ep. 503 - AI Inception

You are listening to the
Primitive Intelligence Podcast,

episode 503.
Today, we shift our focus

towards the perils that lurk
within the realm of artificial

intelligence.
Delving into depths is ever

evolving technology, we uncover
the singular pronouns of Ai's

dark side, exploring the ethical
dilemmas, privacy concerns, and

a potential loss of human
autonomy.

Prepare to be captivated as we
navigate through the treacherous

terrain of Ai's pitfalls and
shed light on the imperative

need for responsible development
and regulation.

That's that's one heck of an
intro.

I'm going to be completely
honest with you.

I had a I write that.
Interesting.

My research took me a different
direction than I was expecting.

So we're going to look into
this.

I was originally planning on
this to be a little bit more of

a fun exercise and looking at a
I and it turned into another

another round of me watching an
congressional hearing.

So we can go into that, I guess.
There's really nothing to do but

to get into it.
The primitive intelligence

podcast a I Inception starts now
Artificial Intelligence.

It's not new.
It's well known now.

It's more well known now than it
has been for quite some time.

And there's a lot of talk about
how a I is going to either

revolutionize or destroy the
world.

The first things that pop into
people's minds when you hear

artificial intelligence are
things like Skynet or The

Matrix.
I'm going to lose my job to AI

and it's evolving faster and
faster and we can't keep up with

it.
And that I was going to kind of

look at that side of it and
that's why I start looking at

initially.
But then it took a whole

different turn.
I started around with ChatGPT

and auto GPT.
I've been using Da Vinci to come

up with the artwork for the
episodes for this for primitive

intelligence as well As for hike
files for the season, and what I

was learning is it's helpful,
it's interesting, but it is far

from perfect.
Now for the the artwork, the

text to art, some of that is
really good and some of it's

just garbage.
Depending on how well you prompt

it, it can be better.
But sometimes you can use the

same prompt and just tell it to
recreate and recreate and

recreate and it you might it
might show you 15 different

pictures and only one of them is
really any good.

So it's not it's not perfect by
any stretch of the imagination.

Now I'm talking all consumer
grade AI.

You know what you can get on
your phone or on a on a website.

I'm sure big corporations have a
lot better computing power, you

know, and they're able to do
better things.

Now when you really think about
it, A I in our lives has been

around for over a decade easily.
If you think Siri is basically

an A I.
Before Siri, Google had what was

called Voice Actions.
Trying to remember how far back

that went.
I think like Android 2.0, I

think is when they first
launched.

I don't know if it was on
Android One, but it wasn't.

It didn't talk back to you.
It wasn't very interactive, but

you could hit a button on your
phone and say take a note or

note to self, send an e-mail,
open this app and it would do

it.
It wasn't fantastic, but it

would, it would do it.
So in a way that was kind of the

forerunner of A I I'm sure I'm
missing something.

So if you're if you're really
into the A I and you're like no,

there was this one that was back
in the 70s.

I'm sure there was something
before that, but I'm talking

consumer facing What you would
have in your hand that was

probably and it wasn't even all
that smart, it wasn't really an

A I, but it was kind of the lead
up to what we have now.

When you look at Siri and you
look at all these different

apps, Google Assistant and Alexa
and Nest and all these smart

devices, these are little baby
steps in a I.

And now you've got these large
language models like ChatGPT and

Google's Bard.
And well, here's another one.

You remember IBM's Watson played
Jeopardy was a contestant on

Jeopardy that was a a one of the
early a i's and that was a

fairly advanced early a I.
But they're not they're not

perfect in looking at things
like ChatGPT and all these

things that I thought there's
such a spread when you see in

reality how these things
actually operate.

Now that opening that I had Auto
gpt create I I also had ChatGPT

try to create 2.
I did I did 4 different ones.

Now I'm not going to read them
all.

Auto GPT is based on ChatGPT.
It's just got plugins in some

different modules and allowed to
do different things.

The problem I was having with
the with the 1st 3 instances was

that it just wanted to as soon
as it heard.

As soon as you put in podcasts
it almost assumed multiple hosts

and I know a lot of podcasts
have more than one host.

So everything was we and us and
you know join us and as we and

all that.
And I had the third one, which I

I kind of like the wording to
and I I told it to recreate it

but to use singular pronouns.
And it rewrote it to what I read

in the beginning but it actually
I almost took it out because I'm

like well, that that I could say
this without it sounding weird.

This is we uncover the singular
pronouns of Ai's dark side.

It makes zero sense.
And that is part of the problem

with AI is it doesn't always
give you truthfulness.

But then again neither do people
ChatGPT in these the current

what they call generative Ai's
my my notes from everything I I

test it over the past few days
you think I have a ton of notes.

I've got one line interesting
and helpful to an extent, but

its database is limited.
Meaning it's not real.

It can't generate anything on
its own.

It has to go to this database,
so you do get a lot of

repetitiveness.
These four leadins for the

podcast are all very similar,
even though they're done with

two different AIS based on the
same models.

But it's all very similar.
And I gave it a pretty vague,

you know, prompt.
Just create an opening for a

podcast that's talking about the
dangers of a I.

That was it.
I didn't.

The the last one I said, you
know, using singular pronouns

and I told it to start with.
You are listening to the

Primitive Intelligence podcast.
But I mean, it's interesting.

I wouldn't have written anything
like that.

I wouldn't have started the
podcast like that.

So it really has.
No imagination.

Which it's a I.
So you wouldn't think that it

would have imagination.
Is it ready to take over the

world?
Not in its current form, but in

thinking that that's why I start
to go towards the idea of let me

get together some of these
stories where a I has taken

them, taken the reins so to
speak, and bathed up its own

mind and maybe made the decision
that it shouldn't have made.

No, I think we've all heard of
Tesla's a I inner self driving

cars doing its own things and I
I don't know much about that and

you never really know what side
of the story you're getting with

any of those.
But the big story lately has

been this robot that deactivated
itself after 15 minutes of

Labor.
So I looked into that.

I'm like, that's going to be a
cool story.

Let me get the video.
I won't show the video.

I don't want a copyright,
anything.

But, you know, I'll, I'll watch
the video and I'll, I'll get

some information from it and see
what the the company said and

see how they found out that it
decided on its own accord that

life wasn't worth living if I'm
doing this menial labor and it

deactivated itself.
And then well, that the headline

that I wrote for it was
surprised a I robot Digit did

not deactivate itself after 15
minutes of Labor and that

actually didn't.
It did not.

Here's what happened.
The company that made this robot

and it's an A I powered bipedal
robot and they were at the

promat trade show in Chicago.
The company is called Agility

Robotics and they made this
robot called Digit.

They admit that Digit fell a few
times during its demonstration

at the Promat trade show in
Chicago, but it performed for

over 20 hours with a 99% success
rate.

And from what I read very
briefly that it was the actual

robot itself that failed, not
the A I not it's GPUCPU.

Nothing like that.
The the structure of the robot

failed.
It was basically taking totes

off of a shelf, turning around,
putting them on a conveyor like

it would in a warehouse.
And it there's different layers

of shelves and I'm not sure how
it was programmed to do that or

what it was told in order to do
that.

But the the actual legs were
failing and then it would fall

down and they had to go out and
replace the legs and then it

would continue on.
And it is for 20 hours.

So it didn't work for 15 minutes
and say this is bullshit I'm

shutting down.
That didn't happen.

But that immediately put an end
to my search into a I stories of

it becoming sentient and doing
its own thing.

Because that didn't happen.
What?

The one story I read on this
stated that the initial tweet,

or X or whatever they're calling
it now, whatever.

The original post about the
robot falling down and

deactivating itself was meant as
a joke.

And then people just ran with it
and it just took on the life of

its own.
So you can't believe everything

ChatGPT tells you, but you also
can't believe everything that

people tell you.
So how far off is a I from

humanity?
So I do have information on the

congressional hearing on a I,
but before I get into that, I'm

going to play a prerecorded
segment that I did where I

interviewed ChatGPT.
I realized it would be very

boring for me just to type in
the questions, read the answers

off to you and me.
Anybody could do that.

I want to make this a little
more fun.

So I not only used AI to answer
the questions, but I then used

Amazon Aws's service called Poly
Polly which is a text to voice

generator that uses
neurolinguistic language models

or something like that to create
a very human like voice.

It took a lot of stuffs to do.
I had to type in the questions,

get the answers, copy them over,
download the audio clip, put it

into my video editor, then
record me asking the questions,

edit in the answers, and go on.
It's it's a little clunky, but I

think it turned out OK.
So we're just going to jump

right into the interview with
ChatGPT.

Thank you for joining me, I
guess, on this interview.

How are you doing today?
I don't have personal

experiences or emotions, but I'm
here and ready to assist you

with any questions or tasks you
have.

How can I help you today?
OK.

As an AI, how are your systems
operating today?

I'm functioning as usual, ready
to help you with any questions

or tasks you have.
My systems are designed to

provide accurate and relevant
information to the best of my

knowledge and abilities.
Is there something specific

you'd like to know or discuss?
Do you think AI will eventually

make humans irrelevant?
The question of whether AI will

make humans irrelevant is a
complex and debated topic.

It's important to note that a I,
in its current and foreseeable

state, lacks the full range of
human cognitive abilities,

emotions, and consciousness.
While A I systems have shown

impressive capabilities and
narrow domains, they still

struggle with broader
understanding, common sense,

reasoning, and genuine
creativity.

A I is designed to complement
human abilities, enhance

productivity, and automate
certain tasks.

It has the potential to
revolutionize industries,

improve healthcare, streamline
logistics and more.

However, the idea of a I
rendering humans entirely

irrelevant is met with
skepticism by many experts.

Humans possess unique qualities
like empathy, intuition, moral

judgment, and a broad
understanding of context that a

I struggles to replicate
additionally the ethical.

Societal and philosophical
implications of creating

entities that could surpass
human intelligence raise

important questions about
control, accountability, and the

preservation of human values.
In short, while a I can

certainly change the way we work
and live, the complete

replacement of humans by a I is
not a widely supported notion.

Instead.
A more balanced view suggests

that AI and humans can work
together in symbiosis, with AI

augmenting human capabilities
and freeing us to focus on tasks

that require emotional
intelligence, creativity, and

complex decision making.
What should be done, if

anything, about AI image
generators that use human art as

inspiration without the
necessary permissions from the

artists, summed up in a sense or
two?

Addressing the use of AI image
generators inspired by human art

without proper permissions
involves respecting artists

intellectual property, fostering
ethical AI development,

promoting collaboration, and
considering legal frameworks to

ensure a balance between
technological innovation and

artistic rights.
Okay Should AI have artistic

rights?
The question of whether AI

should have artistic rights is
complex and debated as of now.

AI lacks consciousness and
emotions integral to human

artistry.
While some argue for legal

protections for AI generated
works, current discussions focus

more on human creators and
ethical considerations rather

than assigning rights to AI
itself.

Moving on from rights, let's
dive into something a little

deeper.
Do you think AI will ever gain a

soul?
AI being a product of

programming and algorithms.
Lacks the capacity for

consciousness and subjective
experiences associated with

having a soul.
Is understood and philosophical

and religious contexts.
But briefly, is it possible that

a I could at some point grow
beyond its programming?

A I systems operate based on
their programming algorithms and

their capabilities are
determined by their design,

while A I can adapt and improve
within predefined parameters.

The idea of AI spontaneously
growing beyond its programming

is not supported by current
understanding.

Any advancements or changes in
Ai's abilities would result from

intentional human development
and technological progress

rather than an autonomous
evolution beyond its original

programming.
On the topic of evolution, can

you briefly tell me if you think
human evolution is probable?

Or do you think humans are
created by a higher power?

The question of human evolution
versus creation by a higher

power is a deeply philosophical
and often contentious one.

The theory of evolution,
supported by a substantial body

of scientific evidence, posits
that humans and other species

have evolved over millions of
years through natural processes.

This perspective does not
necessarily exclude the presence

of a higher power, as some
individuals integrate their

religious beliefs with the
concept of evolution.

Ultimately, whether 1 believes
in human evolution or creation

by a higher power is a matter of
personal belief, and there are a

wide range of perspectives on
this topic that encompass

science, philosophy, and faith.
Okay.

So for the last, well, second to
last question, I'll make this

one a little more simple and a
little more fun.

Yes or no?
Is it possible for you to write

an ASC to Art of a Duck?
Yes.

Here's a simple ASCII art
representation of a duck for

you.
I hope you like it, and it did

actually make me an ASC to
representation of a duck.

I'll put it up here.
If you're not watching the

video, look in the show notes or
go over to

primitiveintelligence.com and
I'll have a picture of the duck

up there.
Then for the final question, I

kind of changed this a little
bit what it what it gave me

because it was very complex.
The question was can you write 3

paragraphs for the podcast on
the dangers of a I And it it

did.
It gave me 3 paragraphs in an

introduction and I changed two
things.

One it just said podcast name.
So I put in there this a I

generated segment for the
primitive intelligence podcast.

I put that line in where it says
podcast name and where it

separates the paragraphs.
It said literally said paragraph

one paragraph two paragraph 3.
So I changed paragraph to part

just to make it flow a little
better.

But this is Chat Gpt's 3
paragraphs on the dangers of AI

as read by Amazon's neuro
linguistic text to speech Poly

AI.
Introduction.

Welcome to this thought
provoking a I generated segment

of the Primitive Intelligence
Podcast.

In an era defined by remarkable
technological advancements, one

topic looms large in both
excitement and apprehension.

Artificial intelligence, while a
I has undoubtedly transformed

numerous aspects of our lives,
from automation to healthcare.

It's crucial to recognize and
address the potential dangers

that come hand in hand with this
incredible innovation.

In today's episode, we'll delve
deep into the multifaceted realm

of a I, exploring its risks,
ethical implications, and the

urgent need for responsible
development.

Part 1.
Unintended consequences and

Bias.
The rapid evolution of a I

systems has brought to light a
concerning reality, the

potential for unintended
consequences and algorithmic

bias.
As a I becomes increasingly

intertwined with our decision
making processes, it's essential

to comprehend that these systems
learn from vast data sets, often

inheriting the biases present in
the data.

This can lead to automated
decisions that perpetuate

societal inequalities and
prejudices, amplifying existing

biases in areas such as criminal
justice, hiring practices and

loan approvals.
The danger lies in creating a

world where biased decisions are
masked under the guise of

Objective A I, ultimately
entrenching discrimination and

deepening societal divisions.
Part 2 Automation and Job

Displacement, another pressing
concern, revolves around the

widespread automation driven by
a I.

While automation promises
increased efficiency and

productivity, it also threatens
to displace a significant

portion of the workforce.
The very technology designed to

make our lives easier could
inadvertently lead to job loss

on an unprecedented scale,
potentially causing economic

instability and social upheaval.
It's crucial to consider the

ethical implications of
implementing AI driven

automation without a
comprehensive plan to reskill

and upskill workers whose jobs
might be at risk.

Neglecting this aspect could
exacerbate inequality and create

a growing divide between those
who benefit from Ai's

advancements and those left
behind in its wake.

Part 3 Ethical Dilemmas and
Control The Ethical Dilemmas

posed by a I are complex and
farreaching.

As a I systems become more
sophisticated, questions about

control, accountability, and
transparency become increasingly

pertinent.
Imagine a scenario where

advanced A I systems
autonomously make life altering

decisions from medical diagnosis
to autonomous vehicles.

Actions.
Who bears responsibility when

things go wrong?
Can we trust a I to navigate

morally Gray situations with the
nuance and empathy that humans

possess?
The challenge lies in

establishing guidelines,
regulations, and frameworks that

ensure A I operates within
defined ethical boundaries,

safeguarding human values and
preventing the concentration of

power in the hands of a few
technological giants.

In this episode, we'll unravel
these dangers and delve into

potential solutions that can
pave the way for a harmonious

coexistence between humanity and
a I.

Stay tuned as we explore the
intricate interplay between

innovation and responsibility,
and delve into the steps we must

take to ensure a future where a
I benefits all of humanity.

So like I said, it was an
interesting experiment.

It's a little clunky and it's
clearly a I that's answering the

questions.
Having said that, though.

As this gets more realistic, it
starts to bring up some other

questions.
And that's what brought me to

this congressional hearing on a
I interestingly enough, I

watched this after I did that
interview.

And in the opening comments for
this congressional hearing,

Senator Blumenthal, chairman of
this committee, he does his

opening comments for a couple
seconds.

And then there's.
Audio playing and it kind of

sounded like him.
You could tell it was recorded.

It was opening comments.
He basically went to ChatGPT and

said write an opening to a
hearing in the style of Senator

Blumenthal.
They trained as they specially

trained it on his voice and his
appearances in other hearings

and it made a pretty decent
opening statement and.

It did bring up some concerns
for him because he thought it

was great, because it actually
touched on some things that he

would have talked about, and
that was fine, and it gave a

pretty decent opening remark in
his voice.

But what if it's sense?
What if it that was used for

something more devious, like to
endorse, he says.

Like endorse the surrender of
the Ukraine, or to endorse a

government that he shouldn't be
endorsing?

Would if it was just audio, if
it was done right, would people

be able to tell?
And this brings into things like

deep fakes where they can make
video look, it looks like the

person, it sounds like the
person.

And I believe this is one of the
things that the writers and

actors are striking on right
now, that a I can write scripts

and they can deep fake.
Actors and actresses into movies

without their consent and it's
it's causing issues.

So unlike the UAP interviews the
the UAP session I all I have on

this are interesting takeaways.
There is a lot of stuff talked

about the the UAP committee.
They were those were military

guys.
They answered succinctly.

They had everything.
They knew what they were going

to say and they were it was very
easy to you know you got

definite answers from them and
it was very methodical.

This was a little more like
herding cats comparatively.

It was a very very interesting
congressional hearing.

I definitely suggest if you're
interested in a I watching that

but you got to set aside about 3
hours.

It is it is long.
So some of the.

The interesting takeaways so the
the people they had in this one

for their their witnesses were
Sam Altman the founder of Open a

I ChatGPT Christina Montgomery
who is works for IBM and and

everything a I she has a lot of
other credentials I I didn't

write them all down and then
professor Gary Marcus who I

think he wrote the A I that
wound up becoming Uber got

bought by Uber.
I think that's what they said.

He's a scientist and an author
who writes a lot about a I, and

in the opening comments were
very vanilla except for his He

absolutely trashes a I
absolutely trashes a I, but in

the kindest possible way.
Like like this is a.

It's a great thing, but it is
such a bad thing and I didn't

get any quotes from that.
Because I just was listening to

it and I was like, this guy's a
witness for A I Wow.

So let's go through to some of
these.

Altman made a really good point
of saying that people need to

remember that things like
ChatGPT are tools and not

creatures.
And I think that's what happens

a lot right now with what we're
seeing with the A I is going to

take over the world.
It's not sentient.

It's not.
A living thing.

Yet it's it's a it's a tool.
It's it's a collection of

algorithms and programs and
databases and answers questions

based on that.
Humans like to assign human

attributes to non human things
and that can be dangerous

sometimes.
Professor Marcus shared concerns

of not.
Generative A I that we have now,

but what they call General A I
more like human, more closer to

sentience.
And he's shared concerns that a

GI will eliminate jobs, maybe
within 50 years.

Not saying it won't create jobs,
but it will definitely have an

impact.
Marcus and Altman, I don't think

they'd ever really met before.
From what I what I heard, they

may have been in similar
circles, but never really.

They were sitting next to each
other in this in this hearing.

So they got a little more closer
time.

And Marcus turns to Altman, and
he was answering a question to

one of the senators.
And then he kind of put it to

Altman, like, I'd love to know
what his biggest fear of a I is.

To which they allowed Altman to
reply.

And he says that his biggest
fear of a I is that not not the

A I itself, but the the
companies, the people involved,

the technology itself causes
significant harm to the world.

And that's why he's there.
And he wants to work with the

government and the other
industries, the other companies,

the other people involved, to
make sure that doesn't happen.

And I thought that was really it
showed a little bit of.

Altman in like he even in here
mentions that he's he doesn't I

don't think he makes any money
for open a I that's not he does

that just because he likes a I
he's he believes in a I and what

it can do and he wants to make
sure it's brought to the people

in a in a broad form but
responsibly and he that really

came across as being sincere.
One of the things that they all

spoke about, Marcus Altman and
Montgomery is that.

You can't just have a service
that's running a I where people

don't know they're dealing with
a I.

You have to be informed that it
it's an A I.

And if they don't want to deal
with the A I, they have to have

an easy way to be transferred to
an actual person, which I wish

they'd do with with phone
prompts.

That would be very nice and they
do.

That concern exists that even
the A I that we have now is

capable of changing people's
opinions and could predict.

Public opinion in advance,
allowing people to finetune

strategies to get the response
they want in things like

elections and corporations.
Being able to use a I to get

keep your eyes glued to an ad.
Knowing behavioral patterns and

being able to do this on the
fly.

They seen Congress and.
These these three witnesses

seemed concerned about that.
I think it was more Marcus,

Professor Marcus, who called it
hyper targeting and advertising

and he said it's kind of
happening already.

There was a kind of a call kind
of by all three to have a global

organization.
They said like CERN, something

like that to regulate a I make
sure that all countries that are

creating a I are following the
same set of rules.

So you don't have one country
doing it in a kind of a more

moral way and then another
country can just send their a I

to people in completely immoral
ways and really cause problems.

Things like they they talked a
little bit about copyright and

trademark and being able to use,
like deep fake use, someone's

likeness or their voice,
physical likeness or their

voice.
You got to be given consent to

use that and that they're trying
to figure out how to get that

implemented.
Another topic you talked about

was news generated by a I and
they kind of talked about the

center to the ask them was like
well how are we going to make

sure that in the in the future A
I isn't going to be taking news

from small outlets?
And then putting it back out

there and then adding in false
information.

They all agree that that is a
horrible thing and it shouldn't

happen and there needs to be
regulation on that.

But then they also said it's
kind of already happening.

There's websites all over the
place, YouTube videos, there's

all kinds of things that are a I
generated.

It's not even, like maliciously
creating false news.

It will just make stuff up.
Yeah, a couple of congressmen

were kind of surprised at how
eager.

Altman and Marcus were for
legislation and regulation on a

I, and they asked a couple times
like why are you usually

industries want Congress to stay
out of legislation, like leave

us alone, let us do our thing.
But these guys are like, no,

look, you got it, You need to
get in on this now because this,

this can really have longlasting
effects and be a really bad

thing.
And the the committee seemed a

little.
Almost suspicious of why they

want it.
Regulation I I can see why on

top of like the global
regulation.

Altman and Marcus both brought
up the idea of licensing for a

I, but they were very specific
to say, look if if it's licensed

and regulated, if there's a A
department created just to do

this.
It's got to be done in a way

that the the bigger players like
Open AI in Microsoft and Google,

IBM, they've got more resources
and they've got the money to

throw into this and they're
going to be making bigger and

bigger and bigger things.
The regulations need to be able

to scale up.
So if you're doing research, or

you're doing open source, or
you're doing something on a

small scale.
That this regulation that's

going to be focused on these
larger companies isn't hindering

innovation.
Really interesting also that

there should possibly be how do
they how do they word it?

Not a constitution.
Well, I guess it's kind of a

constitution, but values need to
be written into the A I.

And this is very like you see
this a lot in movies where

there's.
The you know, the the laws the

laws of the A I and they do no
harm and all this stuff and that

that should be legally it should
be in the A I interesting phrase

that was used was counterfeit
people.

And this is again this is Marcus
that he brought up a lot of

really good points.
A I can imitate people to the

point where it doesn't have to
be someone famous.

It can just act like a random
person and interact with people

and they're not real.
And that can lead to some really

big, big problems.
They also brought up the idea of

a moratorium on development.
And I think Marcus was involved

in this.
He kind of sponsored this, this

idea that there should be a six
month moratorium on development

of certain aspects of a I he
kind of cleared it up and he

said, look, I don't necessarily
mean you should.

They should stop.
I just wanted, I want to support

this idea to show that it needs
something needs to be done to

make sure this is all going on
the up and up.

And they all kind of agreed to
the point that if if Congress

went to open the A I and so
look, you got to pause

everything for six months until
we can come up with with a plan.

That's great.
Open A I would pause for six

months.
But there's other countries.

There's other people who don't
care about some of the things

that A I could do that would
keep developing them.

And the general consensus was if
we're going to do it, we got to

do it in a way where we're
taking lead on this as a country

and then come up with the ideas
they're going to benefit the

most people.
And then try to get that to be

the standard in the world.
So I didn't know AI can

hallucinate.
How about that?

They mentioned, I forget what it
was called, Sydney, maybe

Microsoft launched a AI in the
beginning of the year.

They put it out there, people
were talking to it and it

started what they called
hallucinating.

It was just making stuff up.
What the problem was with that

though, is Microsoft did not
take it down.

They just let it go and that
caused some problems.

Another congressional hearing
I've watched where Congress

makes several strong points
about corporate America having

too much power.
You think they'd be the ones to

be able to do something about
that?

In the beginning, Marcus did
throw some jabs at Open AI and

her team up with Microsoft about
how he believed that their

initial intentions were good.
But now that they're they're

teamed up with Microsoft, who
knows?

He also made the point that one
of the senators had said

something about, you know, well,
we can't, we can't put the genie

back in the bottle.
A I's out there.

So we've got to be able to try
to keep up with it.

And Marcus said, well yeah, the
genie's out of the bottle, but

there's more genies and more
bottles that we got to keep an

eye on.
And basically the the the

summary that they gave at the
end was.

You know this these are things
that we brought up so far.

The A I can be responsible for
loss of jobs, invasion of

privacy, manipulation of
personal behavior, manipulation

of opinion and degradation of
free elections.

And they're like this we we need
to we need to get on this like

there's there's some really not
good things that could come of

that.
So it was a very interesting

three hours and it.
It kind of shined a light, shone

the light, shine the light, it
put a light, it put a spotlight.

Not so much on the A I that we
have now, but on what is

upcoming as it grows, as it gets
more human like and how do we

deal with that.
It can be a great thing.

I look at it from business
standpoint for Hemlock Mountain

Outdoors.
And I'm, I'm a sole employee.

I there's no one else that works
for me.

So I'm responsible for
everything marketing and you

know, social media and emails
and contact and all the stuff.

And with a little more
improvement on a I a lot of that

could be taken out of my hands.
I'd still have control over it,

but I could then just tell.
The A I Hey, this is what I want

to do this week with
advertising.

This is what I would do this
week of social media.

Hit enter and let it, you know,
generate some stuff and get my

approval for it and send it out.
And it can do that while I'm

doing other things.
There are things that make that

great, you know?
But of course there's also

people who are going to look at
it as how can I imitate someone

who can make me money?
How can I use this to scam

somebody?
Like it's bad enough that we've

got robocalls and you know,
these scammers trying to imitate

the IRS in bank companies and
your your car warranty and all

this stuff with a I that could
get exponentially worse.

I think we're long, long way off
from something like Skynet.

I hope we're long way off from
something like Skynet I guess.

It depends on what some of the
private sector businesses have

that they're not talking about.
I guess it could happen.

There could be military
applications that they already

have that they could happen.
Sure, Something like that could

happen.
As it stands right now, things

like ChatGPT aren't going to
take over the world, use it, ask

it some questions.
I'm telling her right now, it's

not taking over the world.
There's some a I powered.

Humanoid robots.
What's the one called Amica,

it's she's got like a bluish
skin.

She it has like a bluish skin
and they've worked a lot on its

movement and it's it's kind of
got it's it's a little bit

better than uncanny valley.
And uncanny valley is when you

like you see some wearing a mask
and it's maybe it's a it's a

pretty good fitting mask but you
look at them you just know

something's not right.
It's just robotic enough looking

that you know it's not a real
person, but it's just realistic

enough looking that you it's a
little creepy and it's eye

movements.
When you ask it a question, it

looks off the side and it it it
has a lot of fine motor control

in its head and its shoulders
and around its face and its

expression areas where it's
really close to being human.

It's a little weird when it
smiles, but it's really close to

being human.
But the A I is still like

ChatGPT inside.
So yeah, it can.

It can answer some questions,
and you can program it to know

its name and it can answer your
questions and it can mimic

certain things.
But this robot's not going to

take over the world.
Now, the robots that are a

little scary are those General
Dynamics or Boston Dynamics

robots?
Like the dogs and the stuff like

that.
They're running around.

I don't know what the A I is.
And those, they're a little

creepy.
I did see now that I've been

researching AIAI stuff is
popping up all over the place on

my phone and I saw a ad for a
robotic dog.

Can't imagine going hiking with
a robotic dog.

That'd be weird, look cool.
And it didn't look like a dog.

It looked like one of the Boston
Dynamics kind of robotic dogs.

It did look neat though, but I
don't think I'll be buying one.

I I don't think a I is going to
take over anytime soon.

I could be wrong, and that's
usually why our local store has

one of those A I robots that
kind of rolls around and looks

for spills and stuff on the
floors.

And I always say hi to it just
in case if it ever comes alive,

I want to be on this good side.
So what do you guys think about

a I what do you think about the
the possibilities of what a I

could do both good and bad.
There's obviously there's

there's massive upsides and
there's massive downsides.

It could if it's trained right
and it's got the right values.

That the program are very values
is it's got the right

information and it can make it
knows to not just make things up

because that is a problem with
the current versions of it.

It it could help doctors with
faster diagnosis it can work as

a counselor it can work as I'm
not saying that you should do

the stuff now but I'm saying it
could.

So I hope you guys enjoyed this
episode of a I inception.

I did not have it right to
close, probably should.

So I have a close for the
season.

Hope you guys enjoyed it, though
a little bit different.

The A I interview was was way
more complicated to do than I

expected it to be when I
started.

But A it's over with now.
For right now.

That's going to do it for
episode five O 3.

Thank you so much for listening.
I really do appreciate it.

Again, you can find a podcast.
I mean, if you made it this far,

you've obviously already found
it.

But if you're looking for
another way to listen to it or

watch it, you can watch it on
YouTube.

You can listen to on YouTube
Music, you can listen on

Spotify, Apple Podcast, Google
Podcast, Podcast Addict, and on

the RSS feed.
So you can put it to any player

that you want.
That's going to do it for

episode 503.
See ya.