When I say AI in this post, I mean the current wave: LLMs, agents, coding assistants.
I have had a theory going around in my head for a while, and I think it explains something a lot of us have been feeling without knowing how to put it.
It goes like this: what you know today is worth less than it was yesterday.
I looked for an analogy that would help explain it, and the name came from there.
Knowledge inflation.
To tell it I need an example.
Let's take a developer with years of experience behind them.
Their knowledge, measured across two different subjects:
- Building software: sits at 7, on a scale from 1 to 10.
- Psychology, from scratch: 1, maybe 2.
The numbers are there to picture the idea more clearly.
That was the scale.
The scale stretched
With LLMs you can go much further than before, so the scale stopped ending at 10.
That knowledge did not move, it is still at 7. What moved was the scale:
- Before: 7 out of 10, which is 70%.
- Today: 7 out of 15, which is 47%.
The floor rose too
If you open a psychology book today with a model beside you, it gives you a base and gets you to a 4 or a 5 much faster. You start almost from there.
And it holds for any subject. Practically everyone starts higher up.
And this has already been measured: the tool lifts the beginner far more than the veteran.
It makes sense, because it is not giving the same thing to everyone:
- To the beginner it gives knowledge, something they did not have.
- To the one who already knows it gives volume, automating what they already knew how to do.
A 7 in practice
Let's put it in real work, it lands better that way. Here is a rough case, with round numbers:
- Before: five endpoints with their validation, their queries and their tests were a week of work. That was the 7.
- Today: the model ships them in an afternoon and that leaves you at 5. From 5 to 7 comes from your judgment and your vision.
What used to mark the 7 now marks the floor.
The trap
The bar for being an expert was 7. So this happens:
- You start at 5, because the floor rose.
- You look at the 7 and feel like you are two points away.
- It looks close.
But that 7 was the bar on the old scale. Today's bar is much higher up, and the same tool that lifted you to 5 is the one that moved it.
And since it looks so close, going for the shortcut makes sense. If you are two points away, discipline and years of practice do not seem necessary.
The shortcut gets you to 5. From 5 onward it takes what it always took: study, practice, consistency and experience.
What did get better
Getting to a solid base used to take years of study, of learning along the way and of accumulated experience. Today that base comes much faster, and that is great news for anyone who wants to get into something new.
There is a fair objection here: if the tool helps whoever knows least the most, everyone should end up at the same point.
The same thing happened with the calculator. We have all had one for decades and it did not make us equal at math. It erased arithmetic as a difference, and the difference moved to knowing which calculation to run.
The tool does not make people equal, it moves where the difference lives.
That is why two people holding the same thing end up in different places: whoever already knows is the one who knows:
- what to ask it for
- what to check
- what to throw away
The model multiplies what you already bring.
This is not just software
You take your car to the mechanic. You used to go in blind.
Today you show up knowing roughly what they are going to do, you upload a photo to the model and it explains what to look at and what to ask. You show up with better questions.
And that pushes everyone up, the mechanic and you. When the person asking asks better, the person answering has to answer better.
We are heading toward a world where the quality bar has gone up in everything. We are not there yet, but I think we get there in time.
It already happened with degrees
When almost everyone holds a degree, the degree stops setting anyone apart and the requirement goes up. What asked for a high school diploma moved to asking for a bachelor's, and what asked for a bachelor's moved to asking for a master's.
That already has a name: credential inflation. Randall Collins described it in The Credential Society in 1979, and predicted exactly that escalation.
The difference today is that you do not need years of study to hold one. With a model beside you, anyone carries something close to a degree in their hand.
Something similar happens in Venezuela
Think about bolívares.
- Before: a million made you rich.
- Inflation hit: anyone had a million.
- Today: it takes a lot more zeros.
Are you still rich with your million? No.
Your million did not move, it is still a million. What changed is what it is worth.
In the end
Technology always ends up pushing quality of life upward, and this wave goes in the same direction. In time the quality of everything will be better than what we had.
Today you can learn about anything. But the subject that moves you will always win out, the one that keeps you up at night and takes over your weekends looking to learn more.
That is where you will put the time, and you will end up being what you actually enjoy being. Being whatever you want was always possible, just at a smaller scale.
After all, our time here is finite, and it is still the most valuable resource we have.
And if this is a multiplier, imagine what it can do for your profession once you get everything out of it.
If you are starting from scratch, you already begin several steps ahead.