“Let the Devil Take the Hindmost”
Theodore Roosevelt is my political hero.
Not because he was perfect — he wasn’t, and he would have been the first to admit it. And he was very much a man of his time, with all the credits and debits that implies.
What inspires me about Roosevelt is what he did with suffering — and what it made him believe about people.
He lost his mother and his wife on the same day. He was a sickly child who willed himself into physical strength, only to have grief nearly destroy him anyway. He went west not as an adventure but as a man running from a wound he didn’t yet know how to carry.
And he came back believing — more deeply than before — that people matter. That they deserve a fair shot. That a society which discards the many for the convenience of the few is not merely unjust.
It is unstable.
It is eating itself.
We are in that moment again.
The greatest lie of the past fifteen years — told loudly and with enormous confidence — is that a fair society is simply a Darwinian meritocracy. That the market sorts people and their value correctly. That those who are displaced were always disposable. That efficiency is its own moral justification.
Roosevelt heard the same argument from the most powerful men of his era.
He didn’t buy it then.
I don’t buy it now.
And I want to be specific about why — because this is where AI enters the picture, and where the fears so many people carry about it are not paranoid.
They are rational.
An old English phrase captures what is actually happening far better than any boardroom euphemism:
“Let the Devil take the hindmost.”
This is the idea that if you must fall so that I may rise, then too bad for you. This is the Law of the Jungle. “Might makes right.”
This is perhaps the oldest justification for indifference to human suffering ever constructed. And it is alive and well in the way AI is being deployed today — dressed up in the language of innovation and efficiency, but underneath, the same ancient calculation:
Some people are worth protecting.
The rest are on their own.
When the dominant message from the enterprise AI market is efficiency — and when efficiency is defined almost exclusively as headcount reduction — people are not misreading the signal.
They are reading it correctly.
The message being sent, whether intended or not, is that human beings are a cost to be eliminated rather than a capacity to be amplified.
That AI is not a tool in human hands but a replacement for human hands.
And that the people making these decisions have quietly concluded that large numbers of their fellow citizens are, in the language of the spreadsheet, “non-essential.”
That is not an efficiency argument.
That is a values argument.
And it is one with a historical track record we should all be deeply uncomfortable with.
Because here is what that argument always omits:
People are not inputs.
They are the network.
They are the circulation.
They are the multiplier.
An economy that generates enormous wealth while systematically excluding the people who built it is not an effective or efficient economy.
It is a terminally ill one that has not yet received the diagnosis.
The productivity gains flow into GPU clusters rather than grocery stores, mortgage payments, and college tuitions.
The “Circle of Life” breaks.
And when it breaks at scale, the multiplier does not merely stop. It runs in reverse.
You begin to see what I call ghost GDP — economic output that appears in the accounts but never reaches a human life.
But GDP that never touches a human life isn’t really GDP at all.
Because nations are people.
AI is the phase shift of our era — as significant as the railroads, the steel mills, and the trusts that TR ultimately broke up.
The productivity gains are real.
The effectiveness gains are real
The potential is real.
But potential is not destiny.
The railroads did not automatically produce the Square Deal. Someone had to fight for that.
Someone had to feel society beginning to shake and decide to act before the crisis made the choice for them.
The fears people carry about AI are not fears about technology. They are fears about what the people controlling the technology have already revealed about how they see them.
And those fears deserve a real answer.
Not reassurance.
Not retraining programs offered as consolation prizes.
But a genuine structural commitment to fairness.
Fairness is the antidote to many fears. That’s something that Roosevelt really understood.
Not fairness as sentiment or charity.
But fairness as structure — the load-bearing wall without which the entire building eventually collapses.
I realize that this post is not what I’m known for. Which is weird because in a very different time and place it was what I was really known for.
I’m not writing this as a political statement, though one aspect of politics is what you believe put into action.
I’m writing it as someone who has spent a long time watching what happens when the people making decisions stop seeing people, and start seeing only positions and costs on a spreadsheet, ready to be eliminated.
The fears are rational.
The stakes are real.
And the window for getting this right — before the crisis makes the choice for us — is not permanent.
We are not disposable.
We are the entire point.
