July 20, 2021

T vs Pi

Before returning home, I was struggling with a few questions, one of them is: What should I prepare to be well-positioned when I come back home?

I was excited to return home. Among other choices, the contribution of knowledge was prevailing. I was thinking about Fick Law at that point. In a simple term, Fick Law states that the rate of mass transfer is positively linear with the difference (delta) between the source and target. I was trained and did well in the United State in one the most prestigious center of environmental biotechnology, working close with my advisor, indisputable well-known and respected in his field, the room for me to contribute would be unlimited. I saw myself know a couple of things that would be helpful to folks back home.

So I was struggling with another type of question, would be better if I am an expert in one deep domain or should I am more in a multiple-domain expert?  Here comes to T (one specific area of expertise) and Pi (π) models of knowledge. T would be considered a longtime standard for scientists, whom has developed a deep knowledge in one narrow domain, and other basic skills to transfer that knowledge. π model implies that multiple disciplinary knowledge (or at least two area of expertise) would be better.

here is no right choice given specificity of each individual . I was gear toward applied research and there was no goal to be a professional scientist in mind. I was looking myself as an engineer, and practitioner in my area. So, I planned to learn another domain of expertise. Fast forward today, my decision did not lead to my imagination that I can be powerful and resourceful, but I have no regret to learn more and the see how powerful the tools I have learned.

Here is how the π model looks like. One should possesses one deep technical knowledge, for example, environmental engineering, another prong with tools to facilitate the knowledge from that deep and narrow field.  The second area serves as a booster, in which manual efforts can be assisted by computing and automation. We saw many example of this. Engineers are no longer draw a design sketch by hands but the assisted graphic software such as CAD, SolidWorks. Doctors are lesser giving readings based their own experience but with the AI built-in the pre-screen some diagnosis. The bar would present more universal concepts and skills such as cultural, social, ethical, moral.

π sounds rather perfect and also simplistic, right? For sure, mastering more areas requires not only time but the extra effort and compromise. You are doing well with an expertise in water chemistry, why do you spend time to learn about embedded system such as Arduino? Excel is just find to calculate standard deviation, to make graph, to yield regressions at ease, why would you have a second to think you should learn R, Python just to carry out pretty much the same tasks? One thing scientists hate listening to juniors' talk is "reinventing the wheels". That mentality, whenever we have to do something that other people do well, perhaps professionally, we grow grotesque that we are wasting our time. We hate others for the same manner. You should do the REAL work.

Problems we are facing or trying to figure out are getting more complex. A Nobel winner these days are those who work in renowned research institute, where supporting structures are world-class. There is no lone wolf spent a few years in the lab that finally found one thing fundamentally change humankind's conception. Listening in a complex problem requires a large amount of vocabulary to understand why each team member with their own deep technical field explain quite differently. If you are not familiar with the terms, certainly, you can ask for a clarification a few times, but it is growing old. Eventually, you need to know other people's term and jargon. And if you are in a small group of researchers, you will need to search around, Internet or nearby groups, to learn the second prong with the simple goal in mind: helping yourself doing your job faster, less error, and most often automatically.

Or should we learn even more, develop multiple-pronged models. Most of ours are experts in multiple areas. The question is how expert are you compared to the field?

Achieving a timely goal with limited resources is a norm. While working in a garment factory, the word "innovation" or "think outside of the box" appears quite often. Manufacturing is very specific with timely work. If you are watching a line worker carrying his/her task then passing that part a long the line before reaching for another part, you have a visual of a factory. That is a place to do, to produce, and not much to think or innovate. For managers, they have more time for problem solving and preventive approach but that comes after having every other issues arising from daily production solved. Battling a question between innovation and just carrying out the daily activity at a time exhausted. Managers need to both, by the way. If you talk to the direct reports, they want to have problem solved so their plate is less crowded but also have a longer plan to reduce repetitive problem. Talking to senior management, they too want to know what is a strategic plan to grow the team, to innovate technology and improve productivity, and by the way ensuring that their is no problem to the production line. Related to searching for another domain of knowledge, we all have time but we don't have time for unnecessary tasks. That underscores the extra effort anyone taken to learn a small thing beyond the job description. Scientists enjoy luxury of free thinking more than others and grow more idealistic to solve a problem. Learning and mastering a set of tools adulterate their primary field, wasting away the productive time, and bluntly, there is someone to do that types of job.

So even desirable by the outcome, there is not much incentive to learn another area of knowledge such as programming, automation or presentation (at expert level) for a hard-core scientist. No one will receipt a Nobel prize because they know ten programming languages while researching cell metabolism or solving climate change with implementation embedded systems for automatic monitoring.

What I have not tell you about the Fick's Law is that the delta of substance is the key, another key is the flow rate. For example, if we have one cubic meter per second carrying the same delta and another is one litter per second, the first one has a mass rate of transfer 1000 thousand times greater. Unfortunately, the flow is neglected in engineering calculation. It is not hard to increase or decrease the flow rate of medium. Choose different pump, turn the valve is sufficient to adjust the flow. We concentrate much on the concentration because in the final product, the mixture of substance and water, is critical for the next reaction or delivery to the end users. In reality, the reality that I have, adjusting the flow is not easy. Certainly, I can join a university for teaching to have that flow, or even opening some sort of workshop to archive the same thing. But that is another nuance of my long story. At this point, investing in Pi model is costly and expensive but I have a sense of being luxurious and the necessary with Pi model to solve complex and multi-disciplinary problem.

* * *