Categories
Community Internet Culture Politics

Day 1903 and Ranting About Bentham

The tyranny of small differences can be the most vicious. I love vendettas in fashion and venture as they are connoisseurs of grievance.

Small communities with insular structures simmer embittered for years. You always know where someone, who is otherwise quite close to you, has committed a venal sin which cannot be forgiven.

But many times these small differences are actually the stuff of the breach. Once crossed you can never return. The opening cannot be closed without a great sacrifice. And these sacrifices are your character.

I am as well versed in the ridiculous schisms of my own affinity groups. As libertarians I’ll go on about the Cato libertarians, I’ll support my an-caps but I I’ll blood feud with the rest.

I feel this way about rationalists and the way they have introduced utilitarianism to Silicon Valley. And I want to be sympathetic here because there are aspects of effective altruism that are perfectly reasonable at first. I like prudent spending and reducing suffering with effective allocation.

But utilitarianism, taken to its end, has issues that anyone who has read Jeremy Bentham has to grapple with. The means do not justify the ends. We are all struggling with the horrors of the problems this creates in a modern society.

I saw the value of the manufactured meme campaign of effective acceleration as it oddly ended up dragging us to the middle. That was the intention and it achieved it. One can have many disagreements in the details.

However I do not think that political actors as far apart as Steve Bannon and MIRI agree on anything philosophically except “we want control over artificial intelligence so the people who are lesser than me can have no say.”

I cannot see how opposing forms of populist control can travel together without fear for character.

Everyone tries to be agreeable right up until coercive violence from Leviathan is required. And I guess some of you don’t think too hard about hard power huh?

I happen to find the request to have so much control over your fellow Americans to be an offensive view.

You think so little of the citizens of your own country when our core constitutional values require us to have so much more responsibility for ourselves?

I do think it is actually a moderate viewpoint that I believe in all of us. I believe in Americans no matter how stupid we can be. Remember that whole being a libertarian thing. I think personal responsibility requires more and Americans have delivered more despite our many failures.

I recognize that my personal stance here is not the final stance, especially as something of an outlier but because we have checks and balances, I know my involved citizenship demands that I declare where I stand.

Which is why the right to compute law that Montana adopted was a largely uncontroversial and popular when it was a bill. Before politics got involved, regular citizens, who were not whipped into a froth or frenzy, could understand that participating in the digital economy is crucial to living in the modern world.

It impacts our first, second, and fourth amendment rights directly because it demands we answer questions about property.

The wider existential issues on artificial intelligence do not get to be more important than our existing jurisprudence nor the opinions of our citizens.

The way we legislate and the value of our system of government, both state and federal, have a part to play. It’s funny the libertarian is making this argument I know, but it is a good revealed values exercise. Don’t get trapped by charlatans who have already declared that the ends justify the means. We both know they don’t.

Categories
Politics Preparedness Startups

Day 1887 and Trust Me Bro

I don’t feel terrific so I’ll keep this as a ramble but with some links. I fear some sort of Rubicon has was crossed, which we will only recognize in hindsight, in the fight for property rights in America versus the tyranny of failing state capacity.

I am referring to the contract dispute between foundation model developer Anthropic and American Secretary of War Hegseth that boiled over on Friday night as some sort of Murderbot surveillance subplot parsed only by lawyers, policy wonks and anyone who remembers the patriotism of Chelsea Manning.

Right before America and Israel went all Epic Fury on Iran over the weekend, somehow it was felt a messy public fight between defense contractor and the civilian oversight of our military was ok? Good thing everyone is too distracted to care.

Funny how one can go from cheering on our military’s capacity for innovation one week and the next to be waving one’s arms screaming “no oh no no no” but that’s just politics as run by the instincts of reality television I suppose.

The best writing on this so far has been Dean Ball (who worked in this Trump administration on its AI position statement not so long ago) who feels that Hegseth took an approach on renegotiating a contract with a frontier AI model developer that has consequences.

Loosely and I am missing much Anthropic had its previous contract to parse classified documents in the Department of Defense (now the Department of War) negotiated by the prior administration. Some aspects of this contract wasn’t going to work for the Trump administration. Anthropic didn’t want to change it and Hegseth threatened to call Anthropic a supply chain risk unless it complied. They didn’t and Sam Altman signed a deal.

Wrapped up in this is that of course we all eventually comply with Leviathan who has the monopoly on violence. I don’t claim any expertise here but lots of people understand defense contractor law.

On its face Anthropic looks like the moral high ground and the position all capitalists and private property respected would prefer. However, it’s in Dean’s words, a bit weird for a company to tell the state that a private corporation has a say in policy as opposed to our democratic institutions.

Anthropic is essentially using the contractual vehicle to impose what feel less like technical constraints and more like policy constraints on the military

This is all complicated by the model developer having wishes for frontier AI development to be nationalized and for their work in particular to be safeguarded by a different political administration. It has become a right versus left thing as it was bound to do.

Now Anthropic gets to enjoy the full throated defense of its industry peers and even conservative policy makers like Dean, because our current state capacity is not what I’d call excellent when it comes to our civilian government capacity. Honestly our military still seems to be crushing it.

Secretary Hegseth looked a bit like he was over reaching in his command that he get his demands met. Americans are very touchy about surveillance of if’s citizens and much less touchy about autonomous weapons so it’s natural to be a bit suspicious if you are a millennial who remembers the Global War on Terror. If you are Zoomer who doesn’t know who Chelsea Manning is please go ask Claude.

However it’s not at all clear that Hegseth’s original negotiations were out of hand as of course it’s the state who decides legal use not the private company and we are meant to have laws, judicial process and all the rest. It’s just that we don’t. So what on earth do we do about it? Trust me bro is not a policy.