darkkindness

Thanks for the link, I was wondering if Scott touched on this. Seems to be an adjacent topic but it's relevant nonetheless.

my tldr of that post: in the context of ideological/political theory, mistake theorists are engineers who want results, conflict theorists are warmongers who want to win.

Perhaps seeing bias in everything is just one of those warmongering traits -- because trust me, it's really hard to win against a neutral point of view.

I like this. Polarization is a pretty good characterization. People are just not treating neutral views as neutral anymore.

Food for thought: perhaps people treating everything as fake news is an unfortunate adverse effect of all the messaging about the dangers of not recognizing fake news.

Thanks for your viewpoint.

As far as I understand, outrage culture is about shunning popular folks for their imperfect human morality. Could you (or someone else) elaborate on how this is related to this, well, let's call it erosion of neutrality?

meta: Over these past four years I've seen a growing number of people interpreting the most nuanced, neutral statements as 'apologist', 'biased', 'lobbying', 'spam', 'politically motivated'. I don't think this has always been the case, especially on HN.

It's not just skepticism, that's for sure. How do I put it? It's like assuming there are sides when there are no sides. I'm curious, is there a word for this?

Just for fun and to make a point, I threw your reply into Talk to Transformer.

> This is something so many people miss when trying test the limitations of GPT-2. It just doesn't make sense to test it on strings of text that nobody ever writes. To me, the best way to evaluate the usefulness of GPT-2 is to compare it to some actual test that validates a lot of its claims. So... let's do just that.

It might be just chance, but gee -- is this text referring to its own generation as a test to convey a point? The self-referentiality is formidable.

I use Firefox on Android with NoScript and an adblocker.

Yet I'm uniquely identifiable. One culprit is screen size putting me at <0.01%.

Does that make defeating fingerprinting on mobile hopeless for the casual user?

Edit: more info. All JS is blocked, and I have privacy.resistFingerprinting. The page doesn't detect my adblocker. Still, there are just too many things I can't change:

- hardware concurrency: 1.7%

- audio formats: 0.2%

- navigator properties: 0.2%

- audio data: 0.1%

I was surprised at this one:

- Media devices: Unique

What are media device identifiers for, exactly? Why does the browser supply it without JS?

It's really weird to evaluate GPT-2 based on its ability to say things no reasonable person would ever say. If I were born in Cleveland I wouldn't be jumping to proclaim my fluency in English. If I told you I left my keys out at the pub, I wouldn't immediately repeat myself and say that my keys are now at the pub. If I'm talking about two trophies plus another trophy, I'd probably try to end it with some punchline rather than saying there's three trophies.

A lot of the things we write assume the reader can make connections on their own. That's a writing skill. It's the reason why Hemingway's famous "For sale: baby shoes, never worn" is so impactful. As such I've found GPT-2 to be incredible at writing fanfiction.

This is insightful. It also massively calls into question if economic models on prices (CPI?) could ever keep up with reality, if trends like shrinkflation didn't make that impossible already.

The conversation Bill's touching on -- the tradeoffs between different notations -- is really valuable but I think it misses something a lot of us desire in a syntax: compositionality. Take the provided "onion" notation, loops, and the "new" syntax. They all look something like this:

  ┌────────┬────────────────────┐
  │myreduce│┌─────┬────────────┐│
  │        ││mymap│┌────────┬─┐││
  │        ││     ││myfilter│x│││
  │        ││     │└────────┴─┘││
  │        │└─────┴────────────┘│
  └────────┴────────────────────┘

(The math one looks more like this:)

  ┌────────┬────────────┐
  │myreduce│┌────────┬─┐│
  │        ││mymap   │x││
  │        ││        │ ││
  │        ││myfilter│ ││
  │        │└────────┴─┘│
  └────────┴────────────┘

But none of them provide the same kind of "putting pieces together" feeling as this:

  ┌────────┬─────┬────────┬─┐
  │myreduce│mymap│myfilter│x│
  └────────┴─────┴────────┴─┘

which we see in the wild as this:

  (myreduce ∘ mymap ∘ myfilter)(x)

or this:

  x | myfilter | mymap | myreduce

or this:

  myreduce mymap myfilter x