Kemp

• Kemp 21 Jul 2025 22:19 UTC

  4 points

  −2

  on: Mo­ral re­al­ism—ba­sic Q

  For those who give significant weight to the possibility that morals are real, in the “actually have some effect on the universe” sense...

  This sense of “real” is contested in contemporary philosophy. As a sketch, here are a few things that are often considered quite plausible:

  1. There are abstract objects and abstract facts

     1. Plausible examples: numbers (or sets), shapes, propositions

  2. Abstract objects are real

  3. Moral facts are abstract facts

  If (1-3) are plausible, then it’s also plausible that (4) moral facts are real. I’m not exactly offering this as an argument, but merely a set of (what contemporary philosophers consider to be, at least) plausible theses. A philosopher who accepts (4) will usually be referred to as a moral non-naturalist.

  What IS it, if not “just” our beliefs and preferences about each others’ behaviors?

  This is an interesting question for moral non-naturalists; but they would respond that this question is no more mysterious for morality than it is for mathematics (or any other abstract domain). And they’d certainly reject that moral facts (and mathematical facts) are “just” reflections of our beliefs and preferences.

  I find the premise that it’s anything but a social consensus of human preferences to be … incoherent.

  Of course, many philosophers will reject 1-4 above, as well as this thought of yours. Such philosophers think that moral facts simply reduce to natural facts. E.g., facts about wrongness reduce to facts about units of pain, in the same way that facts about water reduce to facts about H2O. If they’re right about that, then all we need to do is figure out how to measure pain-units—or whatever natural property morality corresponds to.

• Kemp 21 Jul 2025 4:02 UTC

  1 point

  0

  on: Plato’s Trolley

  On the other hand, it’s possible that objective morality exists but is not empirically obtainable knowledge in nature. If that was the case, the only other way I can imagine for that knowledge to work is by some kind of enlightenment or grace—an inherent inner knowledge that we all possess, or can possess if we achieve the right state of mind, not from observation of the outside world, but by introspection.

  
  Mathematical knowledge is not empirical. By your reasoning, does mathematical knowledge therefore “work by some kind of enlightenment or grace”?

• Kemp 19 Jul 2025 21:46 UTC

  1 point

  0

  in reply to: nwm’s comment on: Notes on spaced rep­e­ti­tion scheduling

  How many such cards do you usually make? In theory, there’s an infinite number of forms for card content Z. And I personally find it too costly (cognitive effort, time) during card creation to brainstorm the various ways in which I might encounter Z in the future.

• Kemp 19 Jul 2025 2:46 UTC

  1 point

  0

  on: Notes on spaced rep­e­ti­tion scheduling

  I’ve recently been wondering whether introducing randomness to flashcard content might also improve… something

• Kemp 15 Jul 2025 16:38 UTC

  3 points

  2

  in reply to: Gordon Seidoh Worley’s comment on: What is David Chap­man talk­ing about when he talks about “mean­ing” in his book “Mean­ing­ness”?

  Let me try:

  On a busy sidewalk, your eyes lock for an instant with those of a cute stranger coming toward you, and then they pass. You stop and look back over your shoulder and see that they have done the same. You can see that this is purposeful—even if it’s not exactly clear what it’s purpose is—and an attentive third person would see the same.

  Hmm

• Kemp 16 Jun 2025 22:44 UTC

  3 points

  0

  on: De­bate ex­per­i­ments at The Curve, LessOn­line and Manifest

  Good post!

  Another way to diminish/​remove status from debates is to shift from adversarial to collaborative modes. I’d like to see more experiments on “collaborative” debates. Here’s an idea, pulling from this community: Crux Speedruns. Participants with opposing view on A must work together to find the crux of their disagreement as quickly as possible. Their team’s time is added to the speedrun leaderboard.

• Kemp 16 Jun 2025 22:29 UTC

  3 points

  0

  on: How LLM Beliefs Change Dur­ing Chain-of-Thought Reasoning

  Interesting, thanks for the work and the writeup. Two naive questions:

  1. Imagine we were able to run this experiment on humans. Introspectively, when reasoning through problems I often find myself “jumping” between potential solutions. For example, say you’re testing my beliefs about the optimal next move in a game of chess. If you truncated my chain of thought by the same method used here, would we also be tempted to say that my beliefs are “wildly fluctuating”?

  2. Related to the first question: on what grounds do we consider some intermediate element of a chain of reasoning to be belief? Here’s a reason why we might not: beliefs are generally thought of as stored cognitive dispositional states of system. The stored dispositional states of LLMs are encoded in its weights. Weights aren’t changing during inference time. So: beliefs aren’t changing during the chain of thought.

  Thanks