Episode Transcript
Transcripts are displayed as originally observed. Some content, including advertisements may have changed.
Use Ctrl + F to search
0:00
At Highland, we're all about celebrating
0:02
little wins and
0:04
little ways to innovate digital processes.
0:07
There's no customer pain point too small for
0:09
us to help with. Maybe
0:11
that's why more than half of
0:14
the Fortune 100 looks to Highland
0:16
to connect their content and data,
0:18
improve processes, and turn little efficiencies
0:21
into big wins for their customers
0:23
and clients. Highland
0:25
Intelligent Content Solutions for
0:27
innovators everywhere at highland.com
0:44
Hello, everybody. Today, I have the
0:46
opportunity to speak with Chris
0:49
Olsen, who's CEO of the
0:51
Media Trust Company. His
0:54
company is involved, occupies
0:57
the forefront of attempts to
1:00
make the online
1:02
world a safer place.
1:04
He mostly works with corporations to do
1:06
that, mostly to protect their digital assets.
1:08
But I was interested
1:11
in a more broad ranging conversation discussing
1:14
the dangers of online
1:16
criminality in general. A
1:19
substantial proportion of online
1:22
interaction is criminal. That's particularly true
1:24
if you include pornography within that
1:26
purview because porn itself constitutes
1:29
about 20 to 25 percent of Internet
1:31
traffic. But there's all sorts of criminal
1:33
activity as well. And so Chris and
1:36
I talked about, for example, the people
1:38
who are most vulnerable to criminal activity,
1:40
which includes elderly people
1:42
who are particularly susceptible to
1:45
romance scams initiated
1:48
on dating websites, but then
1:50
undertaken off those sites and
1:54
also to phishing scams
1:56
on their devices that indicate, for example, that
1:58
something's gone wrong. with the device and that
2:00
they need to be repaired in a manner
2:02
that also places them in the hands of
2:04
criminals. The sick and
2:06
infirm are often targeted with
2:08
false medical offers. 17-year-old
2:11
men are targeted with
2:13
offers for illicit drug
2:16
purchase and juvenile
2:18
girls, 14,
2:20
13 that age, who are interested in
2:22
modeling careers, for example, are frequently targeted
2:24
by human traffickers. This is a major
2:27
problem. The vast majority of elderly people
2:29
are targeted by criminals on a
2:31
regular basis. They're very well identified
2:33
demographically. They know their ages, they
2:36
know where they live. They
2:38
know a lot about their online usage
2:40
habits and they have
2:42
personal details of the sort that can
2:45
be gleaned as a consequence of continual
2:47
interaction with the online world. And so
2:50
I talked to Chris about all of
2:52
that and about how we
2:55
might conceptualize this as a society when
2:58
we're deciding to bring order to
3:00
what is really the borderless, the
3:03
ultimate borderless wild west
3:05
community. And that's the
3:07
hyper connected and
3:10
possibly increasingly pathological
3:12
online world. Join
3:15
us for that. Well,
3:17
hello, Mr. Olson. Thank you for agreeing to
3:19
do this. We met at the
3:21
presidential prayer breakfast not so long
3:23
ago and we had an engaging
3:25
conversation about the online
3:27
world and its perils. And I
3:29
thought it would be extremely interesting
3:31
for me and hopefully for everyone
3:33
else to engage
3:36
in a serious conversation about, well,
3:39
the spread of general criminality
3:41
and misbehavior online. And so
3:44
do you wanna maybe start by telling
3:46
people what you do and
3:48
then we'll delve more deeply into the general
3:50
problem? Great, yes. And thank you, Jordan.
3:52
Thanks for having me. I'm the
3:54
CEO and founder of the Media Trust
3:56
Company, not intended to be an oxymoron.
3:59
Our primary... My primary job is to help
4:01
big tech and digital media
4:03
companies not cause harm when
4:05
they monetize audiences and when they
4:08
target digital content. So
4:10
let's delve into the domains of
4:12
possible harm. So you're working with
4:14
large companies. Can you give
4:17
us who, like what sort of
4:19
companies do you work with? And then
4:21
maybe you could delineate for us the
4:24
potential domains of harm. Yeah,
4:27
so I work with companies that
4:29
own digital assets that people visit.
4:31
And I think maybe to set
4:33
a quick premise, cybersecurity
4:36
is a mature industry designed
4:38
to monetize the CISO, the
4:40
chief security officer, generally
4:43
protecting machines. So there's a
4:45
mindset geared to making sure
4:48
that the digital asset is
4:50
not harming servers, the
4:52
company or government data. Our
4:55
difference is that we're helping companies that
4:57
are in digital. So think big media
5:00
companies. We're helping them protect from harming
5:02
consumers, which is the difference between digital
5:04
crime, which is gonna target people, and
5:07
cybersecurity, which is generally targeting corporates and
5:09
governments and machines. So
5:12
now does your work involve protection
5:17
of the companies themselves also
5:19
against online criminal activity or
5:21
is it mostly aimed at
5:24
stopping the companies themselves from what would
5:27
you say, mostly I
5:29
suppose inadvertently harming their consumers
5:32
in pursuit of their enterprise
5:35
and their monetization? Yeah, so the
5:37
great question. And I think that's
5:39
where the heart of the matter
5:41
is. So our primary job is
5:43
to watch the makeup of what
5:45
targets digital citizens' devices. The
5:48
internet is made up of roughly 80% third-party code.
5:51
And what that means is when a consumer's
5:54
visiting a news website, when they're checking sports
5:56
scores, when they're visiting social media, the
5:59
predominance... of activity that's running
6:01
on their machine is coming from
6:03
companies that are not the owner of
6:05
the website or the mobile app that
6:08
they're visiting. That third-party
6:10
code is where this mystery
6:12
begins. So who actually controls
6:14
the impact on the consumer
6:17
when they're visiting an asset
6:19
that is mostly made up of source
6:21
code and content coming from other companies?
6:23
So our job is to look at
6:25
that third-party content to discern what is
6:28
good and bad based on company policies,
6:30
based on what might be harming the
6:32
consumer, and then informing those
6:34
companies what is violating and how they
6:36
can go about stopping that. What
6:39
sort of third-party code concerns
6:43
might they face or have they faced? What
6:45
are the specifics that you're looking for? Maybe
6:48
you could also provide us with some
6:50
of the more egregious examples of the
6:52
kinds of things that you're ferreting
6:55
out, identifying ferreting it out and
6:57
attempting to stop. Yeah,
6:59
so I think putting any digital
7:01
company into the
7:03
conversation is critical. So
7:06
we're talking about tech support
7:08
scams and romance scams targeting
7:10
seniors. That is an
7:12
epidemic. If you're a senior
7:14
and you're on the internet on a regular
7:16
basis, you're being attacked, if not daily, certainly
7:18
of every week. That
7:21
is now a cultural phenomenon. There's movies being
7:23
produced about the phenomenon of seniors
7:26
being targeted and attacked online. It's
7:29
teens. So a 17-year-old male is
7:31
being bombarded with information on how to buy
7:34
opioids or other drugs and having them shipped
7:36
to their house. If you're
7:38
a 14-year-old female and you're interested in
7:40
modeling, you're being approached by human traffickers.
7:43
The sick and infirm are frantically searching
7:45
the internet for cures. While
7:48
that's happening, they're having their life savings stolen.
7:51
So our job is to watch
7:53
that third-party content in code, which
7:55
is often advertising. It's basically
7:58
a real estate play on the internet. and what keeps
8:00
the consumer active on the digital asset to
8:04
find that problem and then give
8:06
it back to the company. I can
8:09
jump in quickly in how we go
8:11
about doing that. So we become a
8:13
synthetic persona. We've been doing
8:15
this for not quite two decades, but getting on
8:17
19 years. We
8:20
have physical machines in more than 120 countries.
8:23
We know how to look like a senior citizen, a
8:26
teenager, someone with an illness, and
8:28
then we're rendering digital assets as those
8:30
personas, acting more or less as a
8:33
honeypot to attract the problem that's
8:36
coming through the digital supply chain, which runs
8:38
on our devices. And I think that's gonna
8:40
be a key part of this conversation as
8:43
we go. Most of that action
8:45
is happening with us. And
8:48
so it's difficult for tech companies and media
8:50
companies to understand fully what's happening to us.
8:53
That's the point of their monetization, right? That
8:55
moment in time. So our job is to
8:57
detect these problems and then help
8:59
them make that go away. Right,
9:02
okay. So you
9:05
set yourself up as a replica
9:08
of the potential target of the
9:10
scams, and then you can deliver
9:12
the information that you gather about
9:14
how someone in that vulnerable position
9:17
might be interacting with the company's
9:19
services in question to keep the criminals
9:22
at bay. Let's go through these
9:25
different categories of vulnerability to
9:27
crime that you described. I
9:29
suspect there's stories of great
9:31
interest there. So you started
9:33
with scams directed at seniors.
9:35
So I've had people
9:37
in my own family targeted
9:39
by online scammers who
9:41
were in fact quite successful at
9:43
making away with a good proportion
9:45
of their life savings in one
9:47
case. And I know that seniors
9:51
in particular, who
9:53
grew up in an environment
9:55
of high trust, especially
9:57
with regards to corporate entities.
10:00
they're not particularly technologically
10:03
savvy, they're trusting.
10:05
And then you have the additional complication,
10:09
of course, in the case
10:11
of particularly elderly seniors, that their cognitive
10:14
faculties aren't necessarily all that they
10:16
once were. And
10:18
they're often lonely and isolated too. And
10:20
so that makes them very straightforward
10:23
targets for especially people, for people
10:25
who worm into their confidence.
10:29
You talked about, was it romance scams
10:31
on the senior side? It
10:33
is romance games on the senior side.
10:36
Okay, so lay all that out, tell
10:38
us some stories and describe to everybody
10:40
exactly what they would see and how
10:42
this operates. Okay, so a senior is
10:45
joining a dating website, just
10:47
as a teenager or someone
10:49
in middle age would do, they're looking for
10:51
romance. There are people on
10:53
the other side of that, that
10:55
are collecting data on that senior,
10:58
potentially interacting with them. Once
11:01
they get enough information on that particular
11:03
senior, they're gonna start to find them
11:05
in other ways. Send me emails
11:08
and information, let's move out the dating site.
11:10
They're gonna start calling them on the
11:13
phone. As that starts
11:15
to evolve, it's that information collection,
11:18
getting them to do certain things online that sucks
11:20
them deeper and deeper in. From
11:22
that moment forward, they become very much
11:24
wet and emotionally oriented towards that person
11:27
that they're involved with. And
11:30
the theft goes from there. Right,
11:32
so you go on a dating
11:35
website, as say someone in
11:37
your 70s, you're
11:40
lonely and looking for companionship.
11:44
There are scam artists
11:47
that are on those dating websites
11:49
as well, who must
11:51
have what? I
11:53
suspect they probably have keywords and profiling
11:55
information that enables them to zero in
11:58
on people who are. likely
12:01
targets. Do you know how sophisticated is
12:03
that? Like, do you think that the
12:06
criminals who are engaged in this activity,
12:08
how good is their ability to profile?
12:10
Do you think they can identify such
12:12
things as early signs of cognitive degeneration?
12:15
I think this is organized crime and
12:17
they have their own algorithms and processes
12:19
to identify people. I
12:22
also, to your earlier point, people
12:24
believe what they see on computers.
12:26
They're following what's being provided to them,
12:29
which makes them relatively easy marks. So
12:32
once that process starts, they're reeling them
12:34
in. If they lose a fish, that's
12:36
no problem because they're going after so
12:39
many in any given day. They
12:41
also have infrastructure in local
12:43
markets to go deal with
12:45
people personally. So this is a
12:48
very large criminal organization that has a lot
12:50
of horsepower to identify and then attack. Right,
12:54
okay, so do you have any sense?
12:57
See, I hadn't thought about the full implications
12:59
of that. So obviously,
13:02
if you were a psychopathic scam
13:04
artist, posing
13:06
as a false participant
13:10
on a dating website would be
13:12
potentially extremely fertile ground, not only
13:15
for seniors who could be scammed
13:17
out of their savings, but you
13:19
also mentioned, let's say,
13:22
younger people who are on
13:24
the website who might be useful
13:26
in terms of human
13:28
trafficking operations. So
13:31
do you have any sense, for example, of
13:33
the proportion of
13:36
participants on
13:39
a given dating platform that are
13:41
actually criminals or psychopaths in disguise? Because
13:43
here, let me give you an example.
13:45
You undoubtedly know about this, but there
13:47
was a website, can't
13:50
remember the name of it, unfortunately.
13:53
I believe it was Canadian that
13:55
was set up some years ago to
13:58
facilitate illicit affairs. And
14:01
they enrolled thousands of
14:03
people, all
14:06
of whose data was eventually leaked, much
14:09
of that to great scandal. The
14:11
notion was to match people
14:13
who were married secretly with other people
14:15
who were married to have illicit affairs.
14:18
They got an awful lot of men
14:20
on the website and almost no women.
14:22
And so they created tens
14:25
of thousands, if I remember correctly,
14:27
fake profiles of women to
14:29
continue to entice the men to maintain what
14:31
I believe was a monthly
14:35
fee for the service. Ashley
14:37
Madison, it was called. Right,
14:40
and so obviously... Starting
14:45
a business can be tough, but thanks to
14:47
Shopify, running your online storefront is easier than
14:49
ever. Shopify is the global commerce platform that
14:52
helps you sell at every stage of your
14:54
business. From the launch your online shop stage,
14:56
all the way to the, did we just
14:58
hit a million orders stage, Shopify's there to
15:01
help you grow. Our marketing team uses Shopify
15:03
every day to sell our merchandise, and we
15:05
love how easy it is to add more
15:07
items, ship products, and track conversions. Shopify helps
15:10
you turn browsers into buyers with the internet's
15:12
best converting checkout, up to 36% better compared
15:15
to other leading e-commerce platforms. No matter
15:17
how big you want to grow, Shopify
15:19
gives you everything you need to take
15:21
control and take your business to the
15:24
next level. Sign up for a $1
15:26
per month trial period at shopify.com/JBP. Go
15:28
to shopify.com/JBP now to grow
15:31
your business, no matter what
15:33
stage you're in. That's shopify.com/JBP.
15:39
Our dating website would
15:41
be wonderful hunting grounds for
15:43
any kind of predator. And
15:46
so do you have any sense of what proportion
15:48
of the people who are participating
15:51
on online dating sites are actually
15:53
predators, criminals? I
15:55
don't know what percentage of the
15:58
participants on the sites. are predators,
16:00
but where we come in
16:02
in our expertise is
16:04
that everyone that is visiting is
16:07
giving information into sort of the
16:09
digital ecosystem. And so
16:11
the issue from there is that
16:14
they're then able to be targeted wherever
16:16
they go online. So there's information that's
16:18
being collected from the site that they're
16:20
visiting. That is then moving out
16:22
into the ecosystem so that wherever they go,
16:24
they're being pulled back and targeted. In
16:27
an example, like in Ashley Madison, a
16:30
criminal may be able to get the
16:32
digital information about the people whose data
16:34
was stolen, come back to
16:36
them six months later, coming from
16:38
another website via email or SMS
16:40
text, and then press the attack
16:42
at that stage. For
16:44
us and becoming a digital persona, our
16:47
job is to look like someone
16:50
based on the information that sites
16:52
have collected about them. So we
16:54
look like an 85-year-old
16:57
grandmother living in a senior
16:59
community. When you become that type of
17:01
profile, no matter who else is engaging
17:03
with you online, the algorithm and the
17:05
content that is gonna be served to
17:07
you is coming from criminals,
17:10
regardless of their activity on that particular
17:12
site that you're visiting. It's simply based
17:14
on who you are. So
17:17
artificial intelligence has been around in
17:19
use for digital media
17:21
and targeting people for 2010-11. So
17:25
the initial, initial use case was
17:28
collecting data on us. That was
17:30
the key initial step for AI
17:32
utilization. The second step
17:34
was then turning that around and
17:37
targeting people better, right? So AI
17:39
was first used to collect information,
17:42
make things interesting behind
17:44
the scenes for people. Second, creating
17:46
better audience segments, which enable
17:49
that targeting. This
17:51
third phase that's happening today, you
17:53
see chat GPT and the LLMs
17:56
being used in regular use. The
17:58
third big stage. is writing content
18:01
on our devices on the fly. So
18:04
regardless of where the criminal actor
18:06
is, regardless of how they're moving
18:08
into the ecosystem and what initial
18:11
buying point, they're able to
18:13
find that person, write content on the
18:15
fly that's particularly tailored to what the
18:18
digital ecosystem knows about them to
18:20
create the situation where they then respond and
18:22
the criminal activity can occur. Right,
18:25
and so what that implies is well
18:27
then, I suppose is that we're
18:32
going to see very sophisticated LLM
18:35
criminals, right? Who
18:38
will be able to, this is the
18:40
logical conclusion of what you're laying out is
18:43
that they'll be
18:45
able to engage, huh, so I just
18:49
saw a video, it's
18:52
gone viral, it was released about three weeks
18:54
ago that portrayed the
18:57
newest version of chat GPT
19:01
and it's a version that can see you through
19:04
the video camera on
19:06
your phone and can
19:08
interact with you very much
19:11
like a person. So they
19:14
had this chat
19:16
GPT device interacting with
19:19
a kind of unkempt, nerdy
19:22
sort of engineer character who was
19:25
preparing for an interview, a job
19:28
interview and the chat GPT
19:30
system was coaching him on
19:33
his appearance and his presentation and
19:35
I think they used Scarlett
19:37
Johansson's voice for the
19:40
chat GPT bought. It
19:42
was very, very flirtatious,
19:44
very intelligent, extremely perceptive
19:48
and was paying attention to this engineer
19:51
who was preparing his
19:54
interview like
19:56
a, what would you say,
19:58
like the girlfriend of his dream, would if
20:00
he had someone who was paying more attention to
20:02
him than it was ever paid attention to him
20:05
in his life. And so
20:07
I can imagine a system like that set up
20:10
to be an optimal criminal, especially
20:12
if it was also
20:14
fed all sorts of information about
20:16
that person's wants and likes. So
20:19
let's delve into that a little bit. How
20:22
much of a digital footprint do
20:24
you suppose? How well are each
20:26
of us now replicated online
20:29
as a consequence of the criminal
20:32
or corporate aggregation of our online
20:35
behavior? So the
20:37
typical senior, for example, how much
20:39
information would be commonly available to
20:43
criminal types about, well, the typical
20:45
senior, the typical person, typical 14-year-old
20:47
for that matter? Right. The
20:50
majority of their prior activity
20:53
that they've engaged in online.
20:55
So corporate digital
20:57
data companies know a highly, their
20:59
job is to know as much
21:02
about us as possible and then
21:04
to target us with information to
21:06
maximize profit. That's the core goal.
21:10
Criminals have access to that data and
21:12
they're leveraging it just like a big
21:14
brand advertiser would. So
21:16
they know it's a grandmother and they're going to
21:18
put in something that only runs on the grandmother's
21:21
device, which makes it very, very difficult for big
21:23
tech and digital media companies to see the problem
21:26
before it occurs. I think
21:28
another thing that's really important to understand
21:30
is this is our most open border.
21:33
So we've got an idea of national
21:35
sovereignty. There's
21:38
lots of discussion on whether or not our
21:40
southern border is as secure as it should
21:42
be. Our actual devices,
21:44
our cell phones, our televisions, our
21:46
personal computers are open
21:49
to source code and information coming from
21:51
any country, any person at any
21:53
time, and typically resolved
21:55
to the highest bidder. Right.
21:58
Right. So the. Digital world
22:01
the virtual world is
22:03
a. Is
22:05
it a lawless frontier i mean i guess
22:07
one of the problems is like if i'm
22:10
targeted by a criminal gang in nigeria. What
22:13
the hell can i do about that mean that
22:15
the case i mentioned to you of my. Relative
22:18
who is scammed out of a
22:20
good proportion of their life
22:22
savings. That gang
22:25
was operating in eastern europe we
22:27
can more or less identify who they were but.
22:30
There is really nothing that could be
22:32
done about it. These
22:35
are people who are operating well out
22:37
of any physical proximity but also even out
22:39
of hypothetically the jurisdiction
22:41
of well say lawmakers
22:43
in canada police in police services
22:46
in canada and so. How
22:50
lawless how
22:52
is it how should we be conceptualizing the
22:54
status of law. In
22:58
the in the online
23:00
and virtual world. Yeah
23:02
i and i think this is where
23:04
where the major rub is so i'm
23:06
gonna walk back in and talk about
23:09
cyber security is an industry first so
23:11
cyber security is relatively mature. It
23:14
is now geared to monetizing the chief
23:16
security officer the chief information security
23:18
officer what that means
23:21
it's providing products and services designed.
23:24
To protect what they are paid to
23:26
hold dear which is the corporate
23:28
asset, so the machines and the data for
23:30
the corporation. If you
23:32
are part of the government, which is where we're going to
23:34
go in the conversation, then your job
23:36
as a cio or a czo is to
23:39
protect government machines. Governments will
23:41
tell you that they're protecting you right to
23:43
protect you from digital harm what
23:45
that means today is they're protecting your
23:47
data. On the dmv website
23:49
that that's basically the beginning and the end
23:52
of cyber security and digital
23:54
protection. There's a legislation
23:56
which is occurring coming from attorneys
23:58
general from from state. from
24:01
the federal government in the US to a degree,
24:04
other countries seem to be further ahead, seeking
24:07
to protect people from data collection.
24:09
And that's your GDPR in
24:12
Europe. Many states in the
24:14
United States are putting
24:16
some rules in place around what corporations can
24:18
collect, what they can do with the data. The
24:21
predominant use case is to provide a consumer
24:23
with an opt-out mechanism. Most
24:25
consumers say, okay, I wanna read the content,
24:28
they're not doing a whole lot with the
24:30
opt-out compliance. So that's not
24:32
been a big help
24:34
to your typical consumer, but
24:36
it's really the mindset that's the problem
24:38
and the mindset of corporate and government
24:40
that is at issue. And
24:42
so governments need to tactically engage
24:46
on a 24 seven basis with
24:48
digital crime in the same way that
24:50
they're policing the street. So
24:52
the metaphor would look like this. If
24:55
grandmothers were walking down the street and being
24:57
mugged or attacked at the rate that
24:59
they're getting hit online, you would
25:02
have the National Guard policing every
25:04
street in America. The
25:07
government needs to take step forward. And when I
25:09
say the governments, that is governments need
25:11
to take a step forward and do
25:13
a better job at policing people tactically.
25:16
And that does not mean that they're going after
25:18
big tech or digital media companies. It
25:21
means that they're protecting people with the
25:23
mindset that they're gonna
25:25
go ahead and cooperate with the
25:27
digital ecosystem to do a better
25:29
job, to reduce overall crime. Right,
25:32
so your point appears to be
25:34
that we have
25:37
mechanisms in place, like the ones that are
25:39
offered by your company that
25:43
protect the
25:45
corporations against the liability that
25:48
they would be laden with
25:51
if the data on their servers
25:53
was compromised. But that is by no
25:56
means the same thing as
25:58
having a police force. that's
26:00
accessible to people, individual people who are
26:02
actually the victims of criminal activity. Those
26:04
aren't the same things at all. It's
26:06
like armed guards at
26:09
a safe in a bank compared to police on
26:11
the street that are there
26:13
to protect ordinary people or who can be called. Have
26:16
I got that about right? Yes,
26:18
and digital crime is crime. So
26:20
this is when you're stealing grandmother's money,
26:24
that is theft. We don't
26:26
need a lot of new laws. What we
26:28
need to do is actively engage with the
26:30
digital ecosystem to try to get
26:32
in front of the problem to reduce overall
26:35
numbers of attacks, which reduces the number
26:38
of victims. And to date, when
26:40
we think about digital safety, it's
26:43
predominantly education, and then
26:45
increasing support for victims. Victims
26:48
are post-attack. They've already had their
26:50
money stolen. Getting in front
26:52
of that is the key. We've got to start
26:54
to reduce digital harm. I've
26:56
been doing this for a good number of
26:58
years, and the end of that conversation does
27:00
reside with local and state
27:03
governments. And ultimately, the federal government
27:05
in the United States is gonna
27:07
have to find resources to actively
27:09
protect beyond having discussions about legislating
27:11
data control or social media as
27:14
a problem. Okay, so I'm trying to
27:16
wrestle with how this is possible, even
27:19
in principle. So now you
27:21
said that, for example, what your company does
27:23
is, and we'll get back into that, is
27:25
produce virtual victims, in a
27:28
sense, false virtual victims, so that you can
27:30
attract the criminals, so that you can see
27:32
what they're doing. So I
27:34
presume that you can report on what you find
27:36
to the companies so that they can decrease the
27:40
susceptibility they have to exploitation by
27:42
these bad actors. But that's
27:45
not the same thing as actually tracking
27:47
down the criminals and holding them responsible
27:49
for their predatory activity. And
27:52
I'm curious about
27:54
what you think about how that's possible, even
27:56
in principle, is first of all, these criminals
27:58
tend to be, or can't even, easily be
28:01
acting at a great distance in jurisdictions
28:04
where they're not likely to be held
28:07
accountable in any case, even by
28:09
the authorities, or maybe they're even
28:11
the authorities themselves. But also, as
28:14
you pointed out, more and more, it's
28:17
possible for the criminal activity
28:19
to be occurring
28:21
on the local machine. And
28:23
so that makes it
28:26
even more undetectable. So I don't
28:29
see, I can't
28:31
understand easily, you obviously in a
28:33
much better position to comment on this, how
28:36
even in principle, there
28:38
can be such a thing as, let's say
28:40
an effective digital police force. Like even if
28:42
you find the activity
28:45
that someone's engaged in and
28:47
you can bring that to a halt by
28:49
changing the way the data is handled,
28:52
that doesn't mean you've identified the criminals
28:54
or held them accountable. So what,
28:58
if anything, I can't understand how that
29:00
can proceed even in principle. Sleep
29:04
is a foundation for our mental and physical
29:06
health. In other words, you've got to have
29:08
a consistent nighttime routine to function at your
29:10
best. But if you're struggling with sleep, then
29:12
you've got to check out Beam. Beam isn't
29:14
your run of the mill sleep aid. It's
29:17
a concoction carefully crafted to help you rest
29:19
without the grogginess that often accompanies other sleep
29:21
remedies. A bunch of us here at The
29:23
Daily Wire count on Beam's dream powder to
29:25
knock us out and sleep better through the
29:27
night so we can show up ready for
29:30
work the next day. Just mix Beam Dream
29:32
into hot water or milk, stir or froth,
29:34
and then enjoy before bedtime. Then wake up
29:36
feeling refreshed without the next day grogginess caused
29:38
by other sleep products. Dream contains a powerful
29:41
all-natural blend of Reishi, magnesium, Eltheanine, Apigenin, and
29:43
Melatonin to help you fall asleep, stay asleep,
29:45
and wake up refreshed. And with it now
29:47
being available in delicious flavors like cinnamon cocoa,
29:49
chocolate peanut butter, and mint chip, Better Sleep
29:51
has never tasted better. And today, listeners of
29:54
this show get a special discount on Beam's
29:56
dream powder. Get up to 40% off for
29:58
a limited time when you go to shopbeam.com/Peterson
30:00
and use code Peterson. at checkout. That's shop
30:02
at beam.com. And use code Peterson for up
30:04
to 40% off. So
30:10
the digital ecosystem is made up of
30:12
a supply chain, just like every other
30:14
industry. There are various steps that a
30:16
piece of content is gonna go through
30:19
before it winds up on your phone. So
30:21
it's running through a number of different companies,
30:24
different cloud solutions, different servers
30:27
that put content out. Okay,
30:29
they're intermediaries. And
30:31
so a relationship between those digital
30:33
police with the governments and those
30:35
entities on a tactical basis is
30:38
really the first step. Seeing
30:40
crime and then reporting that back
30:42
up the chain so that
30:44
it can be stopped higher and higher up
30:47
towards ultimately the initiation point
30:49
of where that content is delivered. So
30:52
it seems fantastic, but it is
30:55
possible. Well, the criminals need
30:57
to have, they need
30:59
to use intermediary processes in
31:01
order to get access to
31:03
the local devices. And so
31:05
you're saying that I believe
31:07
that those
31:09
intermediary agencies
31:11
could be enticed, forced,
31:14
compelled, invited to make
31:17
it much more difficult for the criminals to
31:19
utilize their services. I guess that's, and that
31:21
that might actually be effective. That does
31:23
that, that still doesn't, does
31:26
that aid in the identification of the actual criminals
31:28
themselves? Because I mean, that's the advantage of
31:30
the justice system, right, is you actually get
31:32
your hands on the criminal at
31:35
some point. Yes, and I think
31:37
ultimately it does. So
31:40
you have to start and you have to start
31:42
to build the information about where it's coming from.
31:45
You then have to cooperate with the private entities.
31:48
Our digital streets are managed and made up
31:50
of private companies. It's not a government run
31:53
internet. All of the information that's said to
31:55
us, at least in Western society, is
31:57
coming from these private companies. So
32:00
I think rather than having an antagonistic
32:02
relationship between governments and private companies where
32:04
they're trying to legislate to put them
32:06
into a position, that may be appropriate
32:09
for certain rules and regulations. It may
32:11
be appropriate to raise the age of
32:13
accessing social media from 13 to 16 or 18. And
32:18
that is a proper place for
32:20
the government to be legislating. On the
32:22
other hand, an eye towards
32:24
reducing crime is critical. And
32:27
the ethical and moral mindset among
32:30
all of the parties, and that's
32:32
governments through our corporations, has to
32:34
be solely on protecting people. And
32:37
I think that's something that is significantly
32:39
missing. It's missing in the
32:41
legislation. It's missing in cybersecurity.
32:44
It's not something that we've engaged
32:46
in as a society. So
32:48
there are a few countries, and
32:50
I think even a few states in the US,
32:53
that are looking at a broader whole of
32:56
society approach. That whole
32:58
of society approach is a mimicking
33:00
of how the internet and the digital
33:02
ecosystem works, which is certainly a whole
33:05
of society activity, right? So
33:07
it is the thing that influences and affects
33:09
all of us every single moment
33:11
of every single day. Engaging in
33:13
that, looking across the impact of society
33:15
and doing better via cooperation
33:18
is a critical, critical next step.
33:21
How often do you think the typical
33:23
elderly person in the United States
33:26
say, is being
33:29
successfully, no, is being
33:31
first communicated with by
33:34
criminal agents, and then how often
33:37
successfully communicated with?
33:41
What's the scope of the problem? The
33:43
scope is, if you're a senior citizen, in
33:47
particular, if you're a female senior citizen, roughly 78
33:49
to about 85 years old, we
33:52
see that two and a half to 3% of
33:55
every single page impression or app
33:57
view is attempting to,
34:00
to target you with some form of crime or
34:03
influence that's gonna move you towards crime. So
34:06
it is highly, highly
34:08
significant. In
34:11
some ways looking at, this is shooting fish in a
34:13
barrel to make a dent. So
34:15
you're concerned that the legal system isn't gonna be
34:17
able to find the criminals. There
34:20
is so much to detect and stop and
34:23
so much room to turn them off
34:25
quickly, right? That we
34:27
can gain a significant reduction in
34:29
digital crime by working together
34:31
and considering society as a whole instead of
34:33
the different pockets and how can we legislate
34:35
or how can we try to move a
34:37
private company to do better on their own?
34:41
Okay, so let's now let's, okay. So we talked
34:43
a little bit about the danger that's
34:45
posed by one form of con game
34:50
in relationship to potential criminal
34:53
victims and that was senior romance
34:55
scams. What
34:57
are the other primary dangers that are posed to
34:59
seniors? And then let's go through your list. You
35:01
talked about 17 year olds who
35:04
are being sold online access to drugs.
35:06
That includes now by the way, a
35:08
burgeoning market in under
35:11
the table hormonal treatments
35:16
for kids who've
35:19
had induced gender dysphoria. So
35:21
you talked about seniors, 17 year olds who
35:24
are being marketed illicit drugs, 14
35:27
year olds who are being enticed into let's
35:29
say modeling and people who are sick and
35:31
infirm. So those are four major categories. Let's
35:34
start with the seniors again, apart from romance
35:36
scams, what are the most common forms of
35:38
criminal incursion that you see? The
35:41
most common form is the tech
35:43
support or upgrade scam. And
35:46
essentially the internet knows that
35:48
you are senior. When you're
35:50
going to a website that you and I would visit, instead
35:53
of having a nice relationship with that site and
35:55
reading the content and then moving on to something
35:57
else, you're getting a pop-up or some.
36:00
of information that's telling you there's something wrong
36:02
with your computer. You either need to call
36:04
a phone number or you need to click
36:07
a button, which then moves you down to
36:09
something else that is
36:11
more significant. This
36:13
is happening millions and millions and
36:15
millions of times per day. And
36:18
it is something that we can all
36:20
do something about. Attempting to educate
36:22
seniors to try to not listen to the
36:25
computer when it's telling you to do something
36:27
is not working. So that-
36:29
Well, no wonder. I mean, look, to
36:32
manage that, it's so
36:34
sophisticated because once
36:36
you've worked with computers for 20 years, especially if
36:38
you grew up with them, you
36:40
know when your computer, your
36:43
phone, is telling you something that's actually
36:45
valid and when it isn't, it doesn't
36:47
even look this, a lot of these
36:49
criminal notifications, they don't even look right.
36:51
They look kind of amateurish. They
36:53
don't have the same aesthetic that you'd
36:56
expect if it was a genuine communication
36:58
from your phone but
37:00
man, you have to know the ecosystem to be able
37:03
to distinguish that kind of
37:05
message from the typical thing
37:08
your phone or any website might ask
37:10
you to do. And educating seniors, it's
37:12
not just a matter of describing to
37:14
them that this might happen. They
37:17
would have to be tech-savvy cell
37:20
phone users and it's hard enough
37:22
to do that if you're young, much
37:24
less if you're outside that whole
37:27
technological revolution. So I can't see
37:29
the educational approach. The
37:32
criminal is just gonna outrun that as fast as
37:34
it happens. So yeah, so that's pretty,
37:36
so 3%, hey, that's a lot.
37:40
That's about what you'd expect. Yeah,
37:42
it is highly significant. And
37:45
I think getting in front
37:47
of this problem requires cooperation
37:49
with states, moving
37:51
that tactically to have the idea of a
37:53
police force looking at digital. And I think
37:56
one of the things that both sides, whether
37:58
it's private companies or states, It's
42:00
the most advanced AI at your
42:02
fingertips. Expand your world
42:04
with meta AI. Now
42:07
on Instagram, WhatsApp, Facebook, and
42:09
Messenger. Yeah,
42:13
for us, there
42:16
is a path that leverages that content
42:18
to bring it to the device. And
42:20
I think understanding that mechanism and how
42:22
it's brought forward versus looking at the
42:25
content, and I'll give you an example
42:27
of what's happening in political advertising as
42:29
we speak, understanding
42:32
the pathway for how that content is
42:34
delivered is ultimately how we get back
42:36
to the criminal or the entity that's
42:39
using that to perpetrate
42:41
the crime. The actual creation
42:43
of the content is incredibly difficult to
42:45
stop. It's when it moves out to
42:47
our devices that it becomes something that
42:50
we need to be really paying attention to. So
42:53
in political advertising up to October
42:56
of this past year, our
42:58
customers asked us to flag the
43:00
presence of AI source code. So
43:03
the idea there was they didn't want to
43:05
be caught holding the bag of being caught
43:07
being the server of AI generated
43:10
political content, right?
43:12
Because that just, it looks bad in the news.
43:14
Someone's letting someone use AI. It's going to wind
43:17
up being disinformation or some form of deep fake.
43:20
By October, we essentially stopped
43:22
using that policy because we
43:24
had achieved greater than 50% of
43:28
the content that we were scanning had some
43:30
form of AI. It may have been to
43:32
make the sun a little more yellow, the
43:34
ocean a little bit more blue, but
43:36
using that as a flag, right? To
43:39
understand what's being delivered out, once
43:41
you get over 50%, you're looking at more than
43:45
you're not looking at. That's not a good
43:47
automated method to execute on digital safety. So
43:51
as we move forward, we have a
43:54
reasonably sophisticated model to detect
43:56
deep fakes very much
43:58
still in a test mode. but it's
44:00
starting to pay some dividends.
44:03
And unquestionably what we see is
44:06
using the idea of deepfakes to create
44:08
fear is significantly greater
44:10
than the use of deepfakes. Now
44:13
that's limited to a political advertising
44:15
conversation. We're not seeing a lot
44:17
of deepfake serving in
44:19
information or certainly not in the paid
44:21
content side, but the
44:24
idea of fearing what's being delivered
44:26
to the consumer is
44:28
very much becoming part of a mainstream
44:30
conversation. Yeah,
44:33
well, wasn't there some
44:35
insistence from the White House
44:37
itself in the last couple of weeks that
44:39
some of the claims
44:42
that the Republicans were making with
44:44
regards to Biden were a
44:47
consequence of deepfake audio, not
44:49
video, I don't think, but audio? If
44:51
I got that right, does that story ring a
44:53
bell? And I think where
44:56
we are at this stage in technology
44:58
is very likely there is plenty of
45:00
deepfake audio happening around the candidates. So
45:02
whether you're Donald Trump or Joe Biden,
45:04
or even local political
45:07
campaigns, it's really that straightforward.
45:10
I think on the video side, there are gonna be people
45:13
working on it left and right. I
45:15
think it's the idea of using that as a weapon
45:18
to sow some form of
45:20
confusion among the populace. Some doubt, right?
45:22
Some doubt is gonna be dramatically more
45:25
valuable than the actual utilization of deepfakes
45:27
to move society. Oh, that's, you do,
45:29
eh? So you do think
45:31
that even if the technology develops to the
45:33
point where it's easy to use, so
45:36
you think that it'll be weaponization
45:38
of the doubt that's sowed by
45:40
the fact that such things exist.
45:43
And we've been watching this
45:45
for a very, very long time, and our
45:47
perspective is coming at this from a digital
45:49
crime and a safety
45:52
in content. Safety in content typically
45:54
means don't run adult content in
45:56
front of children, don't serve weapons
45:59
in New York. they're not gonna
46:01
like that. Don't have a
46:03
couple walking down the beach in Saudi
46:05
Arabia. Right, their ministry of media is
46:07
gonna be very unhappy with the digital
46:09
company that's bringing that kind
46:11
of content in. I have the beholder save
46:14
content, drugs and alcohol, right? Targeting the
46:17
wrong kinds of people. So we
46:19
look at this from a lens
46:21
of how do you find and remove
46:24
things from the ecosystem? If
46:26
we continue down the path that we're on today,
46:28
most people won't trust what
46:31
they see. And so we're discussing education.
46:33
They're gonna self evolve to a point
46:35
where so much of the information that's
46:37
being fed to them is just gonna
46:39
be disbelieved because it's gonna be safer
46:41
to not go
46:43
down that path. I'm
46:46
wondering if live events, for
46:48
example, are going to become
46:50
once again, extremely
46:54
compelling and popular because they'll be the only events
46:56
that you'll actually be able to trust. I
47:00
think so. Frank, cause I mean, you're- I
47:02
think it's also critical that we find a way
47:05
to get a handle
47:07
on kind of the anti-news and get
47:09
back. The entities
47:11
promoting trust in journalism, that
47:16
is a very meaningful conversation and it is something that
47:18
we need to try to get back to. It's
47:20
much less expensive to have automation
47:22
or create something that's gonna create
47:25
some kind of situation where people
47:27
continue to click. That's
47:29
a terrible relationship with the digital ecosystem.
47:31
It's not good for people to have
47:33
that in their hand. And
47:36
with the place where digital crime
47:38
is today, if you're a senior
47:40
citizen, your relationship is often net
47:42
negative with the internet.
47:45
Right, you may wanna stick to calling your kids
47:48
on voiceover IP where you can see their
47:50
face. Lots of different ways to do that
47:52
in video calling, but doing other
47:54
things on the internet, including things as simple
47:56
as email, it
47:59
may be more dangerous. about
1:16:00
the machine. How are
1:16:02
you feeling about your chances of control
1:16:06
over or our
1:16:08
chances for that matter of
1:16:10
control over online criminality and
1:16:12
how successful do you believe
1:16:15
you are in your attempts
1:16:17
to stay on
1:16:20
top of and ahead of
1:16:22
the criminal activity that you're
1:16:24
trying to fight? For
1:16:26
our customers that prioritize digital safety,
1:16:31
the vast majority of what might run
1:16:33
through to attack someone is
1:16:35
being detected and removed. They
1:16:38
need to have the appropriate mindset. They need to
1:16:40
be willing to go up onto the
1:16:42
demand source to remove bad
1:16:45
activity that's gonna be coming down. You don't
1:16:47
just wanna play whack-a-ball. You have to engage
1:16:49
in that next step. Those
1:16:52
that do are very successful and
1:16:55
create safe environments. It is not possible to
1:16:57
make this go away. The
1:16:59
pipes, the way that the internet works, the way the
1:17:02
data targeting works, it's just not something you can eliminate
1:17:04
entirely. But there are companies that
1:17:06
are in front of this that will
1:17:08
withhold millions of dollars in
1:17:10
revenue at any given moment to
1:17:12
prevent the possibility of
1:17:14
targeting something and having something bad
1:17:17
happen. But there are a
1:17:19
lot of companies that are not willing to go that
1:17:21
far. I think right now in
1:17:24
some of the bigger companies, we see a lot
1:17:27
of risk towards this, who's gonna
1:17:29
win the chat GPT, who's
1:17:31
gonna win the LLM race. There
1:17:34
is so much at stake in
1:17:36
that from a competitive and revenue
1:17:38
perspective. The companies
1:17:40
that can monetize that the best
1:17:44
are going to start to leap forward. When
1:17:46
you're looking at the world from a, how does
1:17:48
my technology win versus how do I safely get
1:17:50
my technology to do the things that I want?
1:17:52
That's when you start to run a lot of
1:17:55
risk. We're in a risk
1:17:57
on phase and digital right now. But
1:18:00
your earlier claim, I think, which is
1:18:02
worth returning to, was that over
1:18:06
any reasonable period of time, there's
1:18:08
the rub, the companies
1:18:10
that do what's necessary to
1:18:13
ensure the trust
1:18:15
of what you say, to
1:18:20
ensure that their users can trust
1:18:22
the interactions with them are going
1:18:24
to be the ones that are
1:18:27
arguably best positioned to maintain
1:18:30
their economic advantage in the years to come.
1:18:33
And I think- That's the problem. Yes,
1:18:36
and those that are willing to
1:18:38
engage with governments to do
1:18:40
a better job, to ultimately find the bad
1:18:43
actors and take them down, they're
1:18:45
going to be a big part of making the ecosystem
1:18:47
better, rather than insulating and hiding
1:18:50
behind a sort of risk legal regime
1:18:53
that's going to not want to bring data forward to clean
1:18:55
up the ecosystem. Okay, okay,
1:18:57
okay. Well, for anybody watching
1:18:59
and listening, I'm
1:19:01
going to continue my discussion
1:19:03
with Chris Olsen on the Daily Wire
1:19:05
side of the interview. I'm
1:19:09
going to find out more about, well,
1:19:11
how he built
1:19:13
his company and how his interest in
1:19:16
prevention, understanding, and preventing online crime
1:19:19
developed, and also what
1:19:21
his plans for the future are. And
1:19:23
so if those of you who are watching and listening are
1:19:26
inclined to join us on the Daily Wire side, that
1:19:28
would be much appreciated. Thank you
1:19:30
to everybody who is watching and listening for
1:19:32
your time and attention. And thank you very
1:19:34
much, Mr. Olsen, for, well, fleshing
1:19:37
out our understanding of the
1:19:40
perils and possibilities that
1:19:42
await us as the internet rolls
1:19:44
forward at an ever increasing rate.
1:19:46
And also for, I would say,
1:19:48
alerting everybody who's watching and listening
1:19:51
to the, what
1:19:53
would you say, the particular points of
1:19:55
access that the online criminals
1:19:57
have at the moment. when
1:20:01
we're in our most vulnerable states,
1:20:04
sick, young, seeking,
1:20:08
old, all of those things, because
1:20:10
we all have people, we all know people
1:20:13
who are in those categories and are looking
1:20:16
for ways to protect them against the people that
1:20:18
you're also trying to protect us from. So thank
1:20:20
you very much for that. Thank
1:20:22
you, thanks for having us, me. You
1:20:25
bet, you bet. And again, thanks
1:20:27
to everybody who's watching, listening to the film
1:20:29
crew down here in Chile
1:20:31
today in San Diego, thank you very much
1:20:34
for your help today, guys, and to the
1:20:36
Daily Wire people for making this conversation possible.
1:20:38
That's much appreciated. Thanks very much,
1:20:40
Mr. Olsen. Good to talk to
1:20:42
you. Thank you. When
1:20:45
you find a deal on your
1:20:47
favorite thing in the McDonald's app
1:20:49
and order it, does that technically
1:20:51
count as online shopping? Save
1:20:54
money with the app.
1:20:56
Ba-da-ba-ba-ba. At participating McDonald's,
1:20:58
prices may vary.
Podchaser is the ultimate destination for podcast data, search, and discovery. Learn More