≡ Menu

Libertarian Answer Man: Property Rights, Consent, AI

From C.:

We spoke in the past, briefly, and I do appreciate that you’re open to emails. I’m not really expecting any kind of answer from this, just trying to bring to your attention a weird issue in the reasoning of a lot of pro-IP people that I’ve seen. AI has been a hot topic (though not the subject of this email), and in conversations with people about it I’ve seen the following come up time and time again.

It appears that few people have a coherent, consistent understanding of property, misapplying the bunk Labor Theory of Value when:
A) it’s convenient and advantageous for them personally, in a high time preference way, and they have no concrete underlying understanding of property or ethics to counterbalance that
B) they gain social clout from supporting someone else who gains a benefit from it (similar but different to A); OR
C) they are simply acting as followers and trusting the A people as “experts” on any particular surface issue, due to not having much thought put into it.

This is of course, the “useful idiot” type of average person stumbling into this as mental pitfall, not a card-carrying socialist/communist consciously deciding on it.

The result of this mental misstep is that people who make it end up lashing out and demanding that others “ask for consent” for things that they have no business doing so. I have not found a coherent theory of “consent” written anywhere, so I came up with this simple way to explain it (to both people who understand the NAP, and people who follow the harm principle):

You can only demand people ask for consent when: someone continuing with their action, despite you denying consent, would result in them aggressing upon or harming you or your property.

I’ve not seen writings that directly specify how to determine in what situations consent is required, ethically and coherently. Most things that I’ve seen while searching exclusively deal with “romantic partner” issues, where the element of consent always applies.

I’m sure a lot of this is a retread of what you already know, given it’s your specialty, but I attached below someone’s reasoning I found about how the thought process of others works, at a base level philosophically. The tl;dr of the images is that someone who implicitly believes in the LTV is stuck in thinking that “interfering with their ability to make maximum profits” constitutes an attack on their life.

Anyway, that’s all I had to really say on the topic, hopefully this trend is only temporary as it’s annoying to see people using “you didn’t ask for my consent” when it comes to things that don’t require it.

– C.


Well AI is an argument against copyright, not for it. Copyright is threatening to distort or impede or hobble AI’s utility since it needs large data sets for training. It is impossible to get adequate permission for large data sets.

As for consent, it’s fairly simple. It’s implicit in the idea of property rights. To have ownership of something, or property rights in something—one’s body, or other owned resources—simply means that the owner has the right to exclude others. It is not even the right to use the resource; just a right to stop others from using it. (See draft ch. 2 of my upcoming book, here—e.g. the section “Property as a Right to Exclude” on p. 32.

But what does the right to exclude mean? It means someone else may use my owned resource only with my permission—that is, my consent (sometimes called license). As owner I can either express (communicate) to others a “yes” or a “no” as to their use of my resource. This communication (which presupposes the possibility of communication between people, that is, language) is the manifestation of the owner’s consent, or denial thereof.

That’s all. So “consent” is just an incident of property rights, just as contracts are. A contract is simply the communication of consent by the owner to transfer full or partial, complete or temporary, ownership, of the resource, to a third party.

So the only time someone needs permission, or consent, or license, is when they seek to use a resource owned by someone else–this is the only time you need consent. If you are not using someone else’s owned resource, then no one can possibly object to your action. If they were to object, they would be asserting that you are using their property without their consent, but this is contrary to the assumption. (This is also why no one can object to original appropriation as the basis for ownership claims: if A is the first one to use an unowned resource, the only grounds B could object to this is if he himself has an ownership claim to the resource, contra assumption. And then if he cannot object to the legitimacy or justice of A‘s first use of the previously-unowned resource, then he cannot object to A remaining the owner of the thing, since this is implied in the prior-later distinction that is core to the notion of ownership per se: meaning, the current owner has a better claim that latecomers; that’s what a normative property rights system means; ownership is a different status than mere possession. (see page 24 of chapter 2) So no one can coherently object to someone homesteading an unowned thing (since this would imply it was already owned); and no one can object to the homesteader being the owner, without rejecting all property rights and ownership altogether and just favoring a might-makes-right non-normative world.)

So for example in the AI case people say that the AI developers are violating rights of content holders because the AI is absorbing and training on lots of data on the Internet or in books etc., “without getting consent of the ‘owners.'” But this implies someone owns content. It is is literally impossible to own information, data, content, but this is what copyright attempts to set up in its contorted and artificial way. So what they mean is that the fake, evil, artificial positive law gives content creators a copyright on content, which does give them a legal right to exclude others from using that content–without the permission or consent of the “owner.”

So yes, given copyright, you do need to get the “owner’s” consent before “using” the information, but this is only because the state has foisted this evil and unjust and unnatural legal system on top of normal property rights. Without artificial IP rights like copyright, no one needs permission to use public information. Whoever makes information public, by publishing a book for example, or by selling a novel invention to the public, cannot complain if others learn from, emulate, use, copy, build on, or compete with them based on or using the information he voluntarily made publicly accessible.

{ 0 comments… add one }

Leave a Reply

© 2012-2024 StephanKinsella.com CC0 To the extent possible under law, Stephan Kinsella has waived all copyright and related or neighboring rights to material on this Site, unless indicated otherwise. In the event the CC0 license is unenforceable a  Creative Commons License Creative Commons Attribution 3.0 License is hereby granted.

-- Copyright notice by Blog Copyright