1/ I think PG is wrong. But, I think this take and his previous ones in the same vein illustrate an error that, i) everyone is prone to; ii) no one is trying to solve because it demands thinking 'intersubjectively' (very deliberate usage here), which is real damn hard to do.https://twitter.com/paulg/status/1201164712106434560 …
-
-
8/ (There's also the matter of applying some non-linear transformation to the weights prior to sampling because attention is finite and diffuse attention often precludes robust social signal detection — that's why attention-hijacking works so well for agenda setting.)
Show this thread -
9/ Anyways, that objective function? It's well-characterized by ideology, the cognitive tool every one of us uses (whether we admit it or not) to navigate our socio-political information environment. It's a tool for judgment.
Show this thread -
10/ Paul Graham's objective function as venture capitalist with a near religious following is probably quite different than yours and mind. That's not me saying he's wrong generally, but his beliefs are subject to very different selective pressures than mine and most yours.
Show this thread -
11/ Given OPs use of 'neutral' and the previous posts which mistakenly conflated evolving language with "a dramatic shift to the left," it seems like his prescribed objective function demands, i) an unbiased sample for attention allocation; ii) identity-less presentation.
Show this thread -
12/ The former I want desperately. (And am slowly working on.) But, it's a remarkably challenging socio-technical problem where good solutions are mostly less wrong ones with lots of uncertainty. (E.g. descriptive representation w.r.t. one trait isn't going to get the job done.)
Show this thread -
13/ The latter — identity-less presentation — is something I increasingly think *everyone* gets wrong.
Show this thread -
14/ PG drew some pretty bad inferences because social cognition — being faster and more integrated — caused him to process certain words as identities and out-groups, coloring his downstream perceptions and resulting quality of deliberation.
Show this thread -
15/ And guess what? The same mechanisms cause you to make similar mistakes, as they do for me. And, that *is* a problem because identity cues are so terribly reliable now that careful deliberation affords little boosting.
: "Just take the shortcut, it's fine."Show this thread -
16/ Polarization begets segmentation; segmentation begs polarization. The set of shared beliefs vanishes rapidly. Perception of ideological bias becomes omnipresent.
Show this thread -
17. I'm not sure what the solution to that is. I have some things I'm working on, but seriously — it's hard and, despite mediums of new scale, it's not a new problem.
Show this thread -
18/ But, imagine if PG had said, "listen, I'm bad at processing information given identity cues [as is everyone]. I just start reading between non-existent lines and get terribly confused and angry [as does everyone]. Can you help me evaluate what's going on more effectively?"
Show this thread -
19/ I would think this is a wonderful project (okay, I'm obviously biased), and I wouldn't care if PG had said it — I think that need does exist! But, because PG did say it my first instinct is some variant of, "shut up we don't need a Verrit by Peter Thiel" solution.
Show this thread -
20/ And that's a problem because I'm almost sure technologists *not* in his particularly sycophantic orbit have the same reaction, which creates this weird deterrence space around the need.
Show this thread -
21/ I'm not saying we should strip identity from political discussion. That would be foolish. But, presentation does matter and if one presentation inhibits reception of important information (as it did for PG) while another facilitates it, wouldn't you want the latter?
Show this thread -
22/ Or to put it differently, if I had written this thread under an alt in the vernacular of HackerNews bros — some boisterous mixture of NNT, PG, and Ayn Rand — would you have been as likely to listen to me?
Show this thread -
TL;DR(A): "If I owned a newspaper, I would simply report an unbiased sample of harms subject to correction by social intervention."
Show this thread -
TL;DR(B): I'm pretty sure PG just wants the "paper of record" to reflect his priors and identity, as do we all.
Show this thread -
P.S. To make this more concrete, consider a lot of the discourse around "white privilege." During 2016, there were a few clips interviewing some random white rural person reacting angrily to the premise of white privilege.https://twitter.com/generativist/status/1201291209563656192 …
Show this thread -
People used to thinking in that framework reacted by shitposted derisively and vitriolicly about the "rubes" reaction. You saw the same response to Hillbilly Elegy, too. (I hesitate to use the later as an example, because I didn't think it was good.)
Show this thread -
Thing is, if you're talking about the unconditional expectation of white v black people in the US, it's hard to deny the disparity technically captured by the term white privilege. (Unless you're explicitly racist, as a fuckload of people are.)
Show this thread -
But, the distributions are wide. No, not every black person is worse off than every white person. If you present everything as if they were, you're going to alien a lot of people who quickly learn to associate "white privilege" with "bullshit." Presentation matters.
Show this thread -
Oh, P.P.S., this *is* a problem I am working on from various angles, so if you're a technologist who wants to build or a VC/angel who doesn't want to create the next panopticon / prior confirmation engine, my DMs are open.
Show this thread -
P.P.P.S related for my
#phdlife-heavy followers, https://twitter.com/profmusgrave/status/1201192398124191751?s=21 …https://twitter.com/profmusgrave/status/1201192398124191751 …Show this thread
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.