"Maybe we were never paying for intelligence. Maybe we've always been renting computation."
Yesterday, I was taking a dump while watching SpaceX's 13th test flight.
As one does.
Somewhere between watching another rocket leave Earth and seeing yet another post about Grok getting better, I found myself thinking about something completely different.
Not AI.
Not the latest model.
Not benchmarks.
Not who cooked who on X yesterday.
But... computing.
What does computing actually mean in an AI economy?
For years, computing felt invisible. I wrote some code, deployed it somewhere, and trusted that the cloud would magically do its thing. CPUs became faster, GPUs became more powerful, data centers became larger, but I never really stopped to think about what was actually happening underneath all of it.
Then AI happened.
Suddenly every prompt had a price.
Every generated image.
Every line of code.
Every research session.
Every "thinking..." animation.
Somewhere in the world, a warehouse full of GPUs wakes up just to answer a question that started with someone typing a few words into a textbox.
That's when something clicked for me.
The interesting part isn't that AI is expensive.
The interesting part is that compute has quietly become the product.
We like to say we're paying for AI.
I'm not sure we are.
I think we're paying for access to an absurd amount of computation that's been wrapped inside an intelligent interface.
The intelligence is what we see.
The compute is what we pay for.
If every token costs compute... are tokens slowly becoming a currency?
The more I thought about it, the weirder the entire industry became.
Microsoft invests in OpenAI.
Amazon pours billions into Anthropic.
Google competes with Anthropic while also investing in it.
NVIDIA sells chips to every single one of them.
They're partners.
They're competitors.
They're customers.
They're suppliers.
At the same time.
It feels less like companies fighting each other and more like one giant ecosystem where everyone is simultaneously helping and competing with everyone else.
It's honestly funny when you stop and think about it.
Everyone is connected. Everyone is competing.
What happens when everyone depends on everyone else?
I don't think we're witnessing an AI race anymore.
I think we're witnessing a compute race.
Models will continue getting better.
Interfaces will continue changing.
Benchmarks will continue being broken every few weeks.
But underneath all of that, the real question feels much simpler.
Who owns the infrastructure capable of producing intelligence at scale?
Maybe that's why countries are suddenly racing to build data centers.
Maybe that's why energy has become part of AI conversations.
Maybe that's why GPU supply chains make headlines.
The conversation isn't really about AI anymore.
It's about everything underneath AI.
As software engineers and students, I don't think our job is simply to keep learning the newest framework or whichever model released yesterday.
AI already contains an unbelievable amount of human knowledge.
Our job is different.
To build with it.
To question it.
To innovate alongside it.
Most importantly, to make sure we don't slowly outsource the parts that make us human in the first place.
Our curiosity.
Our creativity.
Our culture.
Our judgment.
Because if intelligence becomes abundant...
Those might end up being the only things that remain scarce.
I don't really have a conclusion.
I'm still trying to make sense of all of this myself.
But I keep coming back to the same question.
If computation is becoming the world's most valuable resource, what does that make software engineers?
Top comments (0)