Weekly Thoughts
What’s up, everyone? I have some more disjointed thoughts and arguments for you. Hope you enjoy!
—Garrett
This quote seems to have grown younger as it’s aged: “If you are everywhere, you are nowhere.” —Seneca
An apology of those in certain circles (of Hell):
That we should do something implies that we can survive doing it.
We wouldn’t survive being outcompeted by those who ‘race to the bottom of the brainstem’ by monetizing moral outrage, invidious comparison, reductive sexiness, and the coldest semblance of community; who, caught in a related arms race, train and deploy unaligned AI faster than you can query, ‘how can I make a big bomb in my garage?’; who immolate the environment on the altar of capital; or who otherwise recklessly endanger, not just human life, but life in general.
Thus we should not be outcompeted by them.
In order to not be outcompeted, we must ourselves risk an infinity of harm.
Therefore, if we should not be outcompeted, we must also risk an infinity of harm.
So we must also risk an infinity of harm.
Trump is a bulbous excrescence of id.
In principle, AI could become more expert than human experts. But I think this possibility assumes that incentives will allow it to optimize for truth and reason. After all, it may be the case that some among us will be willing to pay more for untruth and unreason, especially targeted falsities and casuistry, than friends of humanity will be able to pay for the opposite. So, even in the presence of smarter entities, human experts may still represent the practical limits of expertise as opposed to the limits which could be realized in a less corruptible world.
I think we should all have taken at least one ethics class, lest we fail to justify our bad behavior.
If my present self has reliably been a disobedient servant and an overdemanding master of my past and future selves, respectively, perhaps I should negotiate better with future me so that I won’t keep ruing the failure to meet my future past self’s expectations.
I haven’t scrutinized to what extent the argument of my essay on Hume’s law relies on premises that are consequentially similar or even identical to the premises of the following unsound argument. To that extent, however, the argument of the essay is also wrong.
The use theory of meaning: what language means is determined by how it is used in practice.
People use descriptive statements of the sort, “that’s unfair,” “that hurts,” “my employees do not know their place,” “how disloyal,” “you idolater,” to mean, in essence (but not always), “that is morally wrong,” or “how morally wrong,” “that person is morally bad,” or interchangeably, “you should not do that,” “you should not act like that,” “you should stop being yourself,” etc.
Therefore, by the use theory of meaning, descriptive statements of that sort do frequently mean something normative.
Hume’s law prohibits deriving a normative conclusion from any sequence of purely descriptive premises.
Yet, by the use theory of meaning, certain descriptive statements mean certain normative statements.
If certain descriptive statements mean certain normative ones, then certain descriptive statements imply certain normative statements.
Therefore, by the use theory of meaning, certain descriptive statements imply certain normative ones.
Hence, if the use theory of meaning is true, Hume’s law is false.
Thus Hume’s law is false.
This argument is wrong. It exploits the fact that, per the use theory of meaning, what certain apparently descriptive statements mean can, depending on context, be either purely descriptive or purely normative or both normative and descriptive. Yet, if we grant that the use theory of meaning is true, these apparently descriptive statements are, again, depending on context, in fact purely descriptive or purely normative or both descriptive and normative. So when they are fixed as descriptive by the context in which they were used, they do not mean anything normative and therefore do not imply anything normative. This would be the case if, say, an insufficiently novocained patient said, “that hurts,” to their dentist. They wouldn’t mean that what the dentist is doing is morally wrong.
If, alternatively, the dentist persisted in their metallic manipulations without numbing the patient up more, perhaps the patient’s repetition of her plaint would take on an additional moral, or normative, valence, as in, “what you’re doing is wrong.” In this case, the statement implies something both descriptive (that causes me pain) and normative (that is wrong/you shouldn’t do that). However, supposing again that the use theory of meaning is true, the statement is not implying anything that wasn’t already inherent in its content. If the statement means something normative according to the use theory of meaning, then it is already normative—in this case normative and descriptive—despite appearing to be merely descriptive, being as it is in the grammatical mood (indicative) which a speaker uses to express facts. Here therefore, a normative and descriptive statement implies its own normative content, and Hume’s law stands.
Finally, we can imagine a situation where the patient, having repeatedly and vainly appealed to the android’s presumed capacity for sympathy, now only intends the moral valence of their statement. In this case too, Hume’s law stands: supposing that the use theory of meaning holds, the statement, while flaunting its apparent descriptiveness, is purely normative. And Hume’s law doesn’t bar any statements, normative or otherwise, from implying themselves.