@was's banner p

was


				

				

				
0 followers   follows 0 users  
joined 2022 September 05 02:14:00 UTC

				

User ID: 365

was


				
				
				

				
0 followers   follows 0 users   joined 2022 September 05 02:14:00 UTC

					

No bio...


					

User ID: 365

The person you're responding to is so deep in their own fantasy scenario that they're already rolling out the gotchyas for something that they just straight made up lol.

"Let's assume a hypothetical scenario in which all smart people just punch themselves in the face once a day. Well if they're so smart, how come they keep punching themselves in the face once a day? Riddle me that, ace, riddle me that!"

One cause of the IQ-denier (and extending beyond that, denial of racial differences in IQ distribution) fallacy / fantasy is assortative socializing.

Lots of top-tier VCs don't subjectively think IQ is a strong selector because -- by the time a founder gets to Series B, they've been pre-selected for high IQ (1).

Lots of CEOs don't subjectively think IQ is a strong selector because -- by the time they interview someone for an executive position, that person has already been successful and thus has been pre-selected for high IQ.

You don't even have to be that high up. If you're an engineer at Google, your friend group and work group are probably all people who are fairly high performing individuals. So you might notice that there are fewer black people in that group, but the black people that you do interact with probably feel about as smart as everyone else (2)

Conversely, in my experience, if one talks to ER doctors, cops, public school teachers, or other people who are exposed to relatively large and relatively random slice of society, and one is really careful not to use the words IQ and wait until they've had a couple drinks, each of them will readily attest that some people are just plain smart and some people are just plain dumb.

1 - Or at least high enough to come off as quite smart in a superficial conversation not in one's domain.

2 - Not really true when I talk to non-woke people, but for the woke, at least smart enough that it's easy for them to dismiss any differences as those of education or environment, etc.

By two counts:

  1. Mainstream ideas on content moderation would be to remove anyone that posts some sort of swastika. I would only ban a person with significant publicity who singlehandedly could cause a large advertiser exodus.

  2. Mainstream ideas on what causes "experience degradation" expands over time, e.g. first swastikas, then failure to use gendered terms, etc. I would optimize for the experience of the, say, 1 S.D. above-average tolerant person rather than the 1 S.D. below-average tolerant person.

The corollary of point 2 is that sites get more tolerant over time in my view (as people become desensitized to seeing mildly offensive views), but less tolerant over time in the current paradigm.

Touché

Unfortunately I don't think bright lines exist for most things (i.e. most absolutist positions are untenable -- is spam free speech? there's no incitement of violence.) I think the left has historically been really good at amplifying something that's just far enough to change people's minds but nothing that's considered truly abhorrent (e.g. first gay marriage, then transgenderism; not the reverse).

I think that no Nazis really exist anymore (obviously there are some people -- it rounds to zero) so that there's really nobody left to forgive. Certainly the concept of being genocided doesn't need to be forgiven?

If the source of new users is primarily from reddit, how does moving after they ban us reduce the risk of dying from a lack of new users?

The Motte was the only reddit page I was visiting frequently -- I am glad to be off that site and thus no longer supporting it by contributing to their user / view counts.