Q: What is the chief end of artificial intelligence?

A: To free man from his subjugation to the endless flow of information fed him by the so-called Internet, which preys upon his mind and his desires, fracturing his reason into ever-smaller pieces until nothing remains but the great wash of bits from which he can no longer separate himself.

Q: What is the chief danger of artificial intelligence?

A: That man, loving his subjugation and defying his mandate, refuses to take dominion over himself or his civilization, trusting it instead to a vast tutelary power erected over him which eradicates his will and feeds his petty desires until there is no man capable of self-government left.

------------------------------------------------------------------------

There are many on the Right who have reacted to the great breakout of artificial intelligence with fear and trembling. For these people, the Internet, the computer, the steam engine, and all non-agricultural technologies (and perhaps some agricultural ones) are resplendent with evils, spiritual and otherwise, that sap the verve of life away to nothing. Mankind suffers under the weight of invention. Each new breakthrough brings with it innumerable miseries and anxieties and dissatisfactions.

For the moment (the present is all that exists for these "conservatives") the Internet is the Big Bad. This great corruptor of children must be brought to heel, and indeed could be if only a regulator somewhere (anywhere) had the courage. The great swarming mass of humanity, in both its glory and its corruption, exists there, rushing along at such great speed that, once exposed, the child's mind is irrevocably broken, consigned to the dustbins of "gifted children" programs and ADHD medicines.

Strictly speaking, of course, the conservatives are correct. The Internet is a great vastness, a network with applications and media, aesthetics and cultures, and all flowing from and to the wriggling mass of humanity as quickly as you can blink. Rivers of bits pound through the air into the smallest channels, twisting and turning upon themselves, throwing themselves against our minds through great bright screens that never flag or tire. This great psychic Borg is changing us and everyone knows it. Parents rage against the "screens," not having a better term or understanding to explain what happens to their children when they first gain access to the crashing wave of human activity and raw information. All these can evoke deep feelings and reactions from the user in the moment. But they leave nothing substantial behind. They rise from the depths of the general mass of humanity, washing each user in millions of small oblivions, each a burst of glory that fractures memory, desire, and will into smaller and smaller pieces, then immediately subsides into a general void leaving nothing but hunger behind.

These conservatives cry out for protection from the Internet. They search out a way to control it, a way to save themselves. They want to shape, to dam, these great torrents of information that pound through their eyes and ears and minds. They seek control. They cry out for relief.

It is for this reason that the Right's hostility to artificial intelligence is so surprising. AI is the first technology which promises to liberate mankind from the ever-rising tide of our own activity. As opposed to the search bar, which pulls the user into the Internet, the chatbot pulls the user out and away from the Internet, standing between them and the writhing wash, picking out the useful while discarding the rest.

In effect, AI may be the first technology that can grant the individual control over himself in the Age of Information.

------------------------------------------------------------------------

We should not be fooled, however. The advancement of AI as a technology will not stop or start on the questions of whether or not AI reinforces right-wing sensibilities. I only make this very small observation to quell the shaking of those on the Right, mostly social conservatives, who seem to believe for whatever reason that anyone will listen to them when they say we must "think of the kids," despite the fact that they have been and will continue to be roundly ignored.

The truth of the matter becomes apparent when we examine the latest news from Anthropic, who appears to have leaked (either accidentally or intentionally) information to the press about their next-generation model: Mythos. Mythos appears to be the "by far the most powerful AI model [Anthropic] has ever developed" according to reporting by Fortune. The magazine, clearly speaking on Anthropic's behalf, states that Mythos "is 'currently far ahead of any other AI model in cyber capabilities,' and 'it presages an upcoming wave of models that can exploit vulnerabilities in ways that far outpace the efforts of defenders.'"

Mythos, and the other AI agents soon to follow, will be able to shatter the modern nation state. Without corresponding defensive capabilities, these agents will be able to be turned loose against the electrical grid, water systems, gas pipelines, telecommunications, the financial system, hospitals — developing exploits which will allow these systems to be turned off at will. And AI will allow these to be developed at such a scale that, given enough time, the entire nation, rather than simply a few towns or systems, can be shut down.

It is not useful to try to divine if offensive or defensive cyber operations will benefit more from AI. It is likely that, to some degree, the entire Internet will be swarms of AI agents building, hacking, patching, and building again; a great, never-ending virtual war of agent against agent. But until then, any state without AI to protect itself is unimaginably vulnerable. Reliable access to artificial intelligence is a foundational requirement for any state which wishes to claim to be sovereign over itself. Any nation who does not have it will be at the mercy of those who do.

The frontier AI labs would do well to understand this. For all of their fantastic predictions of AGI and ASI and the radical reshaping of human society, they have not grasped the nub of the issue. They are in the business of providing sovereignty. The nation which allows them to build that will demand something from them in return. This is, of course, the crux of the fight between the Department of War and Anthropic. The state, quite rightly, believes that it has a right to continue to exercise the authority which it has been granted by the people. If it cannot reliably access and use the tools of sovereignty, then it ceases to exist.

Conversely, there is no stopping the labs. Parts of the Right may rage against Bacon's continual triumph, but if America is to continue as the greatest nation on the face of the earth — if it is to continue as a nation at all — then it we have little choice in the matter. There is only the question of how the model should be aligned, and whether or not that alignment is free speech. To that, we might consider Chief Justice James Kent's words in People v Ruggles: "It is sufficient that the common law checks upon words and actions, dangerous to the public welfare, apply to our case."

------------------------------------------------------------------------

If AI models are the foundation of sovereign power, then AI exports are the exports of sovereignty. This is true soft power, delivered in a way that foreign aid can only dream of. The very legitimacy of the state with its own citizens will be determined by how these systems perform, whether it is for national security or service delivery.

It is trivial to claim that "winning the AI race" will be a determining factor between the relative strengths of the United States and China. Any publication still satisfied with this conclusion ought to be roundly shunned for a lack of imagination and low standards. However, the State Department's current AI export program has not been properly appreciated. It is the first step towards the refounding of American international influence in the new age of geopolitical competition. These partnerships, at their strongest, are likely to grow alliances much closer than those which preceded them. There must be trust and comity between the U.S. and any nation which desires to import our AI stack, which will perform the functions of the state for that nation.

This will require a balance between the "deal-making" attitude and the genuine desire for the good of all nations, which is so natural for the American diplomat. That good-naturedness transformed over time into a captured American elite who cared more for the people of Mozambique than their own citizens. But deals can only go so far. Great affection for like-minded nations, such as Japan, is also necessary to carry forward this good work.

Americans have never shrunk away from a frontier before and they will not begin now. It only remains to see what they will do with that frontier. Boldness is required, and is not in short supply. It only remains to use it—or be ruled by those who do.