As AI absorbs the execution, the human side becomes the premium part.
Taste. Presence. Judgment. Conviction. Lore. The work only a person can do is exactly the work that gets scarce when machines do everything else. The move is not to automate yourself out of the picture. It is to use AI ruthlessly on the work only machines should do, and protect the time for the things that make you worth following.
The Human Premium.
When something is in abundance, demand for the other aspects rises. As AI floods execution, what becomes scarce and premium is the human side.
Doing the work used to be the hard part. Writing the code, drafting the copy, building the deck, running the numbers. All of that is cheap now, and getting cheaper. The value does not vanish, it moves. It moves to the parts a machine cannot do well: the call you make, the thing you say no to, the room you are actually standing in, the life you have actually had.
Which is why automate everything is the wrong instinct wearing a clever hat. Hand over the human parts as well and you have quietly made yourself the easiest person in the building to replace. The point of the tool is not to do less work. It is to buy back the hours for the work that stays valuable.
Five things AI cannot reliably do.
Protect these. Delegate the rest.
Taste
Curation. Refusing. The editorial judgment that decides what is worth doing.
The feed picks your inputs and the model picks your phrasing, unless you decide, on purpose, what is worth doing in the first place.
Conviction
Calls made. Stakes taken. Defended opinions you would hold under public pressure.
A model hands you the average of the internet dressed as nuance. Only a person names the call publicly and stands behind it with the receipt.
Lore
Lived experience worth telling. An integrated life, captured, never staged.
The flex is the life, not the post about it. A lived life is the one input no AI can manufacture.
Synthesis
Original work. Connecting unrelated domains. Writing the framework first.
Reposting a summary is not thinking. Connecting a lesson from one world to a problem in another, and writing the rule down first, is.
Presence
Real rooms. Real people. The voltage of being fully there.
Parasocial reach is not the same as being in the room. The voltage of full attention, in person, is the thing a screen cannot fake.
Not anti-AI. Not AI-maximalist. The third path. AI as the multiplier that protects and amplifies what is uniquely human.
There are two loud camps. One says AI replaces you, so brace for it. The other says AI is everything, so hand over the keys. Both are wrong in the same way. Both treat the person as the bit you subtract.
The third path is stronger than either extreme. Point the tool at the work only machines should do, so the taste, the conviction, the lore, the synthesis, and the presence finally get the time they deserve. It is not about surviving AI. It is about weaponising the human side with AI as the multiplier. The trade is beautiful, and it is worth defending out loud.
The thesis is not an essay. It is the operating manual.
It runs across the companies I build and operate, each the same shape in a different domain. Monkhouse & Partners, the advisory that gets major companies genuinely up to speed with AI, useful and applied, not hypothetical. Knightox, where AI is used to deepen relationships in community, not replace them. Threshold, where aviation is binary and AI sits as a copilot on trusted data, never the pilot in command. Underneath it all, Monkhouse AI: the personal system that absorbs the mundane operating work so the human hours get protected.
The life is the receipt. The companies are the proof.
The thinking it is built on.
People judge warmth before competence
Fiske, Cuddy, Glick and Xu
We size up whether someone means well before we size up whether they are any good. An argument that cuts against the grain has to feel warm before it gets to be clever.
A model of (often mixed) stereotype content, 2002→Charisma can be taught
Antonakis, Fenley and Liechti
Managers trained in a set of specific behaviours were rated measurably more charismatic months later. Presence is a craft, not something you are born with.
Can charisma be taught? Tests of two interventions, 2011→The new thing stops feeling new
Brickman and Campbell
Excitement drifts back to baseline whatever caused it. The novelty of a tool wears off and only its usefulness compounds. AI gets no exemption.
Hedonic relativism and planning the good society, 1971→Price changes how good something actually feels
Plassmann, O'Doherty, Shiv and Rangel
The same wine, labelled more expensive, did not just get rated higher. It registered differently in the brain. Value is partly made in the perceiver, not only in the product.
Marketing actions can modulate neural representations of experienced pleasantness, 2008→Stories move people, arguments do not
Green and Brock
People change their minds when a story carries them somewhere, not when they lose a debate. Show the thing happening rather than explaining that it happens.
The role of transportation in the persuasiveness of public narratives, 2000→People need autonomy, mastery and connection
Deci and Ryan
Motivation holds up when people have some control, are getting better at something, and are doing it with others. Automate all three away and the work stops being worth doing.
Self-determination theory, 1985 onward→The argument, in full, every Thursday.
Still Human, Thanks for Asking. Five minutes on enterprise AI and the human side of it.
Subscribe