A Conscientious Objection

 Well, I have to hand it to Anthropic. Despite being a profit-making entity, they were actually willing to take one on the chin this past week in their conflict with the Pentagon—and all for a point of principle. 

Of course, I can't applaud them for wanting to work with the Pentagon in the first place. But they did have a certain moral line they refused to cross. 

I can't say I forgive them for allowing their technology to be involved in earlier illegal U.S. military actions—in Venezuela and elsewhere. But at least they balked when the Pentagon wanted to cross their handful of remaining ethical red lines. 

Like that other conscientious objector Olaf, of whom E.E. Cummings wrote in his poem, they eventually said to Pete Hegseth: "There is some shit I will not eat." 

Upon which Sam Altman promptly sidled up with knife, fork, and plate at the ready. "I'll eat anything," he seemed to imply. 

As a result, Antrhopic actually lost business. For making a conscientious choice. And I have to admire them for it—not least because it is so unusual among today's companies to do so. 

John Ruskin once wrote—in Unto This Last—that the root of honor lies in self-sacrifice. 

That's why every society honors soldiers above anyone else. Not because they sign up to kill, but because they sign up to die. 

This is also, Ruskin writes, why no modern society honors the businessperson. It is not because they provide no useful services to society—they do. 

Rather, it is because their ethos generally precludes self-sacrifice. 

Anthropic's actions in recent weeks, though, cut against the grain. They actually required them to forego excess profits for the sake of a moral stand—however belated. 

They engaged in real, if limited, self-sacrifice. 

And so, we honor them. My hat goes off to you, and all those others whose "noble hearts recoil at war."

It may seem strange, as Wyndham Lewis once wrote, to say "So far, but no farther!" 

Why make your AI available for abductions and killings but not surveillance? 

But life is full of ethical line-drawing exercises of this sort. And even if the line's location proves ultimately to be arbitrary—it is far better that it exist somewhere than otherwise. 

"It is quite simple;" Lewis goes on (in his book The Demon of Progress in the Arts): "beyond a certain well-defined line—in the arts as in everything else—beyond that limit there is nothing. Nothing, zero, is what logically you reach past a line[.]" 

That is where Anthropic was headed, if they hadn't pulled up. It's where OpenAI is eagerly headed now—nada, zero, the moral abyss—

A place where everything is permitted, and the sky is the limit. But that sky is "nothing, and is nowhere, and is endless." (To quote Philip Larkin.)