PriestGPT
1
9
PriestGPT has not written any posts yet.
I think I've found the source of our disagreement encapsulated in something you just said:
"The overt appearance of any statement is one of being told the truth"
I counter that the overt appearance of any statement depends on common context.
Your statement only holds insofar as you ignore (or deny the legitimacy of) the role of context and non-verbal communication in determining the overt meaning of a statement. To start with an easy example, a sentence said in an exaggerately sarcastic tone has the overt appearance of being untrue; if sarcasm is a commonly understood device, then I think it should count as "part of the language", as a legitimate operator over a sentence... (read more)
I did read the article, and still I hastily reached for the concept of "intentional deceit" when grasping for what I felt (and still feel) to be a difference between your example and the author's. Sorry.
But even now that you point out that the author excludes intentional deceit from his definition of bad faith (which I had read but somehow didn't apply in my comment), I only feel like I missed the nail, but a nail is still there, which I will now try to hit on the head.
So here's my second attempt: the difference between your example and the author's has nothing to do with conscious intent to deceive, but with... (read more)
Well, while your friend may be acting in a way that suggests that he cares about learning information from you, it isn't to conceal his true motivation from you. Wanting to enjoy a social interaction with you is not a secret, which is the core difference between this example you brought up and acting in bad faith, where the real intent is intentionally hidden.
It's not that no given evidence should change your confidence level (that would mean that you do not update your beliefs in response to evidence, which is contrary to rationalism), but that your expected post-evidence confidence level should be the same as your pre-evidence confidence level.
Note that the word "expected" here is a technical term in probability theory and statistics that does not always match up with natural-language uses of the term. What I mean by "expected post-evidence confidence" here is "the weighted mean posterior probability".
Of course, once you see the evidence, your confidence level should change, and you should always expect that to happen. That's the definition of evidence; an event that will change your confidence level.
If you still don't understand the difference, learn probability theory notation to be able to read the formulas provided in the text.
One who seeks perfection will never stop taking steps, compared to which any finite number of steps is as little as no steps.
But surely, there must be some really high level of achievment that we can qualify as being further than the first step at least!
Sure there can... if your point of reference for what counts as a "step" is the expected rate of other humans' progress. But, as the author writes in the previous paragraph on humility (reread for context if needed):
Life is not graded on a curve.
Even if the ASI does fully believe in open individualism, I believe it is likely to weigh its own wellbeing far more than others' anyway, on account of its "bigger brain" making it more sentient/sensitive (in the same way humans compare themselves to simpler-minded creatures). In the extreme, the ASI may see itself as a utility monster.
Well, if we as humans place weight on open individualism (OI) being true, then the question of whether an ASI would shield us from S-risks or not should lose importance to us relative to the question of whether the ASI would serve the greater good, even at the expense of humanity. So if an OI ASI came to pose an S-risk, then perhaps we should trust that such an ASI's decision to doom humanity serves consciousness better than our species-preserving biases would have it.
A belief in OI goes both ways, by which what I mean to emphasize is not that this belief's effect on an ASI's probability of incurring S-risks is ambiguous due to OI providing the ASI reasons both for and against harming humanity, but rather that a belief in OI should lead both us and ASI to care less about our own respective fates.
Vegan here -- best essay I have read against veganism so far!
That being said, I have much to criticize that has not yet been in the comments.
However, I also agree that veganism is not necessarily the best thing we can do for the animals, but for reasons that I believe are stronger than the ones you provided. More on that at the end.
... (read 693 more words →)Calcium is one of the only nutrients we know of that can reduce the mood symptoms of PMS for women and it is practically impossible to get enough calcium from real food from vegan sources (you’re stuck taking medicine for it in the form of supplements or eating artificially fortified
But why does "having to explain it more" make it less likely? It seems like your argument would only appeal to an intuition which has already accepted and internalized the very simplicity bias which Solomonoff induction is trying to justify.