Saturday, January 17, 2015

Open mic night on Rondam Ramblings

Whew, I have just spent an inordinate amount of time responding to various dangling comment threads.  Despite my best efforts, I fear I may have left some points or questions unanswered.  If I did, I apologize.  If there's anything you'd like me to respond to that I haven't, please post it in the comments section of this post.  (If you're picking up a thread from another post, please include a link.)

My thanks to Luke and Publius for ongoing lively discussions!

13 comments:

Publius said...

Cognitive Distortions

Filtering: We take the negative details and magnify them while filtering out all positive aspects of a situation. For instance, a person may pick out a single, unpleasant detail and dwell on it exclusively so that their vision of reality becomes darkened or distorted.

Black and White Thinking: In polarized thinking, things are either “black-or-white.” We have to be perfect or we’re a failure — there is no middle ground. You place people or situations in “either/or” categories, with no shades of gray or allowing for the complexity of most people and situations. If your performance falls short of perfect, you see yourself as a total failure.

Over generalization: In polarized thinking, things are either “black-or-white.” We have to be perfect or we’re a failure — there is no middle ground. You place people or situations in “either/or” categories, with no shades of gray or allowing for the complexity of most people and situations. If your performance falls short of perfect, you see yourself as a total failure.

Jumping to Conclusions: Without individuals saying so, we know what they are feeling and why they act the way they do. In particular, we are able to determine how people are feeling toward us.

Violence and Context
The social context determines whether violence is considered "acceptable" or "unacceptable." Refer to:

Robert Sapolsky, Human Sexual Behavior III & Agression I

Publius said...

Naïve realism

The illusion that:
1. We see reality as it really is – objectively and without bias
2. The facts are plain for all to see
3. Rational people will agree with us
4. Those who don't are either uninformed, lazy, irrational, or biased.

Publius said...

demagogue
noun
1. a person, especially an orator or political leader, who gains power and popularity by arousing the emotions, passions, and prejudices of the people.
2. (in ancient times) a leader of the people
verb (with object): demagogued, demagoguing
3. to treat or manipulate (a political issue) in the manner of a demagogue; obscure or distort with emotionalism, prejudice, etc.
verb (without object): demagogued, demagoguing
4. to speak or act like a demagogue.

demagoguery
noun
1. the methods and practices of a demagogue.

Ignoratio elenchi
noun
1. (Logic) a purported refutation of a proposition that does not in fact prove it false but merely establishes a related but strictly irrelevant proposition
2. (Logic) the fallacy of arguing in this way

Publius said...

Rondam Ramblings Meta-Analysis

So I wrote a program to web-scrape Rondam Ramblings posts, remove the stop words, then bag the words by each post - then do a topic analysis to come up with a list of topics that Ron posts about . . . no wait, that was going to be too much work, so I just reviewed a random sample and classified them by topic.

The results are as follows (not rank ordered or anything, or any scoring of frequency of occurrence):

1. Science & Medicine
a. Cold fusion!
b. astronomy pictures
c. hints of Peak Oil
d. a little on climate change
2. Rationalism
3. Politics/Politicians [disappointment in]
4. Government [disappointment in]
5. Religion
7. Travelogues
8. Books
9. Society and culture
10. Investment/start-up advice
11. Flying experiences and airlines to avoid

next time: ideas for new topics

Publius said...

The HRE was neither holy, Roman, nor an empire

Alright, it's been a while [various excuses]. I'm sure everyone has been anxious in anticipation of the "new topic ideas." Well, at least one of them is an old topic, but be that as it may. Below are a couple ideas.

1. LISP - I just don't get it. It's been a long time since I played with LISP; at the time, I found the parenthesis annoying. Humans aren't good at counting ))))))) [or the number of zeros in 10000000]. I don't want to write my own string functions - just give me some string functions!

This brings up the related question - if I don't like LISP[1], what language do I like? Hard question to answer. Nothing is really that exciting. My favorite language is a defunct one, RPL, which was the programming language of an old statistical analysis software package called RS/1. A modern equivalent would be JSL for JMP. I also liked a little known language, Icon, by the late Ralph Griswold - mainly because I usually end up processing string data of some sort. Some Icon ideas did get into other languages (generators, I think, but I can't say I've researched the history of generators). Egison would perhaps be a modern string-oriented language.

One new language I find interesting is Julia. It is said to be "LISP like"! Anyhow, I find the syntax to be pretty easy (easier than Python) and I like the just-in-time compilation and easy multithreading (I want to use all the cores on my multicore microprocessor).

Of course, I don't program for a living - so I never have to sell my code (or make it sell-able). Although lately I've been concluding that all software is crap [oddly, some of the better software is Microsoft Office - I suppose if you never change it, eventually you work the bugs out].

2. Object Oriented Programming. I just don't get it. I'm ready to join the OOP is bullshit movement (see also OOP is Inherently Harmful). I think I could pick up Functional programming faster.

Perhaps the above two are just sour grapes!

3. Manned space flight program: yea or nay

3(A) - New Horizons arrives at Pluto this summer. Looking forward to the pictures!

4. Time Capsules. More on this later.

5. Fine writing - quality paper and is there any ink that won't clog a fountain pen?

6. How the Ariel font induces font rage. (Don't even think of using Comic Sans!) Have you watched the documentary Helvetica?

7. Moon Advertising - clever idea, or crime against humanity?

[1] LISP, perhaps I don't know you at all.

Ron said...

> Humans aren't good at counting )))))))

If you're counting close-parens you're doing it wrong. Your editor should be balancing your parens, not you.

And the key insight to understanding Lisp is that Lisp programs are NOT text, they are linked lists. So you can build Lisp programs by parsing a textual representation of a linked list (that's what S-expressions are) or you can build them up programmatically. So in most languages, writing a compiler is a major exercise undertaken only by coding gurus, but in Lisp it's almost trivial, so whipping up a compiler for a little DSL is something you do as part of the normal order of business. It completely changes how you think about coding.

Publius said...

>If you're counting close-parens you're doing it wrong. Your editor should be balancing your parens, not you.

You might consider, ahem, when I was playing with LISP it was before their were editors that balanced parentheses.

Not that I find parentheses balancing editors that great. You go to try and insert a paren. You click. The editor highlights a paren. No, I want insert. You click again. The editor selects the paren. You type ")". It replaces the paren that was selected. You type ) again. Finally, got that paren inserted! I often turn off paren balancing.

>And the key insight to understanding Lisp is that Lisp programs are NOT text, they are linked lists.

That's helpful. Perhaps I'll plug through Lisp as the Maxwell’s equations of software one day.

Now - how should a programming language count? Suppose we have a list of 3 items. Do we number them 0, 1, 2 or 1, 2, 3?

Ron said...

> when I was playing with LISP it was before their were editors that balanced parentheses

Dude, how old are you? Emacs has been balancing parens for thirty years.

Publius said...

Emacs? How about "ed" on CP/M ...

Ron said...

Good grief, that's like saying you don't like riding in cars because the Model T was too drafty. Things have come a long way since CP/M.

Publius said...

In some ways things have gotten better, and in others they have not. Sturgeon's Law is that "98% of everything is crap."

This can be extended to software - in which 98% of it is crap. Plus the development tools aren't that great (or crap).

Publius said...

Re-reframing theodicy

Yesterday was a record day in Rondom Ramblings, with 44 replies to the Re-reframing theodicy post. It was an all-night dorm session between Ron, Luke, and wrf3 - plus Don adding some comments through the wall. Off to another fast start today, let's see if they can cross 80 replies ...

Publius said...

Overthinking the Problem

A comment on the Theodicy thread referenced this article:

Fundamentally Misunderstanding Visual Perception

.. which is about how college students often hold the belief that vision involves emanations from the eyes (!). [I was hoping for an article on really novel visual illusions]

Curvilinear Motion Problems

The article referenced another paper:

McCloskey, M., Caramazza, A., & Green, B. (1980, December 5). Curvilinear
motion in the absence of external forces: Naive beliefs about
the motion of objects.
Science, 210, 1139–1141.

Unfortunately that paper is paywalled; but this paper is not and it has a representative picture of the problem. Here is another picture of the same kind of problem.

Now, the problem I have with these problems is . . . they say a ball is moving on a curved track (... then exits). Well, isn't it possible that the ball begins to spin when moving through the track? Then it exits spinning - and a spinning ball doesn't follow a straight path? Overthinking the problem

Raven Progressive Matrices

Raven Progressive Matrices are those problems that show a series of graphic pictures, then a blank square (or ? mark). Given a selection of options, you're supposed to pick the next picture to extend the series. Solved via rule induction.

Google images finds many examples of this type of problem. Here is an example, that if you "solve," will supposedly reval that your intelligence is ranked in top 0.1% of humanity.

Here's my beef with these problems. Yeah, I gotta beef. It is simply this: multiple rules can be deduced to solve each problem. So multiple "answers" are possible. The rule I deduced may not match the puzzle writing's rule, but that doesn't mean it's not logical.