Power Law World in the AI Age
Key Points
- The AI era has shifted the world from average‑based norms to a power‑law environment where outcomes are driven by extreme nonlinearity rather than the median.
- Traditional workplace metrics—promotability, software fit, buying committees—are rooted in average performance, but those frameworks no longer apply to talent, product development, distribution, marketing, or business strategy.
- Exponential technological progress means that large, disruptive changes can occur within a few years (e.g., a “country of geniuses” in a data center by 2027), a concept most people struggle to grasp.
- In a power‑law world, the top 1% of performers receive vastly outsized rewards (such as multi‑hundred‑million offers), making the new “average” effectively the 90th percentile for most workers.
- This skewed reward distribution affects not only employees but also individual builders and creators, who constantly see a small elite capture the majority of value and recognition.
Sections
- Untitled Section
- AI Power Law in Product Success - The speaker explains that in a vast market, only genuinely useful, clearly communicated products thrive, and AI amplifies small skill advantages into a power‑law distribution where top performers capture disproportionate value.
- Choosing Your Career Power Curve - The speaker advises dedicating time and passion to a specific power curve—such as product management or engineering—to build networks, capture AI‑driven opportunities, and secure a top‑percent niche before power‑law dynamics intensify.
- Embracing a Power‑Law Mindset - To accelerate any career, product, or team, you must view the world through power‑law dynamics and deliberately invest in the few high‑impact curves that generate outsized results.
Full Transcript
# Power Law World in the AI Age **Source:** [https://www.youtube.com/watch?v=MYK0d5ikeZw](https://www.youtube.com/watch?v=MYK0d5ikeZw) **Duration:** 00:10:03 ## Summary - The AI era has shifted the world from average‑based norms to a power‑law environment where outcomes are driven by extreme nonlinearity rather than the median. - Traditional workplace metrics—promotability, software fit, buying committees—are rooted in average performance, but those frameworks no longer apply to talent, product development, distribution, marketing, or business strategy. - Exponential technological progress means that large, disruptive changes can occur within a few years (e.g., a “country of geniuses” in a data center by 2027), a concept most people struggle to grasp. - In a power‑law world, the top 1% of performers receive vastly outsized rewards (such as multi‑hundred‑million offers), making the new “average” effectively the 90th percentile for most workers. - This skewed reward distribution affects not only employees but also individual builders and creators, who constantly see a small elite capture the majority of value and recognition. ## Sections - [00:00:00](https://www.youtube.com/watch?v=MYK0d5ikeZw&t=0s) **Untitled Section** - - [00:03:06](https://www.youtube.com/watch?v=MYK0d5ikeZw&t=186s) **AI Power Law in Product Success** - The speaker explains that in a vast market, only genuinely useful, clearly communicated products thrive, and AI amplifies small skill advantages into a power‑law distribution where top performers capture disproportionate value. - [00:06:37](https://www.youtube.com/watch?v=MYK0d5ikeZw&t=397s) **Choosing Your Career Power Curve** - The speaker advises dedicating time and passion to a specific power curve—such as product management or engineering—to build networks, capture AI‑driven opportunities, and secure a top‑percent niche before power‑law dynamics intensify. - [00:09:52](https://www.youtube.com/watch?v=MYK0d5ikeZw&t=592s) **Embracing a Power‑Law Mindset** - To accelerate any career, product, or team, you must view the world through power‑law dynamics and deliberately invest in the few high‑impact curves that generate outsized results. ## Full Transcript
One of the greatest truths of the AI age
is that we live in a power law world and
most of us don't realize it yet. I'm
going to explain what that means. If you
are living in any kind of traditional
working world, your entire working
universe is defined by averages. You are
defined relative to your promotability
by the 50th percentile performance for
your particular job role. Your software
competes in the market place
traditionally based on whether or not
it's better than average and is a better
fit for the customer need than an
average or replacement level fit. That
is literally what the entire concept of
buying committees in B2B software is all
about. It's finding the fit that is best
given a narrow distribution of use
cases. We don't live in that world
anymore. We don't we don't live in it
for talent. We don't live in it for
building. We don't live in it for
distribution. We don't live in it for
marketing. We don't live in it for
business. And increasingly in our
personal lives, we don't live in it. But
we don't really think it through. We
aren't used to it. Our brains are not
processing nonlinearity. Well, what do I
mean by that? I mean that people come up
on stage in San Francisco and they say
we're going to have a country of
geniuses that lives in a data center by
late 2027. Most of us have no idea what
that means. And if we do have a sense of
what that means, we don't know how it
changes our world. And we can't compute
the idea that the progress or pace of
change might be that great between now
and the end of next year or the end of
two years from now. It's like 16 months
from now. No way. We can't. No, that's
not happen. What if it does? Regardless
of whether it happens in 2027 or 2028 or
2030, the point is not exactly when the
date is. The point is that we live in an
exponential world. We live in a power
law world. Let me give you some examples
of what that looks like. In a power law
world, the top 1% of performers in a job
family will get disproportionate
rewards. Ridiculously disproportionate.
Like Mark Zuckerberg calls you up and
offers you a hund00 million rewards. By
the way, that really happened. That's
actually happening now, right? Like you
can see in the world that we live in
today, job families are not getting
rewarded according to any kind of normal
distribution, according to any kind of
law of averages. This is an power law
world where the top 1% are reaping
disproportionate rewards. I'm not
talking about Wall Street here. I'm
talking about talent performers in the
workplace, product managers, engineers,
even executives who can lead technical
teams. They are getting rewarded based
on extraordinary ability. People who are
sort of the new average is 90th
percentile, right? People who are below
90th percentile roughly speaking end up
feeling below average because of the
disproportionate rewards that acrew to
the top 10%. This is true for builders,
too. If you're an individual who is
building a project, I hear this all the
time. I guarantee you someone is
building your project and I don't care.
I don't care. The world is a big place.
There's like 8 billion of us. Someone is
building your special project. I don't
care. I care if your project actually
works, if I can use it, and if it solves
a problem for me. And if it does, and if
you can communicate that in a way that I
can understand, well, you have a chance
in the marketplace. There's a power law
world for product distribution, too. If
you are in the 1% of product
distribution as a founder or a builder,
if you're Peter Levelvels, right, that
very famous solo founder and vibe coder,
he can vibe code something over a
weekend and he can immediately make a
lot of money on it. And it's not that
he's cheating at the game. It's that the
game reinforces top performers. It's
that the game is wired so that AI
reinforces tiny disparities in skill
sets and exaggerates them. If you prompt
just a little bit better than the person
sitting next to you, you are going to
get significantly more work done because
you actually have AI as an accelerator.
And accelerating AI in a slightly more
correct direction makes a big big
difference. And that difference is
growing over time because the models
that are powering you are getting better
all the time. We are stacking power
curves here. People don't think about it
that way. I promise you. They don't
think about their products that way.
They don't think about the idea that
there will be disproportionate value
acrewing to a product that leverages AI
correctly, that is distributed
correctly, that has the right product
evolution for where the market is going
and where the market's needs are with
AI, and that those kinds of products
have the chance to be overnight breakout
successes in a way that we haven't seen
before. If you want to know why Lovable
is the fastest company to reach a
hundred, that's why. They figured out
the intersection between traditional
software and AI in a use case that
unlocked a whole new generation of
builders. And tada, $100 million later,
in just a few months, they're off to the
races. The world is a power law world,
and AI is making it more and more and
more of a power law steep curve. We are
not going back to the old world. We are
not going back to the world of averages.
If I can give one piece of career advice
to you, it is believe that you are in a
power law world and act accordingly.
Plan your strategy accordingly. And by
the way, if you think I'm not in the top
1%, this is so disempowering, you have
to realize in a power law world, moving
up gives you disproportionate rewards
wherever you are on the curve. And so if
you move from a 40th percentile
performer to a 70th percentile
performer, you get much more than 30
percentage points in gain there. You're
actually moving steeper up the curve.
And ditto if you move from 90 to 95%. Do
the rewards get greater as you get
closer and closer to the peak of your
profession? Yes. Is it also true, this
is not always popular, but it's true. It
is also true that you get rewards for
just sticking with a subject and being
persistent with it. People think that
moving to the top of the skill chart is
something that requires tremendous luck
and living in San Francisco and
everything else. Look, there's some
luck. There's some San Francisco
connections, but a lot of it is time on
station. It is time focused and obsessed
over your subject that you care about,
that you want to become one of the best
in the world at. and just caring about
that. There are plenty of people who
work at Anthropic, who work at OpenAI,
who will tell you they did not go to
school for this. They obsessed over it.
And so, it is actually easier than you
think to move along this curve. It is a
function of your willingness to put time
and passion in. And as you put time and
passion in, you are going to discover
connections. Those connections are going
to start to become networks. They're
going to give you options as far as
where you want to be in the world. and
you are going to get chances to connect
with companies and products that you
didn't think were possible. But it's a
function of singular obsession with a
particular power curve that you want to
drive. So do you want to ride the
product management power curve? Do you
want to ride the engineering power
curve? And those are big broad ones like
you can get very fine grained. And I
actually think just like I would
recommend a startup focus on a specific
problem. Focus on a specific power curve
you want to drive in your career as a
builder. If you're trying to find a
product niche, it's the same thing. You
are basically looking for a niche in the
world where you can compete to be in the
top 1%. It is it is like a universal
rule. It is how the world works now. And
AI is exacerbating that. AI is pushing
that forward. AI is driving it faster.
It will be more of a power law world
next year than it was this year. So if
you think it's hard now, well, one, you
can control that to some extent, and
two, it'll be harder next year. So maybe
start to take it seriously this year.
Does that make sense? Power laws are
here to stay. You can't control that. AI
is here to stay. That's not going
anywhere. But you can decide where on
that power curve you want to be to a
much greater extent than you might
realize. And that is a fractal insight.
It's an insight for companies. It's an
insight for individuals. It's an insight
for teams. It's an insight for products.
Everything is running on a power curve.
And if we have in our heads that we live
in a world of averages, we're not going
to be competing. and we're going to end
up way way way below where we should be
in that power law world. We're going to
be in the bottom 40% or whatever and
just not achieve the results we want.
And if you think about the big headlines
that happen about why AI didn't get
adopted by this company or or why this
strategy failed with AI, so often if you
dig the mindset and the culture of the
organization that was doing this change
or the even if you're sitting around,
you're getting a drink with someone and
they're frustrated because of sort of
the way AI is changing their job family.
If if you peel that too, if you if you
dig into that insight, what you see is
that people aren't getting to grips with
the fundamental exponential nature of
the problem space now. And if they did,
they would think about the problem space
differently. Now, I'm not here to tell
you that if you're talking with your
friend over drinks and they're two
drinks in and they're complaining about
their job and you say, "It's a power law
world. Nate said so," they're not going
to throw a beer in your face. they
probably will, which would be justified,
but it's still true and there are
probably gentler ways to put it. You
know, you can say something like, "Hey,
I know that we don't have guarantees
now. I know the world has changed, but
let's take a minute and let's think
through what our options are now that we
understand how power laws work." And
they might still throw a beer in your
face, but it's still true. So, power
laws are a thing. There is insight here
for companies. There's insight here for
teams. There's insight here for
individuals. if you're willing to hear
it. It is one of the things that I find
people often do throw metaphorical beers
at me for, but I gotta be honest with
you, that is what is actually happening.
And so, if you want to drive your
career, your product, your team,
whatever, that's what you need to do.
You need to look at the world as a power
law world and you need to look at
investing in specific curves and writing
those curves. Tears.