April 20, 2023 • 64 mins
There are very few, if any, reliable methods of measuring cannabis impairment. Road-side sobriety tests done by trained Drug Recognition Experts are very subjective and workplaces are limited with how they assess employees with safety sensitive jobs. Back by video evidence, Gaize offers real-time impairment screening within 6 minutes. This portable AV device is considered accurate, providing objective eye behaviours when deciding when an individual is impaired on Cannabis. Considered the first impairment detection platform released to the marketplace, Gaize will search for answers by finding the six ways your eyes give you away.
Music by:
Almadura– Tu Rumba- https://www.youtube.com/watch?v=RTzL-G2kIA8
Additional Music:
Desiree Dorion
Marc Clement
Transcripts, papers and so much more at: reefermed.ca
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:00):
Um, when people got together,
joints were passed around.
Marijuana is a drug with potential
health impact.
I have never seen a cannabis
overdose.
It's reefer Madness, The podcast.
Yes.
Kirk.
Kirk, we're back.
Hey, Trevor. How's it going?
Good.
I haven't seen you in person for
(00:20):
a while because you were
a long way away.
Yeah, I had a fantastic
non vacation.
Vacation?
I spent.
Spent four weeks down
in the escape of Baja
and visited some friends down there.
Was basically just lived
in.
Let's pretend I'm from Dauphin,
(00:41):
Manitoba.
Where on the planet is Baja?
Well, the Baja Peninsula, right?
I mean, we know that from the
Westerns in the old days.
Well, right down, right
down on the very tip, the southern
tip is Kabul.
Kabul, Sun, sun.
San Jose.
So you're in the country
of.
Mexico, country of Mexico, it's
(01:03):
called called California,
Big Sur.
So the bar has two states
and Mexico.
And I was in California,
Big Sur, on the eastern
tip. So I was outside
the window at the house I was living
in, and I was looking at the Sea of
Cortez and the Pacific
Ocean.
(01:23):
We were about 50 kilometers away
from the Kabul airport, which
is, you know, the busy tourist area.
And you go on a sand road
and they don't build
up roads there.
They scrape roads down,
so they scrape into the sand.
So we arrived at
around 530,
6:00 in the evening.
(01:43):
Of course, by the time we hit the
road at 630, it's dark
and we wake up in the morning and
we're looking over the Sea of
Cortez and it's like, Oh my God.
So we're in the middle of a middle
of a community of maybe 20 houses,
gravity fed water and
solar panels.
And our job was to house it
and care for a dog.
(02:05):
So Michelle and I basically walked
the desert, walked the beaches,
saw turtles, laying eggs,
threw sticks into the surf,
heard coyotes, you
know, went to farmer's markets
and just hung out in Mexico for
four weeks. It was wonderful.
Excellent.
Well, you're away because, you know,
(02:25):
apparently I don't like warm
weather. So instead I went winter
camping with our friend
Pat, his son and my
son. So that was sort of the long
weekend in February.
And we slept in a tent
in the cold and went fishing.
And of course, you know, instead of
catching fish, we got skidoo stuck
in slush and other people had to
rescue us.
(02:46):
And that was only day one.
And then day two.
Brant Sun is sort of a
we'll call him an outdoor teacher.
He sort of has a
we'll call it
almost like an outdoor daycare slash
school. Anyway, he's outside all day
long, so he decided to take us
on a snowshoe.
But, you know, not down some trail
like bushwhacking the whole time.
(03:07):
So he could collect pine resin.
Then he convinced us that it'd be a
great idea for us old farts
to follow him down a semi cliff
down to a lake.
So, you know, sliding on my
butt down a hill and snowshoes,
wondering if I was going to get up
at the bottom anyway.
Good, good times.
Well, we we had a northern wind
coming down, coming down to
the Sea of Cortez, so we had to wear
(03:29):
sweaters at night.
But, you know, it's always the same.
But so now.
We're sort of updated, though.
You found an interesting guy
with an interesting company doing
some techie
virtual reality stuff.
Want to sort of lead us into who
we're talking to in about today?
Yeah, I'll let him introduce himself
(03:50):
because I like when people do that.
They get they pronounce the name is
correct and all that.
But it's a it's a product.
It's a product that really excites
me. I think it's kind of cool.
So so what can
and and a magic VR
helmet that's going to tell me
whether I'm stoned or not, maybe
kind of something.
Yeah, I can kid.
As an entrepreneur, what
(04:11):
I like about this story is that I
like to talk about entrepreneurs.
This is a this is a fella that was
from Montana, and he's
involved with the government
in Montana. He explains his story
and he was thinking about how
will how will cannabis
affect his community.
And he figured, I love this,
(04:33):
he says. He says that there wasn't
a lot of drawbacks with cannabis
coming on market except
the fact that people could be
driving. So he thought,
what could he do to help with
that? Yeah, you picked up on that.
He didn't.
Yeah, you know, no.
His job is job is to analyze
the effects cannabis will have in
the state of Montana.
And his answer is not a lot.
(04:54):
Not a lot.
But it's going to actually benefit
them, you know, benefit commerce,
benefit business, which I
thought was really quite
interesting.
But however, he did see a drawback
with people driving and handling
heavy equipment and safeties.
So he did some research.
Came up with this gadget called Days
(05:15):
g, I, z
E, and it's real
time measurement of impairment,
specifically the
cannabis.
Now he gets on
to this about
how he hopes to,
you know, get more data into
the system where he'll be able to
expand it to other substances.
(05:35):
But right now, right now,
he's focused. He's focused on
cannabis.
Yeah, and we'll get into this more
afterwards. But there are
some terms he uses like night
stigmas and eye movement.
And you you actually do that
a fair bit in sort of your day to
day clinical work.
But I think
he does it not a bad job of
explaining all that.
And then we'll of course throw our
(05:55):
$0.02 in at the end.
So. So we just let can
take. Yeah.
Yeah, I think so. But a couple of
things so just so people understand.
We talk about nystagmus a lot
in this because it's sort of the
movement of the eyes within stag
misses and it gives
you a window into the neurological
neurological assessment of
(06:16):
your of your client, of your patient
in gaze.
His machine tests
for six and
of the six eye movements, two
of them are very specific to
cannabis. So as you're listening to
this interview, ensure that you hear
that because we do use the word
nystagmus a lot and some of the
movements aren't necessarily
(06:37):
nystagmus, they're how are your
pupil responds to light, which is
not nystagmus.
So, so just be aware
that that we're using the word
nystagmus. But it doesn't describe
all of the six movements.
Does that make sense?
I think so, yeah.
Yeah. So, yeah, let's get into the
discussion and then we'll come
out and get a little deeper into it.
Does.
(06:58):
My name's Ken Letterman, the founder
and CEO of Gaze.
We are creating what I believe
is the first cannabis impairment
detection device that works in real
time.
And basically what we've done is
we've taken the drug recognition
expert eye tests, which are
the tracking my finger tests that
are pretty commonly portrayed on
movies or TV.
And we put those in
(07:19):
a VR headset so programed them to
run exactly according to the
training manual.
And we can very precisely
control the amount of light
that's entering the eyes and things
like that. And so what we do is we
run through maybe the exact same
tests that the cops use
and we capture eye movement data and
video throughout that process.
We provide the video as evidence to
(07:39):
law enforcement and
the eye movement data then gets
analyzed by machine learning and
statistical algorithms to determine
whether or not there is
signs or symptoms of impairment that
we can detect in the eyes.
How old is your company?
We were incorporated in January
of 21, so we're about
two years old almost.
(08:00):
Exactly.
Cool. Cool concept.
I mean, I've been doing some
studying. I'm on your Web page.
Your web page is very informative.
It has lots of studies out there.
And I'm looking at the 1977
study about how,
you know, the difficulties that
individuals have with determining
impairment.
What gave you the spark to say,
(08:20):
hey, I can we can create
this device?
Yeah.
So I was the director of economic
development for the state of Montana
for four years under Governor
Steve Bullock.
And in that role, I
was sort of charged with
understanding what was going to be
the impact on, you know, various
initiatives.
And one of the things that
I was studying was Montana
(08:41):
was a medical cannabis only state at
the time, and I thought it was very
likely would become a recreational
state if we ever saw a ballot
initiative.
Madison's come to pass.
And so what I was doing at the time
was trying to understand what the
impacts would be.
And so I was looking at, you know,
what will be the impacts to the
state economy, what would happen to
tax revenue, what would happen to
the businesses and law enforcement.
And it was all really positive
(09:03):
with this one big exception, and
that was that there was no device to
check for impairment.
And so law enforcement and
business owners were very concerned
that they were going to see this
spike of impaired drivers and
workers.
So I initially approached this
problem from the perspective
of just, you know, understanding
this was, you know, a huge
opportunity one.
And I thought that it was important
(09:24):
to solve this problem in order to
make sure that cannabis legalization
could proceed and proceed safely.
So I looked around at the companies
that existed and nobody was doing
it right. Everyone was trying to
measure impairment by looking at the
amount of THC in the body.
If you look at any of the science,
what becomes very, very clear
immediately is that you cannot look
(09:44):
at the amount of THC in the body and
derive impairment from that number.
There is there's simply no
predictive amount
of THC in the body and
no consistent amount of impairment
is experienced on any number, on
any amount of THC.
So from that point,
it almost became more interesting to
me and I thought, well, nobody
(10:07):
seems like they're doing this right.
And if if it was to be done
right, what would it look like?
And so what I thought it could look
like is what what the tests that
we know work are these
drug recognition expert tests.
They've been shown to be, you
know, reasonably successful
at determining whether or not
someone is impaired on cannabis and
many other drugs, that
(10:29):
they are not widely
used. There's not enough drugs
in the country.
These drug recognition experts are
put through very rigorous training.
It's very difficult.
It's not an attractive thing for
most cops to do.
So there aren't enough of them.
And obviously businesses don't have
access either.
So my thought really was what would
happen if we automated these tests
(10:49):
and it could that be done?
And that really led me down this
path of looking into what tests
were most predictive, of what tests
were most objective,
what are really the shortcomings of
doing this through human?
And
there are some, you know, really
obvious correlations when you start
looking at it from that perspective.
And that is the human tests are by
(11:10):
far the most objective.
The limitations of a human are
obviously, you know, you've got
a lot of opportunity for human
error, you've got opportunity for
subjectivity, bias.
And then these officers are
performing the test based on
memorized test procedures, which is,
you know, probably never a good
thing for trying to get to real
consistency.
(11:30):
So I thought,
you know, we could probably automate
these tests using a VR headset.
And so and if we
did that, we could probably track AI
movement using these very high
precision sensors.
And if we did that, we could
probably use that data and
train machine learning models to
recognize impairment.
And so that's really what we've
done.
Very cool. Now, economic
(11:50):
development really isn't I mean,
it's social science.
It's not a lot of science science.
So how did a social science
guy get into the science of all
this?
AI making an assumption and making
an assumption that you're more into
the social sciences, but not
exactly.
I mean, I've had a couple tech
companies in the past, and
the economic development job
(12:11):
was really a departure from what had
been the norm of my career.
It looked like a really interesting
opportunity. It certainly was.
And I feel like we did some great
work in that role. But that was
that was more than the aberration
for my career.
Okay. So so a little background on
your career would help.
I want to get into the science of
your product, but tell me more about
yourself.
Yeah. So I'm a multi
(12:32):
time entrepreneur, native Montana,
and I
spent some time in a company
doing a turnaround up, a company
that was doing infrared optics
manufacturing.
I had a search engine optimization
company, I had a web app
development company,
I had a marketing company.
And so I've I've sort of had a
(12:53):
couple
different sprints in my career.
And
what I have sort of decided
my specialty is, is thinking about
hard problems and designing
solutions around them and then
building a team to tackle that
problem. So that's what I've done
here. I'm not the scientist
that has done this research,
but I have assembled a team that has
(13:13):
done that and done it successfully.
So I'm more of a business guy.
Well, I congratulate you on this.
So let's talk a little bit about
Gase, then.
Trevor and I, I'm not sure if you're
familiar with our project, but
Trevor Trevor is a pharmacist.
I'm a nurse. What?
We've discussed nystagmus and we'll
we'll define that as part
of our narrative.
But your, your, your device measures
(13:35):
nystagmus.
Now, I guess the question I have
for you is
I noticed on your web page you guys
did a 350
participant study
on your product.
So did that study confirm
your science or was it a study
to help your your your product
learn?
Sure. Yeah. So really, the answer is
(13:56):
both.
That data set is used
as the training data for our
machine learning model.
We're also able to validate that our
machine learning model is effective
based on that same data set.
So it is sort of one in
the same.
I'm really though that was
the most important thing from that
study was to capture a large amount
of training data to create these
(14:18):
machine learning physical models.
Okay, So when someone consumes
cannabis because on your website
you're truly focusing on cannabis
impairment.
So is there a specific
type of nystagmus that happens
with cannabis consumption?
Yeah. So our our product actually
runs through.
So each of the drug recognition
expert eye tests. So I'll list them.
(14:38):
Those include equal tracking,
lack of smooth pursuit,
horizontal gaze of status and
maximum deviation.
So that's the far left and right
peripheries of the vision
causing gaze to stagnant at 45
degrees from center
and vertical gaze is
stagnant. So that's vertical
periphery, lack of
(14:59):
convergence and people are rebound
dilation.
So each of those tests is looking at
a specific thing.
The horizontal gaze, the stag was
test and the nystagmus tests
overall are much more
predictive of impairment on other
drugs, primarily alcohol.
So we are certainly using
that data and our machine learning
model for cannabis.
(15:19):
But it is not one of the tests that
we're finding to be highly
predictive of cannabis impairment.
And certainly I'm sorry, So what
is your gaze isn't in this segment
isn't correct.
So nystagmus is not one of the
indicators of of cannabis
impairment that we're finding.
It is occasionally
present, particularly in extremely
impaired people, but it is not
(15:41):
universally present.
So we're really focused on the
lack of convergence test and the
people. A rebound dilation test like
a convergence, basically
brings the stimulus towards your
nose, the rigid nose, and tries
to get you to cross your eyes.
And the people revaluation test
exposes your eye to three different
light conditions. So that's room
light and then 90
(16:01):
seconds of darkness and allows
your people to fully dilate
and then we expose it to bright
white light and that causes
your pupils to constrict and we
measure how
rapidly they constrict and whether
or not they stay in a constricted
state.
Okay. So to review that.
So your device does.
Did you say five tests
was like six tests?
(16:22):
And out of those six, how
many of them are predictive of
cannabis?
Two are two are very predictive of
cannabis. We're finding some signal
and the other tests, to be clear,
But it is those are not
the tests that we're really focused
on.
Okay. But for the purposes of
defense and law, you want
all those tests in there and then
they will focus up on the two.
(16:44):
Yeah, because really the vision for
Gaze over time is to add additional
substances.
I want to create what I'm calling
the first impairment detection
platform. So a single device that
can detect impairment from any
substance.
And I you know, I always start this
conversation by saying I'm not
opposed to cannabis legalization at
all. I think this is a good thing
and I think it makes a lot of sense
of alcohol's can be legal, Cannabis
(17:04):
should certainly be legal,
but I think we need a way to make
sure it's done safely.
We need a way to ensure that
people are not driving, working
while high because, you know, there
are extremely bad outcomes that come
from that.
So the reason that we do the each
of the tests that the drug
recognition expert does is because
we want our device to ultimately
(17:24):
be able to detect impairment from
any drug or any combination
of drug. And so by
conducting the plurality of those
tests, we can we can determine
if someone's high drunk,
drunk or high,
high and meth, anything.
We just need the training data and
we design algorithms to do that.
It's a very cool it's a very cool
(17:46):
concept. I mean, in health care, I
measure nystagmus as
as an assessment for muscle,
skeletal, neurological, you
know, chronic disease issues.
So walk me through
your selling points to a law
enforcement agency.
How do you how do you say to
a law enforcement agency, here's
how our device will work in
(18:06):
the field?
Yeah. So the
law enforcement use case really is a
bit different than the commercial
use case in that, you know,
obviously there are criminal justice
outcomes that are coming from these
tests. And so one of the things
that we're really focused on for law
enforcement is the generation
of evidence.
That's something that is of marginal
value, all safe to commercial
customers. But for law enforcement,
(18:26):
I think that's of huge value
right now. When a law enforcement
officer conducts a
standardized field sobriety test on
the side of a road or on a drug
recognition expert conducts their
test back at the station,
what evidence is generated
really is only coming from perhaps a
body cam if it was activated or a
dash cam on a patrol car.
(18:46):
There's not any eye
movement data or video that's
generated in this process, even
though these are the most objective
tests that are used.
And so our device
records eye movement from about
an inch away from the eyeball.
And so it is very
up close to very clear video
of what's happening to the
eyes and in the eyes from
(19:08):
a perfectly conducted
very eye test.
So we think that evidence piece is
really, really important for law
enforcement.
The other piece of it, you know,
right now there's no tool that law
enforcement can use to detect a
cannabis driver.
It is really
at the whims of the patrol officer
whether or not they think someone is
impaired on cannabis, whether
(19:28):
they're going to bring them back to
the station for additional tests.
So what we're providing is
really a stand in for with a
portable breathalyzer test
that can be used on the side of the
road to determine whether or not
someone is likely to be impaired on
cannabis with reasonably
very high precision, where our
devices, it's much more accurate
than even the best trained human
(19:49):
drug recognition expert officers.
And so while
we encourage confirmatory
tests to be done every time, I
guess devices predictive of someone
to be impaired by it
is certainly the best technology
available for that purpose.
And so we think that that's the
other really pressing use case.
I think.
Okay. So I guess if I'm
(20:10):
if I am sitting in my car being
pulled over and and
the police officer,
the law enforcement officer thinks
that I'm impaired, he'd ask
me to step out of the car, probably
put me through the, you know, the
walk and turn pass probably the one
one leg test.
But instead of doing his
his subjective test or objective
(20:31):
test of nystagmus, he put the
gaze on at
the roadside. Or would that be
something you would do in the in
at the shop, at the jail?
Yeah, it can be done.
In either case, the vision
here is if if someone.
Is exhibiting signs of impairment in
the roadside.
They are blowing a zero on
a breathalyzer.
(20:51):
The cop has no other tools other
than simply placing them under
arrest. Know that person could
simply be fatigued,
in which case they would probably
be, you know, prone to stop
driving and just go home.
Or they could be impaired on canvas
or another drug. And so.
Gays could be used in a roadside
setting.
Gays. You would require that people
are seated while they're going
(21:12):
through a test, so you'd probably
lean up against the car or sit
in your own car while that was being
conducted.
But the result of that test would be
very clear
data about whether or not someone is
actively impaired.
So so there's the
officer would would key in the
person's name driver's license,
date, all the all the tombstone
information that they would use in a
(21:33):
court.
And then the computer technology
then says, you know, there's
a green light. The guy is a go,
there's a yellow light cautionary,
there's a red light. Oop, the
signals are out, The guy is
impaired.
Is that more or less.
So we run through, as I said, six
different tests. We provide
information about whether or not the
person is passing or failing each of
those tests. And we provide
(21:54):
a predicted substance of impairment
and are confident that that person
is improving that substance.
And so the
outcome for the patrol officer
really is verification that, yeah,
we're pretty sure this person is
impaired, we're pretty sure they're
impairment cannabis.
Let's go ahead and bring it back to
the station and do some other tests.
Okay. But if he pulls up myself, I
mean, I'm in my sixties.
Is there is there a difference in
(22:15):
your study between demographics?
Do like some people have
nystagmus as a
chronic ailment?
So how does your how does your
device quantify
age, demographic sex?
Does it do anything like that?
So in our study, we captured all
this demographic data.
What we're doing right now actually
is going through our our machine
(22:36):
learning model and ensuring that
there's no demographic bias built
in. So, for example, for detecting
more females and males as
impaired, then we need to understand
what's happening there and remove
any any existing bias.
Any time you train a machine
learning model, there's the
potential that you're training in
bias based on what your training
data is.
(22:56):
And so we were making sure that
that is not the case.
And removing bias as we find it,
there is you know, there are many
genetic conditions that can lead to
many different types of abnormal eye
movements.
And so another good reason
for us to do the plurality of the
tests that we use is that
there are there's a reason that
someone can fail a single test.
(23:18):
But when we look at the tests in
a collective,
we can very precisely understand
whether or not the the
eye movement behavior we're seeing
is correlated to a particular
substance.
And so that's, I think,
really, really important because
there are so many reasons
(23:39):
that you could be exhibiting
nystagmus, for example.
There are many reasons that you
could have abnormally
large pupils.
But when you look at, for
example, a failure on lack of
conversions and a failure on pupil
rebound dilation, and then some
perhaps some other indicators from
other points in the test, you can
be very confident that this person
is impaired on cannabis and not
(24:00):
simply tired or
drunk or something else.
Okay.
Your product is still a couple
of years old, so it's truly in its
infancy
and I can see how time
goes on. It's going to become more
robust.
So how is information gathered now?
And I guess the other question is
confidentiality.
I'm going to make an assumption that
(24:22):
if I'm working for snowshoe
up or snowshoe in Montana Police
Department and
I'm using your product, I will have
a database within my department that
the information goes to
that would have the tombstone
information because I need for
evidence.
Does that that it now get pumped up
to the cloud, to your to your
grand device and you like
(24:43):
and then you work with the data.
How does that all work?
Yeah. So the data flow basically is
we've got a we've got three
different absolute built.
The first is a mobile device
app that basically controls the
headset. So the headset is a slave
to the metal device that allows the
the officer in this case or a,
you know, H.R. person or whoever
to press the start button and
conduct the test.
(25:05):
The data from there flows up to our
cloud server, where it gets
processed using our machine
learning algorithm, and that
result then flows back down to
the mobile device to alert the
test administrator whether or not
someone is higher, if they're, you
know, if they're so where they can
go.
So that data is
is there is no
(25:26):
personally identifiable information.
So it's for privacy.
We don't capturing personally
identifiable information in that
data set. And so even in a worst
case, we were hacked or something
that would never be correlated back
to an individual.
The final piece of software
developed is a web app that allows,
for example, a station to look at
the tests that all of their officers
(25:47):
are doing.
And so if you're the chief of
police, you want to understand who
you're which which patrol officers
are using gas mos Most often you can
go to this web app.
You can see where they're conducting
tests, How often, how many of those
were positive results,
download test results, download
evidence, all that.
Okay. So so obviously unit
you need cell service.
If you're doing it in on the scene
(26:09):
or you bring the person back to the
to the building and you have your
Wi-Fi system.
Yeah.
We're looking at ways to
do on device.
So on the mobile device, do the
machine learning processing there.
Unfortunately, there's a huge
variability in processing power
amongst mobile devices.
And so whereas like the latest
iPhone can probably run the model,
I'm not sure that, you know, an
older phone could do
(26:31):
the same. So
we'll probably get there.
But that's a that's a feature state
that we're looking at.
Ironically, I'm thinking of
law enforcement, but you're also
marketing to manufacturers.
So people that, you know, crane
operators, people that are, I
imagine, working in heavy duty.
Heavy duty devices.
Mm hmm.
Yeah. So anybody that's in a safety
sensitive role.
(26:52):
So that would be anyone whose life
is depending on someone performing
their job appropriately.
We're interested in selling devices
into those positions.
Commercial On the commercial side,
the only screening tool that's
currently available really is
urinalysis screening, and that is a
distinctly retroactive way to
look at impairment.
You can tell that someone has
(27:12):
previously used cannabis, but you
cannot tell based on a urinalysis
screening whether or not they're
currently high.
So businesses really have no tools
at all to detect whether or not
someone is currently impaired on
THC and in the
United States.
I'm not sure what the Stats Canada
is, but in the United States, THC
use in states where it's
recreationally legal or
(27:33):
where adult use cannabis is legal
are starting to protect THC use as
a as a class.
And so what that means is
that you cannot, as a business now
perform THC tests and
take adverse action against an
employee based on a positive result.
You have to actually prove that they
were impaired during their
working hours.
And so that's a that's impossible
(27:54):
for businesses currently.
We're really providing, I think, one
of the only ways that that could
reasonably be done.
And so our hope is that this can
really drive safety radically
higher in businesses that are that
have safety sensitive employees.
Yeah, that's I guess
that's where the the skeptics
will say just because
(28:15):
my eyes are behaving in such a way,
how do you prove I'm impaired?
I may be a medical cannabis user.
So if I'm a medical cannabis user
and I'm microdosing my my cannabis
is is Gage going to say
I'm impaired?
No, that's that's a really great
point. So the eye movement
characteristics that we look at are
only present when you have
(28:37):
sort of demonstrable
reductions in your, in your
ability to do
equal tracking or to do
divided attention tasks or reaction
time tasks. And so these are
things that are only
indicative of of impairment
that is, you know,
above and beyond what you'd want to
(28:58):
be operating a vehicle or equipment
under. Okay.
So is there someone
I don't know if this is a fair
question, Can I develop a tolerance,
like if I've been using cannabis for
five years and all
of a sudden my employer says, here,
put this device on?
Well, buddy, I've been working for
you for five years using medical
cannabis.
(29:18):
Is there an is there a tolerance
to the body develop tolerance?
Definitely.
Cannabis has a has a actually a huge
tolerance effect.
So heavy users of cannabis will have
a radically lower amount of
impairment that comes from
using the same amount of THC that
that a infrequent
user can become extremely impaired
(29:39):
on.
So tolerance is one of the things
that our device is particularly
useful for because it is again, it's
only detecting
impairment that
can be objectively
sort of correlated to
reduced ability to perform a
job or to drive a vehicle.
So there's been studies that have
(30:00):
validated this. There's a great one
that came out last year from the
UCSD Center for Medical Cannabis
Research that looked at
what is the sort of impairment
that happens driving
impairment in particular in this
case that happens from consuming
cannabis. And they found results
that were consistent with what we
found.
But basically tolerance has a huge
(30:20):
impact. And so.
A person with a large tolerance
for THC would have to consume
much more cannabis to achieve these
same failure states
that a person that consumes
very little cannabis would do.
I'm wondering if your data
collection captures that.
Like if I'm
if I'm a chronic if I shouldn't
(30:42):
say chronic, but if I'm a medicinal
cannabis user and I get pulled
over, I guess the officer
has to make other determinations of
why he pulled you over and then add
that add that to the database
that your device gives them.
So would you be capturing that?
Do they say in there that what they
have prescribed, would
they have the list of prescribed
(31:02):
medicines somebody is on that you
would capture that?
We don't capture that data, but the
officer would capture that.
And so that's a determination the
officer would have to make.
You bring up a really good point,
though, about tolerance and about
heavy usage with THC.
You know, the current state of of
the situation is that if an officer
pulls you over and they suspect
you're driving while impaired on
(31:23):
THC, don't do
a saliva screening or a urinalysis
screening or in the law enforcement
case, law was probably a blood draw
and they'll look for
THC in your blood.
And if you're above a certain amount
in Canada, I believe it's.
Five milligrams per milliliter if
I'm not.
Yeah. Yeah.
(31:43):
Most other states in
the U.S. that have legislated
legal cannabis have also established
these per se limits.
What's important to know about those
is that there is no science at all
that backs up the idea that
you can have, you know, five
milligrams per milliliter of senior
blood indicates that you're
impaired, that science does not
exist. And in fact, it's quite the
opposite.
There been many studies that have
(32:05):
looked at this issue and verified
that there is no amount of THC in
your body that can be correlated
to some predictable amount of
impairment.
So fundamentally, these five
per milliliter per se laws
are nonsensical and were
created entirely to facilitate
prosecution of US
cannabis users.
(32:26):
And so our product provides
a much more rational path
forward and that we are looking at
actual impairment as it is
manifesting in the body and
not some arbitrary amount of THC.
So a heavy user of cannabis will
always have THC in their body that
is above the legal limit.
And there was actually a case that
came before Washington
(32:47):
State Supreme Court last year
that was
it was exactly this where a guy
had not used cannabis for 24 hours,
was given a blood draw
and was thought to be over the legal
limit. And he challenged the state
on whether or not that was a
rational lie
in what I think is the most insane
court ruling I've ever seen.
(33:07):
The court came back and said, Yeah,
we agree there's no science behind
this whatsoever, but we're going to
stick with it because everything
else that
was the state of Washington versus
Frazier. And it's
it's just, I think, a really
terrible criminal justice outcome.
And it's exactly the kind of thing
that cannabis legalization
is supposed to be moving us away
from. And in fact, it has delivered,
(33:29):
you know, exactly into this really
bad situation.
So I believe that a device
that measures impairment in the body
is a much, much better path forward
and should be supported by the
cannabis industry because of
the reasons I'm describing.
Yeah, no, I agree with you.
If you're impaired,
don't drive.
And what we've what we've learned
(33:50):
through our research, through our
podcast is that some
of the research coming out of
Australia suggests that those
those that use cannabis
will sometimes self
self-regulate themselves.
Yeah, I'm too stoned to drive, man.
Whereas whereas alcohol,
I don't know about alcohol, but the
car can find its own way home.
Right. One. One for the ditch.
(34:11):
Off we go.
So I think there's a behavior
difference between cannabis users
sometimes and alcohol users.
But this leads me to another quote.
This leads me to another question,
obviously, and I shouldn't say
obviously there are, what, 108
cannabinoids in the cannabis plant.
So your research is saying that
right now, if someone's impaired,
(34:31):
it's measuring THC impairment.
I mean, CBD is, we know, does not
cause psychotropic effects, But
is there any core is there any
differentiation in your studies
between the cannabinoids or is it
just impairment?
Yeah, it's just impairment.
So Delta nine, THC,
the people in our study consumed
flour, so they consumed,
(34:52):
you know, kind of dried flower
cannabis.
We've we've done some small
amount of confirmatory testing with
vaped and edible cannabis.
But when you do more work in those
areas, it appears,
however, that the same impairment
manifests across
methods of ingestion.
And so we're very confident advice
for that respect.
(35:15):
You bring up an interesting point,
though. If you start looking at
synthetic cannabinoids,
it becomes a different story.
And there's a there's almost
zero science that has been done.
And looking at how
these synthetic cannabinoids
impact eye movement or
how the impairment can otherwise be
(35:35):
characterized.
And that is
going to be, I think, a very
difficult thing to get
our arms around because there's just
so many of them.
Well, we've often talked about this,
especially early in this project.
We've been doing this about five
years. And very early on we
often talked about how many times
Trevor being the pharmacist fills
(35:56):
a prescription that.
That is a psychotropic prescription
that's going to have some effect on
muscle skeletal movements, going to
have some effect on cerebral
thought processes.
Any any And you watch as the
person, you know, get into the
parking lot takes
one and gets in the car and drives
away. Right.
So we know that people.
(36:16):
But it's okay because it's a
prescribed medicine, right?
It's okay. They can drive on it.
So. Right.
So there's your there's your device.
I mean, nystagmus.
Other substances
create nystagmus.
So as you said earlier, I
guess you're hoping that the
learning technology and your device
will teach you about what other
substances are being abused
(36:38):
or used.
Yeah. So what's what's
really interesting about these tests
is that each class of substance
has a unique way in which it impacts
eye movement.
There's been not as much research
done on on prescriptions,
frankly, but for each class
of drug. So each class of illegal
drugs are regulated substance
like alcohol, opiates,
(37:00):
stimulants, central nervous
system depressants, cannabis,
inhalants, psychedelics.
Each of these has a unique way in
which it impacts movement, and we
believe that we can detect
impairment from at least most
of those, if not all of them.
The work is simply
capture a data set that is
of good enough quality and
(37:20):
large enough number that we can
put our models at it and train it to
recognize that impairment.
Fantastic.
I only got about 5 minutes left with
you here. A couple of real quick
questions. You're in the marketplace
now. People are buying your product.
We are almost in the marketplace
now. We have we have a good
we have a really strong preorder
list. We are, as
I said earlier, we're eliminating
(37:41):
the bias that we could have trained
in during the course of
our clinical trials. So that work is
ongoing and we're doing a couple of
things around
eliminating false negatives and
false positives in our model.
So there are people that for which
it appears, and these are
occasionally high, high tolerance
people and
(38:02):
some others we're still trying to
understand the cause of.
But there are some people for which
the
consumer, the same amount of
cannabis does not have the same
effect. And so we're trying to
understand what are the causes of
that and how do we determine
whether or not this person is
actively impaired or not.
And in some cases, it appears that
some people are just not impacted
the same way by cannabis.
(38:22):
And so that's that's a bit of an
interesting problem.
But we're our intention is to be in
the market by the end of this month
and with a with a model that
is radically more accurate than the
best trained human.
Kirk That was really interesting.
I really enjoyed your chat
with Ken from Gaize.
(38:43):
Because we talked about a little bit
right at the beginning, I'm going to
pick your brain here again in case
the people missed it.
What's nice, nystigma or nystagmus?
What? What? Like what does it look
like? What? What is it?
Nystagmus is when you're
usually what happens is that you you
ask somebody to focus
on your finger and you go the finger
horizontally and you watch the
(39:05):
pupil
go and follow the fingers.
And what happens is that normally a
person's eyes will track
without any variation.
But with nystagmus,
what'll happen is that your eye
will track, but there will be
variations or quivers little
movements of your pupil,
of your eyeball
(39:27):
as it tracks across.
So that's the nystagmus.
Is that is that unusual
quiver motion that
happens when you're tracking.
Okay. And so now
when you said the
horizontal nystagmus
at the extremities where
relate to alcohol, not
(39:47):
not cannabis, that would be, you
know, out when I'm looking way off
to the right or way off to the left,
that I'd get that quiver.
Right.
Is that is that peripheral
vision.
And as you're as you're as your eye
tracks to
the periphery, there'll be a
quivering there'll be a movement.
That's not it should be smooth,
right? It should not be any
(40:08):
movement.
Okay.
So you'll see a small quiver and
it's abnormal,
right? I mean, it's a lot of
assessments on people as you're
looking for something abnormal.
The average person's eyes track
without any alteration.
And the two that he said were
they they think closely related
to cannabis impairment was
(40:31):
loss of convergence.
So I think everybody can picture
that. That's take the finger to the
middle of your nose and you go
cross-eyed and that they aren't
doing that.
And pupil dilation
sort of not as expected.
They said they'd have different
brightnesses of light and there's
sort of an expected how fast that
pupil opens and closes and I guess
it's not doing as normal.
So it does seem to be that the two
(40:53):
out of his six that
were more related
to they think cannabis
impairment. So that was very
interesting.
Yeah. Yeah. But what I like about is
that they're doing the six
and this goes back way back
in time.
The web page.
The web page has
(41:14):
all their research
and some of the research we're going
to put on our web page.
But going to their Web page,
which is
https://www.gaize.ai
You'll find a lot of the research
and they go back all the way to 1977
(41:34):
when the US government
studied roadside testing
and what it would take for people to
be qualified to do roadside
tests.
So what they've done is they've
taken those tests that people do
that the, you know, the police
would do on the roadside, put
it into an algorithm.
And I think what's really cool about
this business is that they're
(41:54):
capturing all these behaviors
you know, your lying eyes, right?
You can't hide your lying eyes.
And they're going to pick up,
over time how
your eye responds
to substances.
I found it interesting, though, that
they're not tracking people's
prescription drugs.
(42:16):
I would think that they would want
that in their database and
some some point
be able to correlate
common eye behavior's
from certain substances.
He did say that he would like it to
expand out to other drugs but that's
a nice segway into I'm going to pull
the usual Kirk and say we have some
other episodes in our library about
(42:36):
driving We have
E60 for Driving Under the Influence
with Dr. Tom Arkel,
an Australian researcher who put
subjects on the road in the
Netherlands with and without
placebo cannabis to see what would
happen. That what is fascinating.
Still fascinates me.
And then more related, this one
Cause Agnostic with Driveable
(42:58):
and kind of like you were saying
Driveable guys, now it wasn't a VR
headset, it was something more akin
to an iPad but, the reason
they call it Cause Agnostic is
they weren't really caring what
impaired you.
They were looking to see basically,
do you still have the reaction times
necessary to perform safety
(43:20):
sensitive things like, you know,
our crane operator or
whoever or driving?
So just interesting that
there's we've talked to at least two
groups now that are using I will
call it, quote unquote, electronics
to see to see
what
cannabis and or other things are
doing to to impair you.
(43:41):
It's all very interesting.
I think it's fascinating.
What I what I also find interesting
is the nuances of this.
You know, our culture is very
comfortable with alcohol.
In the sense that we're very
comfortable to to not drive
when we're drunk. However, we
have a huge issue in our culture
about people driving drunk.
But usually what happens is that
(44:02):
when someone is under the influence
of alcohol, we know
the impairment because we've had how
many decades
of learning.
You know, buddy, you're too drunk to
drive. Give me your keys.
Cannabis is something different.
What I found interesting in
what he's talking about here is that
his machine is measuring impairment,
(44:22):
not the amount
of THC, you know, whereas
other tests, blood tests, saliva
test, they're testing that you have
THC in your system.
Therefore you might be impaired.
Well, his isn't quite measuring
impairment.
It's much closer.
And I think he made a great point
that, you know, unlike alcohol,
(44:44):
which like you said, we have
decades and decades of, you know, a
blood alcohol level of X equals
impairment. Why?
We've established that.
And we've also established
that a THC
blood level of X doesn't mean
anything. You know, if you are
new to cannabis, you might be
freakin impaired.
If you use it every day because of
(45:05):
medical or whatever, you might not
be impaired at all.
And what you'd really
need to measure impairment at the
roadside would be, you know,
a tractor trailer with
a, you know, pull someone off, I'll
throw them in in a driving
simulator. And that's the only way
you're going to actually know if
they're impaired. So this is a good
surrogate marker, but it's still
(45:25):
not, you know, obviously the
drive around driving
simulator to see if someone's
actually impaired is completely
impractical. But this is
this is a nice surrogate marker.
That and big bonus
is it doesn't require any
blood tests.
And back
to what I think you were getting at
earlier is, right now we don't
(45:46):
have, we can take blood, we can
measure a THC level,
but that number is pretty
meaningless.
Well, I interpreted it a little
differently in the sense that
in the model.
The police officer is
driving down the road and he
sees something that
makes him believe he's going to pull
(46:07):
over this driver. And now there
could be a check stop.
It could be erratic, irregular
behavior on the road.
But for whatever reason, the
officer has pulled you over,
comes up to the car, you know,
touches the driver's side
panel on the back.
So, you know, there's witness that
he was there, that old trick and
says, hello, sir.
Driver's Registration.
(46:28):
Now, if it's like,
you know, This Hour has 22-minutes
where they roll down the window and
a cloud of smoke comes out at
you, then he's got reasonable,
right to say, you know, I'd like to
do a test on you if he
like if he can smell cannabis or he
smells or smells a product in
the vehicle or the person's
(46:49):
behavior, or maybe he's
asking some questions.
You know. Sir, have you had any
intoxicants?
You know, I'm a medicinal cannabis
user or whatever.
He can then apply this device.
And my interpretation is that
through the interpretation of
the two primary
two primary tests that
Gaize can predict that,
(47:09):
yes, from the results
of the test, we can
predict this person is impaired.
That's how I interpreted what he was
saying.
So I'm thinking, but before that
happens in a law enforcement
situation, there has to be a
reason for the officer
to have initiated
the test.
Now, from a commercial working
(47:30):
on on an industrial site,
it may be Tuesday
we bring Gaize on to the site
and do arbitrary
tests on people for this for
safety, meeting occupational health
and safety.
And again, if you identify
as a cannabis user or there's a
reason to believe you use cannabis
(47:51):
and they put these goggles
on you this AV goggles
and you fail these two particular
tests, they're saying that you
are impaired.
I really like that.
And you mentioned a couple of times
it's worth repeating.
So again, like you said, getting
away from the law enforcement thing,
an employer with a safety sensitive.
(48:11):
Workforce.
And, you know, we're going to I'm
going to use crane operator as
my fictitious person.
You know, if you
right now, you could make him pee in
a bottle.
But that, again, doesn't really
say much about their
impairment level.
And it also doesn't say what's going
on right now. That might be I smoked
(48:32):
cannabis three days ago.
So the nice thing about
gays is there really
isn't any good tools that
an employer has now
to say whether or
not it's a good idea
for, you know, this person to
go up in the crane.
So this
sounds like a good at least
(48:53):
gives the employer an option
before they you know,
if we're going to do some kind of
random testing or not so random
testing of employees before they go
do safety sensitive work.
Mm hmm.
When we interviewed when we
interviewed Ken, we interviewed him
gosh, I think I interviewed him in
January. This one's been on this
show for a while, and I apologize
(49:14):
for that. But he sent me an email.
I was in Mexico, actually sent
me an email February 13th
and said that he had launched.
So this this product is now
out there.
Good, good. So February 13th,
it went live. So and we'll have a
link to the website on in the show
notes. But yeah, if this catches
your fancy because you're part
(49:34):
of law enforcement or
a safety sensitive employer,
this is a tool that you can
get your hands on now.
Speaking of driving, I know you
were looking up to see
things at some stats
and study. Try and get some current
stats and study about impairment
(49:55):
and driving in Canada and cannabis.
What? What did you stumble on to?
Yeah, well, I receive
updates from Health from the
Government of Canada on new
research, and this is brand new
research out March
2023 in
Health Promotions, Chronic Disease
Prevention Canada.
This document came out from
(50:18):
the Government of Canada sponsored
by and the study is
called Impact of Substance
Related Harms on Injury
Hospitalizations in Canada
from 2010 to 2020.
So over a ten year period of time,
they're studying how
many hospitalizations are a result
(50:38):
of substance
related harms.
Now 2018
was legalization of Canada
of cannabis.
Specifically driving, or you couldharm yourself
in any way.
Any way at all.
It's all substance related
hospitalizations, right?
So, there's
(50:59):
2,108,489
samples in this study.
So, of course, poly-substance
abuse rates the top.
The most people that go into
hospital due to
self intended or
(51:19):
what the terminology you use here,
they use the intended substance
related injuries hospitalization
and unintended.
In all cases, people that
mix substances, opium,
alcohol, whatever you're
mixing with, that's the top dog.
But what I found interesting is
and is that
(51:40):
the female the Intentional
Substance related injuries so
people that that intended
to take the substances
and ended up injuring themselves and
in a hospital the highest
risk group was females between
15 and 19 years old
and females
are.
Are we talking about like self-harm
(52:02):
as in, you know, a suicide attempt
that's probably in there somewhere.
Some of it's in there.
Yeah, some of it's in there.
But you know, but it's just
it's, it's from 14
all the way all
the way to 74 females
rank highest in the age
groups of of intended substance
related injuries.
So for whatever reason they were
(52:23):
taking a substance and hurt
themselves. Unintended
substance related injuries.
And this is fascinating to me the
unintended ones are
these 75 to 85
year old. The highest level, 85
plus. Now, that is people who are
on prescription medications
and.
Took the wrong one or took too many.
(52:44):
Took too many fell down.
Falls, right.
So there's some cool
statistics in here and
they're not going to be surprising
to every one. You know, everyone
because, of course, alcohol.
Right. It's a specific
rates of substance related
motor vehicle collisions
that ended up hospitalized.
(53:05):
The highest among these were 20 to
29 year olds with a hospitalization
rate of 7.2
in 100,000 for men, 3.2,
100,000 for women.
So young boys, young men
are ending up in the hospital,
most for alcohol, for
motor vehicle accidents.
So alcohol
(53:25):
use concurred most frequently
to falls and motor vehicle
accidents. I don't think we're
surprised about that.
But where we might be surprised
about when it
came to cannabinoids.
And here's the quote.
"We found that cannabinoids
account for less than 1%
of all substance related injury
(53:47):
hospitalizations, which
is low relative to the proportions
of Canadians and North American
populations using cannabis."
So, you know, this is
again, I'm not promoting the use
of cannabis, but I guess what I'm
saying is that from the studies,
from our own interviews,
(54:08):
when people were so afraid
of legalization of cannabis that
we're going to have more vehicle
access, we're going to haveaccidents,
It's not happening.
Well, it's not happening.
It's not happening from statistics.
It's not happening from our
research.
And just because I can't resist
your, you know, alcohol related
falls.
So a couple of my favorite books
(54:29):
are Freakonomics.
Well, the whole Freakonomics set.
And now.
I love freakonmics
And now it's a podcast.
And one of the things they quoted in
one of their books was, you know,
how unsafe walking and drinking was
because, you know, far more people
that they weren't promoting drinking
and driving either.
But they said if you look at the
(54:50):
stats, you know, far more people,
you know, get drunk, walk home,
fall and crack their head on the
sidewalk than anything else.
So, you know, drinking and walking
wasn't recommended either.
Yeah, well, I mean, now.
Now.
No, no alcohol.
No alcohol is good for you in
the new studies.
But I
look forward to seeing Gaize grow.
(55:11):
I think this
is a fantastic
product from the perspective
of another, just another
tool to justify,
justifies not the word, I guess
another tool that we can use to just
protect people, Right.
I always say
that one thing about driving, you
(55:32):
know, I have driven a car
in many different vehicles, in
many different lands and
different countries.
And what I have found is that we
all, no matter where you are
driving, everybody wants to
get home, right?
We all we all drive
to get home and we drive
to protect ourselves from
hurting other people because if we
(55:53):
hurt, other people are hurting
ourselves. So I think driving,
I think driving is one of the
biggest communist plots out there.
It's one of the biggest social...
Think about it's one of the biggest
social responsibilities you have.
And we pay to get the license
right. We pay to get the physical.
We pay for the privilege
to do it.
(56:14):
And we do it for all the same
reason to get someplace safe.
So I think these guys
have a fantastic product.
I like the idea that
it's your
lying eyes, man.
You can't hide them.
And I think this is the name of the
episode. You can't hide your lying
eyes. So good on
them.
Yep. No, noninvasive.
(56:35):
You know, we always like an AI
story.
I think this is very cool.
I'm hoping to follow them in the
future as they get more and more
data and maybe more and more drugs
in their list.
But speaking of driving, Kirk,
you and I took a road trip to
Yorkton Saskatchewan,
and and you
talk to a to a random
hockey player in a in a bar
(56:56):
we were at.
Yeah.
One of our what was set up one
to set up this story.
Yeah. One of our buddies
on our team we went to a
Proye Cup which is
well you are.
Fat old man hockey.
Oh, it's some of the best, worst
hockey I've ever seen in my life.
I loved being the trainer
for the local Kinsman team.
(57:17):
It was a lot of fun.
Anyways, I digress.
We were having lunch and
I was again talking
about our podcast with an
acquaintance friend of ours, and he
said, I said we were trying to get
My Cannabis Stories.
He said, I got on my cannabis story
for you and
let's listen. This is a buddy at a
hockey tournament.
Hi, my name's Dan Rojansky, and I
(57:38):
was asked to do a quick excerpt on
what happens to me when I consume or
smoke marijuana.
Essentially, I get the munchies and
everyone is aware of what the
munchies are, except mine are
incredibly violent.
Meaning I eat everything and a lot
of it.
It doesn't matter how much I've
eaten for the balance of the day.
Once I smoke weed, I could
out eat the largest man you've ever
met. I will eat a full large pizza
(57:58):
on my own without breaking a sweat.
And then I'll eat wings.
And then I'll eat something sweet
like a chocolate cake.
It's violent.
So weed is awesome.
But be aware that the side effect
is weight gain because you eat a
lot.
So that was Dan.
Dan from Winnipeg.
It's it's always fun when a
someone you know would be it's it's
a funny story, you know.
(58:19):
Well.
The heart tugging ones are good too
but you know.
Well it's what I like about
it. It's an honest story.
The guy basically saying is
that when he gets high, he gets the
munchies. It's like, oh, gosh,
how many years ago?
I got I got one from an
acquaintance, a colleague up north
that she got high going to a concert
and couldn't drive, you know.
(58:40):
So I like My Cannabis
Stories because everybody has
a different experience with
cannabis. And that's
what makes this such an interesting
substance
to learn because everyone
behaves differently.
Do you want to talk about the
Canada Cannabis Act?
(59:02):
What we are, what we're listening
to today on the radio?
Sure. So not sure when this is
dropping. So it was current when
Kirk and I were talking about it,
but who knows what it'll be by the
time this drops.
But within the last 24
hours, Quebec
said you can't grow pot
in your own house.
(59:22):
And that was the Supreme Court of
Canada ruling.
Yeah, it's going to be interesting
to see how that falls up because
Manitoba has theirs
sitting at the Supreme Court right
now. So that's going to be sent
we're going to be following.
Absolutely.
So I think I better say
I am Trevor Shewfelt. I'm thepharmacist.
I have Kirk tonight because I'm the
registered nurse and we are Reefer
Medness - The Podcast.
(59:43):
Go to our web page.
All the research from
Gaize is,
all the things that Ken was
saying about Gaize.
It's all research.
It's on his web page.
There'll be a link on our Web page.
I will also have the research that
talked about the the stuff
from the Health Promotions
(01:00:04):
Canada that'll be on there.
Go to our web page.
It's searchable.
Episode 101
Reefer Medness - The Podcast.
Music.
As you know, we usually ask our guys
for music and Ken asked
for to Tu umba.
And man, I don't know about you, but
(01:00:24):
I was rocking to it.
I thought it was really.
Driving down the road.
This is a great driving tune.
Its a great driving tune.
Correct.
All right. Talk to everyone later.
Come on back.
Um, when people got together,
joints were passed around.
Marijuana is a drug with potential
health impact.
I have never seen a cannabis
overdose.
It's reefer Madness, The podcast.
Yes.
Kirk.
Kirk, we're back.
Hey, Trevor. How's it going?
Good.
I haven't seen you in person for
(00:20):
a while because you were
a long way away.
Yeah, I had a fantastic
non vacation.
Vacation?
I spent.
Spent four weeks down
in the escape of Baja
and visited some friends down there.
Was basically just lived
in.
Let's pretend I'm from Dauphin,
(00:41):
Manitoba.
Where on the planet is Baja?
Well, the Baja Peninsula, right?
I mean, we know that from the
Westerns in the old days.
Well, right down, right
down on the very tip, the southern
tip is Kabul.
Kabul, Sun, sun.
San Jose.
So you're in the country
of.
Mexico, country of Mexico, it's
(01:03):
called called California,
Big Sur.
So the bar has two states
and Mexico.
And I was in California,
Big Sur, on the eastern
tip. So I was outside
the window at the house I was living
in, and I was looking at the Sea of
Cortez and the Pacific
Ocean.
(01:23):
We were about 50 kilometers away
from the Kabul airport, which
is, you know, the busy tourist area.
And you go on a sand road
and they don't build
up roads there.
They scrape roads down,
so they scrape into the sand.
So we arrived at
around 530,
6:00 in the evening.
(01:43):
Of course, by the time we hit the
road at 630, it's dark
and we wake up in the morning and
we're looking over the Sea of
Cortez and it's like, Oh my God.
So we're in the middle of a middle
of a community of maybe 20 houses,
gravity fed water and
solar panels.
And our job was to house it
and care for a dog.
(02:05):
So Michelle and I basically walked
the desert, walked the beaches,
saw turtles, laying eggs,
threw sticks into the surf,
heard coyotes, you
know, went to farmer's markets
and just hung out in Mexico for
four weeks. It was wonderful.
Excellent.
Well, you're away because, you know,
(02:25):
apparently I don't like warm
weather. So instead I went winter
camping with our friend
Pat, his son and my
son. So that was sort of the long
weekend in February.
And we slept in a tent
in the cold and went fishing.
And of course, you know, instead of
catching fish, we got skidoo stuck
in slush and other people had to
rescue us.
(02:46):
And that was only day one.
And then day two.
Brant Sun is sort of a
we'll call him an outdoor teacher.
He sort of has a
we'll call it
almost like an outdoor daycare slash
school. Anyway, he's outside all day
long, so he decided to take us
on a snowshoe.
But, you know, not down some trail
like bushwhacking the whole time.
(03:07):
So he could collect pine resin.
Then he convinced us that it'd be a
great idea for us old farts
to follow him down a semi cliff
down to a lake.
So, you know, sliding on my
butt down a hill and snowshoes,
wondering if I was going to get up
at the bottom anyway.
Good, good times.
Well, we we had a northern wind
coming down, coming down to
the Sea of Cortez, so we had to wear
(03:29):
sweaters at night.
But, you know, it's always the same.
But so now.
We're sort of updated, though.
You found an interesting guy
with an interesting company doing
some techie
virtual reality stuff.
Want to sort of lead us into who
we're talking to in about today?
Yeah, I'll let him introduce himself
(03:50):
because I like when people do that.
They get they pronounce the name is
correct and all that.
But it's a it's a product.
It's a product that really excites
me. I think it's kind of cool.
So so what can
and and a magic VR
helmet that's going to tell me
whether I'm stoned or not, maybe
kind of something.
Yeah, I can kid.
As an entrepreneur, what
(04:11):
I like about this story is that I
like to talk about entrepreneurs.
This is a this is a fella that was
from Montana, and he's
involved with the government
in Montana. He explains his story
and he was thinking about how
will how will cannabis
affect his community.
And he figured, I love this,
(04:33):
he says. He says that there wasn't
a lot of drawbacks with cannabis
coming on market except
the fact that people could be
driving. So he thought,
what could he do to help with
that? Yeah, you picked up on that.
He didn't.
Yeah, you know, no.
His job is job is to analyze
the effects cannabis will have in
the state of Montana.
And his answer is not a lot.
(04:54):
Not a lot.
But it's going to actually benefit
them, you know, benefit commerce,
benefit business, which I
thought was really quite
interesting.
But however, he did see a drawback
with people driving and handling
heavy equipment and safeties.
So he did some research.
Came up with this gadget called Days
(05:15):
g, I, z
E, and it's real
time measurement of impairment,
specifically the
cannabis.
Now he gets on
to this about
how he hopes to,
you know, get more data into
the system where he'll be able to
expand it to other substances.
(05:35):
But right now, right now,
he's focused. He's focused on
cannabis.
Yeah, and we'll get into this more
afterwards. But there are
some terms he uses like night
stigmas and eye movement.
And you you actually do that
a fair bit in sort of your day to
day clinical work.
But I think
he does it not a bad job of
explaining all that.
And then we'll of course throw our
(05:55):
$0.02 in at the end.
So. So we just let can
take. Yeah.
Yeah, I think so. But a couple of
things so just so people understand.
We talk about nystagmus a lot
in this because it's sort of the
movement of the eyes within stag
misses and it gives
you a window into the neurological
neurological assessment of
(06:16):
your of your client, of your patient
in gaze.
His machine tests
for six and
of the six eye movements, two
of them are very specific to
cannabis. So as you're listening to
this interview, ensure that you hear
that because we do use the word
nystagmus a lot and some of the
movements aren't necessarily
(06:37):
nystagmus, they're how are your
pupil responds to light, which is
not nystagmus.
So, so just be aware
that that we're using the word
nystagmus. But it doesn't describe
all of the six movements.
Does that make sense?
I think so, yeah.
Yeah. So, yeah, let's get into the
discussion and then we'll come
out and get a little deeper into it.
Does.
(06:58):
My name's Ken Letterman, the founder
and CEO of Gaze.
We are creating what I believe
is the first cannabis impairment
detection device that works in real
time.
And basically what we've done is
we've taken the drug recognition
expert eye tests, which are
the tracking my finger tests that
are pretty commonly portrayed on
movies or TV.
And we put those in
(07:19):
a VR headset so programed them to
run exactly according to the
training manual.
And we can very precisely
control the amount of light
that's entering the eyes and things
like that. And so what we do is we
run through maybe the exact same
tests that the cops use
and we capture eye movement data and
video throughout that process.
We provide the video as evidence to
(07:39):
law enforcement and
the eye movement data then gets
analyzed by machine learning and
statistical algorithms to determine
whether or not there is
signs or symptoms of impairment that
we can detect in the eyes.
How old is your company?
We were incorporated in January
of 21, so we're about
two years old almost.
(08:00):
Exactly.
Cool. Cool concept.
I mean, I've been doing some
studying. I'm on your Web page.
Your web page is very informative.
It has lots of studies out there.
And I'm looking at the 1977
study about how,
you know, the difficulties that
individuals have with determining
impairment.
What gave you the spark to say,
(08:20):
hey, I can we can create
this device?
Yeah.
So I was the director of economic
development for the state of Montana
for four years under Governor
Steve Bullock.
And in that role, I
was sort of charged with
understanding what was going to be
the impact on, you know, various
initiatives.
And one of the things that
I was studying was Montana
(08:41):
was a medical cannabis only state at
the time, and I thought it was very
likely would become a recreational
state if we ever saw a ballot
initiative.
Madison's come to pass.
And so what I was doing at the time
was trying to understand what the
impacts would be.
And so I was looking at, you know,
what will be the impacts to the
state economy, what would happen to
tax revenue, what would happen to
the businesses and law enforcement.
And it was all really positive
(09:03):
with this one big exception, and
that was that there was no device to
check for impairment.
And so law enforcement and
business owners were very concerned
that they were going to see this
spike of impaired drivers and
workers.
So I initially approached this
problem from the perspective
of just, you know, understanding
this was, you know, a huge
opportunity one.
And I thought that it was important
(09:24):
to solve this problem in order to
make sure that cannabis legalization
could proceed and proceed safely.
So I looked around at the companies
that existed and nobody was doing
it right. Everyone was trying to
measure impairment by looking at the
amount of THC in the body.
If you look at any of the science,
what becomes very, very clear
immediately is that you cannot look
(09:44):
at the amount of THC in the body and
derive impairment from that number.
There is there's simply no
predictive amount
of THC in the body and
no consistent amount of impairment
is experienced on any number, on
any amount of THC.
So from that point,
it almost became more interesting to
me and I thought, well, nobody
(10:07):
seems like they're doing this right.
And if if it was to be done
right, what would it look like?
And so what I thought it could look
like is what what the tests that
we know work are these
drug recognition expert tests.
They've been shown to be, you
know, reasonably successful
at determining whether or not
someone is impaired on cannabis and
many other drugs, that
(10:29):
they are not widely
used. There's not enough drugs
in the country.
These drug recognition experts are
put through very rigorous training.
It's very difficult.
It's not an attractive thing for
most cops to do.
So there aren't enough of them.
And obviously businesses don't have
access either.
So my thought really was what would
happen if we automated these tests
(10:49):
and it could that be done?
And that really led me down this
path of looking into what tests
were most predictive, of what tests
were most objective,
what are really the shortcomings of
doing this through human?
And
there are some, you know, really
obvious correlations when you start
looking at it from that perspective.
And that is the human tests are by
(11:10):
far the most objective.
The limitations of a human are
obviously, you know, you've got
a lot of opportunity for human
error, you've got opportunity for
subjectivity, bias.
And then these officers are
performing the test based on
memorized test procedures, which is,
you know, probably never a good
thing for trying to get to real
consistency.
(11:30):
So I thought,
you know, we could probably automate
these tests using a VR headset.
And so and if we
did that, we could probably track AI
movement using these very high
precision sensors.
And if we did that, we could
probably use that data and
train machine learning models to
recognize impairment.
And so that's really what we've
done.
Very cool. Now, economic
(11:50):
development really isn't I mean,
it's social science.
It's not a lot of science science.
So how did a social science
guy get into the science of all
this?
AI making an assumption and making
an assumption that you're more into
the social sciences, but not
exactly.
I mean, I've had a couple tech
companies in the past, and
the economic development job
(12:11):
was really a departure from what had
been the norm of my career.
It looked like a really interesting
opportunity. It certainly was.
And I feel like we did some great
work in that role. But that was
that was more than the aberration
for my career.
Okay. So so a little background on
your career would help.
I want to get into the science of
your product, but tell me more about
yourself.
Yeah. So I'm a multi
(12:32):
time entrepreneur, native Montana,
and I
spent some time in a company
doing a turnaround up, a company
that was doing infrared optics
manufacturing.
I had a search engine optimization
company, I had a web app
development company,
I had a marketing company.
And so I've I've sort of had a
(12:53):
couple
different sprints in my career.
And
what I have sort of decided
my specialty is, is thinking about
hard problems and designing
solutions around them and then
building a team to tackle that
problem. So that's what I've done
here. I'm not the scientist
that has done this research,
but I have assembled a team that has
(13:13):
done that and done it successfully.
So I'm more of a business guy.
Well, I congratulate you on this.
So let's talk a little bit about
Gase, then.
Trevor and I, I'm not sure if you're
familiar with our project, but
Trevor Trevor is a pharmacist.
I'm a nurse. What?
We've discussed nystagmus and we'll
we'll define that as part
of our narrative.
But your, your, your device measures
(13:35):
nystagmus.
Now, I guess the question I have
for you is
I noticed on your web page you guys
did a 350
participant study
on your product.
So did that study confirm
your science or was it a study
to help your your your product
learn?
Sure. Yeah. So really, the answer is
(13:56):
both.
That data set is used
as the training data for our
machine learning model.
We're also able to validate that our
machine learning model is effective
based on that same data set.
So it is sort of one in
the same.
I'm really though that was
the most important thing from that
study was to capture a large amount
of training data to create these
(14:18):
machine learning physical models.
Okay, So when someone consumes
cannabis because on your website
you're truly focusing on cannabis
impairment.
So is there a specific
type of nystagmus that happens
with cannabis consumption?
Yeah. So our our product actually
runs through.
So each of the drug recognition
expert eye tests. So I'll list them.
(14:38):
Those include equal tracking,
lack of smooth pursuit,
horizontal gaze of status and
maximum deviation.
So that's the far left and right
peripheries of the vision
causing gaze to stagnant at 45
degrees from center
and vertical gaze is
stagnant. So that's vertical
periphery, lack of
(14:59):
convergence and people are rebound
dilation.
So each of those tests is looking at
a specific thing.
The horizontal gaze, the stag was
test and the nystagmus tests
overall are much more
predictive of impairment on other
drugs, primarily alcohol.
So we are certainly using
that data and our machine learning
model for cannabis.
(15:19):
But it is not one of the tests that
we're finding to be highly
predictive of cannabis impairment.
And certainly I'm sorry, So what
is your gaze isn't in this segment
isn't correct.
So nystagmus is not one of the
indicators of of cannabis
impairment that we're finding.
It is occasionally
present, particularly in extremely
impaired people, but it is not
(15:41):
universally present.
So we're really focused on the
lack of convergence test and the
people. A rebound dilation test like
a convergence, basically
brings the stimulus towards your
nose, the rigid nose, and tries
to get you to cross your eyes.
And the people revaluation test
exposes your eye to three different
light conditions. So that's room
light and then 90
(16:01):
seconds of darkness and allows
your people to fully dilate
and then we expose it to bright
white light and that causes
your pupils to constrict and we
measure how
rapidly they constrict and whether
or not they stay in a constricted
state.
Okay. So to review that.
So your device does.
Did you say five tests
was like six tests?
(16:22):
And out of those six, how
many of them are predictive of
cannabis?
Two are two are very predictive of
cannabis. We're finding some signal
and the other tests, to be clear,
But it is those are not
the tests that we're really focused
on.
Okay. But for the purposes of
defense and law, you want
all those tests in there and then
they will focus up on the two.
(16:44):
Yeah, because really the vision for
Gaze over time is to add additional
substances.
I want to create what I'm calling
the first impairment detection
platform. So a single device that
can detect impairment from any
substance.
And I you know, I always start this
conversation by saying I'm not
opposed to cannabis legalization at
all. I think this is a good thing
and I think it makes a lot of sense
of alcohol's can be legal, Cannabis
(17:04):
should certainly be legal,
but I think we need a way to make
sure it's done safely.
We need a way to ensure that
people are not driving, working
while high because, you know, there
are extremely bad outcomes that come
from that.
So the reason that we do the each
of the tests that the drug
recognition expert does is because
we want our device to ultimately
(17:24):
be able to detect impairment from
any drug or any combination
of drug. And so by
conducting the plurality of those
tests, we can we can determine
if someone's high drunk,
drunk or high,
high and meth, anything.
We just need the training data and
we design algorithms to do that.
It's a very cool it's a very cool
(17:46):
concept. I mean, in health care, I
measure nystagmus as
as an assessment for muscle,
skeletal, neurological, you
know, chronic disease issues.
So walk me through
your selling points to a law
enforcement agency.
How do you how do you say to
a law enforcement agency, here's
how our device will work in
(18:06):
the field?
Yeah. So the
law enforcement use case really is a
bit different than the commercial
use case in that, you know,
obviously there are criminal justice
outcomes that are coming from these
tests. And so one of the things
that we're really focused on for law
enforcement is the generation
of evidence.
That's something that is of marginal
value, all safe to commercial
customers. But for law enforcement,
(18:26):
I think that's of huge value
right now. When a law enforcement
officer conducts a
standardized field sobriety test on
the side of a road or on a drug
recognition expert conducts their
test back at the station,
what evidence is generated
really is only coming from perhaps a
body cam if it was activated or a
dash cam on a patrol car.
(18:46):
There's not any eye
movement data or video that's
generated in this process, even
though these are the most objective
tests that are used.
And so our device
records eye movement from about
an inch away from the eyeball.
And so it is very
up close to very clear video
of what's happening to the
eyes and in the eyes from
(19:08):
a perfectly conducted
very eye test.
So we think that evidence piece is
really, really important for law
enforcement.
The other piece of it, you know,
right now there's no tool that law
enforcement can use to detect a
cannabis driver.
It is really
at the whims of the patrol officer
whether or not they think someone is
impaired on cannabis, whether
(19:28):
they're going to bring them back to
the station for additional tests.
So what we're providing is
really a stand in for with a
portable breathalyzer test
that can be used on the side of the
road to determine whether or not
someone is likely to be impaired on
cannabis with reasonably
very high precision, where our
devices, it's much more accurate
than even the best trained human
(19:49):
drug recognition expert officers.
And so while
we encourage confirmatory
tests to be done every time, I
guess devices predictive of someone
to be impaired by it
is certainly the best technology
available for that purpose.
And so we think that that's the
other really pressing use case.
I think.
Okay. So I guess if I'm
(20:10):
if I am sitting in my car being
pulled over and and
the police officer,
the law enforcement officer thinks
that I'm impaired, he'd ask
me to step out of the car, probably
put me through the, you know, the
walk and turn pass probably the one
one leg test.
But instead of doing his
his subjective test or objective
(20:31):
test of nystagmus, he put the
gaze on at
the roadside. Or would that be
something you would do in the in
at the shop, at the jail?
Yeah, it can be done.
In either case, the vision
here is if if someone.
Is exhibiting signs of impairment in
the roadside.
They are blowing a zero on
a breathalyzer.
(20:51):
The cop has no other tools other
than simply placing them under
arrest. Know that person could
simply be fatigued,
in which case they would probably
be, you know, prone to stop
driving and just go home.
Or they could be impaired on canvas
or another drug. And so.
Gays could be used in a roadside
setting.
Gays. You would require that people
are seated while they're going
(21:12):
through a test, so you'd probably
lean up against the car or sit
in your own car while that was being
conducted.
But the result of that test would be
very clear
data about whether or not someone is
actively impaired.
So so there's the
officer would would key in the
person's name driver's license,
date, all the all the tombstone
information that they would use in a
(21:33):
court.
And then the computer technology
then says, you know, there's
a green light. The guy is a go,
there's a yellow light cautionary,
there's a red light. Oop, the
signals are out, The guy is
impaired.
Is that more or less.
So we run through, as I said, six
different tests. We provide
information about whether or not the
person is passing or failing each of
those tests. And we provide
(21:54):
a predicted substance of impairment
and are confident that that person
is improving that substance.
And so the
outcome for the patrol officer
really is verification that, yeah,
we're pretty sure this person is
impaired, we're pretty sure they're
impairment cannabis.
Let's go ahead and bring it back to
the station and do some other tests.
Okay. But if he pulls up myself, I
mean, I'm in my sixties.
Is there is there a difference in
(22:15):
your study between demographics?
Do like some people have
nystagmus as a
chronic ailment?
So how does your how does your
device quantify
age, demographic sex?
Does it do anything like that?
So in our study, we captured all
this demographic data.
What we're doing right now actually
is going through our our machine
(22:36):
learning model and ensuring that
there's no demographic bias built
in. So, for example, for detecting
more females and males as
impaired, then we need to understand
what's happening there and remove
any any existing bias.
Any time you train a machine
learning model, there's the
potential that you're training in
bias based on what your training
data is.
(22:56):
And so we were making sure that
that is not the case.
And removing bias as we find it,
there is you know, there are many
genetic conditions that can lead to
many different types of abnormal eye
movements.
And so another good reason
for us to do the plurality of the
tests that we use is that
there are there's a reason that
someone can fail a single test.
(23:18):
But when we look at the tests in
a collective,
we can very precisely understand
whether or not the the
eye movement behavior we're seeing
is correlated to a particular
substance.
And so that's, I think,
really, really important because
there are so many reasons
(23:39):
that you could be exhibiting
nystagmus, for example.
There are many reasons that you
could have abnormally
large pupils.
But when you look at, for
example, a failure on lack of
conversions and a failure on pupil
rebound dilation, and then some
perhaps some other indicators from
other points in the test, you can
be very confident that this person
is impaired on cannabis and not
(24:00):
simply tired or
drunk or something else.
Okay.
Your product is still a couple
of years old, so it's truly in its
infancy
and I can see how time
goes on. It's going to become more
robust.
So how is information gathered now?
And I guess the other question is
confidentiality.
I'm going to make an assumption that
(24:22):
if I'm working for snowshoe
up or snowshoe in Montana Police
Department and
I'm using your product, I will have
a database within my department that
the information goes to
that would have the tombstone
information because I need for
evidence.
Does that that it now get pumped up
to the cloud, to your to your
grand device and you like
(24:43):
and then you work with the data.
How does that all work?
Yeah. So the data flow basically is
we've got a we've got three
different absolute built.
The first is a mobile device
app that basically controls the
headset. So the headset is a slave
to the metal device that allows the
the officer in this case or a,
you know, H.R. person or whoever
to press the start button and
conduct the test.
(25:05):
The data from there flows up to our
cloud server, where it gets
processed using our machine
learning algorithm, and that
result then flows back down to
the mobile device to alert the
test administrator whether or not
someone is higher, if they're, you
know, if they're so where they can
go.
So that data is
is there is no
(25:26):
personally identifiable information.
So it's for privacy.
We don't capturing personally
identifiable information in that
data set. And so even in a worst
case, we were hacked or something
that would never be correlated back
to an individual.
The final piece of software
developed is a web app that allows,
for example, a station to look at
the tests that all of their officers
(25:47):
are doing.
And so if you're the chief of
police, you want to understand who
you're which which patrol officers
are using gas mos Most often you can
go to this web app.
You can see where they're conducting
tests, How often, how many of those
were positive results,
download test results, download
evidence, all that.
Okay. So so obviously unit
you need cell service.
If you're doing it in on the scene
(26:09):
or you bring the person back to the
to the building and you have your
Wi-Fi system.
Yeah.
We're looking at ways to
do on device.
So on the mobile device, do the
machine learning processing there.
Unfortunately, there's a huge
variability in processing power
amongst mobile devices.
And so whereas like the latest
iPhone can probably run the model,
I'm not sure that, you know, an
older phone could do
(26:31):
the same. So
we'll probably get there.
But that's a that's a feature state
that we're looking at.
Ironically, I'm thinking of
law enforcement, but you're also
marketing to manufacturers.
So people that, you know, crane
operators, people that are, I
imagine, working in heavy duty.
Heavy duty devices.
Mm hmm.
Yeah. So anybody that's in a safety
sensitive role.
(26:52):
So that would be anyone whose life
is depending on someone performing
their job appropriately.
We're interested in selling devices
into those positions.
Commercial On the commercial side,
the only screening tool that's
currently available really is
urinalysis screening, and that is a
distinctly retroactive way to
look at impairment.
You can tell that someone has
(27:12):
previously used cannabis, but you
cannot tell based on a urinalysis
screening whether or not they're
currently high.
So businesses really have no tools
at all to detect whether or not
someone is currently impaired on
THC and in the
United States.
I'm not sure what the Stats Canada
is, but in the United States, THC
use in states where it's
recreationally legal or
(27:33):
where adult use cannabis is legal
are starting to protect THC use as
a as a class.
And so what that means is
that you cannot, as a business now
perform THC tests and
take adverse action against an
employee based on a positive result.
You have to actually prove that they
were impaired during their
working hours.
And so that's a that's impossible
(27:54):
for businesses currently.
We're really providing, I think, one
of the only ways that that could
reasonably be done.
And so our hope is that this can
really drive safety radically
higher in businesses that are that
have safety sensitive employees.
Yeah, that's I guess
that's where the the skeptics
will say just because
(28:15):
my eyes are behaving in such a way,
how do you prove I'm impaired?
I may be a medical cannabis user.
So if I'm a medical cannabis user
and I'm microdosing my my cannabis
is is Gage going to say
I'm impaired?
No, that's that's a really great
point. So the eye movement
characteristics that we look at are
only present when you have
(28:37):
sort of demonstrable
reductions in your, in your
ability to do
equal tracking or to do
divided attention tasks or reaction
time tasks. And so these are
things that are only
indicative of of impairment
that is, you know,
above and beyond what you'd want to
(28:58):
be operating a vehicle or equipment
under. Okay.
So is there someone
I don't know if this is a fair
question, Can I develop a tolerance,
like if I've been using cannabis for
five years and all
of a sudden my employer says, here,
put this device on?
Well, buddy, I've been working for
you for five years using medical
cannabis.
(29:18):
Is there an is there a tolerance
to the body develop tolerance?
Definitely.
Cannabis has a has a actually a huge
tolerance effect.
So heavy users of cannabis will have
a radically lower amount of
impairment that comes from
using the same amount of THC that
that a infrequent
user can become extremely impaired
(29:39):
on.
So tolerance is one of the things
that our device is particularly
useful for because it is again, it's
only detecting
impairment that
can be objectively
sort of correlated to
reduced ability to perform a
job or to drive a vehicle.
So there's been studies that have
(30:00):
validated this. There's a great one
that came out last year from the
UCSD Center for Medical Cannabis
Research that looked at
what is the sort of impairment
that happens driving
impairment in particular in this
case that happens from consuming
cannabis. And they found results
that were consistent with what we
found.
But basically tolerance has a huge
(30:20):
impact. And so.
A person with a large tolerance
for THC would have to consume
much more cannabis to achieve these
same failure states
that a person that consumes
very little cannabis would do.
I'm wondering if your data
collection captures that.
Like if I'm
if I'm a chronic if I shouldn't
(30:42):
say chronic, but if I'm a medicinal
cannabis user and I get pulled
over, I guess the officer
has to make other determinations of
why he pulled you over and then add
that add that to the database
that your device gives them.
So would you be capturing that?
Do they say in there that what they
have prescribed, would
they have the list of prescribed
(31:02):
medicines somebody is on that you
would capture that?
We don't capture that data, but the
officer would capture that.
And so that's a determination the
officer would have to make.
You bring up a really good point,
though, about tolerance and about
heavy usage with THC.
You know, the current state of of
the situation is that if an officer
pulls you over and they suspect
you're driving while impaired on
(31:23):
THC, don't do
a saliva screening or a urinalysis
screening or in the law enforcement
case, law was probably a blood draw
and they'll look for
THC in your blood.
And if you're above a certain amount
in Canada, I believe it's.
Five milligrams per milliliter if
I'm not.
Yeah. Yeah.
(31:43):
Most other states in
the U.S. that have legislated
legal cannabis have also established
these per se limits.
What's important to know about those
is that there is no science at all
that backs up the idea that
you can have, you know, five
milligrams per milliliter of senior
blood indicates that you're
impaired, that science does not
exist. And in fact, it's quite the
opposite.
There been many studies that have
(32:05):
looked at this issue and verified
that there is no amount of THC in
your body that can be correlated
to some predictable amount of
impairment.
So fundamentally, these five
per milliliter per se laws
are nonsensical and were
created entirely to facilitate
prosecution of US
cannabis users.
(32:26):
And so our product provides
a much more rational path
forward and that we are looking at
actual impairment as it is
manifesting in the body and
not some arbitrary amount of THC.
So a heavy user of cannabis will
always have THC in their body that
is above the legal limit.
And there was actually a case that
came before Washington
(32:47):
State Supreme Court last year
that was
it was exactly this where a guy
had not used cannabis for 24 hours,
was given a blood draw
and was thought to be over the legal
limit. And he challenged the state
on whether or not that was a
rational lie
in what I think is the most insane
court ruling I've ever seen.
(33:07):
The court came back and said, Yeah,
we agree there's no science behind
this whatsoever, but we're going to
stick with it because everything
else that
was the state of Washington versus
Frazier. And it's
it's just, I think, a really
terrible criminal justice outcome.
And it's exactly the kind of thing
that cannabis legalization
is supposed to be moving us away
from. And in fact, it has delivered,
(33:29):
you know, exactly into this really
bad situation.
So I believe that a device
that measures impairment in the body
is a much, much better path forward
and should be supported by the
cannabis industry because of
the reasons I'm describing.
Yeah, no, I agree with you.
If you're impaired,
don't drive.
And what we've what we've learned
(33:50):
through our research, through our
podcast is that some
of the research coming out of
Australia suggests that those
those that use cannabis
will sometimes self
self-regulate themselves.
Yeah, I'm too stoned to drive, man.
Whereas whereas alcohol,
I don't know about alcohol, but the
car can find its own way home.
Right. One. One for the ditch.
(34:11):
Off we go.
So I think there's a behavior
difference between cannabis users
sometimes and alcohol users.
But this leads me to another quote.
This leads me to another question,
obviously, and I shouldn't say
obviously there are, what, 108
cannabinoids in the cannabis plant.
So your research is saying that
right now, if someone's impaired,
(34:31):
it's measuring THC impairment.
I mean, CBD is, we know, does not
cause psychotropic effects, But
is there any core is there any
differentiation in your studies
between the cannabinoids or is it
just impairment?
Yeah, it's just impairment.
So Delta nine, THC,
the people in our study consumed
flour, so they consumed,
(34:52):
you know, kind of dried flower
cannabis.
We've we've done some small
amount of confirmatory testing with
vaped and edible cannabis.
But when you do more work in those
areas, it appears,
however, that the same impairment
manifests across
methods of ingestion.
And so we're very confident advice
for that respect.
(35:15):
You bring up an interesting point,
though. If you start looking at
synthetic cannabinoids,
it becomes a different story.
And there's a there's almost
zero science that has been done.
And looking at how
these synthetic cannabinoids
impact eye movement or
how the impairment can otherwise be
(35:35):
characterized.
And that is
going to be, I think, a very
difficult thing to get
our arms around because there's just
so many of them.
Well, we've often talked about this,
especially early in this project.
We've been doing this about five
years. And very early on we
often talked about how many times
Trevor being the pharmacist fills
(35:56):
a prescription that.
That is a psychotropic prescription
that's going to have some effect on
muscle skeletal movements, going to
have some effect on cerebral
thought processes.
Any any And you watch as the
person, you know, get into the
parking lot takes
one and gets in the car and drives
away. Right.
So we know that people.
(36:16):
But it's okay because it's a
prescribed medicine, right?
It's okay. They can drive on it.
So. Right.
So there's your there's your device.
I mean, nystagmus.
Other substances
create nystagmus.
So as you said earlier, I
guess you're hoping that the
learning technology and your device
will teach you about what other
substances are being abused
(36:38):
or used.
Yeah. So what's what's
really interesting about these tests
is that each class of substance
has a unique way in which it impacts
eye movement.
There's been not as much research
done on on prescriptions,
frankly, but for each class
of drug. So each class of illegal
drugs are regulated substance
like alcohol, opiates,
(37:00):
stimulants, central nervous
system depressants, cannabis,
inhalants, psychedelics.
Each of these has a unique way in
which it impacts movement, and we
believe that we can detect
impairment from at least most
of those, if not all of them.
The work is simply
capture a data set that is
of good enough quality and
(37:20):
large enough number that we can
put our models at it and train it to
recognize that impairment.
Fantastic.
I only got about 5 minutes left with
you here. A couple of real quick
questions. You're in the marketplace
now. People are buying your product.
We are almost in the marketplace
now. We have we have a good
we have a really strong preorder
list. We are, as
I said earlier, we're eliminating
(37:41):
the bias that we could have trained
in during the course of
our clinical trials. So that work is
ongoing and we're doing a couple of
things around
eliminating false negatives and
false positives in our model.
So there are people that for which
it appears, and these are
occasionally high, high tolerance
people and
(38:02):
some others we're still trying to
understand the cause of.
But there are some people for which
the
consumer, the same amount of
cannabis does not have the same
effect. And so we're trying to
understand what are the causes of
that and how do we determine
whether or not this person is
actively impaired or not.
And in some cases, it appears that
some people are just not impacted
the same way by cannabis.
(38:22):
And so that's that's a bit of an
interesting problem.
But we're our intention is to be in
the market by the end of this month
and with a with a model that
is radically more accurate than the
best trained human.
Kirk That was really interesting.
I really enjoyed your chat
with Ken from Gaize.
(38:43):
Because we talked about a little bit
right at the beginning, I'm going to
pick your brain here again in case
the people missed it.
What's nice, nystigma or nystagmus?
What? What? Like what does it look
like? What? What is it?
Nystagmus is when you're
usually what happens is that you you
ask somebody to focus
on your finger and you go the finger
horizontally and you watch the
(39:05):
pupil
go and follow the fingers.
And what happens is that normally a
person's eyes will track
without any variation.
But with nystagmus,
what'll happen is that your eye
will track, but there will be
variations or quivers little
movements of your pupil,
of your eyeball
(39:27):
as it tracks across.
So that's the nystagmus.
Is that is that unusual
quiver motion that
happens when you're tracking.
Okay. And so now
when you said the
horizontal nystagmus
at the extremities where
relate to alcohol, not
(39:47):
not cannabis, that would be, you
know, out when I'm looking way off
to the right or way off to the left,
that I'd get that quiver.
Right.
Is that is that peripheral
vision.
And as you're as you're as your eye
tracks to
the periphery, there'll be a
quivering there'll be a movement.
That's not it should be smooth,
right? It should not be any
(40:08):
movement.
Okay.
So you'll see a small quiver and
it's abnormal,
right? I mean, it's a lot of
assessments on people as you're
looking for something abnormal.
The average person's eyes track
without any alteration.
And the two that he said were
they they think closely related
to cannabis impairment was
(40:31):
loss of convergence.
So I think everybody can picture
that. That's take the finger to the
middle of your nose and you go
cross-eyed and that they aren't
doing that.
And pupil dilation
sort of not as expected.
They said they'd have different
brightnesses of light and there's
sort of an expected how fast that
pupil opens and closes and I guess
it's not doing as normal.
So it does seem to be that the two
(40:53):
out of his six that
were more related
to they think cannabis
impairment. So that was very
interesting.
Yeah. Yeah. But what I like about is
that they're doing the six
and this goes back way back
in time.
The web page.
The web page has
(41:14):
all their research
and some of the research we're going
to put on our web page.
But going to their Web page,
which is
https://www.gaize.ai
You'll find a lot of the research
and they go back all the way to 1977
(41:34):
when the US government
studied roadside testing
and what it would take for people to
be qualified to do roadside
tests.
So what they've done is they've
taken those tests that people do
that the, you know, the police
would do on the roadside, put
it into an algorithm.
And I think what's really cool about
this business is that they're
(41:54):
capturing all these behaviors
you know, your lying eyes, right?
You can't hide your lying eyes.
And they're going to pick up,
over time how
your eye responds
to substances.
I found it interesting, though, that
they're not tracking people's
prescription drugs.
(42:16):
I would think that they would want
that in their database and
some some point
be able to correlate
common eye behavior's
from certain substances.
He did say that he would like it to
expand out to other drugs but that's
a nice segway into I'm going to pull
the usual Kirk and say we have some
other episodes in our library about
(42:36):
driving We have
E60 for Driving Under the Influence
with Dr. Tom Arkel,
an Australian researcher who put
subjects on the road in the
Netherlands with and without
placebo cannabis to see what would
happen. That what is fascinating.
Still fascinates me.
And then more related, this one
Cause Agnostic with Driveable
(42:58):
and kind of like you were saying
Driveable guys, now it wasn't a VR
headset, it was something more akin
to an iPad but, the reason
they call it Cause Agnostic is
they weren't really caring what
impaired you.
They were looking to see basically,
do you still have the reaction times
necessary to perform safety
(43:20):
sensitive things like, you know,
our crane operator or
whoever or driving?
So just interesting that
there's we've talked to at least two
groups now that are using I will
call it, quote unquote, electronics
to see to see
what
cannabis and or other things are
doing to to impair you.
(43:41):
It's all very interesting.
I think it's fascinating.
What I what I also find interesting
is the nuances of this.
You know, our culture is very
comfortable with alcohol.
In the sense that we're very
comfortable to to not drive
when we're drunk. However, we
have a huge issue in our culture
about people driving drunk.
But usually what happens is that
(44:02):
when someone is under the influence
of alcohol, we know
the impairment because we've had how
many decades
of learning.
You know, buddy, you're too drunk to
drive. Give me your keys.
Cannabis is something different.
What I found interesting in
what he's talking about here is that
his machine is measuring impairment,
(44:22):
not the amount
of THC, you know, whereas
other tests, blood tests, saliva
test, they're testing that you have
THC in your system.
Therefore you might be impaired.
Well, his isn't quite measuring
impairment.
It's much closer.
And I think he made a great point
that, you know, unlike alcohol,
(44:44):
which like you said, we have
decades and decades of, you know, a
blood alcohol level of X equals
impairment. Why?
We've established that.
And we've also established
that a THC
blood level of X doesn't mean
anything. You know, if you are
new to cannabis, you might be
freakin impaired.
If you use it every day because of
(45:05):
medical or whatever, you might not
be impaired at all.
And what you'd really
need to measure impairment at the
roadside would be, you know,
a tractor trailer with
a, you know, pull someone off, I'll
throw them in in a driving
simulator. And that's the only way
you're going to actually know if
they're impaired. So this is a good
surrogate marker, but it's still
(45:25):
not, you know, obviously the
drive around driving
simulator to see if someone's
actually impaired is completely
impractical. But this is
this is a nice surrogate marker.
That and big bonus
is it doesn't require any
blood tests.
And back
to what I think you were getting at
earlier is, right now we don't
(45:46):
have, we can take blood, we can
measure a THC level,
but that number is pretty
meaningless.
Well, I interpreted it a little
differently in the sense that
in the model.
The police officer is
driving down the road and he
sees something that
makes him believe he's going to pull
(46:07):
over this driver. And now there
could be a check stop.
It could be erratic, irregular
behavior on the road.
But for whatever reason, the
officer has pulled you over,
comes up to the car, you know,
touches the driver's side
panel on the back.
So, you know, there's witness that
he was there, that old trick and
says, hello, sir.
Driver's Registration.
(46:28):
Now, if it's like,
you know, This Hour has 22-minutes
where they roll down the window and
a cloud of smoke comes out at
you, then he's got reasonable,
right to say, you know, I'd like to
do a test on you if he
like if he can smell cannabis or he
smells or smells a product in
the vehicle or the person's
(46:49):
behavior, or maybe he's
asking some questions.
You know. Sir, have you had any
intoxicants?
You know, I'm a medicinal cannabis
user or whatever.
He can then apply this device.
And my interpretation is that
through the interpretation of
the two primary
two primary tests that
Gaize can predict that,
(47:09):
yes, from the results
of the test, we can
predict this person is impaired.
That's how I interpreted what he was
saying.
So I'm thinking, but before that
happens in a law enforcement
situation, there has to be a
reason for the officer
to have initiated
the test.
Now, from a commercial working
(47:30):
on on an industrial site,
it may be Tuesday
we bring Gaize on to the site
and do arbitrary
tests on people for this for
safety, meeting occupational health
and safety.
And again, if you identify
as a cannabis user or there's a
reason to believe you use cannabis
(47:51):
and they put these goggles
on you this AV goggles
and you fail these two particular
tests, they're saying that you
are impaired.
I really like that.
And you mentioned a couple of times
it's worth repeating.
So again, like you said, getting
away from the law enforcement thing,
an employer with a safety sensitive.
(48:11):
Workforce.
And, you know, we're going to I'm
going to use crane operator as
my fictitious person.
You know, if you
right now, you could make him pee in
a bottle.
But that, again, doesn't really
say much about their
impairment level.
And it also doesn't say what's going
on right now. That might be I smoked
(48:32):
cannabis three days ago.
So the nice thing about
gays is there really
isn't any good tools that
an employer has now
to say whether or
not it's a good idea
for, you know, this person to
go up in the crane.
So this
sounds like a good at least
(48:53):
gives the employer an option
before they you know,
if we're going to do some kind of
random testing or not so random
testing of employees before they go
do safety sensitive work.
Mm hmm.
When we interviewed when we
interviewed Ken, we interviewed him
gosh, I think I interviewed him in
January. This one's been on this
show for a while, and I apologize
(49:14):
for that. But he sent me an email.
I was in Mexico, actually sent
me an email February 13th
and said that he had launched.
So this this product is now
out there.
Good, good. So February 13th,
it went live. So and we'll have a
link to the website on in the show
notes. But yeah, if this catches
your fancy because you're part
(49:34):
of law enforcement or
a safety sensitive employer,
this is a tool that you can
get your hands on now.
Speaking of driving, I know you
were looking up to see
things at some stats
and study. Try and get some current
stats and study about impairment
(49:55):
and driving in Canada and cannabis.
What? What did you stumble on to?
Yeah, well, I receive
updates from Health from the
Government of Canada on new
research, and this is brand new
research out March
2023 in
Health Promotions, Chronic Disease
Prevention Canada.
This document came out from
(50:18):
the Government of Canada sponsored
by and the study is
called Impact of Substance
Related Harms on Injury
Hospitalizations in Canada
from 2010 to 2020.
So over a ten year period of time,
they're studying how
many hospitalizations are a result
(50:38):
of substance
related harms.
Now 2018
was legalization of Canada
of cannabis.
Specifically driving, or you couldharm yourself
in any way.
Any way at all.
It's all substance related
hospitalizations, right?
So, there's
(50:59):
2,108,489
samples in this study.
So, of course, poly-substance
abuse rates the top.
The most people that go into
hospital due to
self intended or
(51:19):
what the terminology you use here,
they use the intended substance
related injuries hospitalization
and unintended.
In all cases, people that
mix substances, opium,
alcohol, whatever you're
mixing with, that's the top dog.
But what I found interesting is
and is that
(51:40):
the female the Intentional
Substance related injuries so
people that that intended
to take the substances
and ended up injuring themselves and
in a hospital the highest
risk group was females between
15 and 19 years old
and females
are.
Are we talking about like self-harm
(52:02):
as in, you know, a suicide attempt
that's probably in there somewhere.
Some of it's in there.
Yeah, some of it's in there.
But you know, but it's just
it's, it's from 14
all the way all
the way to 74 females
rank highest in the age
groups of of intended substance
related injuries.
So for whatever reason they were
(52:23):
taking a substance and hurt
themselves. Unintended
substance related injuries.
And this is fascinating to me the
unintended ones are
these 75 to 85
year old. The highest level, 85
plus. Now, that is people who are
on prescription medications
and.
Took the wrong one or took too many.
(52:44):
Took too many fell down.
Falls, right.
So there's some cool
statistics in here and
they're not going to be surprising
to every one. You know, everyone
because, of course, alcohol.
Right. It's a specific
rates of substance related
motor vehicle collisions
that ended up hospitalized.
(53:05):
The highest among these were 20 to
29 year olds with a hospitalization
rate of 7.2
in 100,000 for men, 3.2,
100,000 for women.
So young boys, young men
are ending up in the hospital,
most for alcohol, for
motor vehicle accidents.
So alcohol
(53:25):
use concurred most frequently
to falls and motor vehicle
accidents. I don't think we're
surprised about that.
But where we might be surprised
about when it
came to cannabinoids.
And here's the quote.
"We found that cannabinoids
account for less than 1%
of all substance related injury
(53:47):
hospitalizations, which
is low relative to the proportions
of Canadians and North American
populations using cannabis."
So, you know, this is
again, I'm not promoting the use
of cannabis, but I guess what I'm
saying is that from the studies,
from our own interviews,
(54:08):
when people were so afraid
of legalization of cannabis that
we're going to have more vehicle
access, we're going to haveaccidents,
It's not happening.
Well, it's not happening.
It's not happening from statistics.
It's not happening from our
research.
And just because I can't resist
your, you know, alcohol related
falls.
So a couple of my favorite books
(54:29):
are Freakonomics.
Well, the whole Freakonomics set.
And now.
I love freakonmics
And now it's a podcast.
And one of the things they quoted in
one of their books was, you know,
how unsafe walking and drinking was
because, you know, far more people
that they weren't promoting drinking
and driving either.
But they said if you look at the
(54:50):
stats, you know, far more people,
you know, get drunk, walk home,
fall and crack their head on the
sidewalk than anything else.
So, you know, drinking and walking
wasn't recommended either.
Yeah, well, I mean, now.
Now.
No, no alcohol.
No alcohol is good for you in
the new studies.
But I
look forward to seeing Gaize grow.
(55:11):
I think this
is a fantastic
product from the perspective
of another, just another
tool to justify,
justifies not the word, I guess
another tool that we can use to just
protect people, Right.
I always say
that one thing about driving, you
(55:32):
know, I have driven a car
in many different vehicles, in
many different lands and
different countries.
And what I have found is that we
all, no matter where you are
driving, everybody wants to
get home, right?
We all we all drive
to get home and we drive
to protect ourselves from
hurting other people because if we
(55:53):
hurt, other people are hurting
ourselves. So I think driving,
I think driving is one of the
biggest communist plots out there.
It's one of the biggest social...
Think about it's one of the biggest
social responsibilities you have.
And we pay to get the license
right. We pay to get the physical.
We pay for the privilege
to do it.
(56:14):
And we do it for all the same
reason to get someplace safe.
So I think these guys
have a fantastic product.
I like the idea that
it's your
lying eyes, man.
You can't hide them.
And I think this is the name of the
episode. You can't hide your lying
eyes. So good on
them.
Yep. No, noninvasive.
(56:35):
You know, we always like an AI
story.
I think this is very cool.
I'm hoping to follow them in the
future as they get more and more
data and maybe more and more drugs
in their list.
But speaking of driving, Kirk,
you and I took a road trip to
Yorkton Saskatchewan,
and and you
talk to a to a random
hockey player in a in a bar
(56:56):
we were at.
Yeah.
One of our what was set up one
to set up this story.
Yeah. One of our buddies
on our team we went to a
Proye Cup which is
well you are.
Fat old man hockey.
Oh, it's some of the best, worst
hockey I've ever seen in my life.
I loved being the trainer
for the local Kinsman team.
(57:17):
It was a lot of fun.
Anyways, I digress.
We were having lunch and
I was again talking
about our podcast with an
acquaintance friend of ours, and he
said, I said we were trying to get
My Cannabis Stories.
He said, I got on my cannabis story
for you and
let's listen. This is a buddy at a
hockey tournament.
Hi, my name's Dan Rojansky, and I
(57:38):
was asked to do a quick excerpt on
what happens to me when I consume or
smoke marijuana.
Essentially, I get the munchies and
everyone is aware of what the
munchies are, except mine are
incredibly violent.
Meaning I eat everything and a lot
of it.
It doesn't matter how much I've
eaten for the balance of the day.
Once I smoke weed, I could
out eat the largest man you've ever
met. I will eat a full large pizza
(57:58):
on my own without breaking a sweat.
And then I'll eat wings.
And then I'll eat something sweet
like a chocolate cake.
It's violent.
So weed is awesome.
But be aware that the side effect
is weight gain because you eat a
lot.
So that was Dan.
Dan from Winnipeg.
It's it's always fun when a
someone you know would be it's it's
a funny story, you know.
(58:19):
Well.
The heart tugging ones are good too
but you know.
Well it's what I like about
it. It's an honest story.
The guy basically saying is
that when he gets high, he gets the
munchies. It's like, oh, gosh,
how many years ago?
I got I got one from an
acquaintance, a colleague up north
that she got high going to a concert
and couldn't drive, you know.
(58:40):
So I like My Cannabis
Stories because everybody has
a different experience with
cannabis. And that's
what makes this such an interesting
substance
to learn because everyone
behaves differently.
Do you want to talk about the
Canada Cannabis Act?
(59:02):
What we are, what we're listening
to today on the radio?
Sure. So not sure when this is
dropping. So it was current when
Kirk and I were talking about it,
but who knows what it'll be by the
time this drops.
But within the last 24
hours, Quebec
said you can't grow pot
in your own house.
(59:22):
And that was the Supreme Court of
Canada ruling.
Yeah, it's going to be interesting
to see how that falls up because
Manitoba has theirs
sitting at the Supreme Court right
now. So that's going to be sent
we're going to be following.
Absolutely.
So I think I better say
I am Trevor Shewfelt. I'm thepharmacist.
I have Kirk tonight because I'm the
registered nurse and we are Reefer
Medness - The Podcast.
(59:43):
Go to our web page.
All the research from
Gaize is,
all the things that Ken was
saying about Gaize.
It's all research.
It's on his web page.
There'll be a link on our Web page.
I will also have the research that
talked about the the stuff
from the Health Promotions
(01:00:04):
Canada that'll be on there.
Go to our web page.
It's searchable.
Episode 101
Reefer Medness - The Podcast.
Music.
As you know, we usually ask our guys
for music and Ken asked
for to Tu umba.
And man, I don't know about you, but
(01:00:24):
I was rocking to it.
I thought it was really.
Driving down the road.
This is a great driving tune.
Its a great driving tune.
Correct.
All right. Talk to everyone later.
Come on back.
Popular Podcasts
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
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
Therapy Gecko
An unlicensed lizard psychologist travels the universe talking to strangers about absolutely nothing. TO CALL THE GECKO: follow me on https://www.twitch.tv/lyleforever to get a notification for when I am taking calls. I am usually live Mondays, Wednesdays, and Fridays but lately a lot of other times too. I am a gecko.