#16: Will AI Replace the “Average” Human?

Alright guys, welcome back to the junction.

Today we're going to talk about normal.

What is normal these days?

No, not you, not me.

But Sam Altman seems to think that he

can, you know, like, replace the normal person.

Average, I believe is what he oh, average. Yeah.

I don't even know what would

you consider the average person, Mel?

Actually, you're says he uses the term median human.

I think about like back to my math class.

Mean median mode.

I always forgot until like the last second, I'd be like,

oh yeah, median is the middle, but so is mode.

Anyway, mode is the most of oh, yeah, the ones

that show up, the did Sam what did Sam Altman

say that got everybody such in a tizzy?

I mean, he literally said he intends

to replace normal people with AI.

So I think that begs that there's a lot

of people that are like, what is normal? Yeah.

What do you mean by average? Right.

So all of a sudden you kind

of put people on the defensive.

Actually when you sent this to me because

you're always kind of eye rolling at clickbait,

I thought this is a little clickbaity.

Little clickbaity.

Yeah, I thought this was interesting.

There's an Oxford researcher in here on the

Futurism article and it says, this is a

hugely problematic leap to make, he says, because

all of a sudden you're assigning agency comprehension,

cognition or reasoning to these mechanistic models.

And I really like that.

It's a really big wordy statement.

But this idea that we can replace a

normal individual, let's just call it the average

Joe is a wild statement, right?

Like, well, you're talking

about feelings of individuals.

You're talking about what they're capable of doing, what

they're incapable of know, just like what makes us

human is now all of a sudden accomplished by

whatever Sam Altman and team are coming up.

You I don't know, how do you respond to that?

How do you respond to saying, hey,

Mel, I'm going to take the normal

marketing person, I'm going to replace them.

What's your response to that?

Well, of course I don't believe

that that's going to happen.

Or at least that's not something

that I want to have happen.

In our capacity here, we are leveraging AI.

So I continue to want to learn how to coexist with

the tools because the more we can do that maybe these

average things are those things that you do learn in textbooks

or in a more kind of a I don't want to

say, like a marketing position, but like an entry level things

that can be repetitive and often manual.

That's what I think that maybe

he's sort of getting at here.

So if there's one particular quote in here about

if you happen to live a median life, you

could soon be out of a job.

So that's got me thinking like,

well, what is a median life?

And this goes back to who's to

say what a median life is.

And if you have traveled at all in the last

year outside of your city, certainly outside of the country,

that looks very different depending on who you talk to.

I just don't see a world in which I'm

just coming back from a vacation and I did

my due diligence and did not work on vacation.

And I got to interact with so

many interesting people and started thinking about

how I could tell their stories.

And there's this just human element that you don't get.

I wouldn't have gotten that if I was just interacting with

some AI server bot that was bringing me my meal.

So I don't know, there's like, the natural propensity for

me to be interested in the other human being on

the other side of the table, whether bringing me my

dinner or I want to understand why you're an expert

in the field of which that we work.

And I don't know.

That's such a loaded question.

As you were talking, I was thinking through all

of the individual things that are possible right now.

I'm thinking like, the voiceovers where they're

replicating voices and they're changing people's faces

and video to look like certain people. Right.

And now we're generating pictures

and content and documents.

You basically have all of the pieces to make

a fake person that works remote feel extremely real.

Like, you get on teams or slack

meet or whatever you're zooming with. Right.

Somebody hops on.

Never met them before, but they just started. Right.

And they talk intelligently.

They start delivering average content.

You're like, Well, I guess they're great.

I don't know how much they're costing the

company, but I can ask this person to

do something and they end up doing it.

Come to find out that that whole thing is AI.

How does that make you feel?

Like you were talking about building a

relationship if nobody told you where you're

talking to this AI person right.

Thing and you thought it was the person the

whole time, how does that make you feel?

I feel like I just got bamboozled.

Maybe you might feel a little like you were

a little fleeced, but you've kind of got me

thinking about well, if this individual is helping to

contribute to those same things that I'm measured against,

or we're working together to deliver, let's just say

in the context of marketing.

Like a collaborative piece, content

piece, like a blog post.

If I can't do it in a timely manner without the

AI person or I can't do it at all, then I

would rather do it in collaboration with the mean.

I'm I was hired to and am incented to develop content.

Like I said, given given a choice between not

having that as a resource or having that as

a lower cost resource, I would take it.

Okay, Mel, so what if you have ten

bots that are producing average amount of work.

Like what they generate is just average.

Is ten workloads worth of work better than one.

Mel who is really superb at what she does.

What's the scale there?

This is maybe a little subjective because fundamentally

average is not something that I you don't

deal in average not particularly well. Okay.

But maybe there is a little bit of a mindset there.

But I think no matter what your craft, I'm just one of

those people, if you're going to do it, do really well.

So yeah, I'd rather have

one superstar than five average.

But that goes against a lot of what there

have been over the last year or two.

A lot of companies that have one

way or another reduced their force.

And you could argue there's been a lot

of terms thrown around about cutting, whether it's

a budget or kind of cutting certain capacities.

And so that's essentially what some of these companies,

the message or the message to the outsider looking

in, they must have deemed that that function wasn't

all that important or maybe the people in those

roles weren't all that effective.

I realize that that is not always the

case and I'm certainly not trying to suggest

that those who were impacted by these reduction

enforced that they don't have valuable skills.

But that is sort of the perception

or even the spin or message.

And so that's kind of what he's saying here is

like if you're just kind of fall into that category

of what a business or leaders or corporate governance could

see as average work, then you are in danger.

Well, I'm glad you brought that up.

I'm scrolling through Sam's posts and on

the 29th he's replying to Ryan Peterson

talking about deploying GPT Four copilot.

It took them 30 minutes to do one task and now

they can do it in 20 seconds with a single prompt.

That sounds awesome.

We hear a lot about that.

And in parentheses Sam's response, he says, of course.

And he's talking about how the systems are much better

at doing tasks than jobs right now and giving people

better tools to do their work faster often leads to

qualitative changes in what they can do.

Well, that makes sense, that's what you're describing.

But then he goes on to say, of course, over

the long run, we expect these systems will be able

to do all of some of today's jobs and aren't

trying to hide the ball on that.

Confident we will find new and much better jobs

when that happens, if he's confident that we can

replace entire jobs with AI, and if you're just

that average Joe, you better start skilling up or

learning some kind of trade because telling you fly

fishing, well, that's the other take of this.

If we just automate everybody's job, every single person's job,

and it's all automated, we got Tesla bots running around,

your cars are driving you everywhere, everybody's going to have

so much free time that they're not going to be

able to make enough money to live.

And that's where you get this kind of UBI idea.

And we're not really talking about that, but it'll

be really interesting to see over the next decade

or so, lots of people are now out of

jobs because they literally have been.

The whole thing is not only just

automated, but is AI the word? Can I say?

They AI'd it. Is that a thing?

No, make it a thing.

It's a thing. Yeah.

It will be interesting to see where it goes.

And that's where I go back to learning how to

coexist with the tools and staying on top of it.

Because where a job can be replaced, it also creates

a new role for someone to manage that AI strategy.

We've talked about this in earlier episodes.

You don't just turn it on and let it ride.

Somebody still has to manage these strategies or

the inputs, and it has to course correct.

So I definitely don't think we're there yet.

But you're right.

Ten years, a lot can change.

Well, you remind me of this other article that I

saw today, and I think there's a couple of versions

of this, but basically one of the professors or somebody

at a university describes that we're now basically teaching these

kids skills that will be obsolete by the time they

before they even make it out the door.

And things are moving so quickly that all of these classes

that we've come up with and basically learning how to code,

things are going to change so quickly that the time and

money that you spend getting that degree are.

Now, I won't say worthless, because there's things beyond

just the classroom that go in with college.

But we're basically teaching kids the wrong thing

because things are moving so quickly in the

industry that now what you've learned is irrelevant.

Like, Mel, how do we solve that?

How do we even sure that these kids are

getting the skills that they need to thrive in

the world that they're about to step into?

So I've talked with parents about this,

like how they're dealing with it.

In fact, I was on a webinar a couple of

weeks ago with a parent who was disappointed that his

kids teachers were discouraging or going so far as banning

the use of chachi PT to get to conclusions to

help them develop inputs for assignments.

And he has had very candid conversations with them in

the home of I want you to understand the tools.

I want you to know how to use them.

And gone so far as to say if we need to have

a discussion with your educators about why and the value that I

see and why we're adopting these tools in our household.

I'm happy to have that conversation. Yeah.

So it kind of goes to I do think I'm not close

to if this is being adopted by certain educators Or Not.

I've read more articles about banning or use,

against which I think a lot of companies

went to in general ven technology.

Initially, we did discourage using the tools on our own

systems with company issued emails because we wanted to, of

course, protect our data and our customer data.

But I think we always kind of knew it was more

of a let's understand it before we go embrace it.

Because we know how we are around here.

It doesn't even have to be a tool.

The second I build a new slide deck, it's like V One.

I've got red text box draft. Yeah.

And then out the door.

The number of times where you've

said, is this ready to go? Can I use this?

Or our sales team is sharing it with a

prospect, and they quick delete that draft box right

before they flip to the next slide.

So, I mean, we knew the minute we sort of rolled it

out or embraced it, we were going to be full speed ahead.

So I think that's where some of those educators are.

My take.

Again, speaking from a place of I don't have

young kids in school, but I do think that

the more that they can understand how these tools

work, that kind of seems where things are going.

That's one of the things that we were talking about

earlier today is that if you're in that mode of,

well, I'm going to wait and see what happens, you

might be backing yourself into a hole.

I was talking about Tesla and Ford and thinking

about the strikes that are going on right now.

Ford and GM and Stalantis.

They're trying to reduce their cost

so they can be more competitive. Right.

And I think people are owed fair

wages for what they're working on.

But Tesla now Is screaming past These guys in Terms

of The AI that They've Built the Neural Engines around

FSD, and how Quickly They've Gone from a nothing burger

to the Largest Automobile Maker in The Entire World.

And They've Done It In A Span

of Less Than 20 Years Or so.

Whereas I was talking about jeep, and jeep's been around

since World War II, and their autopilot type setup can

barely keep me in the lines of the road.

But the point is, you don't want to be the Ford.

You want to be the Tesla.

And Tesla Went Head On Into The Electric Vehicle.

If you're sitting back like Ford.

Well, we're just going to wait

and see how electric batteries work.

Now in five years or so, you as an individual,

or your company that you own and operate, or the

company that you work for, if they're pushing these things

away, they're going to be the Ford in ten years.

Now where they're trying to reduce cost to

still be competitive, whereas the other company that

didn't, they're going to be drowning in money

because they're going to be beating their competitors

left and right because they've automated everything.

And I think that's where your idea Mel, of

being like a copilot really fits in nicely.

Because we're not going to unleash the AI to just

go run a business and check in every two weeks.

Somebody's going to need to automate or

after we've automated it, somebody's going to

need to look over what it's done.

It's going to tell them, well

actually go this direction, right?

You're going to need that one excellent supreme

mind that's going to step in and be

like, let's move these things like this, right?

Like the chess master is going to move the

pieces and you're going to need other people to

review the work and ensure what the AI is

doing is headed in that right direction.

Because the AI isn't the visionary.

I don't know if it will ever be the visionary.

Maybe one day we keep going back to, I think

naturally because we are in the technology space by virtue

of what we do and we're talking about vehicles.

I keep trying to think of scenarios where

you can't quickly automate things like that.

So I go to there have been

amazing advancements in medical care, right?

But there is so much of that system

that is still not even remotely automated. It's great.

There have been amazing advancements made just even in the

last couple of years with the apps and how you

can finally book an appointment on your app and all

of your lab results show up in app. Oh, it's amazing.

But think about the people who check you in

and the nurse who takes your blood pressure.

I know there's technology out there

that can do these things.

But the investment for companies to go invest in

that kind of technology, that seems like it would

be more of an undertaking than continuing to hire

nursing staff that provide care in an environment where

people want that human emotional touch.

Well, I think that's why it's kind of like

a decade of a timeline that we're talking about

because we're going to come out with solutions that

can automate a lot of what people are doing.

But it's going to take a decade for

that company to outgrow the behemoths, right?

Whoever the behemoth is in your space, you're not

going to be able to do this overnight.

It's still going to take time, but it's going to take

a whole lot less time than what it did if you

just tried to throw a bunch of human bodies at it.

Well, if you think about the example of going

to see your doctor, right, and the amount of

time that they feasibly spend or have spent looking

through charts or unfortunately, because they're back to back

appointments are just kind of quick looking at things.

I mean, the last time I went to go

get a physical was a few months ago.

I always kind of geek out over

systems just to see what they're using.

And I could see that there was even just

a simple color coding system to flag, like, oh,

this was high, this was low, right?

So something as simple as that, that's a

technological advancement that is probably not AI.

But can you imagine if they were

layering AI on top of it?

To know Melissa Bell at this age is and

they could look back at your history and they

could pull in risk factors and stuff and all.

The things that they know from practicing medicine and

being in their industry and looking at your chart.

Now they can spend more time actually talking to

you as the patient and making you feel important.

Because I feel like that people always there seems to

always be sort of a like if you're at least

baylor, Scott and White, they seem to have this insane

commitment to providing a high level of service.

I've noticed that the last few times I've been in.

And they have you go around with this

little card and they have a good system

of following up for you to review.

And it's clearly very important to them

that you felt heard and cared for.

If they can help automate or surface information

or bring insights to the providers faster, they

will hopefully be able to see continue to

provide more care to more people.

But then they're spending more quality time with you.

I like your point that customer service

is never going to go away.

There's going to be a hard reject on

this idea that this AI bot can provide

me better customer service than a human.

Maybe like tactically.

But I'm never going to emotionally connect

with this bot over here, right.

I'm way more likely to emotionally connect with

a doctor who's got several kids, right, and

we can talk about the soccer game.

Like the AGI over here.

They don't care about that.

Their kid didn't have a soccer game.

They don't have any they might tell you that.

Yeah, I got to go get my own kid. Hold on.

I'll be right back.

Yeah, no, totally.

I mean, another example, I was overhearing my husband

on the phone the other day with a Wells

Fargo customer support rep, and she was expressing genuine

Care, what sounded like genuine Care because he was

calling about a fraudulent attempt on his grandmother's account.

And she was just so kind and personable.

And maybe the AI will get there because

it's going to ingest all these calls and

it'll feel very sweet and personable.

I don't know.

The only reason I'm calling is because I usually have some

crazy complex issue where something you want a nectar ring and

I need well, I just need a solution, right?

Otherwise, I just do it on the website.

A lot of these help the hold

calls are like, oh, you can easily

go to americanairlines.com and solve this problem.

And I'm like, yeah, I was already there

like 20 minutes ago and it didn't work.

Anyway, I digress.

There was one other thing that

I was going to talk about.

You guys should definitely check this out.

I think it's one of my LinkedIn posts.

Versaill, I think that's how you say it.

Versel.com.

Go check it out.

If you're not technically inclined, that's okay.

Connect with your computer guy because your

computer guy can definitely do this.

But we ended up deploying like

a skin like chat GPT internally.

And one of the things that stuck

out, just I think this was yesterday.

What is today? Monday? This was Friday.

Sorry, I'm losing track of time.

One of the things that stuck out was I gave

him access to this tool to one of our guys,

and it's very much like a chat bot.

And he was like, Man, I

really wish I had this yesterday.

And I'm like, okay. Why?

And he's like, What I did in 6 hours yesterday?

I just typed in the same question

and it did it in 30 seconds.

And I was like, man, that's the reason

why you need to do this now.

Not that he wasted 6 hours, because

I'm sure he learned a lot.

Love him, he's awesome.

But if he could have done it in 30 seconds, what

else could he have done in the 5 hours and 59

minutes and 30 seconds that he would have saved?

And that's the thing that you need to be tracking.

So I think there's two ways

I've gotten the question before.

How are you validating the Use

case for AI and marketing?

And how are you measuring the ROI and

so on the things that you do regularly?

And you know, it took my team 8

hours to produce a blog post on average.

Then when you take a transcript, for example, we love

this Use case and you go write your blog post

off your SME interview and it takes you 2 hours.

Now I know I saved 6 hours.

So now I had a benchmark and then I

had because it already existed where it doesn't exist.

You have to think about how long it would have

taken you right, because you can backtrack that as well.

It's easy to go, okay, I haven't done that

thing yet, but you write out the steps.

It would have taken me 8 hours to

do something that took me 30 minutes.

So that's where he completed the six hour task. Totally.

And you all were able to get to

that conclusion of that was the time savings.

But even though it was frustrating for him to learn

that he spent 6 hours on it, that's critical data.

Like very simple data for you to go.

And this is why we keep going.

This is why we keep building.

This is why we're doing this.

All the things that we've been doing

with our API key, we're tracking costs.

It's been very small and up to this

point, it's cost us less than a dollar. In total?

In total?

Well, maybe not in total.

Last there's definitely been some months where I thought,

like, man, I spent like $5 this month.

Well, it's very little, right?

We were bashing on Sam earlier.

I'm sure at some point this is going to bubble

up and he's going to double our rates or something.

I don't know.

But you take, I think, less than

a dollar and we spent 6 hours. Right.

He spent 6 hours doing whatever it was. Right.

6 hours times whatever your hourly wage is for

yourself or for your employees, man, that's a massive

amount of money in terms of less than a

dollar or 6 hours of time.

It did take you now that's like hard cost.

It took you time to actually go build the thing.

Well, yeah.

Well, on Versal, I deployed it in like, ten

minutes, so it didn't take that much time.

But again, I just deployed a repository that was

already built, so somebody else put in some time.

I'm excited.

I'm ready to get in there.

I'm ready for us to start using

this and totally seeing how it goes.

Well, guys, if you want something like

what we're doing, give us a ring.

You can always email

us at thejunction eventechnology.com.

We're all ears.

Feel free to hit us up in our DMs.

Also, if you want to hop on the podcast and talk about these

topics or tell me how wrong I am or tell Mel how awesome

she is, we would love to have you on the show.

We've got a set of headphones

and a microphone just for you.

Give us a ring, give us a holler,

and yeah, guys, we'll see you next time.

Keep it automated. Peace.