by Tyler Cowen
on June 13, 2014 at 12:30 pm
in Uncategorized |
1. Do imaginary companions die?
2. Albert Hirschman’s Hiding Hand.
3. Should adults read Young Adult fiction?
4. Edward Conard on whether rescuing subprime borrowers would have fixed the economy.
5. Why hasn’t the price of oil gone up more? And should you in fact fear Friday the 13th?
6. Do dogs prefer to avoid being part of the 47%? (speculative, to say the least)
7. How about state licensing for those who write code?
8. Krugman post on wrongness.
1) All but one of the failures mentioned are government projects. But does that stop anyone from thinking insufficient government involvement is the problem? Of course not.
2) IT projects generally fail because of problems with design and the development process, not because of low-level coding incompetence.
3) The heart surgeon analogy is dumb. There are already strict government requirements governing software that, say, performs heart surgery–or does other dangerous things, like flying a plane.
4) The author must own a lot of brooms, since he needs a closet 1/4 the size of a semi truck to store them.
#7.4 I think the number you’re looking for isn’t 4, but 2^(64-16) =3e14…. make that one really small broom! But otherwise 100%.
How many 16 bit registers would it take to hold a 64 bit value? 4.
Sorry, I know this is a really dumb argument, but I had to bite.
16 years ago I had Serious People tell me that programmer licenses were coming in the wake of Y2K.
You might have some state requiring licensing for certain functions within the next 15 years. But this:
Laplante said he expects all 50 states to require software engineering licenses within the next decade, and possibly much sooner.
is total and complete bullshit.
“Thirty states currently require licensure for software engineers working on systems that affect the health, safety, and welfare of the public as well as those offering their services directly to the public.”
This was linked from the article. I wish you were right.
I’ve done work all over North America and no one has ever asked for my coding license. I suspect this is limited to government where no little work gets done anyway. That’s where rackets like IEEE get their juice.
Yeah, but apparently it also covers “those offering their services directly to the public”. The fact that it isn’t much enforced right now is not hugely reassuring.
Who wants to buy my ruby implementation of Quicksort?
It’s only three lines but I’ll selling direct to the public.
Just because you have a licensing scheme doesn’t mean anyone will use it.
I’m a geologist, every state in the union has a geology license, almost no geologists have one and those that do are hydrologists who dig wells and those basically working as geological engineers. The exam is basically a hydology test with some soil engineering thrown in, very few working geologist could pass it without serious studying because it has absolutely nothing to do with what they do.
The professional engineering license can cover many disciplines of engineering, but few engineers obtain it. Really, only civil engineers commonly get licensed. I predict that the % of “software engineers” (what’s that, anyway?) that will become licensed will be about the same as electrical engineers, which is very few.
Of course, The Engineer is a licensed professional engineer.
@Roy, presumably, then there is no legal requirement for most geologists to be licensed. As noted, apparently 30 states already have legal licensing requirements for programmers engaged in many kinds of work. It is all too easy to expand these licensing requirements into other areas, and one could easily imagine that happening. It is in the interests of government (increased revenue and power) and existing programmers (job protection), while the public mostly doesn’t understand software development and can be swayed by misleading fear-mongering, such as in the article.
Re: YA fiction: In my experience some Y.A. fiction is written at about the same level (both in terms of language and plot complexity) as adult fiction. Often what earns it the Y.A. label is just the absence of sex and profanity and a main character who happens to be a young adult.
Absence of sex? You haven’t read any YA in the last couple decades.
I’ve read a few of the more popular series. Harry Potter, Hunger Games, the Eragon books, Garth Nix’s “Old Kingdom” trilogy, Pullman’s “His Dark Materials” series (which has sex, but only very vaguely referenced). But yeah- most of the Y.A. stuff I read has some crossover appeal to adults. A huge swath of Y.A. has virtually no appeal to adults.
Let’s not forget that the (vast?) majority of fiction purchased and read by adults is a paint-by-numbers paperback in one of the following genres: (1) romance (2) mystery or (3) thriller [think Dan Brown , except somehow even worse]. Not everybody is reading Knausgard.
Had their been licensing requirements prior to all of the coding failures he cites, the only difference would have been that the people creating the big coding mistakes would have been licensed.
Who does he thinks get hired to write technical code for a rocket? Sure whoever they are meet the requirements of what licensing scheme would be created.
How about certification instead of licensing?
#8 – nice, interesting post. Even if his admission of mistakes does seem to lead up to the grand finale that he was not really wrong about Britain. But I can’t fault him for that, because he is right that Britain didn’t really implement austerity.
My respect for him went up with that.
+1 – Krugman admits he makes mistakes, and even points some of them out, wow! My respect for him went up +1.
You guys are easy.
All my mistakes are of the right kind except when I don’t listen to my own geniusness too.
I won’t however give my opponents anything because ALL their stuff is wrong for the worst possible reason!
But where was the admission that his supporting the VA as a better model for healthcare seems seriously flawed at this point?
Still, kudos to Krugman for finally admitting to himself and his loyal readers that he (like very other human) is routinely wrong.
I ate my brocolli. I don’t have to read Krugs swill.
#7: oh hell yes, certify and license every programmer you can find … but not software project managers and company executives. i can’t imagine those professionals attaining a higher level of skill and organization than they currently have. i am convinced they know exactly what they’re doing because every time a software company or project runs into trouble, they inform us that America lacks a sufficient quantity of highly skilled programmers…
In the future, AI will probably result in laws that make it illegal to use human written code, within critical applications. Human coding will become a hobby or such, similar to blacksmithying.
Who will write the code for that AI?
Oh, right, those robots who are going to enslave us.
Hold on, let me ask my iPhone how I should respond to this post.
Is that what it told you to say?
It said there’s nothing to worry about.
The next lower turtle, of course.
We’re already in an analogous situation: most programmers don’t write machine code directly: they write (moderately) understandable high-level code, and a compiler converts it into a machine code. [For interpreted code, an analogous situation exists.]
You can ask the same question about regression: who writes the code for the compiler? Well, it’s not magic. The first compilers had to be written directly in machine code, and they did little more than abbreviate some of the repetitive code patterns. Then those compilers were used to make better compilers, and so on.
Likewise with “Automated coding” improvements. Some simplistic “code generator/verifier” will be written by a human, which will be used to write broader generators, until the point were you specify what you want in a very high level language.
I don’t consider this analogous. Converting high level language code to machine code is a relatively mechanistic process, not involving artificial intelligence. The computer simply converts one format to another; it doesn’t decide how to implement things (except in a quite narrow sense when there is compiler optimization, but again, this is quite mechanistic). When you save a Word document as HTML, you don’t claim that the computer was the author of the document.
In some respects, compilers are similar to people’s AI imaginings. “-o3” is certainly as close as we’ve gotten to Kurzweil’s self-improving intelligence.
Yes, and they should be embarassed to watch superhero movies.
Who has time to worry that other people might be reading insufficiently emotionally complex books? Good grief.
I thought #2 was going to be http://i.imgur.com/uaJdSdx.jpg
@#2 – it’s almost as ridiculous, albeit by Sunstein: From the Abstract : “It also notes that the Benevolent Hiding Hand has an evil sibling, the Malevolent Hiding Hand.” – sounds like AlexT and TC talking about their imaginary friend Tyron or something. Economics = social sciences = mush.
Step 1: stop reading YA fiction and only read fiction for adults
Step 2: stop reading adult genre fiction and only read adult literary fiction
Step 3: stop reading adult literary fiction that isn’t serious enough (see: The Goldfinch hullabaloo)
Congratulations, you are now a “serious reader” that no one wants to talk to because you are an incredible stick in the mud.
I only read fiction that is carefully designed not to be enjoyable.
and only eat Unhappy Meals: wheat germ, carrots, whey, lentils, spirulina, oats, and raw, uncooked hay.
Lynda Barry covered this quite nicely with her 20 stages of reading. Spoiler: the last stage is reading whatever you want. http://www.washingtonpost.com/lifestyle/style/lynda-barry-the-20-stages-of-reading/2013/09/13/e82fd970-1cbf-11e3-a628-7e6dde8f889d_gallery.html#item0
#7: There is a bug in human software (coded by whom?) which manufactures an unshakable belief in the efficacy of coercion to better the human mind.
Others have already chimed in on this, but it bears repeating: programming project failures are as likely to be due to problematic design or project management as they are to coding.
Programmers are far from perfect, but bugs are relatively easy to fix as opposed to, say, a fundamental misunderstanding of the customer’s requirements. Or a project that is unrealistically budgeted. Software development is certainly hard and fraught with the possibility of failure, but the difficulties go far beyond coder skill. If you lose a war, you blame the generals, not the soldiers.
“Life-critical” software already tends to be subject to stringent requirements. Licensing programmers will add nothing of value to development efforts. If anything it will slow the flow of new coders into the field and cause shortages and delays.
Lastly, I once had a manager who had this inscribed on his white board:
“Your project can be
– On time
Can I get it faster and cheaper if I just pick one?
No, you can get it faster or cheaper.
#7. The stupid regulatory mentality just assumes that if things aren’t perfect then more restrictions must be the solution. In fact computers have added untold value over the past 50 years precisely because they are a free-for-all.
In the real world, regulations are demanded, or at least permitted by the incumbents. So this will not come to anything until programmers grow the culture or the institutions for protectionism. That will happen eventually, starting in individual domains. I hope it doesn’t happen during my lifetime.
Many programmers support licensing regimes in hopes of making it harder for folks like Zuckerman to replace them with Indian immigrants. I doubt it would work; the immigrants would all be given licenses regardless of competence (they get jobs now without regard to competence) and the license exams, taxes, and disciplinary procedures would only be enforced against white citizen coders because it would be “racist” to enforce them against immigrants. However, the prominent examples of accountants, lawyers, P.E.’s, doctors, nurses, X-ray techs, undertakers, dieticians, barbers, and so-on and so-forth do suggest that restricting entry to your firld is good for your earnings, however bad it may be for other people’s earnings or the overall economy.
Please give an example of where people were given a pass on professional licensing exams because they were immigrants.
Foreign trained doctors is a version of this. Doctors, nurses and engineers are also examples of the education and ojt being far more rigorous than any certification can be. The pe exam having a high failure rate actually proves the point because it is hard because it is non-practical.
It is like certifying professors. You can look for correlations, but the market test and signaling are the ultimate arbiters.
3. To me, the biggest problem with adults reading YA isn’t the adults. It’s the effect it’s having on what’s being published for kids.
Let me rewrite part of Krugman’s mea culpa to better reflect the truth:
“Second, in 2003 I warned about a US financial crisis driven by fiscal irresponsibility, somehow comparable to the crises in Asia a few years earlier. This was, I now believe, because I’m a political hack who roots for the home team. QED.”
Spooky ‘Honey’ Moon Casts Glow on Friday the 13th – Video
Comments on this entry are closed.
Previous post: Is the lack of war hurting economic growth?
Next post: Cities at or near their all-time peaks of excellence
Email Tyler Cowen
Follow Tyler on Twitter
Email Alex Tabarrok
Follow Alex on Twitter
Subscribe in a reader
Follow Us on Twitter
Marginal Revolution on Twitter Counter.com
Get smart with the Thesis WordPress Theme from DIYthemes.