0:04

It's predicted that this year we'll see artificial

0:06

intelligence transform industries

0:09

and redefine human interaction

0:11

with machines , but it's also facing

0:13

challenges to meet ethical commitments . I'm

0:15

Mike Landis To further discuss the benefits

0:18

and potential downside of AI . Ut

0:20

Tyler Radio connects with University Assistant

0:22

Professor of Communication , dr Eric

0:24

Gustafson . Endless possibilities with

0:26

AI and huge ethical challenges

0:28

am I right ?

0:29

Absolutely , and thanks for having me on , mike . I think

0:32

when we get to the conversation

0:34

of ethics , oftentimes

0:37

we've already had those changes

0:39

happen right , and

0:41

the questions surrounding ethics , I

0:43

think , really are at a flashpoint

0:46

, on college campuses particularly and

0:48

then sort of leaking out into other

0:50

industries as well , where we first

0:53

need to figure out what even are those challenges

0:55

. What are the questions that we need to ask

0:59

People in

1:01

a number of different industries ? Very brilliant people

1:03

all around have isolated

1:06

some of those questions . We still have them tinted

1:08

with the different emotions that come from something so

1:10

new and we've

1:13

yet to completely remove ourselves from

1:15

the equation or at least do the best we can

1:17

to assess these things on their own merits

1:20

as opposed to on

1:22

our own fears

1:25

or excitements about them .

1:32

Perhaps the flashpoint of this international conversation came with the demonstrations of what ChatGPT

1:34

was capable of Tell us more about the upside and downside

1:36

of this free-to-use AI system

1:38

.

1:39

Well , one ChatGPT if you've

1:41

ever peeked into it is immensely helpful

1:43

. It's a great tool and a great

1:45

aid for helping to find

1:47

information , to conduct research

1:50

, to learn . It's

1:52

a great tool , but

1:55

it also poses a

1:57

lot of different issues because it replaces those skills that we used to

1:59

have to do on our own . Because it replaces those skills

2:01

that we used to have to do on our own , those

2:03

research skills , those identifying what

2:06

is credible and

2:08

what's not , and while

2:10

with every new technological

2:12

development we often we get a

2:14

new way of knowing or a new way of coming

2:17

to know or creating knowledge , which

2:19

means we lose a pathway . But

2:24

I guess we are in the stage of figuring out is that new way that ChatGPT

2:26

proposes or allows us

2:28

to sort of come to knowledge and create

2:31

knowledge ? Is

2:33

that good , bad ugly somewhere

2:35

in between ? What do we do with it ? Sounds

2:38

like it can be all of those things . Yeah

2:41

, I had a great colleague who once said

2:44

yes , but . Or

2:47

in communication studies we always say it

2:49

depends , it's contingent on context

2:51

. All the time it depends .

2:53

In a day and age in which we are bombarded

2:56

daily with all kinds of information that

2:58

may or may not have been professionally vetted , how

3:00

will we know for sure , going forward , what

3:03

is true and what isn't ? How will

3:05

?

3:05

we know for sure , going forward , what is true and what isn't

3:07

? I'm not sure we will , especially with the upcoming election

3:09

and

3:28

we think about the different sort of campaign messages and that can exacerbate those

3:30

difficulties of identifying what's true and what isn't . I'm not sure we will for

3:32

a little bit . Right now we have different fact checkers or AI checkers

3:35

that have developed rapidly in

3:37

concert with these technologies , but we also

3:39

have programs

3:41

that actually strip AI

3:43

generated content of the markers that

3:45

would be caught by the detectors . So

3:48

in Concert we have all these technologies

3:50

and software specifically

3:52

sort of running together and they're

3:55

leapfrogging well ahead of our

3:57

questions right now .

3:59

I'm thinking of the 1950s

4:01

and 60s sci-fi

4:04

movies in which machines would take over

4:06

the world . They'd start talking to each other and

4:08

then they didn't need humans anymore . I

4:10

mean , is that even technically possible

4:12

?

4:14

I think some of the most interesting

4:17

predictions

4:19

or sort

4:22

of explorations into what the future will

4:24

be seem so far-fetched that

4:26

we can't

4:28

fully grasp them . They seem so fantastic

4:31

, so far off , and yet

4:33

if you transported someone from

4:35

the 50s or 60s to

4:37

here , they'd probably go . What is all of this

4:39

? And I think we will have the same experience

4:42

realization in another 50

4:44

and another 60 . So I

4:47

think all of those at the time of their writing

4:49

seem so far-fetched . And then , all

4:52

of a sudden , in the 50s and 60s we

4:54

were already . You

4:56

know , if we talk about the Turing machine and think

4:58

about sort of this birth of computing

5:01

, this birth of sort of the first artificial

5:04

intelligence , if you will , we already

5:06

had it then , and then it's finally exploding

5:08

now , 70

5:10

years later . I think we'll

5:13

see that with other hosts

5:15

of technologies that are running alongside this

5:18

.

5:19

It's said that security and privacy

5:21

are essential requirements of developing and

5:23

deploying AI systems , but that's

5:25

also the biggest problem facing AI . It

5:28

would feel a little bit like the foxes

5:30

are essentially guarding the hen house at

5:32

this point in time . Is that an overestimation

5:34

of it ?

5:36

I don't think it's an overestimation . I think it

5:38

may be a way of characterizing

5:40

exactly what's always been the case , which

5:42

is , when we look at technological developments

5:44

, especially those that push the frontiers

5:47

of our understanding of the world , we

5:49

often see war

5:52

and security at the forefront

5:54

. One of the technologies

5:57

running alongside artificial intelligence

5:59

that is said to going

6:01

to supercharge it in the next decade or

6:03

two is quantum computation

6:05

, which represents a fundamentally different way of

6:08

computing from classical computation

6:10

. But our original impetus

6:12

for developing quantum

6:14

theory and quantum mechanics

6:16

was to create the atom bomb . It

6:19

also helped us create MRI machines , helped

6:21

us create the transistor , which

6:23

was the fundamental unit for

6:26

classical systems , which

6:28

was the fundamental unit for classical systems

6:30

. But

6:35

oftentimes we do see the foxes

6:37

are the ones funding the research that makes these things possible

6:40

and oftentimes pushing those boundaries .

6:43

And we get to find out about it later . So

6:45

we're just months away .

6:59

We're just months away from what promises to be a very contentious presidential election . You

7:01

mentioned this a moment ago . In theory , a voter in Smith County could get a phone call

7:03

with Joe Biden's voice asking them to vote for Donald Trump . I may sound ridiculous , but technically

7:05

that's possible . Right last semester , I had a student create an AI generated podcast and

7:07

he used the voices of Joe Biden and Donald Trump

7:10

for the two podcast members and

7:12

you would be shocked

7:14

at how good this sounded . It's

7:17

absolutely a possibility . It's more than a possibility

7:19

. It's most likely a probability . And

7:23

to the extent that

7:25

to your earlier question , can we tell

7:27

the difference ? Sometimes not

7:29

. Sometimes , yes , it'll

7:32

develop along with that human touch of the people

7:34

in these campaigns saying , oh

7:37

, if we tweak that , this will sound better and

7:39

maybe no one will know . It's

7:43

really hard to tell , but it'd be very interesting . Yeah

7:45

exactly .

7:46

Deep fakes with video are a little scary

7:48

. I see some of them being used for comedy bits . You

7:50

see those on the Internet . They involve those

7:53

two men , the president and

7:55

Donald Trump . But someone with enough money and

7:57

a determined enough agenda could certainly get

7:59

a lot of myths and disinformation

8:02

out there .

8:03

Absolutely . I think with prior

8:05

elections we've seen that

8:07

explosion of this idea of fake

8:09

news and to our

8:12

college students , to younger individuals . It's second nature

8:14

to them now to question these things , whereas

8:16

some of us who are older had a

8:18

touch more faith in them . With

8:27

video and audio , what we're seeing is fake news sort of pushed to

8:29

its extreme , and it might reverse to its opposite , to use sort

8:31

of a phrase from Marshall McLuhan , a

8:33

media scholar from the 20th century . If

8:36

we take this so far , we push it to its opposite

8:39

. Instead of simulating these voices and

8:42

crafting a message that creates

8:44

trust , we put

8:46

all these messages out there and none of them

8:49

create trust , and so accomplishing the opposite

8:51

of what we thought .

8:53

There's been a call for international collaboration

8:56

and ethical standards to take place . How

8:58

quickly could safeguards be established

9:00

and put into place Should that ? Should that happen

9:02

?

9:03

for the election . I'm unsure for um

9:05

, the legislative

9:07

system in general . There have been talks that have started

9:09

. The European Union just pushed an

9:12

act through that levies significant

9:14

penalties on AI

9:17

, transparency and sort of attempts

9:20

to safeguard these things , but

9:34

legislation always travels slower than technological development . So whether or not

9:36

the right safeguards were put in place prior

9:38

to the election will

9:41

be something we'll see soon .

9:43

Just over 30 years ago , the World Wide

9:46

Web went into the public domain , a decision

9:48

that fundamentally altered the entire

9:50

past quarter century . Are we

9:52

fretting over the unknowns about AI

9:54

in the same way that some did about the Internet

9:57

years ago , or are the concerns about AI

9:59

more substantial ?

10:01

I think if you look to any technological

10:05

development , you're going to find

10:07

anxieties with it . If you go back

10:09

to the ancient Greeks , socrates

10:11

bemoaned literacy because it removed

10:13

knowledge out of the human mind . And how do we

10:15

know if someone's smart if they can't remember it ? Try

10:17

to tell a college student that today . And

10:20

the Internet is a great example too , because we

10:23

find the roots of it first being developed

10:25

by the military to share documents in

10:27

the late 1960s , and

10:29

then in 1983

10:32

, they switched a protocol which was what we consider

10:34

sort of the birth of the internet . And then , probably

10:37

20 years later , we

10:39

see it 20 , 30 years

10:41

. We then see it as integrated into

10:43

every single facet of

10:45

our lives , right ? So I don't think

10:47

the concerns

10:49

about AI are unwarranted

10:53

. I think we're just now

10:55

realizing that this has been a long time

10:57

coming and oftentimes we just

10:59

need to . If only we

11:01

could learn about these things before they exploded

11:04

onto the scene , I guess .

11:06

Any final thoughts you'd like to share about AI ?

11:10

You know AI . It's the next big scary thing

11:12

when we ask ourselves questions about

11:14

it , it's not productive to

11:16

say this is awful or this

11:19

is amazing . It's more productive

11:21

to weigh both of them . Ai

11:23

represents the tip of the iceberg for me .

11:25

Thanks for listening as UT Tyler Radio connects

11:28

with Dr Eric Gustafson of the University's

11:30

Department of Communication For

11:32

UT Tyler Radio News . I'm Mike

11:34

Landis .