How AI Expands Human Capability [Raw Session]

00:00 — How AI Expands Individual Capability

Hey, welcome back to Slow Builds.
So I’ve done a few videos now about how people, in my mind, are underestimating AI and what it’s going to do.
The first couple of videos are around jobs, transportation. Obviously, I include medicine and law and manufacturing, and also the middle class.
The more I think about it, the more I think there is one part of this that people may be underestimating even more than job replacement.
And I’ve done other videos about how AI is widening the path.
I have another video—I’m not sure if I did it or not—about the happy side of AI.
And I’ve done, like, we’re about to see things we’ve never seen before and what it’s going to unlock.
So that’s what this one is more about.
It’s what AI allows one capable person or an extremely small team—like a developer, accountants, lawyers, safety inspectors, doctors, researchers—all these people with expertise that they already have, and how it’s going to unlock their potential.
Because right now, the problem is that expertise has always been attached to human limits: time, attention, memory, travel, fatigue, just the amount of hours and amount of work that needs to get done, and how much capacity a single person can handle, how many locations they can visit, how many files they can read, and how many clients they can deal with.
And I think AI is starting to separate expertise from those limits.
And once you start looking at it that way, every industry starts to look a little bit different.

01:45 — Separating Expertise From Human Limitations

So the core idea is, like I said, AI separates expertise from human capacity, human limitations.
Today, experts are constrained by all these things.
Like I just went through: how many hours are in a day, how much attention one person has, the memory, what systems you’re using.
AI starts removing all of that.
Not all of them. The person still has to know what they’re doing.
They still need judgment.
They need the experience.
They still have responsibility for the result.
Someone has to be accountable, but the amount of work surrounding that judgment can start shrinking.
Then every example becomes the same story.

02:31 — Accounting Practices Can Handle More Clients

So let’s take accounting for one example.
An accountant can handle more clients, not because they are working 18-hour days.
AI now can handle more intake, categorizing, filling out forms, reviewing transactions, flagging and preparing questions, doing follow-ups, organizing everything ahead of time.
So then the accountant is just left more as a reviewer and advisor, the final stamp of approval, and that changes the size of the practice they can operate.
Maybe they take more clients.
Maybe they can serve smaller clients who once were unprofitable.
Now all of a sudden, because it takes less time, now, you know, it is profitable, or at least it breaks even and allows them to help serve smaller clients, help out the less fortunate, or the ones who normally can’t afford to have that high expertise.
Now they can be managed and helped through a different system.
Their capacity expands.

03:29 — Small Law Firms Gain More Capacity

And in law, obviously, one lawyer all of a sudden doesn’t become smarter.
They just spend less time organizing and searching.
Less time reviewing documents, drafting up first versions of arguments.
They can spend more time making legal decisions, more time building those arguments that they need to put out there, and having them reviewed faster.
Actually, they’re more or less using AI to help them create the arguments, and they can review them to quickly dismiss or modify them.
More time understanding the actual case that they need to go through.
Think about a small law firm trying to take on a class-action lawsuit.
The barrier is not only whether they have a good lawyer. It’s whether they have the time and the staff to handle the discovery and research and the massive load of data that’s going to come their way, and the amount of witnesses and victims or defendants that they have to go through, interview, categorize, and capture all the information.
AI can help reduce all that burden.
It can search the documents and find related stuff, inconsistencies, summarize.
They still have to argue the case.
They still have to verify everything.
But now a small firm suddenly has the capabilities and the capacity to do more with fewer associates.
They don’t need to have a massive support staff to help them go through these massive cases.
Maybe now a small firm can compete with the larger firms without requiring the heavy overhead and cash flow and capital just to get through it.
The expertise did not necessarily change.
The capacity around it did.

05:24 — One Developer Can Operate Like a Small Team

And this goes, for me, all about software because that’s what I do and how I live.
So I see this directly in my situation.
As one developer, I don’t just write code faster.
And I’m not a great designer.
I have ideas. I don’t know how to articulate them and put them out there.
But now AI can help me.
It does help me write the code faster.
It does help me come up with the designs.
I can give it examples, and it can prototype stuff for me.
It can help me build the architecture because I suck at architecture.
But now it helps me push things out to my servers, my host, using Telegram with OpenClaw, my Ubuntu server.
It’s building stuff for me.
It’s connecting it all, setting it up, doing whatever it does.
And it makes it work because it does the deployments, does the tests, it does the walkthroughs.
It helps me find all the things that normally take me forever or take me having to deal with my team or a team, and now it’s doing it for me.
It’s doing marketing.
It’s doing copy.
It’s doing UI.
So now a developer all of a sudden can compete with a team.
They don’t need a full team to start, to get an MVP up and running, and see if something’s viable.
Now all of a sudden, all that cost is lower.
It changes what one person can realistically own and operate.
The important shift is not only faster code. It’s that one individual can start functioning more like a small organization, a small team.
And then as those products grow, they can build more products, they can build on the products, they can hire people to help make the product more robust.

07:12 — Individual Capacity Becomes Organizational Capacity

So then this is where I think the bigger idea appears in this whole video idea right here.
AI is not only helping people finish tasks faster. It’s changing the size of the operation one person can manage.
One accountant starts operating like a large practice.
One lawyer starts operating with the research capacity of a big firm.
One developer starts operating like a small product team.
One inspector, like a health inspector or safety inspector, starts overseeing an entire region.
One doctor starts monitoring and reviewing far more information.
Individual capacity starts turning into organizational capacity.
That can create more opportunity.
Smaller firms may be able to compete.
Individuals may be able to build things that used to require capital and staff.
Experts may be able to serve more people.
But it also changes staffing.
It changes pricing and competition.
It changes how many people are needed around each expert.
I do not know exactly where that lands, but I think that shift is much bigger than saying AI makes people more productive.
Because it’s what you do with that productivity.
You optimize your time.
You open up more doors.
You create more opportunities.
But what do you do with it?
Do you need people to help you with it?
Do you need different skills now from different people to help you expand?
Do you take on more clients, or do you become more specialized?
There’s a whole new world opening up for just individuals alone, I think.

08:42 — Safety and Health Inspectors

So to bring this, like I mentioned, to safety and health inspector-type things.
Safety is a useful example because it moves outside the normal office work that I’ve been talking about with papers and computers and coding.
Imagine safety officers are responsible for construction sites, industrial facilities, restaurants, food production.
Right now, the person has to physically go to those places.
They walk through.
They check everything.
They look for hazards.
Then they leave.
Maybe the location does not get inspected again for weeks or months.
Now imagine a safety officer spends time training a system on what they look for.
And I’m sure this already exists to some extent, but let’s bring it to a whole new level with cameras and drones and Optimus.
Let’s just put a robot in there.
The specific machine is not really an important part.
The important part is that the officer’s expertise can be turned into a repeatable inspection process.
Your AI profile.
I’ve had a whole video on that.
Imagine taking one person, their knowledge, their skills, their likes, dislikes, their nuances, all the different things that they notice and figure out.
And now they can repeat that endlessly across multiple locations.
Let’s take that a step further.
Let’s take multiple health inspectors, all their profiles merged into one.
So that’s where a company has a health inspector profile that is built up of many profiles.
That’s a whole other video.
Let’s table that for a different time.
But we bring that back into the system.
So in this case, the Optimus, the drone, the whatever is over there, and the system is continuously checking the basics.
It’s watching for blocked exits, missing safety equipment, unsafe movements.
It’s checking the temperature, equipment, all the procedures being followed.
Then the safety officer reviews the exceptions.
They do not have to personally watch every minute of every site.
They are alerted when they need their judgment, when their expertise is required.
That means that one officer might oversee an entire district, even more, possibly.
And the surprising part is that safety could improve.
Instead of one inspector every few months, the location is effectively inspected every single day, every minute of every day.
The human expert is still responsible for those calls, but the repetitive observation becomes continuous.
AI starts separating expertise from physical presence.

11:08 — Medical Research and Pharmaceutical Discovery

Let’s bring that into medicine.
It’s a natural flow.
This may be where people underestimate the impact the most.
Start with the research side of it, with pharmaceuticals and medical research and scientists and laboratories.
Drugs, clinical trials, imaging, proteins, patient outcomes, research papers.
A human team can only examine so much.
They can only hold so many variables in their head.
AI can move through huge amounts of data.
It can recognize patterns that we would never, ever find.
It compares outcomes, relationships, finds combinations, tries all the different situations with the different variables in controlled situations.
It manages all this stuff, and it does—I don’t even want to say a number—but the amount of compute and comparisons and patterns and whatever else it can do and equations is unbelievable compared to a human.
And that could speed up how quickly researchers rule things out and how fast they find new things and find solutions.
Understand why something failed.
Understand why something worked.
The breakthrough still has to be tested, verified, repeated, and approved, but the search itself may move much faster.

12:32 — Doctors, Nurses, and Continuous Monitoring

And outside of medicine in the physical world, then bring that to the same as the safety inspector.
Now nurses and doctors all rely on monitors, tests, notes, observations, checks.
But humans cannot continue to watch every patient at every single moment.
Nurses move around.
Doctors manage many cases.
Everybody’s tired.
Everyone has limits.
Now imagine AI continuously watching the data.
Heart rate.
Oxygen.
Temperature.
Blood.
Comparing that with lab results as they come in.
Medication.
Their movement.
Just overall reviewing the entire patient’s life and current situation and all the things that are happening at once for every single patient.
It looks for all the patterns and combinations.
Maybe one number by itself looks normal, but several small changes together suggest the patient is deteriorating.
There’s a problem.
A nurse may eventually notice it.
AI may be able to flag it much sooner.
The system alerts a nurse or doctor.
It does not make the final decision.
It tells them, “Hey, you need to pay attention over there. Something changed. This pattern looks unusual. Review this now.”
The nurse does not disappear.
The doctor doesn’t go away.
They are alerted sooner with more information.
That allows us to focus on patients who need them the most.
It removes some of the constant mental load of trying to watch everything at once.
We may still need more nurses.
We do need more nurses, and we do need more doctors.
It doesn’t magically solve the shortage that we have.
It helps alleviate their stress, their workload, the patient stress.
It allows them to monitor better.
It increases their capacity.
So it could allow the people we already have to do more work with less stress.
Less time reviewing routine information.
More time caring for people.
More time making decisions.
More time responding when something actually matters.

14:34 — The Broader Pattern

So the broader pattern here is the accountant reviews instead of manually processing everything.
The lawyer decides instead of manually searching through everything.
The developer builds instead of waiting for a full team.
The safety officers handle more exceptions instead of driving to every location.
Nurses respond to alerts instead of trying to watch everything.
The researcher follows promising leads instead of manually searching through every possible combination.
The expert remains important, maybe even more important, because the final judgment still matters.
Responsibility matters.
Trust still matters.
Experience still matters a lot.
You need to know what you’re looking at.
You need to have that past history of seeing it before, knowing what happens if it’s ignored, knowing what happens if you do the wrong thing.
But the limits around that expert still start changing.
For hundreds of years, expertise has been limited by one person’s time.
I think AI is starting to break that relationship.

15:40 — The Optimistic and Uncomfortable Sides

Now there’s an optimistic side to this.
Better care, faster research, safer workplaces, experts serving more people or taking on more tasks, more products getting built, more cases being taken on.
But there’s also an uncomfortable side.
If one person can now do more, what happens to the people who used to support that work?
Does that accountant grow the practice and hire more people, or handle more clients with the same staff?
Does the law firm take on more cases and expand, or increase revenue without expanding the team?
Does the developer use leverage to start a business, or does the company decide it needs fewer developers?
Does a hospital use AI to reduce stress and improve care, or use it as an excuse to stretch staff even thinner?
The technology does not answer those questions.
The organizations using it do.
That’s why this is not automatically good or bad.
It’s leverage, and leverage depends on who controls it and how they use it.

16:46 — What One Person Can Do With AI Beside Them

So I think this is the part people may be missing.
We keep looking at AI and asking what it can do.
I think the bigger question is what a person can do once AI is beside them.
What can one accountant manage?
What can a lawyer take on?
What can a developer build?
What can inspectors oversee?
Doctors—what can they notice with nurses?
What can researchers discover?
I do not think AI is replacing expertise.
I think it’s removing some of the limits around those experts.
And when that starts happening across every profession, it does not only change jobs.
It changes businesses, competition, research, how widely one person’s knowledge can be applied.
It changes what one human is capable of producing.
That is another reason I still think people are underestimating AI.
Not because every prediction will come true.
Not because all of this happens tomorrow, because it’s going fast, but it still takes time.
But because the relationship between expertise and human capacity is already starting to change.
And I do not think we really fully understand what follows from that yet.

17:53 — Closing

Like, we are on the cusp of—we’re in the infancy of AI.
I’ve done a video on that, how we’re watching it grow, how we’re seeing it morph and change, and not knowing where it’s going to go, where it will end up eventually.
Even then, it’s still going to evolve.
We’re seeing how people are going to use it.
There’s a happy side and there’s a bad side.
But I think there’s more upside than down.
And I think this is one of them, where that doesn’t just widen the path for individuals who don’t need to be an expert because AI can be their team.
This is taking an expert and having AI increase their capacity and capabilities.
And it’s an exciting time that we’re going through.
An exciting time.
Alright, thanks for watching.