Through Public Rambling, I arrived at this piece of “I don’t know nothing about the history of informatics at all”.
Some decades ago there was this idea in informatics that visual programming languages would make programming useless, programmers jobless and empower users to do everything they needed with a computer.
The funny thing is, we are in 2007 and programming is still one of the best technical careers in most places (from the point of view of employability). The problems with jobs in programming have all to do with outsourcing and nothing to do with users replacing programmers (in fact, (programming) complexity is growing, not shrinking as informatics enters more and more areas of our existence).
People that think visual programming tools will make programmers redundant fail to grasp the basic fundamental insight about manipulating computing systems: You can put a “visual”, “easy to use” interface on top of the computing system, but, at the end of the day people using those systems will have to have in their head fundamental concepts about data structures, algorithms, etc… It doesn’t matter how do you present the system to users/programmers, at the end of the day, what matters, when manipulating computational systems, is the conceptual framework inside your head. Do you have the right concepts or not? If you have them you will program in Python, Perl, Java, Visual Basic, Your_Invented_Thingy, Visual_Programming_Stuff, etc… very easily, if not things will be difficult or even impossible.
Well to be honest how you manipulate the computational system does matter (that is why most people use scripting languages instead of C or Assembler), but history has more or less proven that visual programming environments happen to be some of the worse, less productive that exist (with some small exceptions).
There is one way to make biologists make good use of computers and that is for biologists to learn the basic CS concepts. Like they have to have the basic math/statistical concepts, and that is accepted. I don’t see that happening as most biologists (surely not all) try in as much as possible to learn the least informatics at all.
Don’t think I am ranting (and please accept my apologies for the rant format) because I see my “job” (I am a CS guy) disappear, it is exactly the opposite, while this mentality of “easy to use”, “programming redundant” is around there will be an infinite space for people like me to just produce “easy to use” tools for doing a simple task. Why? Because most people lack the conceptual framework to assemble, from existing bits and pieces, and automatically process information in any trivially novel way and that leaves a lot of space for informatics guys/gals.
The problem is that, until people in biology understand that they need to grasp fundamental computing concepts the field will progress at a much slower pace that it could. I would love to see the day where I would be jobless because people control the computational infrastructure to their needs, but for now what I see is precisely the opposite.
In a self-centered way I am only grateful to the state of things…
PS – That being said, I will come back to the article/interview because I think it really addresses interesting issues in Bioinformatics that deserve to be discussed.
2 Comments to "Bioinformatics, visual programming, “unemployment”"
Please share your thoughts
Filed in: bioinformatics
Hey Tiago,
I think you mis-understood me (it is easy to do as I am seldom clear). You see I lecture computational biology and algorithmics to biologists (and computer scientists). I am very commited to the idea of biologists learning CS ideas. In fact the University of Auckland has a Bioinformatics degree program that teaches computer science, mathematics, statistics and biology all to final year (you have to be an extremely good student to master all of those areas!). The point is: not every biologist can be a top notch programmer. Understanding CS principles is very different from being able to write 100,000 lines of object-oriented statistical inference software. I want to lower the barrier that currently exists to scientists doing sophisticated science. I would not call Geneious a “visual programming environment” — it is a productivity tool (like Eclipse is for programmers). At some point in the future it will probably have scripting — but you should be aware that most biologists are not going to write scripts. At least not *this* generation of biologists. I have been programming since I was 8, and I will continue to do so till I am 80 — but IMHO you are short-sighted if you think that everyone will need to know how to program to do science in the future. Maybe I pushed your button by saying that I want to program myself out of a job. That was just hyperbole. Of course there is a growing need for computer scientists — that just reflects that exponentially growing data-rich world we live in.
Alexei
Tiago I feel we are confusing here between Bioinformatics programmer and a Bioinformatics scientist in the same sense as a Computer programmer and Computer scientist.
Dr. Alexei, I feel the bigger challenge still is data integration where Dr. Stein’s focus is, in comparison to analytical tools.