May 2, 2025 • 12 mins
John Reites and Jeremy Franz explore the practical applications of AI, focusing on its use in data analysis and research. Jeremy shares his extensive experience with AI, discussing the evolution from classical NLP to modern large language models (LLMs) like GPT-3. They delve into various use cases, including how AI can assist in synthesizing information, generating insights from data, and transforming complex data into engaging presentations. The discussion emphasizes the importance of understanding AI as a tool that can enhance productivity and decision-making in various fields.
Click here to message/text me your insights and ideas for future episodes
Thank you for joining Inclusion Criteria: a Clinical Research podcast hosted by me, John Reites. This is an inclusive, non-corporate podcast focused on the people and topics that matter to developing treatments for everyone. It’s my personal project intended to support you in your career, connect with industry experts and contribute to the ideas that advance clinical research.
Inclusion Criteria is the clinical research podcast exploring global clinical trials, drug development, and life‑science innovation. We cover everything clinical research to deepen your industry knowledge, further your career and help you stay current on the market responsible for the future of medicine.
Our episodes discuss current industry headlines, career tips, trending topics, lessons learned, and candid conversations with clinical research experts working to impact our industry everyday.
Watch on YouTube and listen on your favorite podcast app. Thank you for supporting and sharing the show.
Please connect with me (John Reites) at www.linkedin.com/in/johnreites or www.johnreites.com.
The views and opinions expressed by John Reites and guests are provided for informational purposes only. Nothing discussed constitutes medical, legal, regulatory, or financial advice.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
SPEAKER_00 (00:07):
We're spending time today talking about use cases in
AI.
And again, this premise ofpractical AI is really
important.
And the reason Jeremy is on isone, because I work with Jeremy
every single day.
Two, because he knows thisinside and out.
He's been working on it for avery long time.
And he won't call himself this,but I will.
This guy's a guru.
Jeremy, tell everybody what yourexperience has been with AI,
(00:27):
what you've been doing, becauseyou've been doing it for a long
time.
But I think bringing thatcontext would be helpful for
people.
Yeah, I mean, I think...
It really starts with trying touse technology to process a lot
of unstructured orsemi-structured information back
in the day with earlier days ofEnvibe.
And we were always looking attechnology and AI and how can we
(00:50):
use this to help synthesizeinformation?
How can you use this to help usanalyze?
How can you use it to help ouranalysts analyze the data?
And we built a lot of softwarearound that stuff.
But on the AI front, we werebanging our heads against the
wall for a long time and usingclassical NLP approaches.
And a lot of it was building allthese manual rules and it was
(01:12):
just, it was very complex.
It took a long time and itusually didn't meet the criteria
of what we considered acceptableto put it in front of customers
and users.
And when this thing called GPT-3came out and There was this
private beta that we applied foraccess for and had to kind of
(01:32):
make our case as to why we had agood use case for it and how we
were going to make sure we hadthe proper safety and protocols
and weren't going to use it foranything harmful.
You ended up getting access tothat program through OpenAI and
trying a lot of the tasks thatwe were using classic NLP
(01:55):
approaches for and it justworked for a lot of them.
And so ever since then, we'vekind of been on this journey of
coming up with harder and hardertasks for AI to tackle and what
would provide value to ourcustomers.
And sort of now the challenge iswhat's the next task that can't
(02:15):
be done with today's models thatprobably will be able to be done
by be performed by the nextgeneration of models.
And we've seen that throughseveral waves of new foundation
models.
And it's been a really fun andeye opening experience for me
kind of seeing that all unfoldthe past four or five years.
I want to give you a lot ofcredit, like, because obviously,
(02:37):
like we work together, we're inthe same company's disclaimer,
that's what Jeremy and I doevery single day.
At the same time, predatingThread, when you were working in
AI, I remember you and I werehaving a conversation.
You were talking about AI.
And again, my brain goes to whateverybody's brain went to five
and six years ago.
Robots, Terminator, autonomousvehicles, like the nerd side of
me went that route.
(02:58):
And then you were explaining tome what you were trying to do
with AI.
And that was a real light bulbmoment for me, actually.
It really started to change myperspective on what AI was and
what it wasn't.
And so I'm hoping actually todaysome people take from you A
couple of things you say thathelp construct how we view AI,
what it looks like, and how wecan use it as a tool to augment
the work we're using.
(03:19):
So before we dig into the workstuff and we talk about use
cases and practicality andclinical research, Jeremy,
you're an expert in this.
This is what you do all thetime.
You know all the models like theback of your hand.
What's the party trick?
You're at the dinner table.
You want people to think you'recool.
You want to show off AI.
Give it to us.
What's the party trick we canall use?
(03:40):
to look cooler than everybodyelse with technology.
Well, I hate to disappoint you,but I don't know if I'm that
cool at parties.
Like, I lean towards really myfavorite use of AI, just my
day-to-day life.
A lot of times, if I'm stuck onsomething, if I'm thinking about
a problem or a decision I needto make, anything like that, I
will just pull out my phone,start an audio recording, and
(04:04):
I'll just speak.
And it will be veryunstructured.
It will be very just...
I may have long pauses.
I may ramble all over the place.
But I can just do that, get itout, transcribe it, send it into
my favorite LLM, have itsynthesize those thoughts, have
it clarify them, and put it in away that is very clear for me to
(04:27):
understand and clarifies my ownthoughts.
With writing, this is kind of acheat for me to use AI to
clarify my own thoughts and andhelp make maybe tricky decisions
or uncomfortable decisions a loteasier.
That sounds really smart andthoughtful and businessy.
Congratulations.
I've got a few party tricks.
I'm going to save some of thosefor other episodes.
(04:48):
The one that I will never forgetwas the first party trick I ever
saw on AI.
I'm at this friend's house.
It's a couple years ago.
I think it was when the OpenAIapp first came out.
We were getting asked because wewere the nerds at the table.
What are these things with AI?
And again, more Terminator, moremacro level stuff.
This friend of mine said, I useit for my wife's grocery list.
(05:10):
And I think we were all like,what are you talking about?
And he's like, well, she had toshop for like 50 or 75 things
for this meal coming up.
And so I loaded it all and toldhim what grocery store I was
going to and laid out this mapto save me like 30 minutes at
the grocery store.
Aisle three, go buy these fivethings.
Aisle four, go to these.
Skip a five.
(05:30):
And I will never forget theamount of people that were like,
I'll buy that.
And for me, I was kind of like,oh, that's pretty cool.
And then after people were like,could you tell me how you got
the grocery list to work?
So for what it's worth, I feellike there's a ton of party
tricks.
And that was one that reallystuck out.
Fifteen gears, though, I want totalk about practical use case of
AI in research and what we'redoing every single day.
(05:51):
And so maybe if you could, canyou get one or two really
practical examples of what yousee sponsors, CROs, different
types of clients that we have?
How are they actually takingadvantage of AI?
How are they actually using itin their companies today?
Probably the number one thingthat AI and specifically large
language models are good at thatget used a ton is just
(06:13):
processing a lot of information.
You can throw tons of text, tonsof data at an LLM and it could
synthesize that.
You can even specify what kindof output you want out of that.
The most common example would besummaries and summarization is
sort of the kind of list appequivalent of an AI feature.
(06:38):
It doesn't mean it's notincredibly useful and powerful,
but I think what's really greatis providing a lot of
information to an AI and thenasking for the output that you
want.
If you're processing interviewdata like we do at Envibe and
you want a few interesting clipsor quotes around a specific
topic, ask for that.
You'll get it.
I think it's just Having an ideaof what you want out of it and
(07:03):
asking for it like you would aska human goes a long way.
I think one of the examples youwere talking about with a client
just the other day.
So I'm just going to say itbecause I thought it was a
really good one.
It's current is you got aclient.
They want to get some insightson the recruitment campaign.
So they load up the survey ofquestions that we're going to
ask, you know, like or scale,but also mostly via their voice,
(07:23):
right?
We want people to respond andtalk to our app with their
actual voice.
You recruit the patients.
They come in.
They talk to the platform.
They give their feedback.
It hits the system.
And then we need to start givinginsights on it, right?
We need to start being able togive insights so that that
client can make decisions on ifthey should change anything, if
the material is working outgreat, right?
(07:44):
Those types of examples.
Because I prompted you for this,pun intended, because I prompted
you.
Why don't you show people whatthat looks like?
I mean, I think the easiestthing to do here is you go in
here and we have this suggestquestions to ask.
button here.
And really all this is doing iswe're telling the AI, hey,
here's all this source data fromthis study that we conducted.
(08:07):
What are some interestingquestions to ask that lead to
unexpected or interestinginsights about this study?
And, you know, you see the firstone here is about the lung bar
puncture requirement from thisinformed consent document.
And, you know, it turns outthere was a highly emotional
reaction to this.
There was a lot of risk This isa scary procedure, and it was
(08:31):
something that had to be donemany times as part of the design
of this study itself.
It's a really quick way touncover something that you
wouldn't have necessarily knownto ask.
Unless you had a good idea aheadof time, you wouldn't have
known, tell me about the lumbarpuncture requirement.
But that's the great thing aboutasking AI for what you want is a
(08:54):
lot of times it can give youwhat you're looking for.
That's my favorite example,actually.
I don't know if I've ever toldyou this before, but sometimes
we look at the data and I'mlike, where do I start?
And I love that AI can promptyou in real time.
And so I think that's a greatexample.
And for me, the number onequestion is, how does it
actually help me make it usable?
(09:16):
Because the data in itself isn'tusable.
I think it's one of the mostpowerful things that AI does.
And I don't know if people haveseen as many examples as maybe
are really out there in the realworld around how to summarize
the data, how to actually useit, how to put it in a
presentation mode that actuallyhelps.
So like I prompted you before,could you just show some
examples, just give people someideas of what they can do and
(09:36):
how I can help them visualize ormake their data useful so they
can present it to other people?
At the end of the day, You don'twant to be sending colleagues a
link to your chat session andhave them read this text
necessarily.
It's not the most effective formof communication.
So one thing that we did is weadded this creative video
(09:58):
presentation tool.
Same sort of idea at play hereis we have the AI suggesting
interesting topics to make apresentation about that we think
would be of interest tostudents.
people using this tool.
You can go in and you can selectone or enter your own.
And then it goes through thiswhole AI-assisted process of
(10:23):
ultimately creating a videopresentation with slides, with
voiceover and splicing in, youknow, voiceover content from the
actual participants whoresponded to your study.
And the user has control overit.
You can make it exactly how youwant to make it, but The great
thing about AI is that we canprovide a happy path for you and
(10:45):
we can suggest, you know, thekey findings here that we think
are most relevant to the topicthat you provided.
And you can go through thiswhole step process that you see
at the top here, where after thekey findings, then you select
the actual quotes that you wantto use to support those key
findings.
And then it'll write the slidesfor you.
(11:06):
It'll write the voiceover andyou click a button.
And then a few minutes later,you know, you get a a video
presentation that you candownload and share with your
colleagues.
The important thing for us wasto get out of just being text.
You know, we want to be morethan just text in a chat
session.
And how do we take advantage ofthe richness of our own data?
(11:29):
You still have a human incontrol, but AI is heavily
assisting and can be anassistant to varying levels
depending on how much the userwants to control in this
situation.
I do like the end.
I mean, it's always, I likepresenting my stuff, but some
people don't.
And so I think it's kind ofinteresting to me that one of
the most practical use cases isactually AI presenting the data
(11:52):
and the findings, right?
Doing a voiceover, doing anavatar.
One of my favorites is when youdid one of those and someone was
like, oh, I really, I don'tremember the name of it.
They were like, I really loveAmber in the corner.
You know, she's a great.
We were like, who's Amber?
We don't know who Amber is.
And the whole time we're like,oh, that's the AI.
And people look at you like,that's not an AI.
That's Amber on the team.
(12:13):
We're like, no, that's an AI.
It's for your voiceover.
I really appreciate these,Jeremy.
There's lots more to talk about.
We'll come back and talk aboutmore of these.
But I think just givinglisteners just a sense of what
kind of practical AI is outthere is helpful, right?
Because this stuff ismeaningful.
It's supportive.
It can help your work.
And it doesn't have to beoverwhelming.
Absolutely.
(12:35):
Yeah.
Thanks for having me on.
Well, hey, thanks for being ontoday.
I appreciate it.
I'll catch you soon.
Thanks, Jeremy.
We're spending time today talking about use cases in
AI.
And again, this premise ofpractical AI is really
important.
And the reason Jeremy is on isone, because I work with Jeremy
every single day.
Two, because he knows thisinside and out.
He's been working on it for avery long time.
And he won't call himself this,but I will.
This guy's a guru.
Jeremy, tell everybody what yourexperience has been with AI,
(00:27):
what you've been doing, becauseyou've been doing it for a long
time.
But I think bringing thatcontext would be helpful for
people.
Yeah, I mean, I think...
It really starts with trying touse technology to process a lot
of unstructured orsemi-structured information back
in the day with earlier days ofEnvibe.
And we were always looking attechnology and AI and how can we
(00:50):
use this to help synthesizeinformation?
How can you use this to help usanalyze?
How can you use it to help ouranalysts analyze the data?
And we built a lot of softwarearound that stuff.
But on the AI front, we werebanging our heads against the
wall for a long time and usingclassical NLP approaches.
And a lot of it was building allthese manual rules and it was
(01:12):
just, it was very complex.
It took a long time and itusually didn't meet the criteria
of what we considered acceptableto put it in front of customers
and users.
And when this thing called GPT-3came out and There was this
private beta that we applied foraccess for and had to kind of
(01:32):
make our case as to why we had agood use case for it and how we
were going to make sure we hadthe proper safety and protocols
and weren't going to use it foranything harmful.
You ended up getting access tothat program through OpenAI and
trying a lot of the tasks thatwe were using classic NLP
(01:55):
approaches for and it justworked for a lot of them.
And so ever since then, we'vekind of been on this journey of
coming up with harder and hardertasks for AI to tackle and what
would provide value to ourcustomers.
And sort of now the challenge iswhat's the next task that can't
(02:15):
be done with today's models thatprobably will be able to be done
by be performed by the nextgeneration of models.
And we've seen that throughseveral waves of new foundation
models.
And it's been a really fun andeye opening experience for me
kind of seeing that all unfoldthe past four or five years.
I want to give you a lot ofcredit, like, because obviously,
(02:37):
like we work together, we're inthe same company's disclaimer,
that's what Jeremy and I doevery single day.
At the same time, predatingThread, when you were working in
AI, I remember you and I werehaving a conversation.
You were talking about AI.
And again, my brain goes to whateverybody's brain went to five
and six years ago.
Robots, Terminator, autonomousvehicles, like the nerd side of
me went that route.
(02:58):
And then you were explaining tome what you were trying to do
with AI.
And that was a real light bulbmoment for me, actually.
It really started to change myperspective on what AI was and
what it wasn't.
And so I'm hoping actually todaysome people take from you A
couple of things you say thathelp construct how we view AI,
what it looks like, and how wecan use it as a tool to augment
the work we're using.
(03:19):
So before we dig into the workstuff and we talk about use
cases and practicality andclinical research, Jeremy,
you're an expert in this.
This is what you do all thetime.
You know all the models like theback of your hand.
What's the party trick?
You're at the dinner table.
You want people to think you'recool.
You want to show off AI.
Give it to us.
What's the party trick we canall use?
(03:40):
to look cooler than everybodyelse with technology.
Well, I hate to disappoint you,but I don't know if I'm that
cool at parties.
Like, I lean towards really myfavorite use of AI, just my
day-to-day life.
A lot of times, if I'm stuck onsomething, if I'm thinking about
a problem or a decision I needto make, anything like that, I
will just pull out my phone,start an audio recording, and
(04:04):
I'll just speak.
And it will be veryunstructured.
It will be very just...
I may have long pauses.
I may ramble all over the place.
But I can just do that, get itout, transcribe it, send it into
my favorite LLM, have itsynthesize those thoughts, have
it clarify them, and put it in away that is very clear for me to
(04:27):
understand and clarifies my ownthoughts.
With writing, this is kind of acheat for me to use AI to
clarify my own thoughts and andhelp make maybe tricky decisions
or uncomfortable decisions a loteasier.
That sounds really smart andthoughtful and businessy.
Congratulations.
I've got a few party tricks.
I'm going to save some of thosefor other episodes.
(04:48):
The one that I will never forgetwas the first party trick I ever
saw on AI.
I'm at this friend's house.
It's a couple years ago.
I think it was when the OpenAIapp first came out.
We were getting asked because wewere the nerds at the table.
What are these things with AI?
And again, more Terminator, moremacro level stuff.
This friend of mine said, I useit for my wife's grocery list.
(05:10):
And I think we were all like,what are you talking about?
And he's like, well, she had toshop for like 50 or 75 things
for this meal coming up.
And so I loaded it all and toldhim what grocery store I was
going to and laid out this mapto save me like 30 minutes at
the grocery store.
Aisle three, go buy these fivethings.
Aisle four, go to these.
Skip a five.
(05:30):
And I will never forget theamount of people that were like,
I'll buy that.
And for me, I was kind of like,oh, that's pretty cool.
And then after people were like,could you tell me how you got
the grocery list to work?
So for what it's worth, I feellike there's a ton of party
tricks.
And that was one that reallystuck out.
Fifteen gears, though, I want totalk about practical use case of
AI in research and what we'redoing every single day.
(05:51):
And so maybe if you could, canyou get one or two really
practical examples of what yousee sponsors, CROs, different
types of clients that we have?
How are they actually takingadvantage of AI?
How are they actually using itin their companies today?
Probably the number one thingthat AI and specifically large
language models are good at thatget used a ton is just
(06:13):
processing a lot of information.
You can throw tons of text, tonsof data at an LLM and it could
synthesize that.
You can even specify what kindof output you want out of that.
The most common example would besummaries and summarization is
sort of the kind of list appequivalent of an AI feature.
(06:38):
It doesn't mean it's notincredibly useful and powerful,
but I think what's really greatis providing a lot of
information to an AI and thenasking for the output that you
want.
If you're processing interviewdata like we do at Envibe and
you want a few interesting clipsor quotes around a specific
topic, ask for that.
You'll get it.
I think it's just Having an ideaof what you want out of it and
(07:03):
asking for it like you would aska human goes a long way.
I think one of the examples youwere talking about with a client
just the other day.
So I'm just going to say itbecause I thought it was a
really good one.
It's current is you got aclient.
They want to get some insightson the recruitment campaign.
So they load up the survey ofquestions that we're going to
ask, you know, like or scale,but also mostly via their voice,
(07:23):
right?
We want people to respond andtalk to our app with their
actual voice.
You recruit the patients.
They come in.
They talk to the platform.
They give their feedback.
It hits the system.
And then we need to start givinginsights on it, right?
We need to start being able togive insights so that that
client can make decisions on ifthey should change anything, if
the material is working outgreat, right?
(07:44):
Those types of examples.
Because I prompted you for this,pun intended, because I prompted
you.
Why don't you show people whatthat looks like?
I mean, I think the easiestthing to do here is you go in
here and we have this suggestquestions to ask.
button here.
And really all this is doing iswe're telling the AI, hey,
here's all this source data fromthis study that we conducted.
(08:07):
What are some interestingquestions to ask that lead to
unexpected or interestinginsights about this study?
And, you know, you see the firstone here is about the lung bar
puncture requirement from thisinformed consent document.
And, you know, it turns outthere was a highly emotional
reaction to this.
There was a lot of risk This isa scary procedure, and it was
(08:31):
something that had to be donemany times as part of the design
of this study itself.
It's a really quick way touncover something that you
wouldn't have necessarily knownto ask.
Unless you had a good idea aheadof time, you wouldn't have
known, tell me about the lumbarpuncture requirement.
But that's the great thing aboutasking AI for what you want is a
(08:54):
lot of times it can give youwhat you're looking for.
That's my favorite example,actually.
I don't know if I've ever toldyou this before, but sometimes
we look at the data and I'mlike, where do I start?
And I love that AI can promptyou in real time.
And so I think that's a greatexample.
And for me, the number onequestion is, how does it
actually help me make it usable?
(09:16):
Because the data in itself isn'tusable.
I think it's one of the mostpowerful things that AI does.
And I don't know if people haveseen as many examples as maybe
are really out there in the realworld around how to summarize
the data, how to actually useit, how to put it in a
presentation mode that actuallyhelps.
So like I prompted you before,could you just show some
examples, just give people someideas of what they can do and
(09:36):
how I can help them visualize ormake their data useful so they
can present it to other people?
At the end of the day, You don'twant to be sending colleagues a
link to your chat session andhave them read this text
necessarily.
It's not the most effective formof communication.
So one thing that we did is weadded this creative video
(09:58):
presentation tool.
Same sort of idea at play hereis we have the AI suggesting
interesting topics to make apresentation about that we think
would be of interest tostudents.
people using this tool.
You can go in and you can selectone or enter your own.
And then it goes through thiswhole AI-assisted process of
(10:23):
ultimately creating a videopresentation with slides, with
voiceover and splicing in, youknow, voiceover content from the
actual participants whoresponded to your study.
And the user has control overit.
You can make it exactly how youwant to make it, but The great
thing about AI is that we canprovide a happy path for you and
(10:45):
we can suggest, you know, thekey findings here that we think
are most relevant to the topicthat you provided.
And you can go through thiswhole step process that you see
at the top here, where after thekey findings, then you select
the actual quotes that you wantto use to support those key
findings.
And then it'll write the slidesfor you.
(11:06):
It'll write the voiceover andyou click a button.
And then a few minutes later,you know, you get a a video
presentation that you candownload and share with your
colleagues.
The important thing for us wasto get out of just being text.
You know, we want to be morethan just text in a chat
session.
And how do we take advantage ofthe richness of our own data?
(11:29):
You still have a human incontrol, but AI is heavily
assisting and can be anassistant to varying levels
depending on how much the userwants to control in this
situation.
I do like the end.
I mean, it's always, I likepresenting my stuff, but some
people don't.
And so I think it's kind ofinteresting to me that one of
the most practical use cases isactually AI presenting the data
(11:52):
and the findings, right?
Doing a voiceover, doing anavatar.
One of my favorites is when youdid one of those and someone was
like, oh, I really, I don'tremember the name of it.
They were like, I really loveAmber in the corner.
You know, she's a great.
We were like, who's Amber?
We don't know who Amber is.
And the whole time we're like,oh, that's the AI.
And people look at you like,that's not an AI.
That's Amber on the team.
(12:13):
We're like, no, that's an AI.
It's for your voiceover.
I really appreciate these,Jeremy.
There's lots more to talk about.
We'll come back and talk aboutmore of these.
But I think just givinglisteners just a sense of what
kind of practical AI is outthere is helpful, right?
Because this stuff ismeaningful.
It's supportive.
It can help your work.
And it doesn't have to beoverwhelming.
Absolutely.
(12:35):
Yeah.
Thanks for having me on.
Well, hey, thanks for being ontoday.
I appreciate it.
I'll catch you soon.
Thanks, Jeremy.
Popular Podcasts
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by audiochuck Media Company.
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.