Uncanny Valley: The Big Interview with Katie Drummond and Lareina Yee

This is Uncanny Valley, The Big interview.

I'm Katie Drummond.

[soft playful music]

Today's episode was sponsored

by McKinsey & Company.

And in this branded episode,

I'm sitting down with Lareina Yee,

Global Institute Director

and Senior Partner at McKinsey and Company

Yee has nearly 25 years

of experience advising clients on growth,

technology and transformation.

She works with organizations across industries

to deploy these new capabilities

to realize greater productivity,

improve the relationship

between people and machines,

and spark innovation in their businesses.

As one of McKinsey's leading experts in AI,

Yee spearheads McKinsey's technology research,

including its latest report

examining how people use or don't use AI in the workplace.

Hi Lareina.

Thank you for being here.

Thank you.

So before we dive into the main conversation,

we usually like

to start these with a little quickfire round,

so like a warmup before a workout.

Are you game, ready?

I'm totally game.

Okay, let's do this.

What is the most active text thread you are on?

Oh, with friends,

geeking out about the latest releases in technology.

What is a piece of tech that changed your life?

Oh, one that I love is the touch screen,

which I think is an unappreciated.

[Katie] Yeah.

Piece where we were able to really move to the iPhone,

the smartphones, Android,

and that changed the human relationship to phones.

And after that, we had the app stores and everything.

And so for me, the watershed moment was re-imagining

how we interact with the phone.

Yeah, I mean, it's so wild.

My daughter is eight years old

and can't conceive of a world

before she was able to like touch and tap things.

Well, exactly.

And so it was interesting,

because a couple years

after the iPhone made its moment,

I remember an executive was saying

that he was looking at a child in preschool

who took a piece of paper and kept going like this

because why isn't the piece of paper going like this?

[Katie] Yeah.

And my most recent kind of more current example

of that is, I'm from San Francisco, so Waymo is everywhere.

[Katie] Yeah. Over 200,000 rides a week.

And my colleague's daughter saw a taxi cab,

something you'd see in New York all the time.

And she said, daddy, what is that?

And he explained to to her and she said,

Oh, I get it.

It's a Waymo with a person.

No.

Yes.

That's where we're going.

That's where we're going.

That's where we're going.

Okay, what does the algorithm know about you?

Well, if it's Netflix, it knows I like British drama.

Exactly.

It does quite know.

If it's the algorithm for an LLM,

it knows that I'm a researcher.

[Katie] Interesting.

It knows I love asking questions.

I use a deep research function all the time.

What is one prediction you feel comfortable making

within a very short timeframe?

The next three to six months, what will come true?

What we are using is the worst version

of what we'll have.

Interesting.

What is a piece of tech you wish existed, but doesn't yet?

Digital twins for consumers. Oh, tell me more about that.

So what's very cool in manufacturing

and really complex businesses,

is this whole concept

that you could have a digital twin of a factory floor

or a new building, a skyrise that you're creating.

But what if we had that at a consumer level

for your room,

for your house, for anything that you wanna design?

What if we had that capability as consumers?

Oh, I love that.

Oh, I really like that.

I have many ideas floating

around in my head right now about how I would use that.

Now the most searched for thing

for McKinsey on the internet is McKinsey report,

but these days it's actually McKinsey AI report.

What is that?

Well, we've invested a ton in research

and it makes me so personally proud that my teams and I,

it's making a difference.

And I think what people care about in our reports

is that we're demystifying all these technology changes

and then saying, what's happening?

What does it mean and is it relevant for you?

[Katie] Hmm.

And so it's everything from our AI explainer series

where we geek out and explain why a context window

is such a big deal,

to really sophisticated things

like the economic potential of AI

where we sized it at 4.4 trillion

of economic value over time.

Wow.

We're one of the first companies to do that.

And so just putting a stake in the ground

to help executives look two corners ahead.

Fascinating, okay.

There's something that we have in common,

McKinsey and Wired,

and that is that at Wired we think

that technology obviously is central to our lives, right?

It's not like a, an IT concern, it's not sort

of relegated to certain industries or certain people.

I mean, it is ubiquitous.

It is the main subject in the boardroom today,

let alone sort of in in people's everyday lives.

Do you think that that is true of every industry?

And I'm curious sort of when did you see

that transformation happen from technology

as sort of something that was over here in the IT department

to technology being sort of the subject matter

for the CEO and for the C-suite and their team

to be talking about and tackling?

I don't think it's an exaggeration

to say technology is the oxygen for growth for businesses,

as well as performance.

So we do have that in common.

For me, the light bulb moment that IT had moved out

of the kind of the back room to the front room,

was I was sitting listening to Mary Barra.

Mm.

And it was one of her first kind

of public discussions

and she was saying how she feels the need

to hire more software engineers

and how when she thinks of a car,

that's essentially software in the dashboard.

[Katie] Yep.

Now this was, you know, eight, nine years ago.

This was a while ago, but for me it clicked that something

so common in my life, an automobile, is not a chassis.

It has many elements that I experience,

but really what it is is software.

[Katie] Yeah.

So that was a aha moment for me.

But I mean, AI has accelerated all of that.

So you are obviously a leader in the firm's tech practice.

How do your teams get hands on with the technology

that they are investigating, researching,

offering recommendations and guidance around,

I mean, cloud migration,

implementing Agentic AI, robotics?

Like how do they actually get into the weeds on this stuff

and understand it so that they can then, you know,

articulate that obviously for someone external?

So I think it's a mindset and an activity.

The mindset piece, is for me,

and I think the people I work with, which I really enjoy,

is we wake up every day trying to learn new things.

And it's amazing to be in the midpoint of your career

and feel like you are learning so much.

I think that curiosity is just basic.

On a activity level,

we have something called our AI labs group,

and that's where we pull together these technologies.

Sometimes we get beta versions

or if something's just released, we put it in the lab,

hands-on keys, we test it, we learn, we play with it.

And we do so, so we can figure out

ahead how might it be useful

for what our clients are trying to achieve.

And those are not technology questions,

those are business performance questions.

Those are the most traditional questions

we've been looking at for over a hundred years.

And so I think it's pulling together that practicality

and then coming back to what our work is,

how do we help clients improve their performance?

Yeah, whenever anyone asks me

sort of how they should think about AI,

I always tell them like,

well, to start you need

to actually use it.

Exactly. Like you need to use

the tools.

And so it sounds like what your team is doing

is they're getting very hands-on as often as possible,

so that they can then translate that technology.

I mean, I have a little story.

So I was sending out,

after the I/O conference for Google,

I was sending a team an example of a just kind

of a cheeky Veo video that people made.

And then they wrote back

and I woke up the next morning, they're like, okay,

after your email we went into a little spiral

and we had so much fun.

We taught ourselves how to use Veo.

We made a video and we featured something in your office

as the main character.

Oh no.

And it was like this sarcastic kind of team video

and it was just a lot of fun.

[Katie] Yeah. They were playing.

[Katie] Yeah. But they were learning

and then they sent it back.

And then of course, this is all just

the collaboration, the exchange.

And it was fun.

And what I loved about it is,

so quickly they learned how to use something

that had just been released to the public

and they were also incorporating in a way

that wasn't serious.

Yeah, it doesn't always have to be serious.

Sounds like that team is very busy right now.

I mean, I'm just thinking about sort of new model releases

and I could probably list five in the last two weeks, so.

I think all of the technology companies are on fire, yes.

Very busy.

So where the rubber hits the road on tech

obviously is with people.

And a quote you have gone back to previously

to illustrate this is,

I will quote it directly,

On every dollar of technology,

we need to invest three

to five in human beings.

So you basically said, you know,

the tech investment is the easy part.

So tell me more about that perspective

and sort of where you're coming from with that.

I mean, if you think about technology,

and sometimes we can get wrapped up in the science,

which in and of itself is super fascinating.

But ultimately, why do we want all this technology,

we want it for humans.

[Katie] Right.

We want it for the advancement of humankind.

And that is the promise.

But that's also the hardest part.

The easy part, let's take an agent,

an AI agent, everyone's talking about that,

the easy part is to stand up the agent.

Right.

I have a call center agent.

The hard part is to actually figure out how is

that a good coworker?

How does that fit in with a team?

What are the new types of activities

and jobs for the other people

who are doing that work before?

What are the new responsibilities for a manager?

Do we all feel comfortable?

Is it safe?

All of the human questions

on how we rewire our work and our interactions

and our relationships with machines, that's much harder.

The easiest piece is to buy the software.

Right.

Yeah, I mean, there is so much conversation right now

around sort of automation in the workplace

and in sort of enterprise.

What would you tell people who are listening to this

about how they should be thinking about that, right?

I think there's a lot of fear and a lot of anxiety

around this idea that LLMs,

Agentic AI, it's coming in, it's going to take my job,

it's going to automate away what I do.

But what you're saying,

and I completely agree with you by the way,

is that no, no,

these are tools for people to be better at what they do.

How do you get that across?

I mean, how should people be thinking about that?

Well, this is a leadership moment.

[Katie] Yeah.

And the technology isn't deterministic.

The technology releases all this power.

We need to decide how to design it into our lives,

how to design it safely.

How to actually create, not just efficiencies,

but more innovation.

We need to figure out how to do that.

And that's something

that is about our leaders being engaged in this

and spending the time.

And I also think there's some expectation setting, which is,

if you think the technology decision is the big decision,

it's not, it's about how do you execute that?

How do you implement that?

How do you actually think about the second

and third order effects?

And that's why leaders are paid the big bucks,

is to think about that and to steer us the right way.

Right.

So you've published a report

that builds on Reid Hoffman.

He's obviously, a very well known venture capitalist.

He has this concept, Super Agency.

So what is Super Agency?

It sounds very powerful.

What is it and how are you thinking

about the way AI will change jobs?

So we were really inspired

by Reid Hoffman's book on Super Agency.

And we wanted to take that to the workplace,

and say, what are the infinite possibilities?

Assume the technology works, then what?

What does that open?

And so we explored lots of questions.

First, we explored, where is everyone today

on that spectrum?

How much are they investing?

How much value are they experiencing?

What are the different segments

and attitudes between the employees and the leaders?

And then also, what are people starting

or how are they starting to reimagine?

So it was an incredibly fun piece of research.

So it's really sort of sky thinking

across every facet of the way we live.

I guess the difference about it now,

or what sort of feels distinct is that

it feels attainable, right?

Yeah. It sort of feels

like we're talking about this idea of Super Agency.

We're talking about Agentic AI.

We're talking about sort of all of this rapid transformation

and some of these ideas that felt sort of utopian

and unrealistic feel like if we play our cards right,

are within reach.

Am I understanding that right?

Absolutely.

The pace for me is nothing I've seen before.

So maybe just give you a tidbit on that.

When I started at McKinsey, we would do strategy projects,

something that we're very famous for,

and the timelines would be three to five years.

And sometimes long range planning would look

five to eight years out.

I haven't done a strategy of that length in quite a while.

Now people are like, 24 months is a long term outlook.

Yeah.

Because there's so much uncertainty overall,

in the world, in the markets.

But also the tools.

The technology you have

to create a better outcome are changing so fast

that you can't overplan for it.

[Katie] Right.

You can only realize it's gonna get better

and it's causing us to be a lot more dynamic,

but also the courage to dream,

because some of that is available within your tenure

as CEO or board member or CFO.

Right, right.

It's wild.

So one of the major findings of this report was

that actually employees are quickly adopting AI.

They're maximizing workflows.

They're, you know,

attaining new levels of productivity.

It's leaders that need to get with the program.

Am I right about that?

It's the C-suite that's struggling here,

it's not the actual employee.

Well, there are a couple

of interesting nuances.

So one is Reid separates kind of four segments,

and then we put some data to it.

And so you imagine on a spectrum,

you've got your most optimistic folks,

two, your gloomers and doomers,

the ones who are least excited about AI.

And one fascinating tidbit is across all four segments,

they would all self profess,

no matter how they feel about AI,

that they're investing in gaining basic

competence in their personal life.

Okay.

So no matter what their outlook is,

they're investing to give themselves the skills.

And then to your point on how people feel in the workplace,

employees are highly likely,

much more so than their managers,

to have already accepted that about 30%

or more of their work will be augmented by AI

over the next three years.

And the next thing they say is,

so the big gap they see is the investment in them.

Interesting. And they're saying,

employers, will you invest in me?

And what does that mean?

Give me access to the tools.

Let me play with it.

Let me be part of the innovation squads or teams.

Help me upskill myself.

Tell me what skills

and give me access to learn those skills.

That's an incredible message for leaders

to say their employees are ready,

they see what's gonna happen,

and they're saying, please invest.

And then the further question was,

do you think your company will invest?

And it was very lukewarm.

A lot of employees said,

it's not clear how much my employer will invest in me.

Why do you think that is?

Because they don't see it in their day-to-day lives today.

[Katie] Huh?

They don't have access to all the tools.

How many training programs are there?

How much is being embedded in the workflow

so that they can actually,

within a week or two, not years,

use AI to be better managers.

Right.

Where do you see that sort of hold up at a,

at an executive level,

at sort of a leadership level in implementing some of that?

Right?

And sort of moving quickly to address

that very big ask of an employee base, which is a,

as you say, like a pretty exceptional thing for a group

of employees to be saying,

hey, we know that this is here.

We know that this is gonna change the way we work,

help us use this technology.

But the blocker being the C-suite

is really interesting to me.

Where do you see that coming from?

So I think it has taken a while for executives

to realize this is not a sideline sport.

And in all fairness to leaders,

there have been many technology trends

that have been hyped up and haven't lived to the promise.

[Katie] Right.

So CEOs are asking the right questions.

You're saying, okay, this is interesting,

but is this a science experiment?

Is this three, five years out,

or is this something real?

And so I think there's that.

There's also really practical things

that leadership teams were working through.

How do you make it secure?

How do you make sure that the privacy

of the data of their own employees

as well as their own IP is protected?

So there are also practical things

they were working through,

but that being said, this is moving so quickly.

[Katie] Yeah.

It cannot be a sideline sport.

And a strategy cannot be,

someone else will lead and I will be a fast follow.

Gotta speed up.

So you have also stressed

that C-suite leaders need to have tech fluency, right?

That needs to be a core competency, a core leadership skill,

which may not have been as important relatively recently.

Your latest report talks about sort

of a lag in that competency.

What is going on there with that sort of lag

that you're seeing in the C-suite around, you know,

tech literacy and competency?

Yeah, mainly people need to play with it.

[Katie] Yeah. You mentioned this earlier.

[Katie] Totally.

One of the gifts, there are many gifts,

but one of the gifts about LLMs is,

whilst they are incredibly hard to create,

and that's a very rarefied talent set,

they're incredibly easy to use.

[Katie] Very user friendly, yeah.

Yes.

Any language, any education level,

the ability to ask a question is the best way to start.

And so the hesitation to make this a formal leap?

It's not, just start using it.

[Katie] Right.

Just incorporate it into your day.

And this is where having access

as employees is really important,

because it's hard to do that

if you don't have access to the tools.

Do you have any advice for someone who's listening.

Maybe they're executive, you know,

they're in a senior management role.

What are some little tips that you give people to like

jump in and start getting their feet wet

if they're still a little bit nervous

or sitting on the sidelines?

Aside from the kind of more personal experience

of jumping in and asking the questions,

I think one of the things we've seen is very effective

for management teams is to create the reimagine aha.

And what that is, is something immersive,

something that feels experiential

where they're saying, I get it.

I get how this can open up different frontiers

for my business.

I get how I could reconceptualize supply chain.

I get how I could re-envision personalization.

I've always wanted to do X.

I now realize I could.

I didn't even know I could do

that scientific step differently.

And so instead of sort of listening to a podcast

or as much as I like,

this is probably not a nice thing to say,

but I think it's not as much

of an academic learned activity,

which is typically how we've as adult learners have tried

to get our heads around things.

I think you have to feel it.

And so oftentimes what we do is we'll bring CEOs,

their management teams, their boards, to Silicon Valley,

and instead of talking about tech,

we go see the tech,

we'll take them to Stanford.

[Katie] Oh wow.

We'll hear from professors

and they'll think about their business,

they'll bring their ideas already.

They'll use it as a catalyzing moment to say,

what ideas do we already have?

And then they'll kind of discuss and experience it together.

And oftentimes through that, there's an aha moment.

So I think, you know, you said the word play.

Technology,

this technology, it's fun.

[Katie] Yeah, it is fun.

I completely agree.

From your vantage point at the McKinsey Global Institute,

what frontier technology trends

do you think leaders are underestimating right now?

I think we're underestimating how long it takes

to actually apply and implement.

[Katie] Oh, interesting.

Right now we have mostly

this rapid rate of innovation and release.

Something we've talked about.

[Katie] Yeah.

But it takes real time to digest that.

[Katie] Yeah.

And so I do think that the challenge

that we're not seeing,

and this will be probably the difference

in timelines where we start

to see the economic potential really play itself out,

is how fast are we adopting and using it?

That's one of the things.

On the actual trends themselves,

I think people are fully aware.

If we look at Google searches

and other types of media mentions,

they're fully aware of Agentic AI.

Yeah.

They're fully aware of AI.

That message is home.

What people are starting to talk about is, wait,

what's the second bounce on AI infrastructure?

Wait, what does that mean for data centers?

What does that mean for energy?

What does it mean for scarce resources?

And those conversations are starting to open up.

As one fact, if we think about data centers,

they're growing 33% per year right now.

And those are kind of the,

think of that as the fuel

for the hungry, hungry hippos of AI.

[Katie] Right. Models.

By 2030, 70% of all data center capacity needs

will be AI data centers.

Wow, that is wild.

So you're essentially talking about sort

of the quote unquote knock on effects of AI

and it becoming much more ubiquitous,

that that creates then new opportunities

from a business point of view, whether it's, you know.

[Lareina] Absolutely.

Data centers, a different type of infrastructure,

the energy needs,

all of that stuff is sort of the next frontier.

Exactly.

And so we will redefine the technology stack,

the components and the opportunity,

if you're on the technology side,

is there's a whole bunch of innovation

at each of those layers.

How do you think about that?

But for other industries,

we have to be mindful of all of the pieces here.

It's not just the use of the LLM,

it's all of the infrastructure that it pulls through.

[Katie] Right.

And it's also the change in all the applications

and different things that we use today

to power our businesses.

But it may also be human things, how we're changing art.

One of the things I've been completely fascinated

about is Refik Anadol and how he's changing visual art.

[Katie] Yeah.

And here in New York, his exhibition in the MoMA,

the number of people who went

and saw generative AI art for the first time.

So there's so much happening.

But on the business side, we do need to think

through the second and third bounces.

So how do you help those business leaders

then actually connect those trends, right?

These technology trends to real business value.

How do you keep them out of this like murky middle

of indecision that they might get stuck in?

We suggest that companies imagine

that they are rewiring their business.

They're not just adopting technology.

And that has three layers.

One, is strategic alignment,

two, is the operating capabilities,

and three, is change management.

And strategic alignment,

that's something we've been doing for decades.

And it's not about technology,

it's about what business outcome

are you trying to achieve and is it meaningful?

What is it worth and for what time?

Asking that, getting your top team aligned,

that's the first step.

Can't skip it.

The next one is how do you get that operating set

of capabilities from talent, to technology, to data?

How do you make sure that all of that is in place?

And then the third layer is where the humans are,

the change management.

And as we've talked about,

we know that the change

is so much harder than the invention of the technology.

Obviously we are in a moment

of incredible transformation, right?

Technologically powered transformation.

What are the key differences between a tech transformation

that stalls and one that succeeds?

So one of the things that we've seen is

that there's a lot of stalling.

Something I often call pilot purgatory.

Pilot purgatory?

[Lareina] Yeah. That's really good.

So the reason I talk about pilot purgatory,

is if we look at the vast majority

of AI that's been deployed,

what we find is that a little under 50%

of those AI investments are stalled in the pilot mode.

Then some set of those have moved to production.

And when we ask executives in our report,

they said about 1% said they had actually

achieved full scale.

Wow, 1%.

1%.

92% of companies all say they're investing more

in generative AI in the next two to three years.

And yet today, the vast majority

of their investments are still in that pilot mode.

And so without moving beyond pilot purgatory,

without moving to production, without moving to scale,

you're not gonna move the needle on any of this.

You're not gonna get the productivity benefits,

the innovation benefits.

And certainly that's just,

that tends to be you're just implementing like a feature,

a task, as opposed to actually re-envisioning your work.

So how do you then partner with leaders, right?

To align an organization around that change

and to make sure that that is a change

that sticks, right?

That you really like nail the landing.

So there's some broad experimentation

that of course you should do.

And we're not suggesting you should have

a hundred percent success rate.

Actually there's a lot of joy and failure,

but there should be two

or three big change the game investments that you're making.

And you should have a view of what you expect that to be.

For example, I would like

to change the entire loan processing approach.

How can you re-envision that?

I would like to re-envision sales in a world

where I actually can do touchless sales,

but you have to take something.

And oftentimes you take a domain or a workflow

and you say, how can I actually infuse all

of these technologies to make that radically different?

And what is that worth and what's the timeline

and the accountability that I'm gonna drive my team

to create that.

You can't just put the flowers out there.

[Katie] Right.

You actually have to know what garden you're developing.

These are big gardens though.

I mean you're talking

about sort of big scale transformation, like big ideas.

Yeah, I mean with what the technology can do,

why would you be looking at individual activities

when you can actually change fundamental workflows?

Amazing.

Well, Lareina, before we end,

we always like to play a little game around here.

It's called control alt delete.

And the way it works is, I wanna know what piece

of tech you would love to control.

What piece would you alt, so alter or change?

And what would you delete?

What would you vanquish from the Earth

if given the opportunity?

Now I understand that you might have a little spin on this.

I am totally open to spin, so let's hear it.

Can I reframe?

Absolutely.

Okay, my reframe is, I like shift seven.

Okay, tell me why.

So shift seven on the computer keys is the ampersand.

[Katie] Okay.

Which is the and.

And the reason why I like that

is because I feel like that is the world we're in.

It's always an and.

It's AI and humans.

It's our traditional sets of things we do, and innovation.

So I think it's very powerful to say and.

In a world where people are thinking scarcity,

I think you have to think, how do you make it more?

How do you widen the pie?

So I'm a big fan of the and.

Okay.

Shift seven. We're ending.

We're ending on shift seven.

Thank you again so much for being here.

This was great.

Thanks Katie.

[soft electronic music]