A new graduate, a gen AI artist, came to me on onboarding and admitted he didn't understand his own role with AI. The world was changing around him, and he had no idea what his part in that massive change was supposed to be. That was genuinely shocking to hear from someone starting their career right at the centre of this shift.
It revealed the real gap: not a skills gap, but a framing gap. Nobody has told the newest members of the workforce what they are actually doing when they work with this technology.
Your real job: build little AIs
Here is the reframe that needs to be communicated to every knowledge worker. Your job is to make a record of your decisions, and then use AI to create an AI that does your decision. You automate a micro decision. You create tiny little AIs, in an event-based architecture, that act on the information coming to you.
That is the whole game. The value you add is no longer just making the call — it is capturing how you make the call, and encoding that judgment so it runs without you in the loop every time.
Your job is to make a record of your decisions and use AI to create an AI that does your decision — you automate a micro decision, one tiny AI at a time.
AI will not take all the jobs
The counter-message matters too. The claim that AI will take all the jobs is absolutely nuts. Books belong to the same technology class — information infrastructure — and they were a net benefit, not a catastrophe. AI is the same class of thing, and I'll explain that comparison in more depth in the future.
Like books before it, this is infrastructure that amplifies what people can do rather than a force that replaces them entirely.
Depth plus AI means working across disciplines
What does this unlock in practice? When you have subject matter depth and you combine it with AI, you can work across a lot of parallel disciplines and coordinate far better with your team. The technology handles the scale; your judgment handles the direction.
So the advice to the new grad — and to anyone feeling that same uncertainty — is to stop asking whether AI will replace you, and start asking which of your decisions you can record and automate next.
Transcript
M I onboarded a young gen AI artist and he told me he didn't understand his own role with AI that the world was changing but he didn't understand his role in that massive change. And this was quite shocking to me. New grad Gena artist didn't understand his role with AI and so there needs to be a reframing that needs to be uh, communicated of what you're doing with this technology. In a knowledge worker role your job is to make a uh, record of your decisions and use AI to create an AI that does your decision. You, you automate a micro decision is what you do and you create tiny uh, little AIs in event based architecture to automate these decisions on information that you would get. So AI will take all the jobs is absolutely nuts. Uh like books, it's a net benefit. They, they belong in the same technology class information infrastructure and we'll talk about that and explained that in the future. Um so really when you have subject matter depth uh and you combine it with AI, you can work across a lot of parallel disciplines and coordinate a lot better with your team.
Frequently Asked Questions
According to the speaker, your job is to make a record of your decisions and use AI to create an AI that makes those decisions for you — automating micro decisions in event-based architecture.
No. The speaker calls that claim absolutely nuts and compares AI to books: both are information infrastructure, and like books, AI is a net benefit.