The good delusion: has effective altruism broken bad?

I'm not saying I'm a good moderator; there are other prerequisites.

But it's probably inevitable that Hacker News will wither and go away (or at least stop being a place that attracts people like me). Eventually, somethign will replace HN like something replaced Slashdot (which had fantastic comments with not that much moderation, back in the day) and Usenet.

Why can't it be conscious merely while the computation is running? You aren't constantly on, you're ostensibly only conscious 16-20 hours every day, while you're not sleeping.

> Continuity of learning: In addition, the AI needs to be continually update its "mental model" of the world—in effect, constantly "learning" and re-training its neural network on the input it receives.

I'm not sure what learning has to do with consciousness, and I'm absolutely certain you don't either. I frankly find it hard to believe that Zen Buddhists are not conscious during their meditations, despite having empty minds and not "constantly updating their mental models of the world-in-effect".

> But I, personally, believe they are both necessary for it to be even worth starting to consider whether a given AI might be. And I believe that consciousness is a prerequisite for sentience.

See, this is what I'm talking about. That's a fine personal belief, but that belief does not rise to the level of expert testimony of the facts even if you are a machine learning expert.

People have made all sorts of assumptions about the human mind that they have never questioned but take as obvious fact, and then pontificate that AI is clearly not doing this, therefore AI cannot share such and such properties with humans. But the fact is, most of those assumptions are just wrong. Without a mechanistic model of how these systems work in the brain we are not going to make a valid comparison against what AI is doing.

I agree in general that we can set some bounds, but those bounds are neither as firm as, nor are they placed where most people think.

Because if all the numbers are made up, you don’t actually know what is an “altruistic” choice. You can say, gamble with other peoples money, and justify it by donating 10% of the returns to charity. And if the math works out in your head, it’s all good.

> Yes there are such people. I don't think they would be less greedy and exploitative if they didn't have a way to feel better about it.

Lol, we’re arguing about a moral philosophy here, and your argument is that anything bad that people do, keeping in line with the philosophy is irrelevant?

It’s also not that “there are such people”. I’m saying that EA will disproportionately attract the greedy because fundamentally it argues that you can buy your way into altruism. It’s the Catholic Church selling indulgences for the 21st century.

First, I'm sure you're right that far worse crap goes unchallenged. But that's not because it's ok (it's not), nor because moderators are choosing to go easier on them (we're not). Rather, it's for the same reason that only a small portion of speeding drivers actually gets pulled over and ticketed: there are way too many to catch them all. In other words, when you see a post that ought to have been moderated but hasn't been, the likeliest explanation is that we simply didn't see it—there are far too many posts for us to read them all. You can help by flagging such cases or emailing us at hn@ycombinator.com.

Second, I believe you that you're posting in good faith and with good intentions. Unfortunately, intent isn't enough—it doesn't communicate itself on the internet, especially not in these tiny little blobs of comment text. You have to include enough information to disambiguate your intent (https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...).

It's common for commenters to assume that their posts will land with readers in a way that is consistent with the intent they have in mind. Unfortunately, they often don't—because while you have your intent in your head, the rest of us don't (I don't mean "you" personally, of course—it's the same for everybody). This is what leads to the 'rebound' phenomenon where someone posts a snarky one-liner or otherwise breaks the site guidelines, and then when a moderator asks them not to, they reply with the thoughtful, substantive comment they actually had in their head*. The problem is that all that good information doesn't communicate itself. You have to spell it out explicitly.

* A case of this from earlier today: https://news.ycombinator.com/item?id=33623956