This note is for informatics students who take modules in intelligent systems, robotics, advanced computing, telecommunications, and the rest listed here.
Soft skills and core skills: When I was a university student, many career advisors wanted us to develop soft skills like making presentations, personal relations, teamwork, communication skills, etc. Still today, I don’t dispute the importance of soft skills, because they are needed to translate what we conclude based on scientific methods to lay people. However, it is wrong to believe that soft skills are all that is important, and that learning Math is just a way to train the mind like learning Latin.
Survivors in the recent economic crisis: First, lets pay a little attention to the recent Global economic crisis. Out of all the Western countries, Germany was one of the least affected. But Iceland went bust! London financial hubs went on their knees. Wall street went nose down, but most Silicon Valley high-tech companies like Apple, Intel, Google, and Microsoft were not that humiliated, nor did European manufacturing companies like Siemens, Ericsson, BMW, Audi, Rolls Royce, Mercedes. Social media companies like Facebook and Twitter rose from nowhere, yes, during the crisis. Can you guess why? Think about who believed in an economy based on manufacturing, high tech products, software based on advanced machine learning techniques, training employees with good math and scientific skills, and who believed too much in the ability to sell something even if the product was toxic, and who gave importance to short term boosts in the bottom line (the profit and loss line) disregarding long term risks.
For companies like Apple, Google, Microsoft, iRobot, BMW, Audi, Rolls Royce, etc, the brand of your lipstick, the perfection in your tie-knot, attractiveness of the animations in your power point presentation, and the charming English accent in your polite speech didn’t come too high in the priority list, though they are certainly important. Instead, soft skills like finishing a product with unbeatable quality, user friendliness and efficiency, strongly underpinned by core skills like the ability to use cool math to solve machine learning problems, to design robust controllers to maintain stable interactions with uncertain environments, and to design algorithms that work efficiently on mobile computers mattered more. Though stock markets crashed, smart software companies that could reduce the computing time to finish a stock trading cycle flourished.
Moreover, employers in more stable high-tech companies didn’t believe too much in ready-made employees to suit their specific needs. They correctly knew the limit of universities to provide a general set of skills to think and learn in a broad class of industries. So, they took the baton from the universities where it should be taken and groomed their young staff to be productive in the specific organizational culture and vision.
I gave this brief background to help you to decide which kind of companies you should stake your life in.
Why concurrent computation is important in a rapid prototyping SME: Now, lets take one or two examples of the math you learn to see how they are important in real life. Take what you learn in concurrent computation. Imagine you are employed in a SME that provides 3D printing services. Different customers send you orders in different intervals with different levels of urgency, quantity, and value. A company can employ a good gambler with nice communication skills to handle this situation. They can keep giving very creative excuses for missing deadlines, and market the service well using valued soft skills. Well it can work in the short run. But if the industry becomes profitable, other service providers also enter the market. Imagine one of them employs a graduate who has taken real-time systems and control. The 3D printer is not a software code, but the employee has learnt to compute total time taken to finish a batch of components if batch manufacturing is interrupted to produce more high valued and urgent orders. The employee then knows how to use mathematical tools to schedule different orders in the most optimum way so that none of the finished product dispatches violate the deadlines while high valued and urgent orders are fast tracked at just good enough phase too. Moreover, the mathematical tools will also help them to decide which orders should be manufactured in-house and which should be outsourced with what terms in the contract. Very soon, the business with gamblers will go bankrupt and you will be happy to have taken real time systems and control in the university and be employed in a place that knows the value of the math you learnt.
Why learn about probability and random variables: Lets take probability and random variables you learn in robotics. You may learn the theory and do a simple lab class to see the theory in practice. However, it doesn’t mean that the lab class is designed to “train” you to suit company A, B, or C. In fact it is designed to give you as much freedom as possible to choose from among a broad array of employers.
Take a simple example of a mobile robot that looks like a cart. Imagine two identical motors drive the two identical but independent wheels, and that there is a passive caster wheel at the back to keep the platform flat on an even floor. If both motors were given electric power supply from the same battery, how do you think the robot will move? Most students first predict that the robot will move along a straight line because it is driven by two identical wheels and two identical motors powered by the same battery. Alas! The robot doesn’t go on a straight line even if you run it on a perfectly flat floor! Not only that, the robot never follows the same path across repeated trials with same initial conditions. We may use this kind of exposures to get you to think about uncertainty of dynamic interactions with the environment, and mathematical tools you can use to deal with random variables.
For instance if I ask you to run the robot 10-times from the same starting point with the same initial orientation and speed, and ask you to make the best guess of the average position of the robot after some 10 seconds, what information can you use to make a guess? Well, your soft skills of making guesses and convincing me with a proper presentation can help to survive for some time, but if your wrong guesses cost you money, what happened to some banks during the financial crisis will happen to you. However, if you notice that the position of the robot after 10 seconds can be used as a random variable, and that you can collect historical data to construct a probability distribution for the position of the robot after 10 seconds, you can use the techniques of function approximation and computing the expected value of random variables with a known probability density function to make a good guess.
You can even go on to compute the risk you have to take to predict another position if you get more money if the robot goes through that location. Then you can take the distribution of monetary rewards for different locations on the floor together with the knowledge of the distribution of the position of the robot after 10 seconds as a random variable to compute the location of maximum monetary expectation. Now, you have experienced enough math about random variables and probability distributions to be productive in any industry that has to survive in an uncertain World by effectively dealing with random variables. A smart company will take you from where you are to give an opportunity to study past cases and data to familiarize you to the special cases they have faced. This is where you have to use what you learnt in the class and labs to extend the skills to more specific industrial applications, and I am sure you will.
We know that even giants like Facebook has failed people they shouldn’t have failed. But later those who failed the interview showed their value through practical demonstration of their core skills, that proved to worth billions! Those stories bring home the message that the employers too should go out of the box when interviewing computer science (CS) graduates. It maybe that CS graduates are not great talkers, but great at coding and math that can turn things around. Therefore, do not be discouraged when some people over-emphasize the value of soft skills and let down the value of math and other core skills. You will be safer in a disciplined place of employment that honors core skills. Having said that, I strongly suggest you to participate in seminars held by humanities and culture departments. I myself benefited by participating in a recent multidisciplinary experiment titled “translocations“, where we got a chance to comment on a segment of a stage show that we had not seen before. This sudden “translocation” was a situation where you have to interpret some new experience to a totally different audience. So, such experiences do matter. However, you will do well even in such situations if you had learnt the fundamentals of reasoning in your core subjects.