#21: Artificial Intelligence's Speedy Progression: Do we trust it?

Welcome back to season two,

episode one of the junction.

Yeah, here we go.

Did you miss it?

Actually, I really did.

I was looking at our podcast page on the Apple

app, and it was like, no episode since December.

I was like, why haven't we been on the.

Did you chew through it?

Did you actually listen to our

episodes after they would release?

I'd listen to some of them.

I really struggle to hear my own

voice, and that just bothers me.

So I've listened to some of them, but then I

had my fourth child, and that was a whirlwind and. Yeah.

So I've just been missing talking about

a lot of these cool things. Yeah.

I mean, I was disappointed that you didn't finish your

chase GPT bot that, like, we still, you know, we

could still ask questions of while you were out.

Oh, totally.

Get us closer to figuring out how

we automate paternity leave or something.

It's funny you mentioned that

because we're actually building now. Somebody.

Did I tell you this?

Somebody listened to one of our podcast episodes where

we talked about, like, an HR bot emailed in.

Now we're building it for them. Yeah. Oh, yeah.

Which is super awesome.

And we kind of approached it as well.

We don't know what you don't know, and

you don't know what we don't know.

So we're going to make it up.

Um, and it's actually really cool.

Um, so we're.

We're figuring out what that looks like.

We're adding in some additional, you know, one off

things, but we might, uh, end up productizing it.

Um, you know what?

We don't have to automate on that one.

The lead source.

I know where it came from, the podcast.

Let's not fight over that one. That's awesome. I'm.

I am very jazzed to hear that it's still moving along.

Um, there's been going back

to listening to old episodes.

I think what makes me, like, cringe the most if

I let too much time go by and don't on

these episodes, which I'm all caught up, but it's the.

So much has changed and so much of what

I know has changed, but it doesn't matter because

it's still changing all the time anyways.

So there is a lot of really cool stuff happening.

I think that I'm just glad we're back on our cadence.

So, to all of our listeners who still tuned in and

asked us when our next episode was coming, we're back.

Thanks for joining us.

We're back and we.

We committed to recording every week.

Um, but we didn't set a number of episodes. Right.

So season two, you know, might be like, what, 1012?

I don't know, DVD TBD. It.

Um, but we really liked the idea of having seasons.

I'm working on this, like, being flex thing.

You know, it's hard when you're like, type a.

Oh, and you like to organize and plan out your life.

You mean you like, set up a whole, like, talk track

for this episode and you're not just winging it like me.

Right, exactly. Yeah.

So Sam. Right.

Was out for a hot minute.

I think we talked about it.

It was all over the news.

Now he's on the board.

I think that came out just a few days ago.

Well, and they've just kind of stabilized.

Like, they've got their leadership more in position

because when that happened, there was some drama

going on with the CTO, I think, and

there was a lot of stuff shaken out.

There were emails from Elon Musk that

I saw, some of which were redacted.

But, you know, there was definitely

a lot of noise circling.

Oh, for sure, yeah.

Well, it's when you have that amount of

money in a industry that is literally exploding.

I've been watching the Nvidia

stock just go straight up.

People are going to have real hot

takes in that kind of industry.

This was a little bit lesser known, wasn't as

widely talked about, but Sam went and talked to

some people about $7 trillion investment, something like that.

And he was trying to get a lot

of zeros, a whole lot of money. Right.

And I think ultimately there's a lot more

in store here that we haven't seen yet.

Well, he says, he promises in a tweet,

he says it'll be worth the wait.

If somebody else, like, if Benioff said, hey,

it's gonna be worth the wait, man.

I'm still waiting from 2016.

When you talked about the customer 360, you know,

we might be waiting for a long time.

You went way back.

Yeah, well, I mean, we really are kind of still

waiting for some of those things that they promised.

But if that was the first and last time you

went to Dreamforce, like, where did you go again?

Actually, that's really funny you say that, because

I think it was the first and last

time that I went to Dreamforce.

When I met Scott, I don't think I went back.

But that was also.

You won't go back until you get those features.

Until you get them, they better deliver.

I'm going to text Sam real quick, tell him 2016.

I mean, it's 2024. I know.

Eight years removed.

Yeah, I think it literally was 2016.

Well, fortunately for us, Sam Altman and all

these godfathers of AI are moving a whole

lot quicker than what it would seem. Absolutely.

On salesforce.

So it, it will be interesting

to see how things shake out.

They've got, you know, board members from

reputable companies, former Gates foundation, ex Sony

exec, Instacart CEO, star studded.

Yeah, I think they definitely have maybe

better people than they did before taking

out, you know, like the main guy.

And then everybody else, you know,

was about to leave, too.

It's probably not the best move, but I

think there was probably some things we don't

know about that were going on.

Well, here's the deal.

We can talk about OpenAI all day long, but we

know who my favorite large language model is right now.

Is it meta's?

What is it? Oh, shoot. What do they have? I forget.

No, it's Google's Gemini.

You're a Gemini fan, aren't you? Nope. Is it.

No, no.

Starts with a c.

It ends with a clippy.

Clippy from Microsoft Word.

Oh, my gosh.

I would love to know how

many listeners that recall Clippy.

Yeah, I bet you they bring back Clippy.

It's going to be like Clippy GPT, and

it's going to be in your online onedrive. Word, doc.

Pure nostalgia.

Take me back a little. Bouncing.

Everybody's heard about how I'd go back to

the days of a landline in a heartbeat.

But then I also like these tools, so I'm kind of torn.

You can have your landline, and I will hang out

with Claude and we'll see how far we get. Yeah.

Have you ever seen that show on Apple tv, severance?

I haven't downloaded it.

I don't think I've watched it yet.

Are you familiar with concept?

Like, you get.

Get canned and you get money? No.

So you go, you give up your

outside, your knowledge of your outside self.

It's called your outie.

And then the innie is the one that works

all day long and it basically severs your memories.

Yeah, I've had a kid or a few since

maybe I've seen it, but now I remember it.

So I feel like sometimes when I think about

my life and how I come to work and

I'm like, in this insane, fast moving technology environment

where we're just like, give me a tool. Give me a tool.

Let me adopt that AI.

And like, we're preaching all this stuff and then

when I go home, I'm like, hmm, no tv.

I wanna, like, I want my, I don't know,

garden and some chickens and like, a landline.

And I don't really want anyone to be

able to access me until I come back.

I'm pretty sure this morning you said something about

making rice for ten years straight and that was

the only thing that you were gonna do.

Well, no, that was in the context of making sushi.

And some of the most renowned sushi restaurants.

Some of the people start out, they

make rice, rice for ten years.

For ten years. You can get really good.

Think about one skill that you have, and if you

did, ten years of effort, I mean, we're still young.

We can have so many skills in our lifetime. Still.

If you dedicate ten years to

it, you think about how good.

Yeah, but it's like one thing, you know, like, I

mean, you can do it in parallel to other things.

Like, I can be really good at the sushi rice and,

you know, being a co host on a podcast and bacon

sourdough bread, I guess, I don't know, the way you described

it, it was like, they make rice every day.

I mean, that is their job for ten years and that's it.

Yeah. Don't wash dishes.

Just come in, grab the rice, make

it, make it sticky, you know?

I mean, where I'm going with this is, is

there a world in which we can have.

I enjoy having access to these tools, but then I

also enjoy the freedom from, or like, I still have

this weird separation, or call it a severance of, like,

my work life has components of AI intertwined into the

things that I do every day.

And I have not yet, even though I

have downloaded, I have, I've downloaded chat GPT

on my phone and I've, I really wish

Claude, that's, that's who we're talking about here.

I wish they had an app.

Maybe they, they have in the most recent release

and I haven't looked at it yet, but that's

where it hasn't quite landed for me.

I haven't brought AI into my personal

life, like yesterday and the day before.

So I learned how to make, me

and my husband made homemade sushi. Okay, yeah.

What did I do?

I didn't go on chat chipt and ask cha chi

pt how to make the best homemade sushi in the

world or how to make the best sushi rice.

What did I do?

Old school Mel went on YouTube and I googled

it and I was like, what's the difference between

this rice vinegar and this rice vinegar.

And I didn't even think to ask.

Not once in 448 hours did I think I

should ask Claude my fave or chat GPT.

Like, I literally didn't even do it.

There's this weird crossover.

I'm curious if you're a listener and you use AI and

automation in your work and you don't use it in your

personal life, let me know so I don't feel so alone.

There's like.

Cause you, you have totally intertwined.

Like, yeah, it's overlapped into your.

I've, I think I've almost taken

it to maybe a new level.

Like, not in a weird way, but in like a,

it's just natural now, because of how quick it is,

is to answer something with pretty good accuracy, right?

Like, I was, I built, rebuilt my fence in

January while I was spending some time home with

the kids, and I couldn't remember, like, the standard

size of the board that goes across the horizontally.

Not your verdict boards. Right?

And I was like, well, I could google this

and spend 30 minutes trying to figure this out

in the middle of Home Depot hunting pack. Yeah.

Or I could just spin up a voice chat with GPT

and ask, hey, what is the standardized size for this board?

And it was so awesome.

It was like, da da da da. You need to do it.

But great, now I'm done, right?

And it's almost like you have an expert in every space.

That's okay, right?

You got to deal with the hallucinations, but you

can just ask it, hey, I've got this problem.

Give me a quick answer.

Like, we were, Rachel and I were hating on the

Teflon pans, and we asked it which is better?

Just hot, you know, hot take.

What's your take on stainless steel versus Teflon?

And it, like, ran down

all these different things, right?

And within seconds, I knew that I wouldn't like

a stainless steel pan because my bacon's gonna stick.

Why are you cooking your bacon in a pan? First of all?

Because I get up oven. The oven.

Yeah, but the oven takes 20 minutes.

I'm trying to move quick, make it more evenly.

Oh, my gosh.

Otherwise it gets like, okay, well,

we shouldn't have brought it up.

Bacon enthusiast over here. Claude.

Claude, what's the best way to make bacon?

Asking right now. It's in the oven.

What's the best?

375 to make bacon, 20 to 25 minutes.

Drain off your grease, save it for

later so you can cook with it.

Makes for great.

A great cooking fat for your

green beans or your chicken. Yeah.

Telling you. Okay.

It says use the right pan, cast iron skillet.

So that cast iron, not oven.

That's my first thing.

Why I was confused by your prompt. You and Rachel.

So do you only have stainless steel and Teflon?

You don't have any cast iron pans.

We do have cast iron pans.

I also hate those because the bacon sticks.

You need to season it.

All the foodies out there are gonna come after me.

Yeah, you need to type in there.

How do I season my cast iron crayons?

Well, it's funny that you mentioned Claude, because I, uh,

just subscribed literally over the weekend because of the way

that it answers things compared to chat GPT.

Um, the thing that I'm struggling with now is

that it, it is, it's like so confident in

what it tells you, like, even when you ask.

Um, let me back up.

I was trying to do some infrastructure as code

stuff on azure, playing around, just figuring it out,

just with some extra time that I had.

And I asked it to simply, I

think it's called a yaml file.

So write some code for me essentially.

And it was so positive that it was right.

I was like, I don't think that's actually right.

And it reconfirmed its own code.

And then I went and googled it and it was

wrong, you know, and it's, it's like a hallucination, compounded

by the fact that it's hallucinating its own confidence.

But you can tell it, you don't like it.

It's its answer.

Yeah, well, you can thumbs down.

And, well, we're always going with that was when you

ask it questions that you are not fully, that you

either don't fully understand all the concepts, right, or you're

just taking it, you know, as the way that it

describes it, you're gonna find yourself in trouble.

Because the first time I did it, I just was, I

ran with it, you know, I'm just playing around, no problems.

And I spent 3 hours debugging something

that it told me was right.

And I was like, hmm, yeah,

gotta worry about those hallucinations.

So I think there's still, and

that was in Claude and chat.

GPT almost did the same exact thing because I

was kind of playing in both, but I don't

know how they're going to get around that because

it's just so confident in what it tells you.

And I mean, like baking bacon, sure.

Like, not that that's, um, a very

challenging topic, but it could, um, get

to the point where it's too confident.

And what that doesn't. I mean, cooking bacon. Right.

I guess unless you burn it, it really.

You're not going to know the difference. Right?

But with. Oh, I do.

I know the difference.

Your face with code.

It either works or it doesn't.

We'll do it at the office sometime. Okay.

We're gonna bacon cook off competition.

Yeah, I'm here for that.

No, I mean, it's.

It's a really good point, I think, when I.

I don't come from the yes, you

want it to be right for me.

I'm using it more of an idea generator

than a fact validator or, you know, so

I'm probably, like, not as tuned into.

Like, is that factually correct?

A lot of the questions I ask, Claude, are

very, like, just to kind of get my gears

going or to fill in some gaps.

Hey, I've got this job description.

I want the.

This job to include these types of tasks.

What am I not thinking of?

Here's some of my expectations of someone in this role.

Do I have the right credentials listed based

on seniority or, like, things like that, where

it could say, frankly, when I do this

now with Claude, it's very suggestive.

It really hasn't come back and said, you really

need to do this, or you really need to

do that the way that I've been utilizing it.

So I think, for me, I just.

I've enjoyed the usability.

I really like the way that it talks to me.

I don't know if that's, like, you

know, there is something that I've experienced.

It's just. It's just different.

And they just released.

I don't know if anyone else out there

using cloud three, they just released this prompt

library, which I'm still kind of wrapping my

head around some of these prompts.

Like, there are some that you would recognize,

like a career coach, a mindfulness mentor, you

know, like, things that you could go.

I could kind of see how I would use that.

A dream interpreter. Ooh, I didn't see that one.

I have to test that one out.

But then there's a couple other ones where I'm

just like, I don't even know what that means.

So, like, philosophical musings?

Is that what you do?

Like, a tweet tone detector?

Like, you pop in a tweet and it

kind of reads the sentiment of the tone.

Alien anthropologist.

What does that prompt?

Are these basically Personas that they've created, and

they're trying to make it easier for people

to transition into this comfortable place where we're

all using apps in our daily lives.

I mean, forget work.

Some of this stuff is literally, what is it?

I mean, there's just a few, oh, culinary creators. See?

Oh, there you go.

Well, I don't need that already. You're already a pro.

Yeah, but, like, I don't know, I just look at

this and kind of go, is this a way to

make it safer, more appealing, accessible to somebody that is

still trying to understand how AI can enhance their life?

Yeah, I think this is just scratching the surface.

It's basically like giving somebody a prompt.

I think about, like, fake it till you make

it, or, you know, those things, like, well, I'm

gonna start this job being a therapist.

Well, don't do that.

Don't be a therapist. Right.

But like a cash register teller, right?

You don't know literally what to

do until you get trained.

Well, then somebody tells you, well, here's

the button you need to press, right?

And almost kind of feels like

that, like, where maybe might be.

Okay, I wouldn't count it as an expert, but

I think some people were starting to take it

kind of to the next level where they're analyzing

data and then doing, like, kind of automatically doing

something else because of that.

I'm thinking about this tweet that you pulled up where he

was looking for polishing, um, business emails right off of, off

of what he was doing with chat GPT four.

Um, um, like, it's, it's

the playground that you're describing.

Feels more of like, um, like a chat bot.

Like, let's chat back and forth, right?

And there's another version where it's like, hey, go

do this thing and spit back the output.

I think where we're headed is the latter of

those two, and then pair that up with the

next automation or the next input and output, and

the next thing, and then you get your response.

There's an example in the chat GBT

documentation that talks about somebody asking for

the weather in, I don't know, Cincinnati.

I forget what the actual town is, but

chatjpt doesn't know what the weather is, right?

So give it access to these functions that then can,

it can run and go get more data, right?

And I don't know if there's, like, levels of

this stuff, but I think that's where we're headed.

The chatbot is more of a, I don't

know, like an interactive kind of thing.

And I think where we're headed

is something that's less interactive.

Actually, I'm thinking about back to, um, our buddy

who was talking about things being in the loop. Right.

Or there's still a human in the loop. In my mind.

The loop just keeps getting bigger, though,

until the human comes back in. Right.

I think in those conversations, it was like, we'll go

do a couple of things, then let me check it. Right.

I think we're going to get to the point

where it's going to do a whole lot of

things, and then you're going to check it.

Um, and that loop will just

keep getting bigger and bigger.

I don't know if, like you mentioned, I don't

think we'll maybe ever step out of that.

And if we do, we have

some different problems to talk about.

Well, it will be interesting to see

how it changes our respective areas.

Sam Altman did.

He was quoted saying that AI will handle 95%

of marketing work done by agencies and creatives.

You know, I screenshotted that and sent that

in the marketing channel here then just waiting

for someone to go, what hot take.

I do think it's interesting.

He's, I mean, he's talking about the agencies

that marketers use, strategists, creative professionals that could

be outsourcing your content or your design.

Maybe it's video, something else.

There's so many services, right?

We on a different podcast, on an offshoot, in

the thick of it by our founder and CEO.

He interviews other founders, and he recently interviewed the

owner of a design agency, a marketing agency.

And his take was, we're adopting it and we're

learning how to use it, and we want to

know how we can continue to steward our clients.

Which is exactly what I think anyone in any industry

needs to do is it's kind of like make you

know that one person in the room that you don't

understand or that one, you want to get to know

them and maybe be their best friend.

Help me understand what it is that you do so that

I can figure out how we can work together better.

Or if you're going to take on some of my

existing responsibilities, then how does that change my role?

And when I read something like 95% of marketing work

done, that's hard for me to, I think about my

responsibilities and how much of that gets outsourced.

And to your point about the loop, there's a lot of

me still, I think, needing to be in that loop.

And so it's more managing those models and figuring out

the data sources and what's feeding them and how we're

feeding the market insights back into those models.

And so I don't know. Interesting.

I don't know what your take is on that.

I think things will start to blur quite a bit.

You know, maybe most, maybe not. I'll say most.

Some folks are worried that it's

going to take their entire job.

Some folks are like, this will never be anything.

I think it's certainly somewhere in the middle

where a lot of those things start.

Pieces of the loop get pulled in,

and now it's an AI thing, right?

Well, and I mean, it's maybe for some companies,

depending on what the task is, the adoption of

the technology far outweighs, it's like a minimal cost

compared to kind of the hard labor cost.

But there are other occupations that require

full blown, like, you think of, like

the restaurant industry, like the robotics, to

replace people in the, in the industry.

It's astronomical. I mean, think about.

And not every restaurant is owned by

a restaurant group that has, you know,

the same level of investment or funding.

So, I mean, there are definitely industries that come

to my mind that are like American Airlines.

Scott and I talked about this in the last episode.

You know, they recently outsourced.

They're or are working on outsourcing a large

majority of their call center to AI technology.

Oh, yeah.

But, you know, pilots and, sure.

Self driving airplanes and things like that, but,

like, pilots and flight attendants and all the

other staff that make things work.

Like, I still don't see a very

immediate, like, we're not talking three months.

We're going to be interacting

with, like, robotic flight attendants.

It could be somewhere off in the future.

But is that really a priority?

Like, yeah, I. Somebody.

Maybe, maybe it was Sam.

I don't know who said this, but somebody said

we're now in, like, the AI era and not,

this was before Taylor Swift's kind of thing. Right.

But we're now in the stage of humanity.

Being in this artificial intelligence piece.

What I've noticed just tracking all the

events or recounting all the things even

in the last 20 minutes, right.

Is that it's, it's a progression,

and not that it's inequality.

That's a whole different topic.

But people are going to be able to

do things faster, cheaper, better because of X. Right.

Maybe it's because of Claude or

because of this thing, right.

And what, what we've seen in the past

is sort of, at least in America, in

terms of, um, like, companies consolidating. Right.

We're going to see, oh, I think it was, maybe

it was Sam that said this, that we're going to

see the first $1 billion company run by one person.

I forget where that showed up,

but that was really intriguing.

Where? Where?

So, from this weekend, right, I was trying to do

infrastructure as a code, just playing around with it.

I mean, it's not going to be my day job.

But what was really interesting was like,

man, this is really complicated stuff, right?

And I could google forever.

Maybe I could get on upwork and talk, pay

somebody, I'm sure, really good money to teach me.

Or I could just use Claude to

get me, like, just one step.

Just one step, right?

And then I ran into some bugs that I

solved and then I got to the next step.

Well, by the end of it, I actually ended

up deploying something on Azure and I didn't need.

I didn't need to go to upwork. Right.

Like, sure, there's probably.

I mean, all the security people are

like, freaking out right now, but. Well, you meant.

I mean, that's kind of like a perfect segue.

I don't know if you were trying to tee this up into

one of our other headlines that we pulled up for today.

Oh, yeah.

You're talking about paying someone who's an expert in

their field, call it a consultant, or in what

I'm about to talk about a tutor.

So there was this recent article about

an AI math tutorial boosting performance in

Ghana, and it's basically this.

I think it came out of kind of really rose out

of the ashes of COVID-19 but there's this app called Rory.

It's a free AI powered math tutor available

on WhatsApp that basically is helping students in

Ghana, like, figure out, like, they're achieving higher

math scores, right, compared to regular math lessons.

And part of it is accessibility, maybe a lot

of the students, especially as this article is kind

of referencing, maybe aren't in proximity to a school,

but also, I mean, how many of them have

access to, like, a tutor, you know?

So this, there's this in depth paper

that kind of breaks it all down.

But they basically, Oxford researchers from Oxford divided a

thousand students into two groups, one that was using

Rory and the other ones that weren't.

And after eight months, the Rory students had

significantly higher math scores than the control group.

And so that's just one of those things that.

It's another great story of how it's making knowledge

accessible to people that wouldn't otherwise have it.

So if you didn't have the alternative of, I guess you

could have googled it and tried to piece it together.

But what I heard you say

was, I really would rather trust.

Like, if I didn't have these tools, I'd have

to, like, hire someone to tell me, right?

Or hire someone to build it.

What these tools are doing is it's

giving you that assistant or that tutor.

It's like open source knowledge almost, right?

Like it's no longer locked behind people

that have spent years doing it.

Not that that's wrong, but when you do that now, you

expedite just, you expedite a whole lot of things, right?

The progression of technology, the

progression of this, right.

Students learning faster because they

have 24/7 access to.

And now, what does that mean for those local economies?

Can students now enter the workforce that wouldn't have otherwise

been skilled or had the education to do it? Right?

Like, there's a real economic impact there, for better

or for worse, depending on the type of tool

and the knowledge we're equipping people with.

It's not minority report.

There was a movie where, like, all the

very wealthy people lived on a space station

and had, like, this medical thing.

Do you know what I'm talking about?

No, I'm asking Claude.

It was like a giant ring and there's a medical bed

that you could lay in and it could solve any problem.

And everybody on earth was dying.

I forget, but it reminded me

of they are working with robots.

And I don't know if this is actually in there, but

the idea behind the movie, or at least in my thought,

in my head, is that one day we're going to be

trained by robots and the robots are going to be like,

no, you have to do it this way.

And it might be.

That might be good for some factions of society.

It might be hallucinating. Right?

And it's still wrong.

I'm just imagining trying to beat up a robot.

How is that different from what

happens in the world today?

I think those are called cults.

I don't know. You're not wrong.

I mean, it's at the end of the day, like, there's

going to be, humans are flawed, there's going to be flaws

in the technology, and we have to find a balance.

Oh, totally.

Looking at this study that you had,

though, it was really interesting because they're

accessing it with a phone. Right?

And I'm scrolling through the study. On WhatsApp?

Yeah, on WhatsApp.

And I'm just curious, like, there's some screenshots

in here with, you know, math questions.

What is the line segment op,

the radius, center chord or diameter?

And I think it's talking about the radius.

Anyway, I'm just curious how they set all

of this up so that you could.

It's almost like there's a program behind the deal

that is informing you of what to do.

It reminds me of actually, maybe this was in the

thick of it, one of the podcasts where the AI

would set up a workout regimen for you and you

were supposed to do like, well, I've only got 30

minutes to rebuild my deal, right?

Like, there's some kind of underlying program

here, but this is super awesome.

What's really interesting, though, is, um, something that

I don't know how we're going to tackle.

Like, in my head, I can visualize, here's how we

solve this problem, or here's the direction this would go.

I don't know how we're going to solve this.

So, Sora, right?

Thinking, taking, talking about our three month break.

Sora has the ability to generate, I think, up

to 60 seconds of video, realistic video backup.

For the listeners who have not heard of sora.

And what is it think, like chat GPT,

you type in, generate a scene of a

individual walking into McDonald's and ordering a cheeseburger.

And there are image generators that would generate

an image of that, but nobody's been able

to really, like, generate film or video.

Not anything in the long form, for sure.

Not anything more than, like, frame by frame.

And now Sora has come to the table where

it can generate realistic video of somebody doing something.

And you tell it what it should be, right?

And if you watch some of these, it is lifelike.

There's a video.

I'm just on the OpenAI.com sora so r a right.

If you pull this up, there's a woman

walking down a street in Japan or Tokyo.

And if you were just looking this, like, I

don't know, maybe there's an ad right on, on

tv, you wouldn't be able to tell the difference.

But if you zoom in, you can see

like one person kind of splits into two.

Anyway, it's really difficult to tell the difference.

There's some other videos.

Oh, they're amazing.

I like this little close up view of a

glass sphere that has a Zen garden with it.

There's a small dwarf in the sphere who's

raking the zen garden and creating patterns.

Man, I feel like I could go

down a rabbit hole with all these.

Well, you scroll over a couple more and

there's an animated scene close up of a

short, fluffy monster kneeling beside melting red candle.

And now I'm thinking about all the

people that spend a ton of time. Right.

Pixar, Disney, and making animated films.

Not that they maybe have to worry just yet, but it

wasn't like six months ago that they could only do, like,

two or three, maybe ten, like a ten second video.

And now we're already up to 60 seconds. Yeah.

And I think.

Which is about the average attention

span of any human these days.

I don't know what you're talking about, but

I think I wasn't referring to you. Oh, right.

I was thinking, like, zelenials, you know?

Is that the new term?

I don't know, but what did they say these days?

I think Sam said something to the tune

of, this isn't being released anytime soon.

And I think the context was the political

environment, right, where you could make your candidate

look like they were doing something very bill

legal or risque or whatever. Right.

Whatever kind of smear campaign you wanted to

run, you could now generate it in Sora

and it could be all over tv.

There's been talks or articles of people running

robo campaigns in Trump or Biden's voice.

And it's like, if you can't tell what's

real anymore, we're, we're getting into problematic areas.

Um, so I think I'm actually, we've got

all this cool stuff happening, and I think

it's going, and the overall benefit is going

to be more than the potential negatives.

But I can't help but be concerned and

thinking about, like, well, what if somebody picked

a photo of my kid, right?

And now they generated a video of my kid. Absolutely.

Like, I'm already like.

But there was a. What was this?

It was on, it was on Reddit.

I saw this yesterday.

It was a photo of a child in

Africa that had put together a dog.

This is a generated image.

So not real, right.

But took a bunch of plastic bottles and made a cute little

dog out of it, like a, you know, like a sculpture.

And that wasn't really the interesting part.

I mean, you, we can do that now in, like, 30 seconds.

But it was posted on Facebook and the number

of likes and comments that it had just kept

going on and on and on and on.

Like, there's so many people that took it for real.

Didn't question.

There was no questions in the

comment, comments about it being fake.

And, uh, if you maybe think about the people, the

age group that's still on Facebook, um, you know, and

they just take that as what it was.

Like, now you can misinform that entire platform

on something that isn't true, and you could

take that in a variety of different directions. Right.

Um, so I wonder if that's where

more of the jobs are too.

Maybe that's opening up a sector of jobs around

like, ooh, it's like a whole different privacy.

Some sort of like advocacy or detecting the detection.

Yeah, we did do an episode, maybe it was toward

the end of last year, one of our kind of

predictions where we see content creation shift to content moderation.

A lot of what you're describing is information

that needs to be moderated or should be,

should be some level of discernment.

Just like, you know, the search engines sharing that

we're going to knock down AI generated content.

Surely there's another similar set of tools or a whole

market for people to fill seats to solve for that.

Yeah, well, I do have to hand it out to OpenAI.

They have literally, if you scroll down,

there's an entire section about safety.

So I don't want to spread misinformation of my own,

right, but I think it's important that as we build,

we being the proverbial everybody, we, as we build these

things out, thinking about those people on Facebook that just

said, oh my gosh, this is wonderful, we gotta build

for those people, not the people that understand what it

is we're talking about right now.

Let's talk about that next time. Next time.

I think that's a good way to end it. Yeah. Are we out?

Are we up on time?

I just think that we've covered a lot

of ground and there's, there is so much

to unpack in that, what you just said.

And I think it's worth spending some more time on.

What are the implications and what

should we be looking out for?

So from a practical application, I'm scrolling through

LinkedIn or Facebook, or I'm watching x program.

Like me as the everyday person that's listening to

this podcast or that's my friend or my grandparent,

like what should they be looking out for?

And like, not in an entirely nefarious way, but

um, to kind of critically analyze, um, because information

is so readily available and it spreads so fast.

We're in this information overload, right?

It used to be, I heard some stat the

other day, I'm totally going to butcher it.

But we, in one single day, in one, let's

say eight hour day, consume more information than people

of the 17th century did in their lifetime.

Like, there is some studies that have

been done around just the vast amount

of information that we have access to.

And let's assume that you are on your, you know,

in a feed and you're scrolling and you just like.

It's pretty incredible.

And that was just a few hundred years ago.

Are you saying that you, like, you were talking

about severance, you want to go back, you live

like a 17th century lifestyle, but then at work

you're like the 22nd century marketer.

I'm not trying to be uncomfortable, but yeah, there is

this level of, there is a point where I think

it can be, the technology can be somewhat.

We're in this, I've heard it recently described, we

were going to wrap up, but we're still going.

I've heard it recently described as the comfort crisis.

So much so that we have manufactured things that

we used to be like hardwired to do.

So, like, if we used to have

to hunt for our food, oh, yeah.

That inherently created an

environment where we exercised.

But now because we don't have to hunt for

food, we've manufactured these things called gyms to go

exercise because our bodies, like, need it to function.

Yeah.

So we've actually started to manufacture the

things that hundreds of years ago were

more born out of necessity.

So anyway, I'm just wondering where this is

all going with additional AI driving this stuff.

It's happening so fast.

What else are we going to

manufacture for ourselves to like? Interesting.

Take us back.

I like that thought about

the gym entrepreneurs out there.

If you're looking for like the next big thing,

I think it's going to be writing business emails

in a lab that you go to. Right.

And you exercise your fingers or like generating spreadsheets,

you know, just for fun because you're going to

miss having to do that for fun.

Oh, my goodness.

Not that you go to the gym for fun.

Maybe that's our next marketing campaign. I love that.

Generating spreadsheets for fun.

Did you get your twelve month membership

for the excels shop down the street? Really?

Oh, my gosh. I love it.

Okay, well, on that note.

Yeah, well, actually though, last thing, if you heard anything

today that you think you're an expert on and you

want to be on the show and tell us how

right or wrong we are, send us an email@thejunctiontechnology.com.

And we would be so thankful to have you on the show.

Or if you just want to send us a note

and tell us that, you know, hey, y'all should talk

about this or you should talk about that.

Feel free to write out an email to us. Right.

And maybe we'll catch up. Yeah, that'd be great.

Thejunction dashechnology.com in the

meantime, keep it automated.

I.