
Do you see that happening?
And if so, how do you anticipatethat to play out?
SPEAKER_01 (53:13):
Yeah, I uh that's a really good question.
Um and I haven't given it a tonof thought, but I you can
envision a scenario where thatmight be the case.
I really think that if that doeshappen, it's probably probably
more medium to long term.
(53:34):
But um, you know, I as AI mayreally kind of I can envision a
scenario where AI kind of levelsum the compensation structure,
levels the prestige factor forphysicians, um and a lot of that
(54:01):
is you know kind of what drivescompetitiveness.
Uh and so if if that happens,right, if AI leads to say uh
lower compensation forprocedures or something like
that, which I could totally seehappening.
(54:23):
Um yeah, I mean, I think uh thatcould dramatically kind of
change the face of what whatconstitutes a competitive
specialty or not and change umyou know the whole residency
application process is awesomewithout doubt.
SPEAKER_03 (54:43):
Okay, well that's very interesting.
So um do you think that AI willand we've touched upon this
briefly earlier in ourconversation, but um do you
think the demand for physicians,and I know you mentioned that
the healthcare system has a uh away of creating its own demand
(55:05):
in the sense that havinghealthier, healthier people, and
and and this is very apparent inour life expectancy.
We've seen this.
Like I life expectancy, even inmy lifetime, I remember when I
first moved to the UnitedStates, the US life expectancy
at the time was uh in the late70s, today it's in the early
80s, uh, or very close to 80, ifI'm not mistaken.
Um, and I and that this isexclusively uh because uh the
(55:31):
government and physicians haveworked towards making doctors
more accessible, and and andyes, that there's a whole
healthcare and and uh insuranceargument.
I'm not getting into that, butgenerally being able to see
physicians earlier just allowspeople to get healthier.
But do you think AI at any pointwill reduce the demand um for
(55:59):
some specialties across the thethe realm of physicians?
Maybe family physicians, youdon't need to see them that many
um that much.
Uh maybe dermatologists, thatthat could be one.
If I have a skin tag, I put itup on Chat GPT and chat GPT.
Maybe your software is doingthat.
SPEAKER_01 (56:19):
I uh I really think that it's going to change how
physicians practice.
But uh I I don't think I meanI'll just stress again like if I
was going to medical schoolright now, I would not have a
major fear that oh my goodness,we're gonna have a glut of
(56:41):
primary care physicians in 10years because AI's doing too
much of this work.
Um I don't think so.
SPEAKER_03 (56:51):
Awesome.
Well, thank you so much, Dr.
Son.
I know we're out of time, but uhthis was such an invigorating
conversation, an eye-openingconversation.
I think I have some homeworkbecause I'm gonna go and create
uh uh a course for our studentsthat we uh we we actually have
some, yeah, we actually do havea very yeah, a basic course that
(57:13):
just pre-residency, we actuallytell students to get uh a little
more acquainted with AI.
But I think uh I'm definitelygonna push our academic team to
create a program that's in a uhembedded in our curriculum.
But uh it's been veryeye-opening.
Uh, before we leave, I do wantto uh ask you a little bit about
your business.
(57:33):
We already spoke about that, butthe name um and uh how do you
think uh as a school, do youthink we can help in any way?
Do you think uh some of thethings that you're doing can
actually uh help students beequipped with uh in facing the
(57:54):
real world of AI?
SPEAKER_01 (57:56):
Yeah, I'm I'm always uh happy to talk to anyone uh
just in general about medicineand career in medicine, but but
especially AI, if anyone wantsto ever reach out.
Um uh but yeah, I am working onthis company called South A
Health.
It is in beta.
Um, but like I said, it's it'skind of a patient-facing AI
(58:19):
telehealth platform, and and youknow, hopefully if in the
future, if any of your studentsare are looking for for work to
even telehealth remote work, uhlove them to reach out as well.
So after they finish residency.
SPEAKER_03 (58:36):
Thanks again, Dr.
Collinson, for joining us todayand sharing uh your journey from
neurosurgeon to entrepreneur andhow AI is reshaping the future
of medicine.
If you enjoyed this episode,make sure to follow or subscribe
to the podcast on your favoriteplatform, leave a review, and
share it with a friend orcolleague who's passionate about
healthcare and innovation.
(58:57):
You can also download pastepisodes and find updates on our
uh upcoming guests.
And remember, there is noshortcut to getting an MD.
SPEAKER_00 (59:05):
Thank you so much for tuning in to our show.
We hope you enjoyed anotherepisode of Med School Minutes.
If you like our content, pleasefollow us and receive
notification when a new show isposted.
This podcast is brought to youby St.
James School of Medicine.
For a video version of thispodcast, please check us out on
SJSM.orgslash video.
Hello, and welcome to another episode of the Med
School Minutes Podcast, where wediscuss what it takes to attend
and successfully complete amedical program.
This show is brought to you bySt.
James School of Medicine.
Here is your host, Kashik Gua.
SPEAKER_03 (00:20):
Welcome to another episode of Med School Minutes
where we talk about everythingMD related with a focus on
international students,specifically students from the
Caribbean.
My name is Koshik Guha, andtoday we're joined by a unique
guest, Dr.
Colin Sun.
He's trained as a neurosurgeoncertified by the American Board
of Neurological Surgery,specializing in minimally
(00:40):
invasive brain and spine care.
But Dr.
Sun didn't stop at surgery.
He's now moving intoentrepreneurship, applying
artificial intelligence totransform patient access to
physicians.
In this conversation, we'll diginto his transition from
medicine to business, explorewhat real-world AI means in the
neurospace today, the challengeshe's facing, and his vision for
(01:02):
where things go next.
Let's turn to Dr.
Sun now.
Thank you so much for your time,Dr.
Sun.
Really appreciate uh you makingtime to come on our podcast.
Uh, why don't we start by yougiving us a little bit about
your background and uh yourjourney in the field of
medicine?
SPEAKER_01 (01:20):
Yeah, that thank you for having me.
Um I really, really appreciateit and uh appreciate what y'all
y'all are doing with thispodcast in terms of you know,
uh, you know, I'm sure it's amarketing tool, but it's also
seems super educational formedical students and and
especially pre-med.
Uh so my name's Colin.
(01:41):
I am a neurosurgeon.
I'm up in Texas.
Uh I uh uh am also anentrepreneur, so I've gone
part-time clinical.
I've totally jumped on uh kindof the AI hype train and how I
think it's gonna, you know,really, really change uh how we
(02:06):
train physicians, how we educatemedical students, how we access
care, how we pay for care, howwe regulate care.
Um uh and so I have uh a companytrying to get off the ground in
that space.
Um, but still practicingneurosurgery part-time.
Uh I came to neurosurgery verycircuitously.
(02:27):
I went to film school.
Okay out in out in Californiaand tried to be a screenwriter,
and uh uh it quickly becameclear uh maybe uh maybe some of
those things like selling myselfor or maybe even my talent as a
(02:47):
writer were were gonna leave mewaiting tables.
Um and so I started, I was like,well, I'm pretty good at school.
unknown (02:54):
Right.
SPEAKER_01 (02:55):
Maybe I'll do something, uh go back to
professional school and andlooked into um, you know, uh
business school and law school,applied to both, applied uh to
medical school uh for the samereasons.
And while I was doing all that,I shadowed a neurosurgeon and uh
had never been in the OR before,uh, didn't really watch a lot of
(03:18):
ER as a kid.
Um, but uh that was I was like,oh my goodness, this is um this
is really impressive.
So I came into medical schoolpretty much knowing I wanted to
be a surgeon.
It was the whole reason I wasthere, and uh led led to this.
SPEAKER_03 (03:36):
Wow.
So I I do want to focus a littlebit on on about this uh uh film
writing uh aspect.
That is so interesting.
Um and and and one of thereasons I do want to focus on
that is because this would betruly inspirational.
Because um what I hear,especially in the United States
for more mainstream medicalschools, uh, is that if you have
(04:00):
uh if you vacillated on yourcareer, that is typically not
the ideal candidate for a lot ofUS medical schools.
Um however, it seems like youhave obviously turned that
around and you were in LA atsome point, waiting tables,
trying to get into Hollywood.
But then in my opinion, I wouldthink that neurosurgery is
(04:22):
probably should have been a muchdifferent difficult path than
trying to get break intoHollywood.
But you obviously chose theharder path.
But I do want to focus on alittle uh hear a little bit more
about that.
Can you tell us a little bitmore about the how you you went
from film writing or scriptwriting to um uh neurosurgery?
SPEAKER_01 (04:43):
Yeah, so um so I I went to film school.
Um and uh and it was uh it was afour-year program.
I was I was already getting mybachelor's, um, but I was not
looking at medical school.
I was not pre-med.
Right.
In my junior year, I was like,oh my goodness, um uh this is uh
(05:09):
this is not, you know, I'm notnot sure uh trying to do this
long term is for me.
And so like I said, I startedlooking into all these schools.
When I when I started doingthat, um, you know, the second
semester of my junior year, Iactually took a biology class,
which counted as a prereq justin case.
(05:31):
I was also like, you know, Itook like a pre-law class, like
my electives were uh, and thenum when I decided that year, I I
really wanted to try for medicalschool.
Um I did uh I did summer work ata local community college, um,
which allowed me to get a lot ofcredits and then uh more more
(05:56):
kind of prereqs work.
Um, you know, I I hear whatyou're saying that it is so
competitive nowadays, uheverywhere, and that uh you know
I think I definitely had someknocks against me doing doing
some of my prereqs hurriedly,like at community college.
(06:18):
Um, but uh I also you know,maybe it was just the way I kind
of went into the process, but Ialso kind of viewed it as um,
you know, this kind of atypicalpath as uh as a real selling
point, honestly.
(06:40):
I don't remember getting a lotof questions like in interviews,
like uh, you know, people wantto know, you know, as they
always do, why you want to be aphysician.
But I uh I really, you know,when I went into these
interviews or even my personalstatement, um I really kind of
(07:00):
pitched it as uh as like aunique path, right?
Uh this is part of the reasonI'm sure about medicine.
These are the experiences I haduh before deciding on medical
school.
This is how I think my creativeside um makes me more empathetic
(07:22):
uh and uh and ideal for dealingwith patients.
And so um, you know, I I think Ihave lots of friends who took
very non-traditional paths, youknow, time off, uh worked, uh
had careers before beforemedicine.
And I I think sometimes whenyou're applying to medical
(07:43):
school, that can be, you know,when you're selling yourself,
that can be a real benefit.
SPEAKER_03 (07:48):
Okay, well, I mean, uh that's really interesting.
So it almost sounds likespending and and the film school
was, I'm assuming, full fouryears.
SPEAKER_01 (07:56):
It was, yeah.
SPEAKER_03 (07:57):
Yeah.
So film school gave you thetools to write a strong story
that strengthened yourapplication.
It probably did.
Yeah.
Because I I will say, I mean,uh, I've I've spoken to I speak
and interact with a lot ofphysicians, and obviously the
physicians who make it to uppermanagement are very, very strong
(08:18):
people persons.
Um, uh, I think without thatthey can't really get to
business side.
Yeah.
Yeah.
Um however, also I will caveatthis by saying people who are
hospitalists aren't necessarilythe most uh people oriented, uh
generally speaking.
I mean, and I don't mean anydisrespect, I just think that uh
(08:40):
it's just not a skill that uh isabsolutely essential to be a
good physician.
So it's compared to otherprofessions, I would say that
that tends to be a little moreuh rudimentary for most
physicians.
But it it sounds like um yourexperience in film school really
(09:01):
bridged that gap and made you abetter, more holistic uh
candidate.
Would you agree or disagree tothat?
SPEAKER_01 (09:09):
I think I think I did.
Um I am uh uh I'm a prettyintroverted, kind of nerdy, uh
you know, not I mean I've gottenbetter over time, but like uh if
I had if I had been going intomedical school interviews,
(09:30):
having been a pre-med, havingbeen only around pre-meds, just
with my social skills uh comingout of high school, I bet I
would have been a worseapplicant than having gone
through this uh the field schooland um yeah.
Okay.
SPEAKER_03 (09:49):
So uh going back to today, where as you mentioned,
it's incredibly competitive.
Um medical school, generallyspeaking, is incredibly
competitive.
Um and I feel like a lot of themedical schools generally really
focus on reasons to rejectstudents.
And if you don't have a moretraditional pathway, from what
(10:12):
I've seen, at least nowadays, itseems a harder task.
You don't necessarily see peoplenowadays saying, I went to film
school, or as you said, you hadfriends.
Um and and I feel like thismight have been a paradigm shift
sometime in the mid-2000s.
Um and do you think that todaythe programs are probably doing
(10:35):
a disservice by looking atsomebody who's done pre-med,
then they want to go to medicalschool, and then they're looking
for residents, uh, especially inthe uh uh very elite specialties
like neurosurgery, orthopedicsurgery, ENT, etc.
etc.
Uh, what are your thoughts onthat?
SPEAKER_01 (10:51):
Yeah, I can imagine that really being the case.
Um uh you know, for people whoare non-traditional, um you
know, the path I kind of tookand probably would be my advice
uh is to really lean into it.
(11:13):
Um I d I you don't need to bedefensive about it, but I I
think these non-traditionalpaths can certainly uh make
people better physicians.
And if you're able to articulatethat argument like in your
interviews, like in yourpersonal statements, like when
(11:34):
you ask for letters ofrecommendation, like discussing
that with your you know, uh withyour mentors.
Um you know, I I I don't thinkthese are things that should be
you should have I I agree thatsometimes it probably feels like
you have to, but uh that youshould try to excuse.
(11:55):
Um, and I I think they can be astrength, really.
SPEAKER_03 (11:59):
Okay.
Wow.
So, you know, this is somethingwe tell our students all the
time.
Everything in this world isdetermined by storytelling.
And how do you formulate yourstory, your life?
And it sounds like you're reallysaying the same thing that it it
all depends on how you presentthe things.
(12:19):
Um I I have to ask one morequestion about your background.
You did say you looked at uh MBAand law, but you chose not to do
that.
You chose to go to neurosurgery.
Uh what made you go to medicalschool over MBA or law?
SPEAKER_01 (12:35):
Oh, that's shadowing experience.
It was a single time in the inthe OR.
So I had taken the LSAT, uh, I'dtaken GMAT.
I I had applied along withmedical school to a bunch of a
bunch of those programs.
And um I went out and you know,in terms of strength, you know,
(12:56):
knowing what I wanted to do andof course strengthening my
application, uh, went lookingfor actual shadowing
opportunities.
And the one I came across umhere in town through a family
friend, happened to be aneurosurgeon.
And uh, you know, most ofneurosurgery uh is is spine
(13:16):
surgery, especially in privatepractice.
But this turned out to be abrain tumor case.
And I was like, this is amazing,holy moly.
And uh, and so uh yeah, thatthat singular shadowing
experience while I was deciding,yeah, totally totally.
SPEAKER_03 (13:33):
So it almost sounds so it almost sounds like it was
the aha moment when you werethere.
Definitely what I want to do.
SPEAKER_01 (13:40):
Yep.
I went into I went into applyingjust being like, I am good at
school.
Yeah, uh, I was very good inhigh school.
I don't know what I want to do.
It seems safe to stay in school.
Like I went into it very scared.
Uh, and it was, you know, meapplying to all these schools
(14:00):
was very much just driven kindof by fear.
Like, I I am not gonna make itin Hollywood.
Uh, I have to do something else.
School seems safe, but afterthat aha moment, I was like all
in on medicine.
It it totally changed myperspective, yeah.
SPEAKER_03 (14:15):
Right.
I and I can I can definitelyassert that uh you don't
necessarily become aneurosurgeon by fluke.
There has to be some focusthere.
But um, you know, I mean, uh canwe switching gears?
Do you said that you actuallyare an entrepreneur now?
Can you tell us a little bitabout your business?
SPEAKER_01 (14:34):
Yeah, so this is my first company.
It's a startup.
So from scratch, uh, it is atech company, and I'm not in
Silicon Valley.
So we'll see how this works out.
But um, like I said, I I havemaybe I've been one-shotted, but
um, but I'm all in on the AIhype.
(14:56):
Okay.
I really think it is uh gonna betransformational for healthcare.
It's as big as the internet,it's gonna be as big as mobile
phones, it's uh I think really,really gonna alter society over
the next few decades.
And so um uh my company is uh isa telehealth platform.
(15:18):
There are a million of those,but it's AI native.
So connect your connect yourmedical records, right?
Patients can connect theirmedical records, have that as
context for AI, just forinformational conversations.
Hey, what does this mean?
What were my last labs?
But when you need a physician,get connected seamlessly,
(15:40):
basically in chat.
Um so I'm very much like focusedon kind of urgent care, uh
simple primary care.
When we get off, we're in betaright now.
Um but this is this is my firstuh my first entrepreneurial
steps.
And I know you have a lot ofexperience, but it's uh it's
(16:02):
it's different.
Even though I was in privatepractice as a neurosurgeon, so
basically kind of a smallbusiness owner for for a while
with partners.
Uh this is this is definitelysome different skills, skill
sets.
SPEAKER_03 (16:17):
Yeah, right, right.
Um, so you mentioned that youthink that AI will be
transformative.
Obviously, the jury is out onthat one.
Nobody really knows.
Uh, I mean, everybody has thesame feeling, but there seems to
be a lot more apprehension thanthere there is adoption and uh
people embracing this as a newtechnology.
(16:38):
So, as an example, I'll tell youa lot of my uh our students come
in saying that, oh my god, youknow, we're gonna be out of a
job.
Uh AI is gonna take, you know,like everybody seems to pinpoint
on radiology specifically, thatoh, AI is gonna do so much uh
better than radiologists.
And, you know, I've spoken toDr.
(16:58):
uh Patel, who's the CMO of uh uhPrime Health.
Uh we've spoken to uh thepresident of the uh University
of North Florida, and their viewis pretty diametrically
opposite.
That they're saying that AI isactually going to supplement and
make physicians better.
What's your view?
SPEAKER_01 (17:17):
I think it's gonna be a mix, but um but I think
clearly, right, this kind offear-mongering about job losses
and things, at least in the inthe short term, has has been
doomerism.
There was a there was a recentgreat article um looking, you
(17:38):
know, uh some very famous peopleuh in the AI field, obviously
famously, you know, five plusyears ago were like, oh,
radiology is right for thepickings.
It is true that most uh AIsoftware as a medical device
that have been approved by theFDA are AI uh radiology tools.
(18:02):
But um this great article, it'slike two weeks old, but looking
at the average radiologist's youknow compensation, uh, which is
outstripped the averagephysician compensation.
Right.
Like it doesn't it uh AI rightnow is not taking radiology
jobs.
Um and I I think uh you knowthere's a there's a lot that
(18:26):
radiologists do just besidesjust looking at images beyond
just the the technical aspectsof hey, out of the lab, how do
these radiology models perform?
And obviously, you know, an LLMcan't consult, you know, with
another physician, it can'tperform procedures, it can't so
(18:49):
I um you know I would never Iwould not be telling people who
are considering medical school,hey, don't because AI is is
gonna take physician jobs in 10years.
I I just don't think that's thecase.
SPEAKER_03 (19:05):
Right.
Um and and I think uh, you know,we've spoken to some uh leaders
of some large hospital systems,and uh, they've always obviously
echoed the same thing thatradiologists tend to make a lot
more money.
Uh however, their feeling wasthat they're not gonna hire less
radiologists.
SPEAKER_01 (19:21):
Yeah, I agree.
SPEAKER_03 (19:22):
Their expectation is that they're expecting the
radiologists to do more.
Um, a case in point, uh uh thisis a little uh deviated from uh
what uh medicine, but uh I havea friend who works at uh one of
the largest uh truckmanufacturers in the world.
(19:43):
And they are in the cutting edgeof technology as well, where
they're developing AI trucks, etcetera, et cetera.
So I asked him, does this meanthat truck drivers are going to
become obsolete?
And he said that, well, we'llneed less truck drivers, but
what this also means is that ifyou are a truck driver, you just
cannot really drive a truck.
You have to have a morebroad-based skill where you have
to do a little bit of codingbecause if we are doing uh truck
(20:05):
caravans that are AI poweredwith the first, second, third
trucks all AI driven followingthe first truck.
If something goes wrong, and itdoes always, the person driving
the per uh in the first cab hasto be able to service this.
So just having a CDL is just notgoing to be good enough in down
the road.
Um and and I think you know,when I spoke to these uh
(20:27):
hospital system leaders, a lotof them have suggested the same
thing that radiologists will bethere's not going to be any cut
in compensation, they're notexpecting to hire less, but they
are expecting in due course oftime radiologists to be able to
do more.
Um would you what are yourthoughts on that?
I and and we've seen that inGoogle and some of these IT
(20:48):
software companies too, wherethey're expecting the production
to triple quadruple per softwareengineer.
SPEAKER_01 (20:54):
Yeah.
Well, we'll see, even as even asAI writes what you know,
whatever these these techcompanies claim 90% of of code,
right?
But um yeah, I I will tell youthat that is that is probably my
more realistic fear.
Um that uh you know what youknow, as you have these clinical
(21:23):
decision support tools or thesecomputer assisted assisted
diagnosing tools that likeradiologists can use, what it
what's that gonna do forinsurance reimbursement and um
the expectation on like theradiologist work workflow and
work workload, right?
(21:43):
Like, oh the the computer'sdoing some of this for you, read
three times as many films,right?
Um so uh we'll s yeah, that thatmight be how this plays out.
Um I do think that uh as thesetools proliferate, just having a
(22:04):
basic understanding of them, AIin general, um is is gonna be
important for physicians.
Okay.
SPEAKER_03 (22:13):
Uh so uh in your specific field, uh surgery and
neurosurgery more specifically,do you think AI has impacted
your field directly yet?
SPEAKER_01 (22:26):
I do.
I'll give you I'll give you acore example.
So I I even have a little, I'meven more subspecialized in
neurosurgery.
I I do mostly neurovascularstuff, so blood vessel diseases.
Um and so big strokes, likethrombectomies, brain aneurysms,
things like that.
(22:46):
And this is very, very niche,but I'll give an example where
AI has mattered a lot to mypatients.
So I take a a lot of strokecall.
Um in stroke call over the past10 years, uh for large vessel
occlusions, so when a big bloodvessel gets blocked to the brain
(23:10):
and is causing a stroke, there'sbeen this paradigm shift where
we know that much like whenpeople have a heart attack and
the cardiologists go in andreopen the vessel, that
reopening these large vesselsemergently can uh prevent people
from having a stroke or keeptheir stroke small.
(23:31):
And so that uh that is a veryemergent thing.
So people will come into the ERwith a stroke, they will get
these this imaging, a head CT, aCT angiogram that looks at the
blood vessels of the brain, andthey get that um pretty much a
standard of care urgently.
And if you see a blocked bloodvessel, right, you have to
(23:53):
activate the team, activate me,get them over to the angiosuite
to get the blood vessel open.
And there have been aproliferation of tools, even
though radiologists uh areobviously very good at detecting
these blocked blood vessels.
These um these apps can do thatin 15 seconds, right?
(24:17):
What used to maybe take 10 to 15minutes of kind of radiology
workflow.
Oh, the images get pushed overto a radiologist, the
radiologist looks at it, theradiologist calls the ER doc,
the ER doc calls me.
Now uh you get the scan and theAI like pings immediately.
(24:37):
There's a blocked blood vessel,and it just shows up on your
phone.
And and even those 15 or 20minutes, right?
We know from the literature uhcan affect outcomes in terms of
the chance of people doing well.
So that's one place that I useweekly that AI has really made a
difference.
There's no doubt.
SPEAKER_03 (24:57):
Oh wow.
And um has AI reduced theworkflow though for you?
Do you think so?
SPEAKER_01 (25:06):
I think I think it's in that very, very niche
example, yes, it has made usmore efficient.
There's no doubt.
I use a lot of other AI tools,so other popular ones, right?
Um clinical decision supporttools like open evidence, right?
Oh, this is, you know, put in,you know, ask questions rather
(25:26):
than going just to PubMed andsearching.
Uh I do use an AI scribe.
It's kind of plus or minus,right?
These are tools that listen toyour conversations in clinic
with patients and then spit outa note for you.
Um, I think that's a little bithelpful.
But I I would say that um thatthat probably that stroke
(25:47):
example has been the mostmeaningful in terms of for my
patients.
SPEAKER_03 (25:52):
Awesome.
Um, so one thing I do want totouch upon and switching gears
to like more um uh less, youknow, like internal medicine and
family medicine.
One thing I get a lot ofquestions from from students is
that um, you know, a lot ofpatients are using uh ChatGPT to
(26:12):
really diagnose initial problemsthat aren't major medical
emergencies.
That, oh, I have a skin spot,this is the situation, they take
pictures, chat GPT gives a bunchof diagnoses, uh, et cetera, et
cetera, which has made a lot ofum students who want to do
family medicine or internalmedicine really nervous about uh
their professions.
(26:33):
What are your thoughts on that?
SPEAKER_01 (26:35):
Yeah, I th I think that's uh an increasing use
case.
It's even kind of it's even kindof the principle that that I'm
building a company around,right?
Um, but I don't think we arenear uh automating away
physician knowledge work.
Um uh and so I I do if I wasgonna go into a mostly
(27:00):
knowledge-based field, uh evenyou know, primary care or a
medical subspecialty.
I I think uh the future isbright.
Um even with AI.
Uh I don't think uh these toolsare, even though they're
changing rapidly, who knows whatit looks like in five years, but
(27:22):
I don't think they're they'rethere in terms of uh safety.
Um I think there are gonna behuge non-technical things to
replacing physicians (27:33):
societal trust, reimbursement, liability.
I do think these tools, youknow, 10 years from now might
how we practice medicine look alittle bit different.
Um AI is triage or things likethat.
Uh yeah, of course.
(27:54):
But uh I think there is a longfuture for primary care, uh,
even with AI.
SPEAKER_03 (28:01):
Okay.
And um so so obviously we'vetalked a lot about what AI can
do.
Um but I I think that there aresome downsides, and I'd like to
talk about that.
What you feel like the downsidesof AI is.
One thing that I have seen isthat the proliferate and as this
is not just AI, I think it'sjust generally information.
(28:24):
The proliferation of informationum is good and bad.
We we can see some of the uh badconsequences of democratization
of information in society.
But one thing I have noticed isthat there seems to be a lot
more pushback from the patientside in the sense that a patient
will go in and they have somesort of a diagnosis in their
(28:46):
mind that Chat GPT told them.
And the the the physician willcome and tell them that no, I
don't think this is what is it,what it is.
Uh this is probably more likely,and the patient comes like, oh
yeah, the physician doesn't knowanything.
Uh I trust ChatGPT.
And I think that that is anincredibly dangerous aspect, and
(29:06):
I and I dub it as uh the age ofexpertise is over to a certain
extent, because now people thinkthat I could type in a prompt in
ChatGPT and get all theinformation I need, or I can
watch a uh five-minute YouTubevideo that ChatGPT pulls out for
me, and I'm just as good as umuh any physician or any other
(29:28):
specialty profession.
Uh, what are your what so Ithink I feel like that's a
pretty big downside.
Um, I don't know if you'veexperienced that and what your
views on um that is.
SPEAKER_01 (29:43):
Yeah, I to be honest, even when I heard you
talking, I was about to about tosay we've even before you use
the term, we've we've enteredthe age of of anti expertise,
right?
Yeah.
Um I I don't I don't thinkthat's uh You know, I I think uh
that a little bit uh predates uhAI.
(30:06):
Um I mean, not to get intopolitics right, but you're you
see, you know, RFK yesterdaybasically saying, hey, we don't
we uh we shouldn't trustexperts, that's not science.
Um and so I do think there's uhthere's I do think that's uh
there's a lot of that and thatAI might might make that worse.
(30:30):
I do think it's kind of abroader societal trend though,
uh not just AI.
Um you know, even when it'sfrustrating, even when it uh
seems personal, that I have justcome to be like have grace for
patients.
There is nothing wrong withpatients, even if it's it seems
(30:53):
misdirected to you, like guidedby AI, of them going and getting
a second opinion.
You know, I think we have totrust our our colleagues to to
handle it the same the same way.
Um but it can be a frustratingaspect of practice, there's no
doubt about it.
SPEAKER_03 (31:13):
Well, I mean I will I will go out on a limb and say
this is not just for medicine,it's for every profession.
Um people, you know, um I'm aCPA by training.
I I have a lot of colleagues whoare accountants, and they come
and say that they ran it throughAI.
And in some cases, this is true,this does work, but they're
like, oh, my tax refund could beso much larger because AI told
(31:34):
me so.
Um and there are certain nuancesthat they don't understand,
there are certain physicalitiesthat they don't understand.
So, for example, in certaincases, um, you know, an
accountant might know better thefamily that just had a kid, and
AI missed that because theyeither they didn't put it in the
prompt, or you know, I mean it'sthe AI is as good as the prompt.
(31:55):
And I think that that humantouch is pretty necessary for
almost every profession, and Ithink more so for physicians
than than any other profession.
But uh one thing that I'venoticed is that AI is absolutely
terrible at managing chaos.
Um, it cannot handle chaos, andthat's and uh let's face it,
(32:18):
human beings are chaotic, ourthoughts are chaotic, our
thoughts are not organized,logical, and streamlined.
Um, as you mentioned, RFK Jr., Iwould say that his thoughts are
pretty chaotic too.
Um, and and that's what uh tricktrickles down.
Do you see any majordisadvantages of our adoption of
(32:40):
AI beyond obviously the societalimpact?
unknown (32:44):
Yeah.
SPEAKER_03 (32:45):
From a technical perspective.
SPEAKER_01 (32:46):
Yeah, I mean, uh, technical challenges with AI,
where we stand right now.
I mean, with the caveat thatthings are still continue to
change very rapidly, whateveryou think about the newest
models.
I mean, uh, I completely agree.
Um, AI is pretty poor, even whenyou try to design tools around
these models, at going out andseeking information, okay.
(33:10):
Which physicians are great at,right?
You know, uh patient comes intothe exam room, medical student
is interviewing them, theymention something, oh, this
triggers something there, I'mgonna follow up, right?
Maybe this wasn't the chiefcomplaint.
Or you know, AI is amazing atsynthesizing data, um, but not
(33:31):
seeking it out itself.
Uh it's also not very novel.
Um, you know, that maybe that'sgetting better as as as it tries
to get into science, but it youknow, things it has uh not seen
before and not been trained on,it's not it's still not great
with.
Um and so these are these arethings that I think physicians
(33:56):
are still much, much, muchbetter at.
Yeah.
If you give AI all the context,right, yeah, it's it's great at
synthesizing it as long asyou've given it everything it
needs.
But um, yeah.
SPEAKER_03 (34:10):
Okay.
And um do at some point, I wouldassume AI, AI's been
transitioning in leaps andbounds.
Just two weeks ago, there havebeen updates across all the AI
platforms um that haveessentially, you know, replaced
a lot of uh or will be replacinga lot of consultants because uh
(34:31):
as long as you can communicatewith AI appropriately, it can do
uh Excel sheets, it can do procharts and and and all of that.
Um but do you feel that that'llchange at any time in the near
future, in the distant future,where we don't need to go to a
physical person, but we're justsitting and I don't know,
talking to a chat GPT persona?
SPEAKER_01 (34:54):
Um we may get there, uh especially in terms of like
initial care.
You know, you I can I canenvision a you know, a kind of
medium-term future, maybe wherephysicians do a lot of triaging,
a lot of like edge cases.
(35:16):
Hey, the AI is unsure, hey,review the AI's work, um click
in when when it needs more.
It that might happen.
Um I will say, you know, we justin a broader sense, uh there's
there's a lot of talk about AItaking jobs.
(35:38):
Um I uh not just for physicians,but I just think historically,
right?
Um with uh as technology hasbecome you know, everything of
of the twentieth andtwenty-first century, right?
(35:58):
The driving force.
Even as technology has advancedit has it has really shifted the
type of work that humans do.
Um but it's it's not like youknow the internet killed male
mail carriers or something likethat.
(36:18):
And so um I think AI may reallychange how physicians practice
and maybe how uh patients accesscare.
Probably not in the in the inthe very, very short term, but
longer term.
But I don't think um physiciansare gonna become like dinosaurs
(36:42):
at all.
Right.
SPEAKER_03 (36:43):
Right.
I I w I would definitely agreewith you.
Uh one thing I do want to saywhen you talk about jobs, from
everything that I've spoken to,uh everybody I've spoken to, it
really seems like uh these largehospital chains are not
necessarily interested incutting or looking to cut any
physician positions.
(37:04):
What they are looking to changeis they are looking to change
the headcount and looking to uhuse AI and leverage AI as much
as possible, but they're lookingto eliminate a lot of
lower-level administrativepositions.
And um, but but because of that,do you anticipate the nature of
(37:26):
physicians changing wherethey're doubling up and doing a
lot of the functions that uh notall of it, but a lot of the
functions that potentially uh anurse is doing now?
Hypothetically, I'm you know,because uh because uh they might
have more free time now becauseAI is doing some of the analysis
(37:51):
work or some of that uh busywork, as you said, for
radiologists.
They're going through many, manyreports, and something that used
to take 10 minutes is taking 30seconds now.
Uh and and if there is not anincreased patient volume,
they're looking at uh doing morejust like a truck driver is
(38:12):
being considered to essentiallybe a software engineer as well,
or a software engineer is goingto be driving trucks also
nowadays.
Um, do you anticipate that kindof a world where a physician is
also doubling up as uh doing alot of the basic work just
because they have more time?
SPEAKER_01 (38:30):
Um that's a really good question.
I uh I had never considered likephysicians doing more.
I mean, not that this is whatnurses do, but more more like
scut work.
I I I guess I'm probably alittle bit skeptical of that.
Okay um I'm interested in, youknow, I don't I want to be
(38:55):
careful, but um you know, healthhealthcare has you know, we're
all in this to help people, buthealthcare as a business does a
pretty remarkable job of uhcreating its own demand, right?
(39:19):
So, you know, I think a lot ofpeople um look look at medicine
as some sort of fixed thing.
When you're sick, you go to thedoctor, but I uh you know, I'm
not sure that's a hundredpercent how and you know, and
people only get sick so much,but I'm not sure that's a
hundred percent how it works inthe real world.
And so um I think there willalways, you know, in the in the
(39:42):
medium term, uh even if AI makesus more efficient, that there's
there's gonna be enoughpatience.
Um and so I think it's probablymore realistic that like we
talked about, uh, you know, Ithink it's gonna be important
for physicians to know how touse these tools, troubleshoot
(40:04):
these tools, absolutely.
Um and I think like we talkedabout with the radiologist, that
maybe there's demand to be formore throughput, more be more
productive.
I can I can envision that.
SPEAKER_03 (40:21):
Okay.
Um and and I think this kind ofgoes back to the uh the the
pyramid structure of ourhealthcare system, where at
least in the United States, umthey focus a lot more on primary
care in the beginning, so thatspecialists have less let's not
(40:42):
call it a workload, but uh lessacute patients at the latter
end.
Because if you're dealing withhypertension at a relatively
early age at the PCP level, uhyou're probably less uh likely
to develop complications thatyou would be seeing uh at your
level.
And that's generally thepyramid, but uh I think what
you're suggesting that makes alot of sense where um maybe the
(41:08):
level of care is increased tohave a healthier population.
Um I definitely think thatthat's possible.
That and and I and that makes alot of sense to me.
Um so I want to switch gears alittle bit, and since we are an
educational organization, whatdo you think?
And and I have to caveat this bytelling you that education moves
(41:31):
at a glacial pace.
Um, I can tell you right now, somany conversations I've had with
faculty members, programdirectors, um people who are in
various leadership roles invarious uh residency programs
who absolutely abhor AI.
They have clearly said that ifyou are using AI, you're
probably an idiot.
(41:52):
You don't know what you'redoing.
I want people to do all of thesethings hands-on.
So there's still a lot ofpushback, and I think that
that'll stay for at leastanother half decade or so, if
not longer.
Um and even from our accreditorsand the accreditation
standpoint, um, there have beenno mandates on any particular
(42:15):
course that's tech-driven,AI-driven.
Um they are, in fact, when wesay that we want to introduce
some of uh some courses likethis, the general reaction is
that you know, we have acurriculum, we want to stick to
it, we don't want to take awayfrom what we have.
So from the educationalstandpoint, at least uh general
(42:38):
leadership, we don't necessarilysee a whole lot of movement on
what schools should be doing totrain our physicians um
appropriately.
What are your views on whatschools should be doing to train
our the physicians of tomorrow?
SPEAKER_01 (42:57):
I think training people on tool use is uh
absolutely important even rightnow, right?
And and you know, you'retraining young people who are
generally more technologicallyinclined, right?
They're probably learning thesetools already um by themselves.
(43:19):
I uh, you know, uh it'll bereally interesting how AI comes
into education, even medicalschool education.
I am I am pretty sympathetic,and I've heard similar arguments
that that you have.
I'm pretty sympathetic to theidea of using, you know, of what
(43:40):
is what are the costs of usinggenerative AI.
Uh like you know, I thinkthere's a lot of value in terms
of critical thinking andorganizing your thoughts to uh a
simple task like writing.
And when you offload that, whatare what are kind of the
long-term costs?
And so I'm sympathetic to peoplewho really don't like that.
(44:03):
But um, you know, I I'm I'mreally curious to see how, you
know, holy moly, right?
Uh does every does every lectureneed to be live, right?
My goodness, right?
We we asked AI about this topicand it generated this whole
(44:25):
hour-long audio lecture andslides itself and um or uh you
know using it to to do research,things like that.
So uh yeah.
I think it'll be reallyinteresting.
I mean, you're you're obviouslya lot closer to it than I am um
in the education field.
(44:46):
Uh I do think you have to becareful when you're trying to
offload your critical thinkingto it.
I think it is an amazingresearch tool.
It can be amazing for likehelping you learn, um, for
synthesizing stuff.
But if you're like, well, I'mjust gonna have this AI write
(45:09):
this essay for me, um I I don'tknow.
SPEAKER_03 (45:14):
Right, right.
Well, I mean, uh absolutely soso that's just being lazy.
Like if you're disparating AI tocompletely substitute something,
but um so one of the challengesthat we've had is that right now
AI is pretty disparate.
You have ChatGPT, you haveClaude, you have Perplexity, you
have uh Google Gemini, you haveGrok.
(45:35):
They all are different and theyall have somewhat different
answers.
And and we've tested all ofthese.
All the ones, the five that Imentioned, we've uh I have
personally used all of them.
Um and the results are notreliable in the sense that uh
obviously uh I I'm asking itrelatively subjective questions.
(45:56):
I just basically choose theanswer that I like best, that I
think fits my narrative thebest, which is obviously a
little bit of critical thinking.
Um but when you're saying if youwere to hire somebody or or if
you were to get a neurosurgeonresident, a first-year resident,
uh, for your uh in youruniversity, would you care that
(46:21):
they know what these fivesystems are and how to use them
and how to uh create a username.
Does that matter to you as as oftoday?
SPEAKER_01 (46:33):
Uh no, to be honest.
Uh I uh, you know, there wouldcertainly be expectations.
Hey, these are the tools we use.
Uh this is the ambient scribe weuse in our clinic.
This is the, you know, this isthe stroke alert tool we use in
(46:55):
our clinic or hospital.
Um, you know, learn the basicsof these.
But would I care if they do thenuance of, hey, I've discovered
Claude is better at you know atwriting and ChatGPT codex is
better at coding.
(47:15):
Uh no, I wouldn't.
Okay.
SPEAKER_03 (47:18):
So uh when you're when we're talking about AI
education in general, do youthink uh what what would matter
to you?
Like how how to write a prompt?
They have a general idea of howto write a prompt, um, having
some guidelines or guardrailsthat they personally know that
(47:38):
we shouldn't do these thingswith AI.
What what's the flow that youwould care about as an AI
entrepreneur, as a neurosurgeon,when you're hiring somebody,
what would what would matter toyou?
SPEAKER_01 (47:50):
Uh I want them, yeah.
I want them to have berelatively technologically
savvy, right?
All these, even the tools we'rebuilding on top of models, they
all work generally the same,right?
Hey, you're coming in and youknow, uh I'll give you an
example, right?
I mean, this is not a newchallenge.
(48:11):
Uh AI is not new, but mygoodness, looking at my we, you
know, we we recently, well,about a year and a half ago,
sold ourselves to a bighealthcare system.
We were private practice and webecame employed.
And that uh entailed, becausethis was entirely new employees,
(48:32):
entirely new hospital, uh, a newEMR.
And like watching my you know,60-year-old partners try to get
a handle on that, uh, you know,it's a it's a slow, painful
process.
So, you know, these a lot ofthese tools work very, very
similarly.
Being familiar, being able topick up new tools, I think being
(48:54):
able to prompt um and tocritically assess AI's
responses, right?
Oh, that that's like not right.
Um, and to be able to ask it uhquestions and give it context
where that's a lot less commonand it gives you uh answers that
you're looking for, yes, isabsolutely key uh as well.
unknown (49:18):
Okay.
SPEAKER_03 (49:19):
So so generally the concept of prompting, which I
think is really the basis andbackbone.
I don't know uh I've been usingAI for about two years now, and
I think the sophistication thatI needed to create a prompt to
give me meaningful answers twoyears ago has dramatically
reduced today.
Uh back in the day I felt likeit almost needed to follow a
(49:42):
certain formula.
Yeah, perfectly.
Yes, and whereas today I feellike I can just have a just
talking to a person and I cankeep elaborating until I get to
that answer.
So I do think AI is getting moresophisticated, but at the same
time, having that knowledge ofwhat the structure originally
should be like helps me getbetter answers.
(50:02):
So I think uh having a coursethat kind of talks to students,
not necessarily about uh what uhAI is, but generally having a
general understanding.
And you know, especially onething I've noticed is even today
we have this these token limitson AI, which means that if you
(50:23):
keep the same chat window openand you keep asking questions at
some point only as so muchcontext, yeah.
Yeah, uh and and it's at somepoint it starts spewing
gibberish because it's run outof tokens.
Um, this is something that I'vebeen a little surprised to know
that most people don't know.
So the answers that you get inthe beginning of the
conversation versus the answersthat you will get towards the
(50:46):
end, they're a lot looser.
Um AI hallucination is still afactor, which is again, I'm
surprised that a lot of peopledon't know.
But I think having a course thatreally talks about these
concepts, and again, these arechanging so much uh over time
and may not maybe completelyobsolete at some point for
hallucinations.
(51:07):
Um, and and if they have thatkind of a structure and a
knowledge, would that bebeneficial to a person like
yourself?
SPEAKER_01 (51:14):
Honestly, that'd be super beneficial.
Yes.
Okay.
SPEAKER_03 (51:17):
Uh awesome.
My final question about AI isreally do you see AI ever
leveling the playing field?
Because let's face it, today inmedicine, uh, if you're a
neurosurgeon, if you're an ENT,if you're an orthosurgeon, uh,
for one reason or the other,there is dramatically uh a lot
(51:39):
more prestige, let's say.
And and and not to say that Ithink just generally being the
doctor, uh, you know, I have myview that there are only four
real professions in the world.
Um, and I think uh being adoctor is definitely one of
them, no matter what kind of aphysician you are.
Um but I think uh do you thinkAI will ever level the playing
(52:00):
field between the discrepancybetween all these specialties
and doctors?
Number one, number two,obviously, uh in some of the
more elite fields likeradiology, anesthesiology,
surgery, the different kinds ofsurgery, uh there's still a
little bit of stigma for IMGs.
Um I know that stigma has goneaway quite a bit for medicine in
(52:21):
general.
Um, as I said, our school issince 1999.
We saw a lot of that in theearly years.
Uh, I've personally hadconversations with the program
directors who've said that ifyou're from a Caribbean school,
this is not the program to applyfor.
Um, which is, in my opinion,completely baseless because in
the United States we have astandardized system.
(52:41):
You have everybody takes thesame tests.
If you score well on thesetests, I don't see the problem
with with uh applying.
But um but do you see AI everreally leveling the playing
field and uh decreasing thediscrepancy between the various
medical professions and thenwithin the medical profession,
(53:02):
say the DOs, the M uh theCaribbean MDs and the American
MDs?
Do you see that happening?
And if so, how do you anticipatethat to play out?
SPEAKER_01 (53:13):
Yeah, I uh that's a really good question.
Um and I haven't given it a tonof thought, but I you can
envision a scenario where thatmight be the case.
I really think that if that doeshappen, it's probably probably
more medium to long term.
(53:34):
But um, you know, I as AI mayreally kind of I can envision a
scenario where AI kind of levelsum the compensation structure,
levels the prestige factor forphysicians, um and a lot of that
(54:01):
is you know kind of what drivescompetitiveness.
Uh and so if if that happens,right, if AI leads to say uh
lower compensation forprocedures or something like
that, which I could totally seehappening.
(54:23):
Um yeah, I mean, I think uh thatcould dramatically kind of
change the face of what whatconstitutes a competitive
specialty or not and change umyou know the whole residency
application process is awesomewithout doubt.
SPEAKER_03 (54:43):
Okay, well that's very interesting.
So um do you think that AI willand we've touched upon this
briefly earlier in ourconversation, but um do you
think the demand for physicians,and I know you mentioned that
the healthcare system has a uh away of creating its own demand
(55:05):
in the sense that havinghealthier, healthier people, and
and and this is very apparent inour life expectancy.
We've seen this.
Like I life expectancy, even inmy lifetime, I remember when I
first moved to the UnitedStates, the US life expectancy
at the time was uh in the late70s, today it's in the early
80s, uh, or very close to 80, ifI'm not mistaken.
Um, and I and that this isexclusively uh because uh the
(55:31):
government and physicians haveworked towards making doctors
more accessible, and and andyes, that there's a whole
healthcare and and uh insuranceargument.
I'm not getting into that, butgenerally being able to see
physicians earlier just allowspeople to get healthier.
But do you think AI at any pointwill reduce the demand um for
(55:59):
some specialties across the thethe realm of physicians?
Maybe family physicians, youdon't need to see them that many
um that much.
Uh maybe dermatologists, thatthat could be one.
If I have a skin tag, I put itup on Chat GPT and chat GPT.
Maybe your software is doingthat.
SPEAKER_01 (56:19):
I uh I really think that it's going to change how
physicians practice.
But uh I I don't think I meanI'll just stress again like if I
was going to medical schoolright now, I would not have a
major fear that oh my goodness,we're gonna have a glut of
(56:41):
primary care physicians in 10years because AI's doing too
much of this work.
Um I don't think so.
SPEAKER_03 (56:51):
Awesome.
Well, thank you so much, Dr.
Son.
I know we're out of time, but uhthis was such an invigorating
conversation, an eye-openingconversation.
I think I have some homeworkbecause I'm gonna go and create
uh uh a course for our studentsthat we uh we we actually have
some, yeah, we actually do havea very yeah, a basic course that
(57:13):
just pre-residency, we actuallytell students to get uh a little
more acquainted with AI.
But I think uh I'm definitelygonna push our academic team to
create a program that's in a uhembedded in our curriculum.
But uh it's been veryeye-opening.
Uh, before we leave, I do wantto uh ask you a little bit about
your business.
(57:33):
We already spoke about that, butthe name um and uh how do you
think uh as a school, do youthink we can help in any way?
Do you think uh some of thethings that you're doing can
actually uh help students beequipped with uh in facing the
(57:54):
real world of AI?
SPEAKER_01 (57:56):
Yeah, I'm I'm always uh happy to talk to anyone uh
just in general about medicineand career in medicine, but but
especially AI, if anyone wantsto ever reach out.
Um uh but yeah, I am working onthis company called South A
Health.
It is in beta.
Um, but like I said, it's it'skind of a patient-facing AI
(58:19):
telehealth platform, and and youknow, hopefully if in the
future, if any of your studentsare are looking for for work to
even telehealth remote work, uhlove them to reach out as well.
So after they finish residency.
SPEAKER_03 (58:36):
Thanks again, Dr.
Collinson, for joining us todayand sharing uh your journey from
neurosurgeon to entrepreneur andhow AI is reshaping the future
of medicine.
If you enjoyed this episode,make sure to follow or subscribe
to the podcast on your favoriteplatform, leave a review, and
share it with a friend orcolleague who's passionate about
healthcare and innovation.
(58:57):
You can also download pastepisodes and find updates on our
uh upcoming guests.
And remember, there is noshortcut to getting an MD.
SPEAKER_00 (59:05):
Thank you so much for tuning in to our show.
We hope you enjoyed anotherepisode of Med School Minutes.
If you like our content, pleasefollow us and receive
notification when a new show isposted.
This podcast is brought to youby St.
James School of Medicine.
For a video version of thispodcast, please check us out on
SJSM.orgslash video.
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
The Burden
The Burden is a documentary series that takes listeners into the hidden places where justice is done (and undone). It dives deep into the lives of heroes and villains. And it focuses a spotlight on those who triumph even when the odds are against them. Season 5 - The Burden: Death & Deceit in Alliance On April Fools Day 1999, 26-year-old Yvonne Layne was found murdered in her Alliance, Ohio home. David Thorne, her ex-boyfriend and father of one of her children, was instantly a suspect. Another young man admitted to the murder, and David breathed a sigh of relief, until the confessed murderer fingered David; “He paid me to do it.” David was sentenced to life without parole. Two decades later, Pulitzer winner and podcast host, Maggie Freleng (Bone Valley Season 3: Graves County, Wrongful Conviction, Suave) launched a “live” investigation into David's conviction alongside Jason Baldwin (himself wrongfully convicted as a member of the West Memphis Three). Maggie had come to believe that the entire investigation of David was botched by the tiny local police department, or worse, covered up the real killer. Was Maggie correct? Was David’s claim of innocence credible? In Death and Deceit in Alliance, Maggie recounts the case that launched her career, and ultimately, “broke” her.” The results will shock the listener and reduce Maggie to tears and self-doubt. This is not your typical wrongful conviction story. In fact, it turns the genre on its head. It asks the question: What if our champions are foolish? Season 4 - The Burden: Get the Money and Run “Trying to murder my father, this was the thing that put me on the path.” That’s Joe Loya and that path was bank robbery. Bank, bank, bank, bank, bank. In season 4 of The Burden: Get the Money and Run, we hear from Joe who was once the most prolific bank robber in Southern California, and beyond. He used disguises, body doubles, proxies. He leaped over counters, grabbed the money and ran. Even as the FBI was closing in. It was a showdown between a daring bank robber, and a patient FBI agent. Joe was no ordinary bank robber. He was bright, articulate, charismatic, and driven by a dark rage that he summoned up at will. In seven episodes, Joe tells all: the what, the how… and the why. Including why he tried to murder his father. Season 3 - The Burden: Avenger Miriam Lewin is one of Argentina’s leading journalists today. At 19 years old, she was kidnapped off the streets of Buenos Aires for her political activism and thrown into a concentration camp. Thousands of her fellow inmates were executed, tossed alive from a cargo plane into the ocean. Miriam, along with a handful of others, will survive the camp. Then as a journalist, she will wage a decades long campaign to bring her tormentors to justice. Avenger is about one woman’s triumphant battle against unbelievable odds to survive torture, claim justice for the crimes done against her and others like her, and change the future of her country. Season 2 - The Burden: Empire on Blood Empire on Blood is set in the Bronx, NY, in the early 90s, when two young drug dealers ruled an intersection known as “The Corner on Blood.” The boss, Calvin Buari, lived large. He and a protege swore they would build an empire on blood. Then the relationship frayed and the protege accused Calvin of a double homicide which he claimed he didn’t do. But did he? Award-winning journalist Steve Fishman spent seven years to answer that question. This is the story of one man’s last chance to overturn his life sentence. He may prevail, but someone’s gotta pay. The Burden: Empire on Blood is the director’s cut of the true crime classic which reached #1 on the charts when it was first released half a dozen years ago. Season 1 - The Burden In the 1990s, Detective Louis N. Scarcella was legendary. In a city overrun by violent crime, he cracked the toughest cases and put away the worst criminals. “The Hulk” was his nickname. Then the story changed. Scarcella ran into a group of convicted murderers who all say they are innocent. They turned themselves into jailhouse-lawyers and in prison founded a lway firm. When they realized Scarcella helped put many of them away, they set their sights on taking him down. And with the help of a NY Times reporter they have a chance. For years, Scarcella insisted he did nothing wrong. But that’s all he’d say. Until we tracked Scarcella to a sauna in a Russian bathhouse, where he started to talk..and talk and talk. “The guilty have gone free,” he whispered. And then agreed to take us into the belly of the beast. Welcome to The Burden.