1945: the bomb outran the wisdom to govern it

Oppenheimer watched the first atomic device detonate at a site he had named himself, from Donne. The fireball climbed. Then the sound arrived. Then the heat.

And a line from the Gītā came to him. Now I am become Death.

He had unleashed great power. The institutional capacity to contain it did not exist.

Power without wisdom ultimately destroys.

With God seated in the heart, we operate the world machine

Cultivate the heart, and let divine intelligence flow through it into the management of the world machine.

All three words stand in one line of the Gītā, 18.61.

gita.ekrasworks.com
/gita/18/61

Only about a dozen companies can build one

The most powerful thinking machines ever built will concentrate power.

There is not much disagreement on this. The labs say so as readily as their critics do. The concentration has a physical form.

1. a frontier model
2. trained on specialised chips
3. housed in a data centre
4. drawing the power of a small city
5. funded by capital that perhaps a dozen organisations hold

That is why it cannot be done by many. So it concentrates.

Every age has had to rein power in, and bend it to the common good

Our machines are new, so we assume we are the first required to ask this.

It has been a king, a priesthood, an empire, a merchant house, a bank, a bloc of capital.

Constraining a concentration of power is the oldest task a civilization has. The second is to bend it toward the common good.

The Catholic Church actually built something

Fr. Paolo Benanti. Franciscan, trained in engineering, doctorate in moral theology. He coined algor-ethics, and advises both the Vatican and the United Nations.

He can speak to a technologist in the technologist's own patois, while bringing a theological depth that secular ethicists cannot match.

The space did not stay empty

The space the traditions left has not stayed empty. A secular movement has moved into it.

A secular calculus with no need for ideology, a telos, or moral formation.

One tradition has answered in strength, but alone.
Most have not answered at all.
And a secular movement now occupies the ground the others left open.

Religious institutions are slow

Religious institutions are slow. That is true without argument. They deliberate for years and argue internally for decades.

Beside a technology that moves in quarters, their pace looks absurd.

And the slowness turns out to be part of the answer. An institution that bent to every shift in the weather would have dissolved long ago, for having no centre to hold to.

Some questions have to be refused before they can be answered

1. Should we pay a basic income when AI takes the jobs?
2. Is the machine conscious, can it suffer, is it owed rights?
3. Should we build advanced AI at all: build, or pause?
4. Which fairness metric should the system satisfy?
5. Who owns the inputs and the outputs, and what licence is due?
6. Where should the dial sit between safety and access?
7. Which nation should own the frontier model?

A well-framed question carries the seeds of its answer. A monk asked Joshu whether a dog has Buddha-nature. Joshu said mu.

The answer has the shape of a wheel

The centre must be changeless.

A wheel is the oldest machine that is still at the centre and fast at the rim at the same time.

The only way to move fast and not break things.

Repurpose, do not build from scratch

The wheel needs a hub that is already old, and that rules out building one. An institution with standing is an institution with age, and age is the one thing a founding cannot supply.

So the hub is found rather than founded.

Satkāryavāda: the effect is already present in its cause. Pariṇāma: the unfolding of what was there, rather than the arrival of what was not.

Nothing on this view begins.

The three objections, answered

• Interfaith coordination dilutes to the lowest common denominator.
• Religious institutions are too slow for a technology that ships in quarters.
• This is religious overreach. The governance of a secular technology should stay secular.

A coalition dies as a committee when it makes the many decide the fast things. This design never does. The hub is inherited, so there is nothing beneath it to dilute toward, and the rim is delegated, so action does not wait on consensus.

For most countries this arrives as infrastructure

The dominant model of adoption is the import of Silicon Valley systems together with the assumptions built into them.

One civilization's account of human flourishing, arriving as infrastructure in places whose own traditions hold a different account.

That does not stop being colonial when the owner is friendly, or when the owner is your own nation exporting to its neighbours.

Oppenheimer had wisdom but nowhere to put it

Oppenheimer's failure was not a failure of wisdom. He had the wisdom.

It was a failure of infrastructure. One formed conscience, and no institution to carry it.

We have great power to alter the fate of humanity. We may even have wisdom.

But do we have the institutional framework?