THOSE CLOSEST TO THE WORK MUST BE CLOSEST TO THE DECISION (1) On AI, Labor, Proximity, and Human Standing

THOSE CLOSEST TO THE WORK MUST BE CLOSEST TO THE DECISION (1)

On AI, Labor, Proximity, and Human Standing

Artificial intelligence is entering the workplace
with extraordinary speed.

It can diagnose patterns, organize schedules, measure productivity, automate repetitive tasks, assist professionals, predict outcomes, and reorganize work itself.

But before asking what the machine can do,
we must ask another question:

Who gets to decide what the machine should do?

The answer cannot belong only to those
who own the technology.

It cannot belong only to executives.
It cannot belong only to engineers.
It cannot belong only to investors.

The nurse standing beside the patient
knows something the algorithm does not.

The technician repairing the equipment
knows something the executive dashboard does not.

The warehouse worker knows something
the productivity graph does not.

The teacher standing before the classroom
knows something the model does not.

The person closest to the work
possesses a form of knowledge that cannot simply be
downloaded from above.

It is the knowledge of proximity.
And proximity should confer standing.

Those closest to the work must be closest to the decision.

The emerging AI framework negotiated
between Kaiser Permanente and its union partners
offers an important illustration of this principle.

Its significance is not merely that workers receive
protection after technology is introduced.

The more important idea is that workers can participate
earlier—in identifying problems, considering technological solutions, implementing systems, and evaluating their effects.

That reverses the usual sequence.

Too often technological change follows this pattern:
Design first.
Deploy second.
Explain third.
Listen last.

The better order is:
Listen first.
Identify the problem together.
Design carefully.
Deploy responsibly.
Evaluate together.

The difference is not procedural.

It is moral.

When workers are consulted only after deployment,
participation can become little more than explanation.

When workers participate before deployment,
consultation can become governance.

That distinction will matter enormously in the AI age.

Because artificial intelligence is not merely another tool when it begins influencing who is hired, how work is scheduled, how performance is evaluated, which tasks disappear, which workers
are monitored, and whose employment becomes unnecessary.

At that point, technology has entered the territory of power.
And wherever power enters, accountability must follow.

The worker must not become transparent to a machine
that remains opaque to the worker.

If AI measures the worker,
the worker should know what is being measured.

If AI evaluates the worker,
there must be a meaningful way to question the evaluation.

If AI changes the job, those performing the job
should participate in discussing that transformation.

If AI produces greater productivity,
workers have a legitimate interest
in what happens to the gains.

And if an automated system causes harm,
responsibility cannot disappear behind the sentence:

“The algorithm decided.”

The algorithm did not invite itself into the workplace.

Human beings brought it there.
Human beings selected its purpose.
Human beings determined what it would measure.
Human beings decided what authority it would possess.
Human responsibility therefore remains.

This is why collective bargaining may acquire
a profound new function in the twenty-first century.

The bargaining table is no longer concerned only
with wages, hours, benefits, and safety.

Another participant has arrived:

the algorithm.
And workers must have standing before it.
But this principle extends beyond organized labor.

It belongs wherever technology affects human beings.

Those closest to the patient should participate
in decisions about healthcare AI.

Those closest to the classroom should participate
in decisions about educational AI.

Those closest to public services should participate
in decisions about automated government systems.

Those closest to the street should participate
in decisions about homelessness.

Those closest to the consequences should not be farthest
from the decision.

> The Man Carrying His House

Years ago, during my homeless ministry near Stanford,
I encountered a man walking with his belongings.

There was no conference room.
No presentation.
No housing-policy report spread across a table.
There was simply a man on the street carrying
what remained of his life.

A bag in one hand.
Belongings gathered together.
His home, in a sense, had become portable.
Around us stood one of the wealthiest and most intellectually sophisticated communities in the world.

Nearby were laboratories, universities, technology companies, expensive homes, investors, engineers, professors, and institutions capable of processing extraordinary amounts of information.

The community possessed enormous knowledge
about homelessness.

Government could count him.
Researchers could study him.
A database could classify him.
A map could locate him.
A budget could assign a cost to him.

A future AI system might even predict
where someone like him would sleep tomorrow.

But standing beside him, another question came to me:
Had anyone asked him where he believed he should go?

That question changed everything.
Because he knew something none of those systems knew.

He knew what the pavement felt like at night.
He knew which places were safe enough to sleep.
He knew where he would be awakened.
He knew where bathrooms disappeared after dark.

He knew what it meant to carry possessions
because there was nowhere secure to leave them.

He knew how far a person could walk
carrying everything he owned.

He possessed no university degree in homelessness.
But he possessed something the university could not award:

the knowledge of living it.
And suddenly the principle became unmistakable.

We had built systems capable of making decisions
about him while rarely constructing tables
where decisions could be made with him.

That is not merely a problem of homelessness.
It is a problem of governance.

The same mistake can happen in the AI workplace.
An engineer may understand the model.

An executive may understand the investment.
A manager may understand the productivity target.
But the worker understands something
none of them possesses:
what the work feels like from inside the work.

That knowledge must have standing.
Because there is a difference
between knowing about someone's condition
and knowing from within it.

The map knows the street from above.
The homeless person knows the street from below.

The algorithm knows the pattern.
The worker knows the work.

Both forms of knowledge matter.
But when the knowledge from below is excluded
from the decision above,
intelligence becomes incomplete.

That encounter near Stanford taught me something
I have carried ever since:

The person carrying the consequence carries knowledge
about the consequence.

And therefore:
The person carrying the consequence deserves standing
in the decision.

For years on the street,
I have learned how easily institutions can speak
about people without speaking with them.

We discuss homelessness.
We measure homelessness.
We map homelessness.
We budget for homelessness.
We regulate homelessness.

Yet the person sleeping beneath the overpass
may be the last person anyone asks:

What happened to you?

That is the danger of distance.
Distance turns persons into categories.
Proximity gives categories names.
And artificial intelligence can intensify either direction.

It can become the greatest instrument of distance ever
constructed—allowing institutions to classify, predict,
evaluate, and govern millions of people
without encountering them.

Or it can help human beings understand problems more deeply while preserving the voice and dignity of those living inside them.

The difference will depend upon governance.

The machine may see patterns from above.
But the neighbor knows what the pattern feels like from below.

We need both forms of knowledge.

The map matters.
The street matters.

The model matters.
The worker matters.

The data matters.
The person represented by the data matters more.

That is why proximity should become
a principle of AI governance:

The closer a person stands to the consequences of a decision,
the stronger that person's claim should be to meaningful participation in making it.

Not because proximity makes anyone infallible.
Not because expertise does not matter.
Not because every decision can be made by everyone.

But because expertise without proximity can become blind,
while proximity without standing becomes powerlessness.

The AI age needs both knowledge and participation.

Both intelligence and accountability.
Both innovation and conscience.

The future should not be something designed by
a small number of people and delivered to everyone else.

It should be something human beings learn to govern together.
So before the machine enters the hospital floor, ask the nurse.

Before it reorganizes the warehouse, ask the worker.
Before it evaluates the teacher, ask the teacher.
Before it transforms a profession, listen to those practicing it.
Before it makes decisions about the poor, sit beside the poor.

And before we call technological change progress,
ask the neighbor who must live with its consequences.

For the deepest democratic principle of the AI age
may turn out to be extraordinarily simple:

Do not make the people closest to the consequences
the people farthest from the decision.

Those closest to the work must be closest to the decision.

Those closest to the suffering must be closest to the remedy.
And those closest to the neighbor must help determine
what the machine is permitted to do.

Pastor Street Gospel Lee
Street GMC Corps
September 30, 2026

https://www.littler.com/press/press-release/fewer-10-us-employers-feel-very-prepared-ais-impact-labor-relations-littler
Fewer Than 10% of U.S. Employers Feel Very Prepared for AI’s Impact on Labor Relations, Littler Survey Finds
Survey of more than 650 executives reveals how AI, multigenerational workforces and communication challenges are impacting labor relations strategies
September 29, 2026

https://www.eurofound.europa.eu/en/publications/all/collective-bargaining-on-artificial-intelligence-at-work
Collective bargaining on artificial intelligence at work

https://www.facebook.com/steven.g.lee1/posts/pfbid02LW4xPDqzfDcr3CQFm66uRHosaFPZngpprNqqSyRzVenXUYsMwcJBB5Axhff7AuL5l