On Writing

Steven King has written on writing. A decent book, workmanship in nature - like his writing - and a combination of a lite autobiography and a how to and a philosophy of writing and an encouragement book.

The book does not gel together very well, which is likely a result of how it was written - in chucks, well separated in time (which he warns against...), and finished while he mended from being hit by a truck. The autobiography starts the book, to tell the story of one writer. The one thing that stands out is how much is mom encouraged him - encouragement can go a long way.

The how too is short, but as far as it goes decent. He simply points to some simple guidebooks elsewhere, rails against common annoying habits [0], and makes some simple points such that a sentence is simply a verb + noun, and can even be stripped down to just two words: "Fluids flow." [1]

The final part of the book is a "I got hit by a truck and it sucked" discussion on life, why writing matters, what it means to King, and some details of his booze + drug problems [2].

All through the book words of encouragement for would be writers are scattered, and King also includes a program on "how to be a writer" in his section on how too (in a nutshell: read lots, write lots: 3-6 hrs a day of reading/writing, write 5 pp a day, everyday, read at every chance [3]).

The book is a quick and enjoyable read, its major flaw actually underscore one of his warnings (i.e. it not gelling together likely directly links to how the book is written [4], as discussed in the book...), and King encourages all the way through.

On writing is flawed, but cheap and quick. Well worth the effort.

Notes;
[0] Of the type railed against in Orwell's "Politics and the English Language".
[1] The simple aspects King discusses are actually pretty encouraging, as he strips down grammar to the simplest to show the heart of the situation. King once taught basic English, so he has the chops down.
[2] King states that booze does not an artist make, and that while many of his best works came out of a booze and coke fueled fire that... um... booze is bad, and it is a cop out to say an artist is more sensitive and thus needs to blunt life with booze/drugs. Though artists are more sensitive, don't get him wrong.
[3] King gives the would be a day off, but notes he takes no days off. No Christmas. No Halloween. No birthday. And he writes 10 pp a day. Included is a "decent books I've read" list.
[4] The book is somewhat repetitive at times...

Can markets provide disinterested ratings?

Free-market anarchists believe [0] that we can do without government. For example, private rating agencies will spring up to provide a "seal of approval" that currently regulation provides: the Food and Drug Administration will be replaced by some private safety standard company
specializing in rating drugs, car safety will be ensured by a private corporation that tests cars, etc. By having a seal of approval a company can charge a premium, and rating agencies will be kept honest as they make their bread and butter off their reputation as honest and accurate providers of quality. At the same time the costs of regulation are removed, and consumers are given a choice - they can judge for themselves if they want the unrated product, or the more costly product with a seal of quality on it. Win-win.

But is this likely? The accounting firm debacle (remember Arthur Anderson?), followed by the more recent credit rating agency failure, suggests that it is not. One can claim that small number of players is the cause, and that imperfect competition is the problem - we just need more rating agencies to compete, with less restriction of said agencies, but experimental economics suggests you can have a sparsely populated market and still get much of the benefits of "perfect markets" that economists have worked out [1].

A naive public choice model suggests why things will fail: conflict in incentives. Public choice makes the seemingly underwhelming assumption that people are people, and works out consequences of this assumption. Special interests can aggregate enough power to effect rating agencies to meet the special interests needs rather than the institutions involved. Factions within companies, stockholders, and individuals at leveraged positions will have personal incentives to game the system. Further, only the stockholders, or the company itself, will have incentive to pay for ratings: both of which may have reason to have overly rosy ratings [2].

The obvious solution is to have each and every consumer of companies’ products to share in the cost. We currently have this situation - it is called regulation, the costs of which are passed on to consumers and taxpayers [3]. What is the difference between regulation and a rating company? Admittedly, only a matter of degree: special interests groups also have incentive to influence the regulation, but here we will have more competition among interest groups, tending to cancel each other out [4]. This idea underpins James Madison’s theory of republics.

Ratings agencies have incentive to be right, and companies have incentives to have solid (and real) ratings: both survive due to reputation. But there are strong short-term incentives for numerous players within the corporations (both raters and ratees) to game reputation and trust for personal advantage.

Public choice tells us to beware of having overly romantic views of government, as "government failure" can occur due to incentive mismatch. Public choice also suggests that corporations [5] have a similar route to failure: treating an aggregate as an entity with self-interest is a poor way of predicting what will happen, and assuming a romantic view of selfless individuals taking on the persona of that aggregate institution is a limited reflection of reality.

Sure, it may be in a companies “self interest” to maintain “its” reputation, but a company cannot act – only people within that company. And people can be tempted. Yes, most people feel a duty to their country, their corporation, their church, their friends - but they are people, and will have incentives that tend to undercut selfless duty [6]. We see this again and again - sexual abuse scandals in the church and in schools, corporate scandals, and governmental failures. By identifying institutions as selfless entities working for good, and individuals within those institutions as taking up the banner of pure duty to fulfill that good, we set ourselves up for disappointment.

Notes:
[0] Or at least they claim to believe. Often their actions undercut this claim, as in the example of "pure libertarian" professors (Taxing is an evil? evil? And your salary and lifestyle comes from what? Oh, yeah, "evil". You must mean "minor annoyance", excuse me while I discount most of your arguments as you seem to be a bit hyperbolic...).
[1] But make unlikely assumptions such as: infinite number of companies, everyone having perfect information, and people making purely rational decisions. Not that these assumptions are actually seriously thought to be needed, or are even meaningful.
[2] Potential investors do have an incentive to get the truth, but here the free rider problem prevents potential investors from paying for a study - better to have someone else pay for a study and get the results for free. Ditto for consumers, as the next paragraph discusses.
[3] This whole group payment is fundamental to the idea of "insurance" - what we often call insurance is both insurance against risk, as well as a mechanism to level the variation in costs by grouping people together. This pooling payment scheme is a crucial idea that is at the heart of how we run society. There are problems with it - try evenly splitting a bill at a restaurant instead of having each pay for themselves and see the steaks and fancy drinks flow - but there are issues with its absence also - public goods are goods. Consider a society without clean drinking water or roads. The floor of a society dictates how much one can move up; if everyone around you has TB, cannot read, and only the rich can move on a private road you have significant training costs, risks (health, violence, etc. etc.), transaction costs, etc. and one can imagine serious limits on both personal improvement and the rate of social improvement. Being taxed and getting public goods like sanitation is far better than the "freedom" of not having them.
[4] Further, as there is a single "player" transparency, in principle, is easier to maintain. In practice this will only occur to the extent that responsibility is clear and efforts are taken to provide transparency.
[5] Indeed, in general, the similarity between a corporation and our modern society is pretty high.
[6] At the very least, a subset of people will feel this pressure. If they also believe that such behavior is justified they be attracted to such positions and look for such opportunities.

Natural Understanding

Psychologists and cognitive scientists believe that our minds are like Swiss Army Knives - they consist of an array of specialized tools that can be harnessed for various tasks. The base of the tool is called a "substrate" (i.e. the hypothesized location [1] in the brain for the function), and these substrates are used for more than one task - just like you can pound a screw into a board with a hammer if that is all you have the brain uses tools for more than one task [2]. Now take this general understanding of the brain and apply it to how we must understand objects in the physical world in order to manipulate them - it is speculated that we have an "intuitive physics" substrate in our brains [3]. This intuitive physics is one reason physics in school is so difficult - our built in predictions and feelings of how things behave, which work in the real world on the human scale, are actually *wrong*. Our physics teachers do not bother to contrast our built in expectations and how things actually work, and we often cannot move past this. But more importantly than poor teaching is this: if substrates are reused, can we expect our intuitive physics of the physical world to bleed over into our intuitive understanding of the social world?

There is some evidence for this: an interesting study looks at how people expend thought - if they fill out surveys attached to heavier clipboards they will put more thought into their responses. Just like heavy objects require more muscle power, our brains seem to be tricked into thinking that more brain power is required. This is automatic and unconscious. This is suggestive that our intuitive understanding of physics is also used in social and conceptual spaces. And even if the "physics module" of our brain is not specifically resued in a certain situation we may be able to intentionally apply this module to a specific action.

So, what other built in physics understandings do we have? Friction. Inertia. Can we game this natural understanding to guide our actions?

What do we call these terms in social/conceptual spaces? "Transaction costs", and, well, "inertia". We are all familiar with friction - if many steps are required to do something we realize that the friction is simply too great, and we don't bother to start the project or we quickly give up. Even if we know the end is valuable, the effort requried is simply too great [4]. Related to this friction is risk - if 1/20 talks are worth sitting through, then it is most likely not in my interest to go to a talk [5]. Why don't people go to free financial advice talks? They are not free - I will most likely be wasting my time. The "friction" is simply too high - where friction is the wasted energy that depletes ones efforts towards some constructive end. If we really want to improve things for ourselves we act to put friction between us and things we do not want to do, and remove friction from things we want [6]. Don't want to be fat? Don't keep junk food in your house. Block out a time for exercise.

As for inertia, we all know that once we get going we tend to keep going. We can use this intertia principle to trick ourselves into action - working on something for 5 minutes will often turn into an hour, as once we overcome the inertia of inaction continuing motion is easy(er). Or we can do a minimal and small amount of something, and before we realize it we are doing more and have a new habit. The trick here is to use our built in "intuitive" understanding of the mechanical world and apply it to other aspects of our lives, and since the actions will be underpinned with an intuitive feeling of "yeah, this is how stuff works" it will naturally flow.

Want to make better decisions? Apparently just pick up a heavy clipboard and work with it [7]. Build up or remove friction in order to guide actions. Use your knowledge of inertia to get yourself moving. Think about your natural understanding of the physical world, conform your actions to this understanding and you will likely have better results.

Notes:
[1] Sometimes this location is a specific well known area of the brain, well studied by MRI and other imaging studies, and some are not well known or thought to be dispersed over the brain.
[2] Just like we cannot afford all the tools we might like, we cannot have every tool in our brain that we might like. This invariably will lead to using a hammer where a spatula would have been better.
[3] Indeed, it is hard to imagine how we could survive if we did not have an intuitive physics module. We can play catch with ease, just like we can "learn" to count with ease - most of the things that we learn are only learnable as they are largely built in and we simply refine our use of these innate skills.
[4] Consider citations versus hyperlinks. The difference in the effort required to look up an article cited is fairly significant - think about the old fangled days when grandma had to go to the library, look up the book in the index card, hoof over to the shelf, then flip through the bound journal until she got to the article, then read the abstract to see if it is worth reading more. Now we click. An order or two magnitude in transaction costs difference is huge - Hamming's Law: Enough quantitative difference becomes qualitative difference.
[5] Unless I have more information - such as knowing the talk is going to be good via word of mouth, or seeing a well crafted abstract.
[6] An excellent book on this is Nudge: Improving Decisions About Health, Wealth, and Happiness.
[7] We all have heard of the "speaking conch" idea, were a speaker must hold a token and gets the floor. The urban legend is that ancient tribes did this, and this ensures that a free for all doesn't derail discussion. But perhaps the conch was used not to control the audience - but instead to control the speaker; by holding the conch they would more carefully speak due to the weight. Maybe meetings would work out better if people had to hold a 5-10 lb object, and we wouldn't have to sit through mindless and inane chatter...

Kiss of Death: Educational

I have recently been thinking about computer programming, and did some looking into Logo.

There are two main flavors of programming language: languages that closely follow "Turing" machine like style and closely resemble the actual (von Neuman) hardware, and languages that closely follow lambda calculus. The Turing style is very mechanical, gear and switches, and flows. The lambda calculus is very, well, much like calculus with everything being a function. C and Scheme are the cardinal languages representing these two types. Most languages are an offshoot to these two (formally equivalent) ways of doing things, with the von Neuman/Turing style being the most popular as it has (1) less of an entry barrier (initially easier), and is (2) faster as it is "close" to the hardware and thus makes effective use of it [0].

Now that hardware is cheap and powerful, and programmers time is more valuable than computer time, the "gold standard" of (fast!) C is less of an issue and various ways of making better use of programmers time is coming to the fore - for example, Java is popular is it is easy for a so-so programmer to crank out code in teams. In general there are many many languages, as there are many many uses for a computer and thus many niches exist for a programming language to fill (plus programmers tend to like playing around and creating languages). Here I want to briefly talk about one - Logo.

Yeah, you know - Logo. The 40+ year old language that is an offshoot to LISP (the original lambda calculus language), the language that has the little turtle that can draw on the screen. Turtle graphics - were one can draw neat little pictures by having a "turtle walk around the screen". Why did Logo never really take off? Why did it die?

It was educational.

"Educational" is the kiss of death. Do you want to look at irrelevant and pointless discussion? Look at X from an "educational" point of view. For example, "physics education". Or computer languages for education. Anything that enters the niche of education is bound to die a painful wallowing death, most likely scarring many children in its death throws.

There are likely many reasons for this - but take it as an axiom [1].

Logo was a victim of moving into the wrong niche. Logo seems like an interesting language - it can be used to teach a fundamentally powerful way of thinking (functions! recursion!) and has cool graphics too boot. But it was a powerful elixir given to people without the proper skills and incentives and want to learn it, and a watered down weak and mostly pointless version was poured out for the students. "Now after me: type {blah blah blah} {blah blah blah} See the turtle move? Neat - huh?!", "Um, no. What's the point to this?" [2]. Oh good, it is now time to socialize/"discuss"/"explore".

Now, what lessons can we draw from this?

Never, ever, ever do anything for educational purposes or spend time on something that is mainly sold as educational or go into education. You want you kids to be smart? Baby Einstein ain't going to do it. In investing your time never decide to do something because it is educational. EVERYTHING IS EDUCATIONAL. Education is a side effect of action and reflection. Pick to do something that is useful, or fun, or challenging, or orthogonal to what you know, or is part of the basis set of society, or that hot chicks/guys do, or is scary, or seems neat. You will learn. "Educational" things are often infantile versions of something real, where the dumbing down process kills it - or something that means well but is killed by the educational system. The lack of respect this process shows to the intended audience is sad, but the undercutting of actual education is what is truly unfortunate. The creators of Logo must have shed tears. If you love kids and want to teach them, don't move into an environment that is structured against this aim. I'm happy I have had the teachers I had, but you see many of them as empty shells by the time they are done. Maybe private school is a viable option, or writing kids books, or teaching summer camps, or having kids, or teaching Sunday school, or volunteering - but seriously, going into public education is not something conducive to education [3].

If we contrast Logo with BASIC, Logo comes up short. BASIC is a language that is vastly inferior to Logo in many respects, but actually was successful. The language still lives and has a vibrant grass roots community of people playing with it and making programs [4]. BASIC was basic - yet not cute and infantile and dumbed down. Programmers used it. Hobby magazines existed. It was real, not some pablum that was spoon fed and curricularized. Logo entered the maw of the educational system, in the hopes of bringing play and possibility and programming to the people, and it was smothered. BASIC was a pathetic tool compared to Logo, yet it was a real tool and one that had life external to schools.

Maybe Logo was too complex to be expected to thrive and no matter what would have died, but I suspect not. I sense that if the makers of Logo had realized what seems apparent now - that educational = kiss of death - and had targeted Logo to real life, not the holding pen of schools, it might have succeeded. The difference between school and life is somewhat like that of "astroturf" and "grassroots" movements, we don't expect much out of astroturf. It is not an ecosystem with evolution and life and wonder and creation and death and mysteries, it is a plastic replica sold by the square foot and colored with cheap dye. Logo expected to thrive and spark a revolution in the worse possible environment [5].

The key lesson to learn from Logo is this - "educational" is an euphemism for something that lacks vitality and value. Do not invest time and life into "educational" pursuits.

Notes:
[0] The magic of calculus is we can train monkeys to use it - engineers learn it all the time - as it's notational power is so great that you can get by with turning some gears and getting correct answers. But to really get calculus is harder - but very powerful once you do. Lambda calculus is like this: there is a barrier to getting it that prevents mass use. But if you get it you can do amazing things.
[1] For one, we don't select our best and brightest to be teachers. Can one really launch a revolution in thinking using third tier thinkers to propagate it? Is it plausible that "new math" or a dialect of LISP is going to be understood deeply and communicated powerfully by teachers, and then have the revolution spread from the school? Maybe for simplistic ideas that push built in buttons, but really: LISP? New Math? Secondly, budgets are huge and change is slow - so once a toehold is found, evolution is going to die - so if the idea is not perfect to begin with you just locked it in to a straight jacket. Thirdly - politics. Politics is about spending other peoples money - and thus you will see decisions that (1) can be sold to the lowest common denominator as a "good idea", (2) fear of responsibility for taking risks, (3) poor decisions as it is not really your money one is spending, (4) community and political and bureaucratic meddling and bickering, and many other factors that lead to bad decisions. Once something is put into the political domain you might as well hand it over to the USSR or GM to handle - because that is the level of quality you will be getting. The environment is poisoned. Do not expect deep thought, novel insights, or new (positive) revolutions to spring out of education.
[2] I actually was exposed to Logo in this manner - I had no clue about any of the ideas underlying Logo, or why we would want to use it, or how it was interesting. Not until years later did I even realize that I used Logo - mindlessly typing in some instructions to get a turtle to move, and then "exploratory learning" which consisted of letting the students do what they wanted (=talking about non-Logo/turtle stuff=fun but not "learning") after a brief "teaching session" (see above "type {blah blah blah}"). And this is in Canada - you know, the country to the north of the US with a semi-functioning school system! You can't teach a whole subtle and difficult theory of teaching - such as constructionalism or Socratic method or anything - in a few teacher development days or courses and expect useful outcomes.
[3] It really is surprising how badly schools actually teach. I doubt we could intentionally make a worse system - and for the most part every single person involved in the sausage factory of public schools has the best of intentions. As for being a teacher - look at the best writers, do you think they got that way by going to school to learn to be a writer? Sure, some did go to school in writing - but the best writers taught themselves through action. They would be good writers if they went into engineering instead (although they would have less time to spend writing, and some less exposure - so that route would slow them down...). They became good writers by living and practicing writing. Why would we expect an university degree in "education" to produce good teachers? The degree acts as a filter & signal: it, in combination with union rules, protects jobs by reducing supply and it signals that the person with the degree is serious enough about the work to sit through 4 years of courses. If we actually cared about having good teachers would we even have "education degrees"? Or would we select teachers out of the general population of people? I would rather have some former mechanic teach me mechanics, a writer teach English, a mathematician math. Sure, teaching requires certain logistical skills - but certainly not 4 years worth of them. Since teachers get the summer off, why not get new hires at the beginning of the summer break, train them up, and then then apprentice them into the system with training and interaction with other teachers? Oh yeah, I know why. The teachers are of the publicani class - and thus a solid voting block (i.e. have political value), and imagine what the system would be without the current setup (e.g. education degree & unions) - mostly retired people, people changing careers, and others that are nothing like the (typical) current teachers. The system is basically politically locked in - a large voting block would not like to see the change.
[4] And after we "learned" Logo in class, the cool kids would type in a little BASIC program that flashed colors on the screen, or play a little song, or made letters fall. The kids would play and explore and have fun with BASIC - and note my use of cool. Cool - not just nerds, but popular kids would learn the little "viral" programs. The makers of Logo would scream in horror at this, as it was exactly opposite to their intent and dreams. It was to be the other way around! Logo was to be fun, to be explored, to be used like art and against the dictates of the man! Logo was top down and withered as a distributed network of people experimenting and exploring did not exist.
[5] Ironically, Logo may still live - for it now is being used for stuff (versus "in education"), namely it is being used in studies of distributed systems (i.e. lots of "turtles" or "agents" that act and interact).