Putting machines back in their place
Ego to Eureka IV · A Newer Hope · AI Horizons Lisbon · Thursday 15 October 2026 · Unicorn Factory Lisboa
Ego to Eureka is my argument that the hard part of working with machines is not the machines. Across five talks, a podcast and a panel, it comes down to this: building got cheap, judgment did not, and the thing standing between most people and their own good judgment is ego. I am Baz Rao, and this is where the argument lives.
Three steps for Monday
-
Fix the plumbing.
Make what you already run reliable and watched before adding anything.
You can't optimise what you can't describe.
Don't automate dysfunction.
-
Keep the memory.
Every conversation, interview and note, in a store you own, joined to the people and organisations it is about. This is what the CRM never held: not what happened, but why.
Nobody can rent what you remember.
-
Build the engines.
One per job you do every day. Built to a spec, measured from where you start, never finished.
An engine works. But it's never done.
Before you build anything for anyone, point it at your own week.
The argument, in three moves
One. The resistance is not ignorance
Somebody mentions AI in a meeting and the most capable person in the room folds their arms. That pause is not stupidity. It is intelligence protecting itself, and the people who resist hardest are usually your experts, protecting a craft they spent years building.
Two. Expertise became a liability faster than anyone expected
In the beginner's mind there are many possibilities; in the expert's mind there are few. Four years into any discipline you have the right answers, and they have quietly become a wall.
Three. What got cheap was building
The demo that felt like magic in 2024 is the floor in 2026. Anyone can build anything in an afternoon, which moved the interesting questions: what is worth building, and who is actually in control.
Four terms this work needed
-
Ego to Eureka
Ego to Eureka is the argument that the hard part of working with machines is not the machines.
Coined March 2024.
-
Orchestrator economy
The orchestrator economy is the shift from doing work to directing the engines that do it.
Idea set out 4 June 2025. Phrase fixed for the panel of 17 September 2026.
-
Vibe to viable
Vibe to viable is the distance between a prototype that demos well and a system a real team can rely on.
Coined 2026.
-
Build it for yourself first
Build it for yourself first is the rule that you point a new engine at your own work before you offer it to anyone else.
Stated 18 September 2025.
What I do about it
I build these for a living, inside recruitment businesses. I run a company called Sittin' Pretty, which does that work. It has its own site and it does not belong on this one, so this is the only place I mention it.
Who is arguing
I am Baz Rao. I build AI systems inside real businesses for a living and I host Ego to Eureka, which is where I argue that the hard part of working with machines was never the machines. Five talks, a podcast and a panel, all circling the same claim: building got cheap, judgment did not, and ego is what stands in between. I live in Lisbon.
Follow the argument.
Speaking enquiries and disagreements go the same way; I answer the disagreements first.