A fictional recession just crashed the

stock market. And the real story is what

nobody's going to write about tomorrow.

So here's what happened. A Substack post

written as speculative fiction from 2028

wiped out over a hundred billion dollars

in market cap on Monday. IBM created 13%

its worst day in 25 years just because

Anthropic published a blog post about

cobalt. But that wasn't the Substack

post. The Substack post was by

investment research firm Catrini, which

wrote a highly regarded piece about how

bad things could get if labor

replacement driven by AI really takes

hold due to massive AI capability gains

over the next couple of years. Look,

you've seen me cover the individual

sell-offs. This video is different

because the doom narrative that is

driving all of them, there's seven I'm

counting now, it's not been about the

technology, it's been about the

economics. And so I'm going to take a

little bit of time in this video to lay

out the economics of the bear case and

the bull case, the doomer case and the

boomer case for AI. And then I'm going

to tell you the thing that neither of

them is talking about. So we'll get to

that in a second. So first the doom meme

and why it hits so hard. So let me steal

man this out properly because it does

deserve it. Catrini research and Olab

Shaw wrote a piece called the 2028

global intelligence crisis and they

framed it as a fictional macro memo from

2 years in the future. The scenario is

simple. AI capabilities keep

compounding. Companies rationally cut

white collar headcount to protect

margins. Displaced workers spend less.

The consumption hit cascades through

mortgages. The credit ultimately

contaminates the entire financial

system. And so in their fictional

scenario, the S&P drops 38% from its

2026 highs. Unemployment hits 10.2% and

things are very, very bad. I'll cut it

short here. The mechanism they describe

is consistent. It's well constructed and

it's pretty easy to follow even if you

don't have a degree in economics. White

collar workers make up about half of US

employment and drive threearters of

discretionary consumer spending, stuff

you get to spend on because you like it.

The top 20% of earners account for about

65% of consumer spending. These are the

people who buy second homes, who buy

cars, who buy vacations, who buy private

school tuition. If AI structurally

impairs their earning power, the

consumption math gets really ugly for

the whole economy quickly. A 2% decline

in white collar employment could easily

translate into double that, like a 4%

hit on discretionary spending. And so,

Satrini describes what they call an

intelligence displacement spiral. I

would call it a negative feedback loop.

Basically, what they see is AI gets

better, companies cut payroll, savings

go into more AI, AI gets better, and so

on. There is no natural break on the

spiral. And the financial contagion

chain is plausible, too. It's certainly

familiar to anyone who lived through

2008, as I did. Essentially, once you

start to get into the financial

institutions that own AI vulnerable

companies, there is a risk of contagion

because of what they're linked to in the

global credit system. And so in this

case, the mechanism is private credit

because private credit grew from a

trillion dollars in 2015 to over $2.5

trillion by 2026 when private credit

over that decade or so picked up and

rolled up a bunch of SAS companies at

valuations that assumed perpetual

revenue growth. I I was part of some of

those exits. I've seen them at work.

Those assumptions are dying in real time

and that's been part of the sell-off

story. I think the most haunting line in

the piece is this one. In 2008, the

loans were bad on day one. In 2028, the

loans were good on day one. The world

just changed after the loans were

written. I get why this went viral. I

get why it was shared everywhere. I get

why the markets convulsed. The scenario

is vivid. It's simple. It's wellargued.

It's emotionally resonant. It's

plausible. It activates the same dread

that made the big short a cultural

touchstone. the feeling that the system

is fragile, that nobody in charge sees

what's coming, and that the smart money

that's already headed for the exits. But

the thing about doom narratives is that

they are dramatically more viral, not

due to their nuanced analysis, but due

to one of the most robust findings in

human psychology. I'm referring to

negativity bias. We humans, all of us,

are evolutionarily wired to pay

disproportionate attention to threats. A

headline that says AI can crash the

economy generates way more like 10 to 50

times more engagement than a headline

that says AIdriven deflation could cause

real purchasing power for the median

household. You're asleep already. Both

headlines describe potential futures.

One of them is going to get millions of

views and the other one won't. And

that's what I want to talk about because

the asymmetry right now is distorting

the information environment that people

are using that we are all using to make

career and investment decisions. The

doom narrative is not wrong because it

went viral. The fact that it went viral

while the counterveiling evidence barely

registers should make you suspicious

about whether you're getting the full

picture because you're not. I'm going to

give you two different bullcase

arguments. And yes, they both use

economics, but we're going to simplify

it so it actually works. Alex Emis is an

economist at the University of Chicago

Booth School of Business. He read the

same intuitive arguments about AIdriven

demand collapse that Catrini formalized

as fiction and he actually went out and

built a model to figure out what would

happen. I'm going to drop a lot of the

modeling stuff and cut straight to the

chase. When you model the actual

conditions that Catrini describes where

labor share in the economy dramatically

declines very quickly where there is no

consumption that comes back after prices

decline where wealthy capital owners who

own data centers don't end up spending

more. Where interest rates hit the floor

and they can't drop further and there's

no policy response from the government.

Yeah, you kind of get what Satrini came

up with. But what Alex argues is that if

you have all of those in a row, the idea

that you have no policy response is kind

of laughable. And as someone who lived

through 2008 with a divided government

where everyone was fighting tooth and

nail, when things get bad enough, if

things get as bad as the ST treaty memo

argues, yes, government does end up

responding. And the reason why is

entirely selfish. They want votes. They

realize they're in trouble if they don't

get votes. And so they figure out a way

to get it done. But there's a lot of

other reasons to suspect some of the

other assumptions that the Catrini memo

just kind of handwaves aside. And I

think Alex is making a good point. I'll

give you a couple of examples. One of

the things that the Catrini memo doesn't

take fully aboard is the idea that we

might consume more

if we have lower prices in the economy.

That's actually pretty reasonable.

Everywhere you look, you see evidence of

Jieven's paradox. So that's the policy

piece. The other things that the Catrini

memo talks about like labor replacement,

like prices falling and people not

buying more stuff, at least not at

scale. Those are things that might be

individually plausible, but it's sort of

difficult to add them all together and

assume that all of them are correct at

once to make sort of a perfect doom

scenario. One example is the consumption

side. Let's imagine for a moment that

prices are going down in the economy

because AI is making things cheaper. If

that's the case, then people are going

to probably buy more stuff. Now, they

may not buy 10 TVs because the price of

TVs go down, but net net in the economy

overall, if people end up having more

purchasing power, they're going to end

up buying more stuff. And this is not

just about TVs and shoes and the and

sort of the hard goods that we produce

as a society. This is about services

too. And the services case is actually

worth calling out because I think it is

pointed than a lot of the bears want to

acknowledge. And I have to give credit

to Michael Bloke who wrote a direct

response to the Catrini piece when he

saw it and made this argument really

really coherently. I think it deserves a

lot more attention than it's getting,

but again, it's not a doomer narrative,

so it doesn't tend to get the attention.

What he argued is that most of consumer

spending is in services. Think mortgage

services like how to buy a house. Think

tax preparation, think insurance

brokerage, think travel booking. You get

the idea. These are all tasks that AI

agents plausibly make dramatically

easier today because they're

fundamentally a function of complexity.

And so if you're sitting there and

you're like, where can AI agents impact

the economy? It's really plausible that

AI agents will impact the economy first

by making a bunch of those services

cheaper. I would argue that's more

plausible than say replacing all the

cobalt and the ATM machines because

that's something that the stock market

was worried about this week. But

services are really easy to replace.

They're not legacy code. They don't

touch like core of the financial system

kind of stuff. It's like, yeah, is an

agent going to be good at travel

booking? Maybe it will be, and if it is,

you'll use it. If that's true, AI agents

could plausibly compress costs by 40 to

70%. And Michael did these numbers. I'm

not just making them up. And plausibly

return 4 to7,000

in annual gain per median household

taxfree. No legislation. Basically, we

all get more money in our pockets in the

US because AI agents are compressing the

margins of all these services. And the

point is simple. Is that money just

going to evaporate? No. People are going

to spend it. Let's say it goes into home

mortgages. Let's say you pay less for

buying a house in commission because the

services cost comes down. Well, now

you're going to put that money into

furniture, into renovations, into moving

costs. It doesn't disappear. It goes

back into the economy. There's one more

piece in Michael's scenario that I think

is worth calling out. He identifies the

ongoing high trend of business formation

in the US as significant. The Census

Bureau reported 532,000

new business applications in January of

2026 alone, up over 7% from December.

That continues a long-term trend that's

been accelerating since 2021. And

Michael reasonably supposes it's going

to continue. And what he's suggesting is

that essentially oneperson businesses

have more leverage in the economy than

they've ever had before because now they

have the skills, they have the tools,

they have radically lower overhead, and

they have more reach all thanks to AI.

And this is not just theoretical. Know

personally people in my life who have

gone from not coding at all to I'm

setting up a business and I'm making

real money from it. and they feel so

motivated they're starting a formal

business out of it. This is not just one

story that I'm cherrypicking. I know

more than I can count on two hands.

There's a lot of folks out there who are

finding that the conditions the AI

revolution is bringing are ideal for

people who want to strike out on their

own. Of course, it's easy to hear the

bears responding. This time it's

different because if AI is a general

intelligence, it's going to replace

everything at once. So, where will these

entrepreneurs go? Sure, that's a real

argument, and I take it seriously, but

this brings me to the part of the video

that nobody else is talking about.

Whether or not AI displaces labor, the

way the bears describe depends on the

speed of labor displacement outrunning

the speed of technical adaptation. And I

think that is an incredibly

underrepresented part of this

conversation. And if that made no sense

to you, don't worry. We're going to get

into it. Fundamentally, both doomer and

boomer narratives assume that AI

capabilities translate incredibly

rapidly into economic impact. The doom

narrative assumes that everyone's

getting fired. The boom narrative

assumes really rapid technical

adaptation across society. Both assume

the conversion rate from AI can

technically do this to the economy has

reorganized around AI is incredibly

fast. It's not. And the reason it isn't

is the most underrepresented part of

this conversation because capabilities

are not the same as deployment.

Deployment is not the same as adoption.

Adoption is not the same as deep

integration. Deep integration on its own

is still not the same as economic

impact. Social inertia is a massive

force in the economy and it is

dramatically underrepresented in every

AI analysis I've read, bare or bull.

This is what I mean concretely. I'm

going to name kinds of inertia because I

don't want to just throw away a line and

say it's all about inertia. No, we're

going to get specific. Regulatory

inertia. Financial services firms that

want to use AI for compliance work need

approval from regulators who haven't

finished writing the rules. Health care

organizations need to navigate HIPPA and

FDA clearance and institutional review

boards. Government agencies run

procurement cycles measured in years,

not quarters. The cobalt systems that

Anthropic is talking about modernizing

run an estimated 95% of ATM transactions

in the US. Hundreds of billions of lines

of cobalt run in production today,

powering critical systems across finance

and airlines and the government. Nobody

is migrating those to a new codebase

just because a startup published a blog

post, even if that startup is anthropic.

IBM's own CEO, Arvin Krishna, said last

year that their mainframe AI coding

assistant has gotten wide adoption

because it understands existing Cobalt

code bases and decides what to modernize

across those code bases. It's not

replacing them, it's understanding them.

The distinction matters. IBM stock

dropping 13% doesn't change the fact

that their client switching costs are

measured in years of institutional pain.

Not in API calls. But we're not done.

What about organizational inertia? The

Satrini scenario assumes companies cut

headcount rationally and rapidly as AI

capabilities improve. Companies are not

rational actors. In practice, large

organizations don't work that way.

Headcount decisions are filtered through

HR policies, through employment law,

through union agreements, through

severance obligations, through

institutional knowledge preservation,

through management politics, and the

simple fact that most executives have

never managed an AI transition and do

not know what they do not know. The gap

between Claude can technically do the

parts of this job that matter and we've

reorganized our workflows and retrained

our remaining staff and built QA

processes for AI output. and we've

confidently reduced headcount. That's an

enormous gap. I've seen firsthand how

long it can take to go from AI strategy

to pilot program. It takes so long. I

have seen multiple cases where big

company pilot programs are abandoned

because the very piece of AI capability

that they worked on is no longer

relevant because AI has moved so fast

past it. You know what a good example of

that is? Rag. Everyone was excited about

Rag in early 2025. You hear a lot less

about it now because Agentic Search has

gotten better. You hear a lot less about

it now because context windows have

gotten larger. And all of the people

that spend an inordinate amount of time

fine-tuning rag systems for their wikis

are pretty much regretting it. Companies

move slowly. Here's another one.

Cultural inertia. Yes, that's different

from organizational inertia. Most people

still don't use AI in their daily work.

I know lots of those people. They are my

friends. Yes, I have friends who don't

use AI. The adoption curves are real,

but they're way, way slower for most

people than the capability curve on AI.

When Toby Lutkkey, one of the most

technically fluent CEOs on the planet,

running a company whose entire business

is tech, when he has to issue a

companywide mandate in April 2025 saying

reflexive AI usage is now the baseline

at Shopify. When he has to build it in

performance reviews in order to get it

adopted, that tells you something

important about how slowly even

high-erforming organizations change

their cultural behaviors. Toby was

really explicit about this on the

acquired podcast. He said using AI well

is a skill that needs to be carefully

learned by using it a lot. He's right.

He talked about using what he calls a

Toby Eval. Like he applies this first to

himself where he has a personal folder

of prompts he runs against every single

new model release systematically probing

capabilities as if he was a QA engineer

running unit tests. And he says that the

skill of learning to prompt AI well, of

learning to give AI all the context it

needs to write a really coherent answer

without additional search has made him

better at everything else in his job. I

actually will agree with that as someone

who's worked a lot on prompting. I feel

like I'm a much clearer communicator

because I am a prompter. But regardless,

step back for a minute with me. Toby is

Toby. He is a 1enter AI fluent CEO. Do

you think all of the personal work that

Toby put in, all of the cultural work

that Toby put in is something that a

mid-market manufacturing firm in the US

is going to replicate? Is that CEO going

to do what Toby did? No. Now multiply

that mid-market manufacturing firm times

a million. Look at all the other

businesses that are led by leaders who

are not as AI fluent as Toby. Cultural

inertia is real. The last inertia force

I'll call out is trust inertia.

Enterprises do not and should not trust

AI output by default. And the cost of

figuring out how you formally scale

verification systems is really high.

Unless you're willing to put in the

capital to invest in verification as a

competency, you're not going to get to

the point where you trust AI enough to

let it do the kind of high lever work

that Citrini needs you to do for their

memo to come true. And most

organizations don't have the capital for

that kind of investment. And most of

them frankly don't have the stomach

because moving your workforce from I

have to do this work to I have a new

skill and it's verifying the AI at scale

is really really hard. Figuring out how

to do that in a way that helps you go

faster is even harder. And all along the

way you have to build institutional

trust to deploy that AI at scale. You

have to show that you have the

appropriate guardrails, the appropriate

audit trails, the appropriate human

oversight. that takes time that no

amount of benchmark improvements can

compress. Look, these four forces don't

mean that AI is never going to transform

the economy. All they mean is that it

won't transform the economy on the

timeline the stock market is pricing,

frankly, in either direction. The

doomers require a speed of labor

displacement that social inertia simply

won't permit. And the boomers require a

speed of adoption and integration that

organizational reality won't permit.

What actually happens is slower than

both, messier than both, and far more

unevenly distributed than either

narrative allows. Here's how I think

about it. Imagine two curves on the same

chart. The first curve is really

familiar to you if you listen to me.

It's AI capability. It goes up really

fast. Model intelligence, reasoning

depth, agentic endurance, you name it. I

can tell you any number of numbers and

they all go up really fast. Gemini

doubled its reasoning in just three

months. There's an example. The second

curve is societal dissipation. And we

never talk about it and we should. The

rate at which those AI capabilities

actually permeate the economy and change

how work gets done, how money flows, how

institutions operate. This curve is way,

way flatter. It's governed by the

inertia forces I talked about. It does

compound over time, but it starts from a

really low base and it goes really

slowly. The gap between these two

curves, the really fast exponential

curve for AI and the really slow

societal dissipation curve is where we

all live today. And it's the gap that

explains almost everything that seems

confusing about this current moment. It

explains why AI capabilities are

stunning and the economic disruption is

still modest. It explains why the stock

market frankly cannot make up its mind

because it's simultaneously pricing

incredible return on investment for AI

capabilities and also pricing incredible

disaster on the other hand. It explains

why the doom narrative and the boom

narrative both sound compelling. It

explains why a blog post can crash a

stock. But there's something that's much

more important than all of that

narrative explanation that these two

curves do. and that is unveil reveal a

specific and very large economic

opportunity. That opportunity exists for

you and for me and for a bunch of

businesses precisely because this gap is

wide. If AI capabilities were

irrelevant, there would be no advantage

to adoption. Guys, do we see no

advantage to adoption? No, we do not.

It's the gap. It's the fact that the

tools are powerful but very unevenly

distributed, understood by very few, and

integrated by even fewer. That's the gap

that creates asymmetric economic returns

for you and me and for anyone who wants

to invest seriously in their AI

capability set. And that's true for

companies, not just for people. The

people and firms operating at the

capability frontier while the rest of

the economy moves at the dissipation

rate are capturing an outside share of

economic benefit. And because social

inertia is so strong, the advantage that

we're getting here does not erode very

quickly. It persists. It compounds. And

it may persist and compound for a whole

lot longer than a lot of the models

predict because the models are not

really accounting for how slowly

societies tend to change. This is not

the same as saying learn AI and you'll

be fine. By the way, it's more specific

and it's more structural than that. The

capability dissipation gap means that

the economic rewards for early

aggressive adoption are higher and more

persistent than anyone is currently

modeling. The bears assume the gap

closes really fast with rapid labor

displacement. The bulls assume the gap

closes really fast with rapid technical

adaptation. Both are wrong about the

speed. the gap stays wide and while it

stays wide the people on the right side

of it accumulate advantages that

compound with every single model

release. Now the implications of this

gap play out really differently

depending on your scale. Frankly, large

firms are positioned to win on every

dimension except one and that may be the

one that matters the most. Like start

with capital advantage. Large firms have

the money to spend 20 grand a month on

an AI agent if that's what OpenAI wants

to charge. They have data advantages.

They have decades of proprietary

information. They have distribution

advantages. They have existing customer

relationships that create deployment

surface area. And they have the budget

for extensive verification and

compliance infrastructure if that's what

they need. But but but they carry the

full weight of organizational inertia.

Every new AI workflow has to survive

procurement, legal review, security

audit, you name it. So it can take 18

months from this tool will save us 10

million a year. So we're actually saving

the money. The only exception to this is

a highly involved founder like Toby.

Those are the wild cards in the pack. If

you have a really aggressive AI friendly

founder like Toby at a large company,

that can change. Small firms and

individuals and that difference is

blurring. We have the opposite profile.

We lack the capital. We lack the data.

We lack the distribution. But we have

the one thing the big companies don't

and that is speed. The capability

dissipation gap creates an asymmetric

advantage for speed and for anyone who

can collapse the integration timeline.

So a solo consultant who can integrate

AI into their workflow today is

operating at the capability frontier

while their competitors are still doing

quarterly meetings. The practical

huristic is really today. One of the

things that marks people who are AI

native is they think in terms of the

next couple of hours or get it done by

the end of the day. They are not coming

back and talking to me about we'll get

it done in two weeks. They're not coming

back and saying can we do it next month,

next quarter. And in the cases where big

companies can move to that way of

operating which is an enormous cultural

change which hits cultural inertia etc.

they have tremendous advantages. But for

everybody else who's smaller, who's

missing the capital, who's missing the

resources, they do better if they can

get on that speed train, if they can

leverage the advantage of being small. I

think Toby understands this

instinctively, and I think it's worth

looking at a case study from Shopify as

a result. Toby's mandate with AI is not

use AI when it's convenient. It's

demonstrate why AI can't do this before

you're allowed to ask a human to do it.

and he treats model evaluation as a

personal discipline, right? He's running

structured evals on his own on his own

time and growing his test harness over

time. Toby also requires AI exploration

in the prototype phase of every single

project. Not because the output will be

production quality, but because even if

AI fails at the task, you now have an

eval for the next model. That last point

deserves emphasis. When Toby makes a

junior employee test their project

against an AI tool, he's not expecting

the AI to succeed. That's not the goal

here. He's building organizational

muscle memory. He's ensuring that when

the next model release drops, and it

will, his company has a pre-built

evaluation framework that immediately

reveals what's newly possible. He's

investing in the rate of dissipation

within his organization. Most other

companies are trying to run the AI race

with the same tools they brought to

cloud and Toby is busy shortening the

track and focusing on how he actually

can get adoption with teeth. Toby made a

really sharp observation on the podcast

this week that stuck with me. He pointed

out that the best chess game every year

for the past 20 years has been played by

machine versus machine and nobody

watches those games. But everybody in

chess knows who Magnus Carlson is. We

don't actually care about the chess. It

turns out we care about the humans

playing the chess. Toby sees this as the

key insight people get wrong about AI.

The tools are instruments to be played.

They're not replacements for the player.

The craft still matters. The judgment

still matters. What changes is the

ceiling of what a skilled player can

achieve. Look, if you've dug in this

far, I want you to walk away with three

things. First, please recontextualize

that stock market activity. I don't know

what role you play. Maybe you're a

passive investor. Maybe don't invest at

all. The memes are still there and the

memes will get into tech and get into

your company and affect you. The AI

scare trade is creating mispriced assets

and organizational chaos. Some of the

companies getting hammered aren't going

to face real disruption, but on a

timeline measured in years, not in the

weeks that the market is pricing.

Meanwhile, the market isn't pricing in

the buy side of any of this at all. What

do companies do with the savings from a

40% reduction in software costs? What

happens to the $42 billion that gets

redirected from real estate commissions

to home buyers? No one's investing in

that. The doom narrative just doesn't

have a place for it. It doesn't drive

clicks. Second, recontextualize those

doomer narratives, too. The bare case

for AI is built on real economic forces.

Demand side effects of income

redistribution from workers to capital

owners, potential savings gluts. But the

conditions required for a full economic

contraction, the thing that is making

everybody panic right now are really

extreme. And when you get an economist

from the University of Chicago modeling

out these scenarios and basically saying

they are too unrealistic to hold in

practice, don't read that as

dismissiveness. It is some needed rigor.

That is a great antidote to some of the

panic going around in both investor and

tech circles today. The doom narrative

is useful as a policy warning. We should

absolutely be thinking about how we can

support job and career transitions. We

should be thinking about broader capital

ownership. But the doom narrative is not

particularly useful as a career planning

framework or as an investment thesis.

It's it's a meme, right? It's 10 to 50

times more viral than the counter

evidence and you should calibrate

accordingly. Third, and this is by far

the biggest one, map the capability

dissipation gap as it applies to you in

your situation. The most valuable thing

you can do right now is not learn AI in

the abstract. That's 2024 advice. That's

table stakes. The valuable thing to

figure out is where you sit relative to

the exponential curve and the flat

curve. Are you operating at the

capability frontier? Are you testing new

models regularly? Are you integrating AI

into your real workflows? Are you

building evaluation frameworks for your

domain or are you kind of content at the

dissipation rate? You're aware that AI

exists. Maybe you use it occasionally,

but fundamentally you're working the

same way you did 2 years ago. The gap

between those two positions is where

economic value is concentrating in the

next 2 or 3 years. And because social

inertia is so strong, that gap actually

isn't going to close as quickly as

people think. The person who spent the

last year building genuine AI fluency in

their domain is therefore not just

learning a tool. They're building an

asset that compounds. Every model

improvement makes that asset more

valuable, not less. Because each new

capability lands on a foundation of

practical understanding that takes real

time with the model to develop. The

career move right now is to become the

person in your organization who can walk

into a room of panicking executives. And

there's a lot of panicking executives

right now. and say with genuine

authority, I've tested this. Here's what

AI can actually do in our actual

workflow. Here's what it cannot do. Here

is the implementation plan. Here's the

budget and here's the timeline. That

person does not exist in most

organizations. The technical people

understand the models. The business

people understand the workflows but not

the technical side. And the consultants,

they just understand the frameworks and

talk. But nobody can cross all three.

And if you can bring all three of those

together, you have an incredibly

valuable skill set in 2026. The doom

narrative is a useful warning. The boom

narrative is a useful aspiration. We

should study that, too. Neither is a

useful plan for you or me or our

careers. Your plan that that's the one

that matters. It should be specific. Map

which of your problems are reasoning

problems. Point them at that model.

Which are effort problems? Which are

coordination problems? test which models

handle which tasks in your real workflow

like Toby does. Build the evaluation

frameworks that let you immediately

exploit each new model release. You are

trying to collapse the gap between

capability and integration in your

domain because every month that gap

stays wide is a month you're leaving

returns on the table. Stop worrying

about the doomer narrative. Do not pay

too much attention to whatever the next

investor-driven stock selloff is. I am

sure there will be another one. Pay

attention to the capability gap. Pay

attention to where AI is going and how

slowly it's actually getting adopted by

society. That gap is the greatest

generational opportunity anyone in the

workforce is going to see. That is where

you should be spending your time. Best

of luck.