I'm not here to be your friend.

I'm here to be your leader.

I'm not here to tell you what you want to hear.

I'm here to tell you what you need to hear.

It's a mathematical model that is predicting what it should

say based on what you, you ask.

It really doesn't.

Think and process and say, what is the right answer?

What is the nuance?

Which is what we humans still have a leg up on.

I think it's zero that it makes for a better meta.

I think it's close to a hundred that it will actually roll out.

Much of the younger generation prefers to have a discussion with A LLM

because it's better at being empathic.

Whereas some of the managers that they've had, or some of the support

system that they have doesn't listen.

LLMs will always listen.

They'll always have an opinion, they'll always prop you up.

They'll always support you.

Entry music,

it's the Bob and Josh show.

Not too long ago, meta announced that their AI team has a one to 50 ratio,

so their manager has 50 direct reports.

Bob and I are gonna debate that.

We're gonna talk about why that's amazing and why everybody should do it.

And then we will flip the coin and say, maybe you shouldn't do it.

So that's today.

That's today's topic, Bob.

But how do you feel like, how does that, does that like send,

you know, caffeinated vibes?

Like, this is amazing.

It does.

Oh, I get so excited.

Oh, yeah.

Oh, let's look at the balance sheet.

Oh.

Oh, now I'm excited.

so.

I'm gonna come at this from a different direction, Josh.

Uh, it's ratios.

So to me it's ratio is sometimes, and I was, I was looking at, uh, I'm looking

at care facilities lately for my wife, there's one of the, the key I've, I've

written about this in my caregiving blog.

One of the key things, probably the key thing to ask and to focus on is what

are their caregiver recipient ratios.

Now these big businesses, there's no federal regulation and there's

very little state regulation.

So it's up to the companies to be either well behaved or not so well behaved.

it's, it's really, they have the freedom to do whatever they want.

So yesterday, uh, I asked, and I had a memory care facility

and it was eight to one.

Uh, so eight.

So there's eight, uh, re residents to one caregiver.

Um.

So that means in an eight hour day, if you just sort of average it across,

that's, that's a what one hour, uh, of care, uh, that they can provide.

Uh, and I'm, I've been mulling over that.

Uh, and that's the state regulation.

And a lot of folks will go higher than that.

it's, and it's purely a money play, Josh.

It's purely, so to me it's a money versus quality play.

Does that make sense?

Quality of care.

And how much, how much juice can we squeeze out of the

fewest number of people?

and if you do any research into the caregiving industry, RNs, LPNs, uh,

assistants, they're all burned out.

They're all, even in the hospitals.

When Diane was in the hospital, the nurses were, the ratios are out of

kilter, the businesses are driving the ratios and what's suffering.

Quality.

Right?

So now let's bring that back to here.

Meta, trying to create an attractive ratio that makes perfect business,

makes perfect sense to me.

How do we, how do we squeeze more juice?

What do managers do anyway?

the

Yeah.

I mean, nothing.

Yeah.

do, what do they do?

They do nothing.

They sit there, they smoke cigars, you know, and, and they go play golf,

whenever they can during, during the day, uh, and, and go to the bathroom.

That's all they do.

Uh, so, uh, so we, you know, we need to.

need to maximize the doers versus the leaders.

And, and so from a pure bottom line and ai, I think makes

this even more attractive.

know, because now you can have ai, what, grabbing some data or coaching

people individually or giving them performance hints and things.

So I, so AI can actually compliment that so you can handle 50 because

now we've trained you on ai.

Um, and so I'm, I'm actually.

I get it, but I, I worry about the quality.

So in the caregiving, I've actually, there, there's sucky care.

Flat out, say it anyone do the research because the businesses

have driven too far up.

this meta 50 to one strikes me as the same mindset, which is, uh, we're looking at

it as a balance sheet we're looking at it as trivializing the role of leadership.

That's, I know it's managers, but leadership.

We're not talking about bean cam.

I'm talking, you know, what we're talking about here is p

We're not talking about practice.

We're talking about.

Okay.

This is leading individuals, leading human beings.

And then we're minimizing that, or we're, we're creating a scent where a good leader

doesn't have much time to lead people.

And that, that worries me.

It, it pisses me off.

Uh, and, and I worry about it undermining the bottom line.

Right.

And being counterintuitive, it affects the business side of things as well,

Yeah.

Mm-hmm.

So my analog,

so argue with

I, I'm, I'm gonna draw an analog, um, as a, as a father with college

aged students, so many colleges talk about their professor to

student ratio, similar thing, right?

So, and everybody's trying to sell lower ratio is better.

Um, I'm gonna take an opposing stance on this of what you would expect

me to say and what Bob has said.

There's two things that I like about this.

I'm not saying you should do it, but there's two things I like about this.

One.

I appreciate a willingness to experiment and say, Hey, we're

gonna try something crazy.

Let's see if it works.

It seems crazy, but everything's crazy until somebody goes out and

proves that it actually can't work.

I, I personally would not go to the one to 50 ratio.

They are claiming they are already at the one to 25 ratios.

They're, so, they're doubling it

course,

using ai

of, of course they are.

Yeah, why

so.

Why, why not go through?

Why not just have the CEO everyone report to the CEO for Christ sakes?

Let's go

Okay, so the, the, the,

I'm

the, the other thing that I like, and Bob and I debated this before we started

a little bit, is that I read this as a drive towards a flatter hierarchy.

Bob made a good point that that doesn't mean this is exactly

how it's gonna work out.

But I value a hierarchy that is fewer pieces, fewer bits, fewer,

fewer steps from bottom to top.

Why?

Because what I found in those cases, everybody has a greater connection

to the context, greater clarity on where we're going, why we're doing it.

What it means to our customers.

I found that to be easier.

The fewer levels there are in the middle.

So if this truly were a play to drive towards that, then

I think there's some benefit.

Do I think it's gonna work?

No.

Would I do it within my org?

No, because I believe, I believe that a, a leader's job is to make.

Their team better at the craft that they're working with.

And I don't know how, again, we might be proven wrong.

Bob and I have been wrong once or twice in our, you know, 16 years of doing this.

So maybe this is the third time.

Uh, I don't know how I would effectively and with confidence

say, I'm making all of you better.

I'm making a difference here.

I feel like you're just hanging on trying to, trying to make sure things

are getting done as opposed to really being a leader and growing people.

I don't know how you do that.

Again, I've been wrong once or twice.

This might be it.

But let me go back and so if, if this is managers.

Uh, so, so first thing, let's be clear in the article that we're referencing

Meta Casters, it's a Forbes article.

I believe that Josh is referencing.

Uh, they're talking about line managers, so they're not talking

about removing hierarchy.

They may be behind the scenes, they may be doing that, but the article is

focused on line managers, first line managers, where team members report

up to them and they're talking about the ratio of line manager to employee.

And moving that from something.

Historically, I've heard numbers like eight to one roughly, plus

or minus, and you're in trouble when you get to 12 or more to one.

In, in traditional 20 years ago, there were ratio guidance that

you're overloading the manager, okay.

With direct employees.

So, and they're going to 50.

So it's that level.

Now you are, by doing this, you are removing managers at that level.

So if you had a five to one and you go to 50 to one, you're removing

quite a few of those line managers.

But that doesn't mean you're removing directors and senior

directors, MVPs and board members.

Uh, they, you could have, you could have a very heavy weight.

Since those folks are the ones making all the decisions and getting the big bonuses,

they, they may still remain there.

So that's, that's one thing.

I'm fighting for leaders at that level.

So if those folks are just pointy headed managers, I would

support what MED is doing, right?

They're not providing, so to me, the value is not telling people what to do.

It's not tracking their hours and spreadsheets.

It's not signing their vacation requests, not doing sort of the tactical management

stuff that is not what I'm fighting for.

50 to one can do that.

Maybe a hundred to one can do that.

AI can probably do some, do some of that.

Automated, fighting for leadership team development.

Team visioning, right?

Organizational alignment.

And, and, and aligning vision up and down, uh, doing strategy development, being

involved in product strategy, technology strategy, infrastructure, strategy.

It's those things.

Uh, mentoring, and we've talked about it ad nauseum in the MetCast

as the role of leadership is this.

If you start now, that's where I'm coming in and saying 50 to one.

Oh my God, you've just, you.

It's not just about.

The leader, it's about the impact that has on the quality of the

growth of the organization.

And that goes back to my, my caregiving example.

I.

Another thing that I'm thinking about is let's turn the table around and as

an engineer what this means for you.

There are many engineers that I've worked with and I've been a part of

my organizations that would love this.

Because they want to be just left alone, so their capability of being left alone

and allowed to cowboy code a little bit more because there's less oversight.

That opens that up.

Now, in theory with AI and everything along that path,

you can really reduce that.

But it'd be interesting to see the makeup of that team, or if they have an increased

turnover within that team, where they're moving from the one to 25 to the one to

50, um, I am sure that meta will not have an issue trying to attract developers,

especially in today's days and age where, um, it's unfortunately a buyer's market.

So that's, yeah.

It'll be interesting to see how that shakes out.

Um, yeah.

Thoughts on the engineer side and how you think that would be viewed, Bob?

I mean, again, I think if engineers, so if they've had bad leaders,

then it's like, yay, Whoopi.

Right?

Get, get 'em outta my face.

Um.

I think if they've had good leaders, like folks that have reported to you,

Josh, and they've had, they've had a full dose of Josh Anderson historically.

Uh, and then you said, well, we're gonna rip, you know, we're gonna reduce that.

And now I'm arguing that you're a good leadership.

Clearly.

I mean, it's not much of an argument there, you know?

That's true.

Yeah.

We are in agreement.

but I'm just, I'm, I'm just saying it's, so, it's it again,

it comes back to quality.

Right, and I'm, I'm arguing for quality, quality of the leader.

if it's a, if it's a quality leader, I think folks would,

have a problem with that.

Right, because it's not just about the leader, it's about the

effect that that has on them.

It's their growth, it's their mentor, it's their promotability, it's someone,

a sounding board to help them navigate.

Uh, someone's doing a presentation to senior execs.

You're gonna coach them through that.

I mean, part, there's, there's these myriad of moments where you're

developing the team and you have no time for actually the 50 to one.

Is relegating a good leader to the tactical?

Right.

That's what I would, even the good ones.

Then, because they have no time to do the good, the, the, you know, the courageous

stuff, righteous stuff, the leadership stuff, they're gonna be, they're gonna

be, it's just gonna push them all.

So I worry about that organizational growth.

Uh, and I know it's gonna undermine it.

I, I know, I, I know that's the fact they're looking, the, the

leaders, the CEOs and stuff, they're looking at people as being fungible.

They're fungible commodities, they are resources, they are ratios.

And you and I don't look at, well, I, I I don't look at people

that, I look at them as assets.

They are capabilities.

They're the future.

So I, I actually think this is sort of resource-based thinking on

the part of the leaders at Meta, they have the right to do that?

Of course they do.

Of course they do, but I think they're being shortsighted.

So if I'm meta and I want to make this happen, I'm gonna assume they've

made a pretty heavy investment in an AI bot or skill to do a fair

amount of the management work.

And what's interesting, and Bob buckle up now, but.

Much of the younger generation prefers to have a discussion with A LLM

because it's better at being empathic.

Whereas some of the managers that they've had, or some of the support

system that they have doesn't listen.

LLMs will always listen.

They'll always have an opinion, they'll always prop you up.

They'll always support you.

So there is a little bit in the younger generation, not a little

bit there, there, there, there is a.

AI as a support system.

So there's a potential that some of the younger generation might value

this because the thing you and I are trying to solve, they've had

some crappy managers in their day.

Let's build a manager.

Let's build an AI manager that we want and let's see if it works.

So that's my assumption in how they're trying to, if, if I got tasked with

this and for some reason I would like, yeah, we should do that.

That's how I would try, try and solve this problem.

And I would hire a much younger generation that operates in that mode

and in that manner to have a chance.

So I just wanted to put that out there as there's a, there's a,

there's a wild and crazy way where this might work for some people

you don't like it.

I know that.

No, no, no.

It's not that I, yes, it, it will.

But is that, so as you were talking Josh, I was thinking of

radical candor and, and, and ra.

Two of the tenets of

Hmm.

candor are care personally and challenge directly.

My goal is not to make you love me.

My goal is not to blow sweet smoke up your petto.

I can do that.

I leaders can do that.

You and I try to run the fine line between relationship and caring and providing

clear feedback and growth feedback, challenging people blowing smoke.

Uh, we try to do that generationally.

We may modify a little bit.

I'm not here to be your friend.

I'm here to be your leader.

I'm not here to tell you what you want to hear.

I'm here to tell you what you need to hear.

don't know if AI is gonna make that the way you described it, in the way

I think I, I, I think more pleasing.

They're more Right.

And, and I'm not, I'm not trying to be an ass.

I'm not suggesting we be assholes, but what I am suggesting is you

and I run this fine balance.

Uh, of how many hard conversations have you had in your freaking management

career, your leadership career, Josh?

Countless.

Countless cha.

Was it, were you trying to make someone happy?

Were you were trying to, you know, give them an a, trying to give them

a, a prize if it was warranted?

Yes.

Uh, but you were also just trying to be real with them.

You were trying to care and, and that connection is your caring, right?

I'm not mean I care about you.

I care.

And, and I've established that relationship.

Uh, so all, all I'm saying, and that's again, meta is disabling

that with the ratios, they're reducing the ability to do that.

I still think that if, if we're talking about a kind of company to build.

Right.

A kind of leader to build a

Okay.

to emphasize a kind of leader who creates organization if we

think that's going to be more

Okay.

not only in the short term, but more importantly in the long term.

you don't buy that, that's fine.

Meta is successful, right?

They can do whatever they want and they're gonna be successful for while.

Uh, but you remember ge.

Gen General Electric, Josh, remember?

And who was, who was?

Who was the CEO?

Uh, oh my God.

I'm blanking

Is the Jack Welsch guy, right?

Yeah, Jack.

It was Jack Welsh.

for years he was put up on a pedestal, the reality was his bullshit kind

leadership wasn't sustainable.

And, and GE went, GE went down the tube.

Largely.

They went down the tube.

So I'm, I'm just saying they have the right, uh, you and you

have the right to applaud them.

Uh, El Elon Musk.

I don't wanna lead like Elon Musk, right?

You can.

he's, he's inherently successful.

Um, but life is, you know, life is short.

Make some decisions.

Yeah, through my LLM usage, one of the more frustrating bits is

when I'm having a discussion with it and we're working on a problem

together and it's going down a path.

And I say, I believe this path is wrong.

Here's why.

And its response is, you're absolutely right.

So it's like, well, if you knew I was right, which it doesn't, right?

It's a mathematical model that is predicting what it should

say based on what you, you ask.

It really doesn't.

Think and process and say, what is the right answer?

What is the nuance?

Which is what we humans still have a leg up on.

We'll see how long that lasts, but that's the, that is the difference

is the capability of a mathematical model to determine what should be said.

For the betterment of the person on the other side of the screen versus

what the most common response is based on the training model that it has.

Whereas humans have the capability to say like, okay, cool.

Bob's asking me about this, but I really need to tell Bob about

something orthogonal to that because that's what he needs to hear.

And that's a key differentiator.

For us as humans that LLMs have not even come close to being

able to tackle at this point.

So I, what I worry about is some, similar to what Bob said is there

might be a burst of success, but I think it's gonna be capped for meta.

One thing that might happen is they may hire some rock stars into

that team, into that model that they're gonna succeed anyways.

Because they're just freaking rock stars and you get outta the way and let 'em

go, which a good leader would, would do.

But they then also might hit their head on a ceiling because they haven't

been coached or supported as well as they should have that as that as they

move throughout their career, they're likely gonna be some tough lessons that

need to be learned, that need to be.

Supported by a great leader.

So there is gonna be a cap, there is gonna be a top there.

Is there, there, there are gonna be some roadblocks they're gonna run

up against, but there's, there's a likelihood that there will be success

because of the caliber of talent that meta can attract and that those

folks are gonna succeed anyways.

But it's gonna hit a ceiling pretty quickly.

You know what though?

I'm going to call BS a little bit on that.

I,

Okay.

this is anecdotal,

Yeah.

haven't they hired, they spent millions and millions of dollars

on salaries to hire AI experts.

Largely, their AI efforts have been a failure.

They're like in a reduc of it.

talking currently about a 20% layoff reduction.

This thinking may be part of that, or they haven't done it yet, but they're,

they're leaning into an AI driven.

Right, an AI substantiated layoff, but it's also a corrective action

for strategy, like poor strategies, poor hiring strategies, poor

AI strategy, poor leadership.

These, I'm just calling the leaders on this, Josh.

yeah, yeah.

Yeah.

leaders that are 50 to one, right?

AI is our strategy.

It's, it's this combination of effective leaders working with effective teams,

with effective su support structures and making sound decisions, both in

the short term and the long term.

anything.

If anything, they should be firing.

So who are the, who are the people that are making these

silly leadership decisions?

I would argue fire them first.

I mean, we're not talk reduce them.

Re reduce them and start with Zuckerberg or something.

Right?

Right.

Go on an island.

You have a boat, you have a yacht.

You've, you've done your thing.

Get the hell out of there and bring in new leadership.

I'm okay with experiments, but it needs leadership and guidance and

boundaries and principles, and I think they've lost the weight on that.

Yeah, the, the recent track record is not great, especially in the AI space,

and this is targeted at the AI teams.

Um, and to Bob's point.

Yeah.

Is it robbing Peter to pay Paul because they've overpaid to hire

people that now they have to make some difficult choices, and so, uh, yeah.

gotten AI to do some hiring trend

Yeah,

Maybe they

Now I, I,

IA

yeah.

the leaders.

My, my immediate response is.

There's no way this works, but I'm really trying to find ways where

there's something good that's gonna come out of this, where they're gonna

walk away with a result or something.

And what I'm worried about is that there are gonna be some false positives

again, because they have the potential that there might be a rock star in

there that becomes a who knows what.

And so then that, that becomes a thing.

And the thing that worries me.

Which is why I think an episode like this is important is you and I have

seen trends where it's a copycat space,

Yeah,

gonna look at this and say, oh, hmm, met is doing this.

We should do that too.

Hey, let me march into my boardroom and say, we're making

this big change and here's why.

And hopefully somebody in the boardroom is like, what?

What do you That's, that's preposterous and doesn't look strictly at.

The spreadsheet and the numbers.

Hopefully there's someone in there that puts their foot down, is

like, this is, this is a bad plan.

Um, so I hope.

trend I'm seeing um, I'm gonna, I'm gonna share an article with you, Josh, offline.

Maybe we can have a, a look at it in our next, uh, meta Cast.

But we're DU this AI Strat.

To G trending is, is dehumanizing the humans in the

view of leaders towards them.

Or the, the view of organizations towards, it's commoditizing

the humans in, in our mindset.

And then we can reduct them right.

In some way.

Or, or we can be, become more efficient.

and, and that's okay if you're looking at it just an operational

mindset, a balance sheet mindset.

But what about the, and we've said this a few times here on previous,

the innovation, I'm talking about the innovation side of the equation.

The, the multiplying your people where one plus one equals five in your idea

space, in your construction space, in your, you know, architectural space

of whatever you're building, right?

Uh, and, and, and that that's what's going to be lacking.

Uh, and we're sort of leaving, leaving.

We're, we're leaving that thought behind.

Uh, and that, and that's what I'm like again, part of that is leaders.

You could have mediocre, you could have management leaders fine redux them.

But they're not all that way.

What, what are you doing?

Do you want leaders who are building your organization effectively

for now and into the future?

I do.

do, but now we're, we're treating them the same way.

I think that's a shame.

Not every company, but a lot of companies are jumping on that bandwagon.

Yeah.

Um.

I don't feel as if that's new.

It might be enabled in more direction by ai.

Yeah.

with that.

It's accelerated.

Yeah.

and to me it feels like a 10 x or more.

Yeah.

significantly accelerated.

Is it new?

No.

No.

There, there you, there's always been companies, it goes back that

tayloristic, that MEChA are, are we cogs, Are we delivering machines or are

we trying to innovate for the future?

You saw that in the computer space.

Uh, sun Microsystems, apple, right.

Earl early, they, you know, IBM, there was IBM, which was more mechanistic,

and then there were all of these, apple would be a spinoff of that example where

they, they treated things differently.

Okay.

So to wrap,

what percentage, let, like let's, let's, um, if you had to place a

percentage somewhere between zero and a hundred percent that this works

and becomes the norm across meta, what percentage would you put on it?

What do you mean?

will, will it last?

Or is is it gonna roll out?

That

that

So it is rolling out as an experiment in the AI group.

What percentage chance do you believe that it becomes the norm because of success?

Not because of spreadsheet math or executive math.

Um, do you think it's greater than 1% that this actually makes for a better meta?

I think it's zero that it makes for a better meta.

I think it's close to a hundred that it will actually roll out.

Ooh.

Okay.

I am, I, I am gonna say optimistic.

Josh is gonna say it's less than 25% that this becomes the norm across meta.

I think somebody's, somebody's gonna,

See, I,

somebody's gonna say the thing.

You don't think somebody will say it.

Okay.

All right.

Well.

I think, I think that bean counters, the new ai bean counter mentalities

across leadership teams in most corporations is going to drive.

That means what?

I'm equating that to 50 to one becomes the norm.

Yep.

Uh, it's, it's not, it moves from experiment to the norm.

Right.

It's a cost saving measure.

It's an efficiency driving measure.

It's an AI is gonna make up the difference measure, but it's not,

it's, it's not measuring the quality of the leadership and the, and that

will, that will, that will drop.

And I'm not saying they won't be successful,

Yeah,

right?

well they can, they can solve many problems with money.

Um, so.

but that's the, that's the thinking.

say it in a different way.

I think the people that think like you and I, I think we're Don Otes to some degree.

Uh, I think we're less than 10, 10% or less of the leaders in the world.

and that makes me sad,

Damn it, Bob.

We're gonna fix that.

Where, but I, that's why I keep talking about

Yeah.

want to influence, I want to challenge that mindset.

if Zuckerberg was sitting here, I would butt my old tired forehead

against him until I died.

Right?

I, I would, uh, because.

He's, he's essentially wrong.

He's lost sight of people.

and there's all kinds of factors for that.

His bonuses, his board, the stock market, right?

All, all of the stuff that's going on, and he's lost sight.

I think he's lost sight of even his, his principles from

when he started the company.

You, you get a lot of leaders get lost, uh, because they, they, they had much

more balance early, then they lose that.

You can see, you could, Zuckerberg I think is a fair example of that.

Well, and we've talked about in the past, you know, people get pickled by

the environment that they're in, and so the environment that a leader like that.

Spends time in is vastly different than it was when he was a maker,

when he was building the company, when he was growing the company.

What he's surrounded by is vastly different, which therefore shapes the

discussions and the questions and the thought processes and all those things.

So.

So we, so meta casters, we, I hope you enjoyed partial debate.

It's hard for Josh and I I to debate if you followed us for a while.

Uh, but I think we, I think we came at it from different directions.

We tried.

this was a, I think this was a relatively solid, uh.

Topic for us to discuss.

So hopefully you're noodling on it and you're thinking about

ratios and things like that,

Yep.

about where you fall and where your influence curve is.

So from beautiful downtown Carey, North Carolina,

And from beautiful downtown Fuke Wave Arena, North Carolina.

I am Bob

No.

the Killer Whale Galen.

And I'm Josh, the Nimble Dolphin Anderson.

Take care of y'all.

The Nimble Dolphin.

Yes, you are