A colleague just got back from the NVIDIA conference in San Jose…
And he said that despite the hype, the consensus was clear that we’re accelerating.
The tools are getting better, faster than expected.
Agents are handling entire chains of work.
Video, content, analysis, and synthesis are all moving toward a marginal cost of zero.
It’s not theoretical anymore. It’s here.
And if your work lives on a screen, you can feel it.
AI researcher Andrej Karpathy recently mapped this out in a way that’s hard to ignore.
He ran an analysis across 143 million jobs in the U.S., scoring how exposed each one is to AI.
1 being least exposed, 10 most.
- Copywriter – 9
- Data analyst – 9
- Software developer 8–9
- Plumber – 2
- Construction laborer – 1
Higher-income, knowledge-based jobs are more exposed than lower-income, physical ones.
If your job can be done from a laptop, it’s highly exposed.
As Tom Bilyeu put it: “The knowledge economy is the exposed economy.”
Writers, analysts, designers, developers, the very categories that were once positioned as insulated from automation are now sitting closest to the center of it.
There’s a strange inversion in that.
The paths meant to create security, the jobs we were told were safe, are exactly where the pressure is building.
And yet, if you look closely, the data doesn’t point to disappearance so much as transformation.
The people who learn to work with these systems move faster, think bigger, and produce at a level that used to require teams.
Some people will adapt quickly.
Others will keep doing the same thing and slowly feel the ground move underneath them.
What’s more interesting, at least to me, is what happens at the edges of that shift.
As more output is easily generated, a different kind of scarcity emerges.
Elon Musk made an observation recently that’s been sitting with me.
As digital media becomes effectively infinite, the scarce commodity is now live events.
People want experiences that cannot be replicated or replayed.
We’re aching for things that only exist in a specific place, at a specific time, with a specific group of people.
There’s something about the emotional texture of a live environment that doesn’t compress well.
You can document it, record it, try to recreate it.
But you can’t recreate what happens in a room when people are actually paying attention, responding in real time, building something together that didn’t exist an hour before.
You can’t replicate presence.
Gary Vaynerchuk has been saying in-person experiences are becoming a kind of luxury.
That says something about how rare they are now.
Most interaction is trending toward convenience: faster, smoother, easier to access, easier to exit.
And in that context, anything that requires you to show up physically, to stay present, to engage without a layer between you and the other person, begins to carry a different weight.
John Maynard Keynes saw a version of this almost a century ago:
“The economic problem may be solved… and mankind will be deprived of its traditional purpose… We shall once more value ends above means and prefer the good to the useful.”
You can read that now and see a faint outline of what he was pointing toward.
A world where the primary constraint is no longer production, but orientation.
Taste. Judgment. Trust. The ability to move a room.
The question shifts away from how to do more, and toward what is actually worth doing.
This is part of why we do AI Cafés at KINN.
A space for people to sit together and talk about what’s changing while it’s still changing.
There’s something useful about being in a room with other people who are navigating similar questions, hearing how they’re approaching it, noticing what resonates and what doesn’t, and slowly refining your own sense of direction in the process.
Because if a larger portion of the execution layer is going to be handled elsewhere, then the leverage moves into areas that are harder to systematize.
How you see.
What you prioritize.
Who you spend time with.
And over time, the environments you place yourself in start to matter much more than the tools you use.
Talk soon,
Oliver