colllectorof

Seems fairly accurate. I'm just annoyed that many people can't grasp that awful police behavior and awful behavior from other actors are not mutually exclusive. There is lots of unjustified police violence and provocations against perfectly peaceful protestors. There is also systematic vandalism and arson by people who aren't peaceful protestors. And there is massive amounts of opportunistic vandalism and looting. All three can and do co-exist.

This is not a two-sided conflict. More like three- or four- sided mess.

You think the supposed contradiction is between legality and morality of not censoring certain speech. The actual contradiction is between claiming to support people's right to free speech while also supporting policies that undermine one of the main reasons that right is important in the first place.

This is not a matter of spherical freedom in vacuum. This is a practical matter of whether a tiny number of people should be able to effectively control country's political processes because those people happen to own certain technological tools.

Not to mention that to be truly useful a static type would need some sort of literal that would emit a compile-time error if the supplied init data was wrong.

In short, examples above do not demonstrate Haskell's ability to statically enforce rules like "string needs to end with a dot".

Now, you could make a type that always prepends a dot when asked for a string representation, which happens to enforce this specific constraint, but you cannot use a trick like this for most constraints (such as the second example of "only alpha characters").

These are just two functions that wrap and unwrap a string. Your "type" doesn't generate a compile-time error when constructed with the wrong data. This doesn't even handle the simplest use case:

   putStrLn (toString (fromString "something."))

  module Foo = struct
   let first() = 1
   let second() = first() + 1
  end

  module Bar = struct
    include Foo
    let first() = 0
  end

DNS was a passable solution in the 80s, but right now it's absolute shit.

Secondly, OO composition does not solve the problem I described in #3. If object X has 9 methods I want to reuse in object Y, making X a component of Y still requires me to re-implement those 9 methods in Y, even if they are just pass-throughs.

The sad reality is that since so many people are only familiar with the strawman version of OOP rather than the real thing intelligently discussing the advantages and disadvantages of these paradigms is very hard.

This correctly identifies one of the worst problems with how people construct OO programs when they get out of college. They are infected with the idiotic desire to make neat taxonomies. They don't know why. They can't even explain the rules that make one taxonomy good and another one bad. They just feel some immense pressure to arrange things in trees. I've been like that when initially learning Java and it took a while to recover from it. It's like fucking brain damage from bad examples.

Inheritance is not about creating taxonomies.

Namespaces are not about creating taxonomies.

Both are practical solutions to practical problems that occur when writing code.

Inheritance is a result of several insights:

1. It's convenient to be able to sketch out a protocol which a set of objects will follow - without having to work out every damn detail about its implementation.

2. Most of the protocol can be implemented once and then reused.

3. Often you want to use the default protocol implementation with some changes. Instead of re-implementing the rest of the protocol, it's highly convenient to be able to specify only the differences.

#3 is the core reason why inheritance was developed.

But hey, these days many people want companies to stick their nose into everyone's business, police "bad speech", create comfortable "spaces", enforce conduct and so on. Well, you reap what you sow. Any mechanism that is created to do all this "activism" will inevitably be subverted by something like CPC to enforce their interests upon everyone else.

>Big claim. Any proofs?

All this skepticism around Google's capacity to abuse their troves of data and invasive services is a clear indicator that this discussion has very little to do with real privacy. It is mostly a playground for various corporate, political and media shills.

> but the medium does have an implicit model

No, it does not. There is no such thing as universal cognitive model "for books". Every writer has his own. That's why there are so many different structures and styles of presentation. That's why some books are much easier to absorb than others.

Fuck, I don't write books (yet), just articles, and there are tons of different ways to handle this "mental modeling" even for short texts. It's not even always a good thing to pay too much attention to such models. Most writing is not propaganda where the only point is to imprint someone's brain with the desired information and opinions. It's a conversation. Accurately capturing your own mental model of the subject is usually far more valuable than overthinking the mental model of the reader - because you will have many different readers and (unless you're doing propaganda/marketing) you don't want to create something that appeals to a kind of faceless "average consumer".

The fact that modern books (even educational non-fiction) are a "lossy format" is usually by design, even if people don't think about it too much. It is not always the case and it was not always the case. For example, religious texts were designed to be "lossless". Epic poems were designed to be lossless as a verbal form and that continued to shape them when they were captured in writing. You can make a pretty good argument that deliberately lossy writing was a far more recent invention than writing in general.

Something genuinely useful might, perhaps, be achieved by applying GANs to parameters of those 3D models to allow easier customization, but I haven't seen any work in that direction.

[1] https://www.youtube.com/watch?v=KZ9mb3Jylb0

A radical is someone who is willing to use extreme measures to achieve their political objectives. Nothing to do with the status quo. Totalitarian regimes, for example, usually maintain their status quo with radical zeal and measures.

Teaching children to read is an interactive process that has pretty much nothing in common with data steamrolling in modern machine learning.

>And your last point is wrong. ML models are studied and understood much better than human reasoning.

Is that why new ANN architectures are almost universally constructed by trial and error?

We explain how to understand handwriting to every single child that goes to school. It's not the answer you want, but it's the answer that actually matters here.

Trying to equivocate AI and human cognition in this way is completely disingenuous.

Human reasoning is not 100% reliable, but we know very well in which ways it's unreliable and how to deal with it. We have shared biology and millennia of experience trying to empathize and communicate with others.

Not sure I'm sold on LIME and other similar approaches, though. Seems like a lot of deep learning people are all too happy to substitute "intepretability" for actual explanations.

Also, people should have the right to know when machine-learned models are used to make decisions about their lives. They should be able to ask why a particular decisions was made and get that information.

This is real AI ethics.

https://www.youtube.com/watch?v=CEXSSz0gZI4

It is made by a physicist, not mathematician, and it's a great combination of informal explanations and carefully covering the actual math definitions.

I feel that his approach works way better than making analogies.

There is another course from Wildberger. A bit more hardcore, but also has some interesting perspectives:

https://www.youtube.com/watch?v=Ap2c1dPyIVo&list=PL6763F57A6...

You clearly don't understand the concept of whistleblowing. And you did not read the article you're commenting on:

“Cracking the password would have allowed Manning to log onto the computers using a username that did not belong to her,” the indictment says. “Such a measure would have made it more difficult for investigators to identify Manning as the source of disclosures of classified information.” So the goal was to protect Manning’s identity, and Assange offered to assist.

Well, I happen to value Glenn Greenwald's analysis of the impact this prosecution will have on journalism a tad more than unsubstantiated comments from random people on Hacker News.

He has not been under trial yet, so how do you know with such certainty he actually did that?

>Press freedom and protecting your sources is not the issue here;

A lot of very prominent people who deal with politics, activism and journalism disagree with this assessment. On top of that list would be: Glenn Greenwald, Edward Snowden and Noam Chomsky. Their reasoning seems pretty solid. And unlike some media people, they didn't suddenly flip-flop in their positions on such issues.

There is a lot of hypocrisy going on, but this doesn't mean there is no right answer or that the right answer is to invite government fucking censors to control everything.

There are pretty obvious things platforms can do:

1. Instead of opaque algorithms, give people control over what they see.

2. Either do bare legal minimum of moderation or create clear, exhaustive, stable and unambiguous rules for which content is and is not allowed on the platform. Once the rules are set, take a stand and abide by them in all cases.

https://www.thisisinsider.com/social-media-link-depression-l...

https://www.cbsnews.com/news/rise-in-suicide-and-social-medi...

As evidenced by what?

From everything I see, deplatforming creates filter bubbles, which lead to radicalization, which lead to hate.

After the rise of deplatforming tactics in the last 5 years, American society is more radicalized than before.