Garo Garabedyan's Divergent Thinking Blog

About Garo Garabedyan, author of

Garo A. Garabedyan picture


As you might have guessed my surname from the blog address, my name is Garo Garabedyan. Here is the place where I write about news and other things that matter to me. If you have recommended it, I owe you a beer. (Thanks!)

Blogging is committing to the freedom of the web which is to immediately visit any site we want without being dependent on a social network or mobile phone app for it. You have decided to visit this site and you can change your mind and navigate to other sites or stick with the decision to stay on this site and visit pages from it. You are free to change your mind as many times and as often as you want. Making knowledge and news both open and easily available is something we should all strive for where possible. Knowledge has become as rapidly obsolete as universally available. We live in an increasingly uncertain and volatile world. Business has become more global and diverse. Many high-ranked universities have opened their courses for the public, like MIT’s Open Course WareStanford School of Engineering, etc.

Here are a few practical approaches I focus on:

  1. Disruptive innovation – democratizes the product to the point that everyone has an access to it and targets new larger markets. Disruptive innovation is not a breakthrough but transforms a product that is too expensive and complicated that only few people with a lot of money and skills can use it into a more affordable and accessible and a much larger population can have an access to it.
  2. Divergent thinking – While creativity can be defined as the process of having original ideas that have value, divergent thinking is an essential capacity for creativity-an ability to see a lots of possible answers to a question, lots of possible ways of interpreting a question, to see multiple answers but not a single one. This blog was found just after I graduated high school. Since then this blog have provided me with an opportunity to practice divergent thinking with a certain amount of feedback from Internet readers. Read more on divergent thinking. We believe in principle over profit. Not having a profit margin. Not making the same deals or shaking the same hands as everyone else. And all of these is fine by me. We, the bloggers, are not beholden to stake, share or power holders. We answer to no one but to you. We do not operate this way because of the fun of it even though it is incredibly fun. We operate this way because we believe it is the right thing to do.
  3. Iterative process of achieving achievable goals with quick evaluations of the progress.

Image of Bartholomew JoJo Now allow me to get you back in your childhood where important things happened. Bartholomew JoJo “Bart” Simpson, a fictional main character in the animated television series The Simpsons, have rejected making any effort to understand what divergent thinking is, but liked the idea that for any option he thinks of is treated as more smarter. Bart shared that is tired of being always wrong and to fail in attempts to give the expected answer in Springfield Elementary School. As you have focused on this you can now be reminded of a deeply rooted belief in ourselves that a correct answer is exactly one. “In school we are thought that there is exactly one correct answer and you can now realize that I was smart all the time” Bart have called upon all fiction characters.

Co-founding a company with the idea to grow it, if things go extremely well, and to think disruptively to address client’s needs are ideas that have been constantly thrilling me. The knowledge about the processes and goals within a company helped me fitting in various workplaces. Starting a company is not doing whatever you want as many think. The biggest advantage of start-ups would be the speed of delivery.

Projects relying on divergent thinking while in M.Sc. Computer Systems and Technologies, Technical University-Sofia:

  • Final thesis: Hybrid parallel application (OpenMP and MPI) on C++ for biological alignment of multiple sequences, an NP-complete constraint-satisfaction search in a vast combinatorial search space, using a meta-heuristic algorithm based on a metaphor of swarm intelligence. Mentored by Prof. Ph.D. Plamenka Borovska (head of the faculty). It was a challenge to choose and implement a mathematically based function to generate unique multiple sequence alignments and calculate their count. The algorithm was tested with real-world biological sequences on 50 executions on the university cluster and 3 executions (128, 256 and 512 CPUs) on IBM BlueGene/P’s supercomputer.
  • Traveling salesman problem with parallel (MPI) genetic algorithm. Constructing 2 solutions (children) from 2 solutions by randomly selecting at which point in the tour to cross. Mutation in a solution (city tour) is implemented by swapping the places of 2 random cities.

Approximately in each of the all 4 academic years in B.Sc. Computer Systems and Technologies, Technical University-Sofia an university project relying hardly on divergent thinking was completed. Here they are listed:

  • Final thesis: Simulating molecular dynamics with Amber molecular dynamics software package mentored by Prof. Ph.D. Plamenka Borovska marked with Excellent (6). In the final project I was faced with the task with many uncertainties and it was important the time management when switching between research of the theoretical foundations of the product and its practical usage. Molecular dynamics theoretical foundations rely hardly on advanced Physics and its computation approximations. The final project was followed by a course on Application of supercomputer software products in natural sciences on the theoretical foundations and approximations behind the software packages for supercomputer simulations organized by the National center for supercomputer applications, Sofia University, Medical University-Sofia, Bulgarian Academy of Sciences (read the certificate here).

  • A long-term project on Chess game engine using Breadth-First Search with heuristic static evaluation functions was completed running on C++ on a multi-core environment using OpenMP API. An open-source chess engine Beowulf, chess engine of a project winning a world record for computation in 2004, was edited to work on OpenMP. On the final build 1285 EPD tests were executed placing the engine in sophisticated chess situations and waiting for the right move in a discrete amount of time of half of a minute. After revealing the results it seems that the parallel version performs faster by examining many more moves (higher computational speed) but fails short in evaluating the best move. It took the parallel algorithm too much time to find the best move compared with the sequential Beowulf and in some of the test cases the time limit of half of a minute was not enough for the parallel Beowulf to find the correct move while the sequential algorithm succeeds. The full explanation of the project can be read at the blog post about the chess parallel project.
  • A report about “Cross-site Request Forgery (CSRF) on JavaScript Object Notation data (JSON)” was marked with Excellent (6++) in the 6 based grading system. The document of 9 pages was referencing 14 official sources and leading books on the subject and even a bug in the public bug tracking system of Mozilla. A blog post of the report and a few other blog posts (CSRF on changing browser’s stateAjax enabled web unfolds many new CSRF security problems) on the topic followed on this blog.
  • Presented two papers in Bulgarian (the conference’s official language was Bulgarian) titled “Heuristics” and “Foundations and technologies for the Semantic web” in 5-th National Student Scientific-Technical Conference in the city of Sozopol, Bulgaria (27-30 September 2009). In the paper about heuristics I have used the following literature: a) Gerd Gigerenzer and Christoph Engel, Heuristics and The Law, 2004, and b) Stuart Russell and Peter Norvig, Artificial Intelligence- A Modern Approach, 1995. In the paper about semantic web I have used the following literature: a) John Davies, Toward the Semantic web, 2003, b) Software Engineering Radio Podcast No.116: Interview with Jim Hendler about The Semantic Web, c) Marc Ehrig, Ontology Alignment- Bridging the Semantic Gap (Springer 2006, Semantic Web and Beyond). A certificate of attendance to the conference can be found here
  • 1-st place in a programming competition UniDevelop in Technical University-Sofia was won with a C++ project. The project, its documentation in Bulgarian, a newspaper article on the cover of “Technical avangard” (published by TU-Sofia) and a scanned diploma along with an English explaination about the concepts behind the project can be found at the blog post about the C++ project. The commission have granted 1-st place to me motivated by (listed in the news article in the newspaper of the university): Active work on the project even after the marks for the semester; Completely made by himself documentation, containing theoretical explaination of the problem, practical implementation, future development possibilities and used information sources; Well structured and commented source code. The newspaper quoted me saying “I am a great fan of seeking sources and reading. When you want to succeed you must seek to complete the cycle – to read systematically in the direction you need and to know from where to take the necessary knowledge”
  • A report about “Differential equations in the development of Physics” co-authored by assistant professor Sava Donkov, Department of applied physics, TU-Sofia was presented at Days of Physics’2009 at Technical University-Sofia. The slide show presentation was with the same design of Maria Kamenova from the presentation from the previous year. The paper is published in “Days of Physics 2009. (Collection of popular and scientific papers)” ISSN 1313-9576.
  • A report “Mathematics and Physics: Searching for knowledge (Relationships and Influence between Mathematics and Physics)” mentored by assistant professor Sava Donkov, Department of applied physics, TU-Sofia was presented at Days of Physics’2008 at Technical University-Sofia. The slide show presentation to the report was designed by Maria Kamenova. The slides, the content in Bulgarian and a newspaper article about the event in the newspaper of TU-Sofia can be found at the blog post about the report. During the summer of 2009 me and assistant professor S. Donkov were preparing the report as a paper for publishing, but it was not published. The paper focuses on the first theory in the history of science-the ancient Greeks’ theory of Astronomy.

Trying to have more possible choices in the above projects was done with extensive search on the Internet for both solutions (including scientific papers) and reusable components. I have developed the framework for MCUs with having in mind the idea that everyone must be able to easily modify its code. But modifying someone else’s source code was an issue I ignored in my planning for parallel Beowulf chess engine. It happened to be very hard to read and figure out how to enhance procedural programming code (legacy code). One of the most remembering experiences was to be forced to partially revert my own work and work extra on more boring stuff in order to finish the projects which is a natural thing in many real market products.

And Now You Can Mail Me About The Things That Matter to Us

In the vast space of materials to read available on the Internet it is a pleasure for me to read your message. All comments are welcome, because you, like me:), are spending time in writing about any of these topics. I am interested in the conclusions of colleagues on the topics I work on and I would like to thank you in advance for sharing any observations you have made.


In case the form is not suitable for you write to

Written by garabedyan

May 1, 2007 at 19:25