Until now I knew the exploration-exploitation-dilemma mainly from the path finding problem in a partially unknown environment. Exploitation will definitely get you to the goal, even if not in the optimal path. You can choose to spend some resources on exploration, hoping (but not guaranteed to) find a better global solution.
It's interesting to see the same concept applied to what the author calls 'mental load'. When there are mental resources available, not being tied down by anything, I can spend some of them freely on exploration. In stressful situations I always have to fall back to exploitation-only, possibly getting stuck in the local minimum of what I'm doing.
While it is definitely good to leave the mind free of unnecessary dead weight (Did somebody already reply referencing GTD? Can't be long.), it doesn't answer the question of how much time should be spent on exploration to achieve the best results. I find it fascinating when people talk about their own modes of operation and allocation of these mental resources, for example the recent article about the 100:10:1 game design.
> In stressful situations I always have to fall back to exploitation-only, possibly getting stuck in the local minimum of what I'm doing.
The situations where you are with your back to the wall and have to hunker down to meet a hopeless deadline are exactly the situations where exploring solutions benefits the most. Taking 5 or 10 minutes before you start down a path to consider the problem from the outside does not hurt the time to completion, and often benefits it when there are relatively obvious alternate solutions that are more narrowly scoped.
Spending longer than a trivial time on exploration is only really worth it when the possible upside is also very big. Before starting a year-long project I would find it normal to spend a few weeks exploring ideas, instead of just hunkering down to write code and seeing where it ends up.
>> Taking 5 or 10 minutes before you start down a path to consider the problem from the outside does not hurt the time to completion
This is true for modern, civilized world. We humans still suffer from "savannah brain", where main cause of stress was physical danger, and taking even 5-10 seconds to consider your options might result in you ending up dead.
That's why we developed physiological reaction to stress, that with help of hormones such as adrenaline and cortisol shuts down exploratory part of our brain on chemical level.
Following your suggestion of taking 5-10 minutes to reconsider requires serious effort to overcome that reaction and, depending on your individual physiology, might even be not possible at all.
On top of that there's the thing called "stress addiction", which basically means that some people start seeking new sources of stress when their live becomes to peaceful. Which again works against your exploratory brain potential.
Mental load is not the best way to look at performance hits from psychic states. The more central concept is attention. It's the contextual switches, the act of maintaining or allocating attention and the physical effects of stress on the brain that lead to reduced performance. Though short term states (conflicting active representations, attention and working memory[+]) share some mechanisms with long term negative states ( => stress, attentional sinks meant to bias the individual towards resolution that no longer work as well in the modern setting due to reduced personal agency),
The article muddles things by overloading the two concepts.
But regarding exploration vs Exploitation, it is indeed a very important concept. It is most studied in the literature in terms of "Bandit Algorithms" (you can play Go well with a version specialized for trees), imitation learning and reinforcement learning. Exploiting too much does not often lead to optimal long term rewards or minimized regret. A fair amount of recent reinforcement learning research has been how to get an agent to explore complex state spaces more widely.
We can even look at things at the societal level using this rubric. Basic research can be seen as exploring and focusing on product can be viewed as exploitation. It is my belief that things are weighted too much towards exploitation, currently.
[+] "Another possibility, though, is that activation of the internally attended content is attenuated when attentional resources must be diverted to the articulation task. " in http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3594067/
[+] "Engle and colleagues have investigated individual differences in WMC (working memory capacity) in young adults, and have argued that WMC is related to the ability to control attention, particularly under conditions of interference or distraction" in http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2852635/
Relatedly, business schools teach exploration vs exploitation in the context of companies, with a view that a company that does one finds it very hard to do the other.
It's interesting to see the same concept applied to what the author calls 'mental load'. When there are mental resources available, not being tied down by anything, I can spend some of them freely on exploration. In stressful situations I always have to fall back to exploitation-only, possibly getting stuck in the local minimum of what I'm doing.
While it is definitely good to leave the mind free of unnecessary dead weight (Did somebody already reply referencing GTD? Can't be long.), it doesn't answer the question of how much time should be spent on exploration to achieve the best results. I find it fascinating when people talk about their own modes of operation and allocation of these mental resources, for example the recent article about the 100:10:1 game design.