I think I have some kind of brain atrophy happening, because I did 2 masters' degrees & 5yrs of a PhD in what was essentially ancient & modern hermeneutics, & I like this silly PULSE/Obama analogy in m latest news letter better than anything I can remember from all that.
The idea that reading is an upscaling activity, done by a trained network that looks for an adjacent point on a manifold... I dunno, man. Seems simplistic & overly gee-whiz-techno, but I can't stop thinking about this.
I would not want to reduce reading to that by any means. Rather, maybe that's just one thing that reading is.

I probably like this upscaling idea because I hated all flavors of reader response theory, yet never formulated a real critique of it.
What I ended up with in this post is a kind of reader response theory, but you still get to have authorial intent. (I always hated how authorial intent had been problematized out of existence, so in seminar we had to ostentatiously pretend not to care about it.)
We'd do a thing where, in seminar "authorial intent" was not real &/or did not matter, because we were doing some reception theory. Then we'd submit a paper & when the prof was grading it what we intended to say was very much a factor in its interpretation. 😂
If you want to peek under the hood as a very cleverly trained network does a bunch of deliberate upscaling of a written text, & tries to trace out the contours of his internal manifold as he navigates it, here you go: amazon.com/S-Z-Essay-Rola…
Screencap in the OT is from here: doxa.substack.com/p/clubhouse-is…

• • •

Missing some Tweet in this thread? You can try to force a refresh

Keep Current with jonstokes.com

jonstokes.com Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!


Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @jonst0kes

7 Apr
Asians don't count towards "diversity" in AI. A story in three parts.
You can find the paper here: ainowinstitute.org/discriminating…

In it, racial diversity is very narrowly defined in terms of black and latino representation. This entire discussion is broken and the people running it are bad actors.
It would be different if the paper were focused specifically and explicitly on black & latino representation, but it is not. It purports to speak about "diversity" while defining everyone out of it who doesn't fit into one of two categories.
Read 10 tweets
7 Apr
If your theory of the summer 2020 protests doesn’t have a prominent role for the mass trauma that was the global pandemic + lockdowns, it is not a good theory. This is not a good theory.
It’s worth noting that this is exactly like the Chenoweth piece in the WaPo from last year, which treated the pandemic as an obstacle to the protests rather than an intensifier. I don’t know why this keeps happening.
Read 4 tweets
28 Mar
At some point we'll all have to reckon w/ the following issues /in combination/:

1. Society has problems
2. ML amplifies some of them
3. #2 will get worse, making #1 worse
4. Something must be done
5. The ppl yelling the loudest about 1-4 are an obstacle to #4.
We need an entirely separate, competing conversation about fairness, accountability, & transparency in ML that takes as its premise that fixes for #2 are possible & desirable independently of efforts to fix #1, and then proceeds from there.
Not only are the FAccT clique openly opposed to efforts to look for technical fixes, but they're trying to rule out discussion of technical fixes as out-of-bounds entirely.
Read 6 tweets
28 Mar
They’re increasingly being driven around the bend by his refusal to perform an elaborate apology, or even admit error. It’s pretty awesome to watch. One of the guys trying to justice him into submission is a AI higher up at Chase bank lol.
As I predicted, the place there really pushing back hard is on his claim that the problems are fixable. This claim is a direct threat to the entire AI ethics project so it can’t stand. They’ll keep at him over it.
Here’s what I’m seeing in his responses, though: he’s smart enough to see that the larger points re: societal bias are obvious, & he has enough intersectional armor that he doesn’t have to pretend they’re saying something novel/deep/complex/technical. Greatness.
Read 16 tweets
27 Mar
*spits out coffee* This is exactly what I said would happen, how I said it would happen. 😂
Notice the difference between how she handles Carr & how they have handled others in the past who've stepped on this same landmine. Anyway, Carr will apologize for this thread.
They hate fixes so much! Proposing a fix that is not "totally reform all of society" is what gets you targeted for criticism or abuse.
Read 8 tweets
27 Mar
This thread is an interesting artifact, b/c Carr is one of the main figures in "2+2=5" discourse, yet this thread contains exactly & precisely the argument that gets people like @ylecun dragged by the AI ethics brigade.
LeCun even tweeted this thread approvingly, which is no surprise since it backs what was essentially his argument in the big Twitter dustup with @timnitGebru.

Prediction: Carr will have to essentially retract this thread. It won't stand, as-is.
From the perspective of the AI ethics people, this thread contains wrongthink. If LeCun himself had tweeted this identical thread, he'd right now be subject to a massive pile-on led but a handful of big accounts.

So I think Carr will get pressure to do a "clarification" at least
Read 10 tweets

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!

Follow Us on Twitter!