Programming is Easy, Software Development is Hard
thoughtclusters.com
thoughtclusters.com
I think it has something to do with the cognitive load of assimilating language. A lot of what makes math hard is the language of math. Those that understand math find the language to be the easiest part and by far the most mechanical and simple to understand. Yet not everyone can pick up new languages. Most people as I understand it only have room in their heads for one or two. I would wager those that learn math or programming might have room for a little more. To learn a language is rote excercise, but not everyone can retain it for some reason afaik.
I think most people can learn what a "variable" is and might understand assigning a number or a "string" of letters. Yet even understanding arrays, hashes, and such low-level primitives is difficult. Going beyond that into functions, operators, generators, classes, objects, closures, data structures, memory allocation, pointers, references, etc... forget it. It takes a certain kind of person to be able to assimilate all of that and use it effectively.
Why else do we continue to invent languages "for people who cannot program?"
It's hard for a master to think like a student.
I was well into my twenties when I started programming, but I found the logic very natural - if this do that, otherwise do something else.
A friend of mine was skilled in natural languages (she spoke several languages reasonably well) and wanted to learn programming (it's just another form of language, right?), but she gave up after a semester - she found programming languages to be hopelessly nitpicky with nonsense rules and no relation whatsoever to languages used to communicate between humans.
For some segment of the population the basic premises of programming come easy, and it's just a matter of spending time enough to learn the details.
Others struggle mightily with the same basic premises, it seems like they can't really switch to the "extremely narrow minded one step at a time and you can't substitute uppercase letter O for numeric zero even though the look the same" frame of mind that is a prerequisite for doing any kind of programming in any language.
It's possible that programming is learnable by everyone, but it certainly seems to come more natural to a few.
I suspect that the difference is more deep seated than we normally give it credit for, possibly even somewhat analogous to the mix of "fast twitch" and "slow twitch" muscle fibre (http://sportsmedicine.about.com/od/anatomyandphysiology/a/Mu... ) that can give someone a natural advantage in certain kinds of sports.
I find that funny. Has she tried learning Russian?
As far as being "just another form of language;" I think the assumption is quite reasonable. However, like any significantly complex language it's easy to learn enough to get around and order dinner. But to be conversant enough to explain to someone how to perform a complex task in that langauge is quite another. I read somewhere once that people generally only have a certain amount of space in their brains in which to assimilate a language to such an extent.
Similarily, once you are well versed in a programming language the compiler really stops bugging you about trivial errors in syntax and grammar.
However, it still won't forgive you if you make them. Just as I won't forgive people when they make glaring mistakes in their use of English. Yes I can still understand them, but I have the advantage over a computer -- there are probably half a dozen non-verbal cues that I can use to decipher your intended meaning (and is probably why online discussions are so prone to turning into flame wars).
She can't or won't make the mental leap from basic algebra skills to variables and functions. Mentally unrolling loops or recursion, forget it. I don't even bother trying to explain data structures like stacks and heaps. How a programming language [that's mostly human readable] gets turned into machine interpretable information is black magic. She even took an intro to programming class as part of her degree, so I'm not the only teacher who's failed at the topic. It can really be frustrating at times because I have a respectable knowledge of geology and its industries and can conversationally hold my own with her and her peers, but she can't reciprocate at all with my peers.
For my sanity, I've always assumed some people just can't readily grasp the abstract symbolism and recursiveness programmatic thinking requires.
This is perhaps the biggest obstacle for people who are trying to teach others to program - and so I agree with you here.
However, it does seem like there are some people who will just never be able to pick it up - even though they are demonstrably smart in other areas.
There will always be those with exceptional aptitude. But if you must be exceptional to simply understand a subject as broad as computer science, that's a problem with how we conceive the science.
The Romans thought multiplication was really hard, too. Exponents were literally unthinkable. We know better not because we're that much smarter, but because we have a simpler and more powerful way to think about it.
I can conversationally say to fellow computer scientist "I organized the pieces of data into a red-black tree" and that's totally programming language agnostic and conforming to English syntax while still requiring a hefty amount of CS knowledge to grasp the actually meaning of what was said.
There was a study two years ago showing that learning maths literally rewires your brain (noticeable on scans)... so it's not just a matter of adding some more knowledge - it's a different way of thinking.
That's because mathematicians do use algorithms, and concepts like "position" (indexing to us programmers), though they don't have names for them. If you explain programming using the concrete things every educated person knows, it is much easier to teach.
assignment and sequence recursion / iteration concurrency
The conservative right has started a grassroots movement to make traditional measures of intellect and achievement worthy of less distinction. A PhD? Who cares? Nobel Prize in Chemistry? Chemistry isn't that big of a deal, anyone can do it. Plain talkin' people who struggle with things like math and science -- those are the real intellects. Because being an intellect is being someone who can just relate to the every day man. Everything else is just a "hobby" that is no more worthy of distinction than being good at Call of Duty.
Over 2011 expect a full court press on this. Academia, scholarship, science, math, etc... will all be attacked as not particularly interesting. Plain spoken populism will be the new metric for brilliance and the best this country has to offer.
I say we agree that yes, a Nobel Prize winner might not be destined to become a world-class CEO or whatever. The place to hold the line is in the value of their contribution.
Perhaps we can spin this potential trap back towards reality. Really, when someone becomes a flawless violinist, or programmer, or chemist, what they have made is a tremendous sacrifice. Being best in the world at those things means letting other parts of your mind lie fallow.
I think I might have been more fun to have dinner with when I was in punk bands and reading more books. By dedicating myself to tech I'm probably making a bigger difference in the world, but at some cost to my well-roundedness.
We should laud the winners of intellectual pursuits for the sacrifice they have made for the rest of us.
The right would be having less success with its "movement" if science had done a better job of policing themselves and staying scientists. Instead, across a wide variety of fields, they've been getting into politics. Now they are beginning to experience the consequences, and frankly, "the right", bogeyman or otherwise, isn't entirely wrong here. We don't need more politicians, we need scientists.
It doesn't count that it's really important!!, like the global warming scientists. It's still scientists in politics, not merely saying what is happening but what we should do about it and when. It doesn't count that you like what the political scientists are saying, it's still scientists in politics.
The big problem is that when science goes head to head with the "politics" of the day then there are problems. Galileo encountered this, as did even Darwin (he elected to withhold publication to avoid many problems). The end result in the short-term is the scientist is ostracized ... the long-term vindicated.
Now maybe you're asking scientists to do the same today. Don't publish papers or offer recommendations if it will ruffle feathers. I'm sure many scientists do this, but frankly I don't think this is a satisfactory answer.
Most of those people aren't scientists.
Edit: on re-reading, that sounded way snarkier than I wanted. What I mean is, actually do the exercise of trying to find quotes of scientists playing politics. It's not as easy as you seem to think.
That said, his position is not that "plain talkin' people who struggle with math and science" should be lauded but that social skills are not skills that can be learned through formal education. His point is that group dynamics, being able to read people and lead people, is a very subtle art that is learned throughout childhood and adolescence and Chua denied her children the opportunities to learn these skills by refusing them sleepovers and plays and such.
His point is actually slightly different than what you are saying, and this subtlety is important. Because while Brooks is on the right, is one of their more astute thinkers. Brooks is actually saying that formal education is easy. Not intellectually challenging, and largely a playground. Lets take some quotes:
"I believe she’s coddling her children. She’s protecting them from the most intellectually demanding activities because she doesn’t understand what’s cognitively difficult and what isn’t."
"these and other social tests impose cognitive demands that blow away any intense tutoring session or a class at Yale. "
"Chua would do better to see the classroom as a cognitive break from the truly arduous tests of childhood"
He's not simply saying that there exist other skills that aren't taught in the classroom and that these ar also important and difficult. He's saying something much stronger. He's saying the things that people like Chua value (math, science, arts, literacy, etc...) are the easy things -- the things not to be valued. They're the things you do when your child needs a break or to be coddled. These are the things of intellectual lightweights.
The hard cognitive work is the stuff you see at the corner store in Smalltown USA. These people don't have college degrees, PhDs, Fields Medals, but they can do the actual intellectual demanding activities, not the stuff "you coddle your children" with.
Again, he is not saying this and you are also leaving out the quote at the end which is:
"I wish she recognized that in some important ways the school cafeteria is more intellectually demanding than the library." Emphasis mine.
Please read this article from him and then come back here.
http://www.nytimes.com/2010/09/10/opinion/10brooks.html?ref=...
And even that sentence you quote is completely ambiguous. Is he arguing that they're equally demanding? Or that the important ways that matter the school cafeteria is more demanding than the library or vice-versa? You can't tell from that one sentence. But if you read the whole article you can tell what he actually believes. Read the quotes I have from above that are NOT ambiguous. He clearly states that the library is LESS intellectually demanding. No ambiguity, no qualifier.
Read the article in full then come back here.
Although, I do find the below deliciously ironic:
Finally, there’s the lower class. The problem here is social breakdown. Something like a quarter to a third of American children are living with one or no parents, in chaotic neighborhoods with failing schools.
Many of these children have navigated the most treacherous of social environments. Abuse at home, gangs in the neighborhood, and no teaching at school. You'd think they'd be the intellectual highlights of our country. They're not, because everyone really knows that this is not sufficient, nor effective.
This article by Brooks is really saying two things, and is largely unrelated to his later article:
1) The free market doesn't work. People should forego higher paying jobs and do jobs that better help the country. Almost like a large-scale self-imposed socialism.
2) Businesses can't be bothered to train people that don't have college degrees. There are millions of people in the inner city that have the social skills, but not the book smarts or degrees.
Brooks's should put his money where his mouth is and implore businesses to swoop into Watts and hire the available talent. He won't though, because he doesn't actually believe that this is where the intellectual horsepower is. He thinks its at Harvard and MIT, but he'd like you to believe its in Smalltown USA.
You believe this because he makes claims like this: The free market doesn't work. People should forego higher paying jobs and do jobs that better help the country. Almost like a large-scale self-imposed socialism.
So basically, Brooks is part of the conservative right because he is more or less pushing one of Obama's policy proposals [1]?
[1] They might disagree on whether it should be self imposed or government imposed. http://www.barackobama.com/pdf/NationalServicePlanFactSheet....
Yet at the same time isn't willing to ask businesses to step up to the plate.
In other words, he's all about the subtext. He doesn't really believe that traditional measures of achievement are coginitively less demanding than 14 year old sleepovers, but he'll say that so that you value it less.
He doesn't actually believe in self-selected socialism (he loves free markets), but he'll attack Harvard grads so you value them less.
He seems to focus on industrial entrepreneurship, a capital-intensive proposition to be sure. I peg Brooks as a "supply-side" kinda guy, so he would probably point to tax cuts as the divining rod of spurring entrepreneurship in that area. Thing is, you can cut taxes all day, but if the bank won't lend you any cash because they're not lending anything much less to your really expensive first-time venture, tax cuts aren't gonna help.
And, VC cash is hard to find in the US unless you're doing software, consumer, web, mobile, or cleantech. If you're in "old line" tech like medical devices or consumer electronics, forget it. Overseas is where to go. Open up shop in Singapore if you're building the better artificial heart.
How would we know? I am not aware of any mass formal education system, state supported or otherwise, that makes more than a very token nod in this direction. The closest analogue I'm aware of is social skills training for people with autism spectrum disorders. I know that works if the participants apply the theory outside the classroom, I've seen it. Hell, it even helps a little if you don't, because your modelling skills get better once you have some idea what's going on.
Your example of skills training for autism is probably the best. There's also the self-help and pick-up/dating coach industries. Every single corporate team-building workshop. Basically all of psychology.
Social skills can certainly be taught through formal education (and programmed into robots, too). They do need to be practiced also, but claiming they're not like other kinds of information is the same baseless anthropocentrism argument the conservatives keep pulling out to deny that people weren't like apes, evolution didn't happen, god exists etc. etc.
Of course, what you refer to is just marketing, which will result in underestimating the utility of having these degrees. The fact remains, however, that it still matters if you're well-educated, and it's still going to be causal when it comes to earning income or influencing your sphere.
All that changes is the value - elite topics will be undervalued, so it will be a great time for those who are intrinsically motivated to enter academia.
Programming and debugging with existing code is challenging, and it's easier to talk about common ideals than actual solutions.
Identifying problems that are soluble and then solving them without creating too many new problems is a lot more difficult than just solving problems. And also a lot more difficult than just wandering around pointing out problems.
Writing two lines of code that works is also easy. Anyone can do it.
Writing ten thousands lines of codes, however, is complex and hard. A good software developer can manage, or tame, this complexity.
Programming is deceivingly simple. That's what makes it so hard.
There is no induction with no inductive hypothesis.
A well-defined problem. But which one was hard: Stating or solving it?
The attempt of artificially separating the "crafting skill" programming from the rest of software development (which is blurry enough) is an ongoing effort since decades, and still annoying.
I have always been puzzled by people who say that
math is hard. In my opinion, your understanding of most
math that you need is directly proportional to the effort
you put in, and less to do with your intelligence
Other people simply hate math, and as an analogy ... if something tastes like shit to you, you'll never be able to eat as much as someone who loves eating "math" AND you'll probably regurgitate whatever you force yourself to eat.
This also dismisses the research done on the ways our brain works: some people think more with the left-side of their brain, which translates in better logical / rational / analytical / objective thinking, all traits that help at doing math. And other people are downright unable to think with their left brains, further than basic functioning.
Also, consider how theorem-proving or problem-solving works: you've got knowledge of a set of theories you've learned thus far (that set can get pretty big btw) and the possibilities of combining those theories are increasing exponentially. This actually requires raw brain power (well, unless we're talking of applying really basic algorithms, like solving an ecuation of a known type, but that's not math).
When speaking of math / computer-science / programming / whatever, I actually saw people sucking at it no matter how hard they tried.
You define the problem, but how do you solve it? Especially if you already have some code you have to work with.