In his editorial Answers to the Cunning Questions About TRIZ Y. Karasik wrote: “Altshuller … was good at proposing non-trivial heuristics … in a sense that when it works it gives rise to unexpected nice solutions. Whereas … a trivial heuristic … hardly can result in unexpected solutions. Nevertheless, the latter probably works more often than the former. There seems to be a kind of technical contradiction between the strength of a heuristic and the probability that it would help!”
I agree with this statement. However, I think that TRIZ proposes a mix of more and less trivial heuristics. One of the former is the law (trend, line, or pattern) of “Decreasing human involvement with increasing automation”. According to www.mazur.net/triz, www.trizexperts.net/evolutionpatterns.htm, www.primaperformance.com/triz.htm, www.triz-journal.com/archives/2001/02/a/index.htm, www.innovation-triz.com/intellectual/, www.trizexperts.net/Tech1Rev.htm and many other TRIZ publications, it is one of the fundamentals of technical systems evolution.
However, it just means that technical systems take over human labor. This is hardly a non-trivial revelation and some TRIZ articles admit it. Here are two examples.
“Last, but certainly not least, the evolution towards decreased human involvement is again one of these lines whose obviousness at this point in history is legend. Numerous household chores are great examples of this line of evolution.” (www.innovation-triz.com/papers/forecasting.html)
“Systems and products evolve toward a state of less human involvement. This principle is very familiar. We see it everywhere, from automated tellers to automatic bill-paying. There is certainly no reversal of the trend to automate in industry. (www.memagazine.org/backissues/nov05/featu res/invondem/invondem.html)
Being a trivial heuristic, it “hardly can result in unexpected solutions”. In fact, I could not find even one.
TRIZ repeatedly claims that its principles were discovered by analyzing many thousands of patents. Attention: “Patents are granted for any new and useful industrial or technical process, machine, manufacture, or chemical composition of matter, or any new useful improvement thereof.” (www.silo.lib.ia.us/specialize d-services/patents-trademark/patent-definition.htm). They are not granted for increasing human involvement, so if such trend exists, it wouldn’t be reflected in patents. Patents are a biased source to make a generalization about “decreasing human involvement”. This generalization is not only trivial, but it’s also based on a mistaken research technique.
So, the question is, is the law of “decreasing human involvement” correct? After all, if it is a “law of evolution”, it should be a part of the “theory” regardless of the fact of being trivial and useless.
Historically, trivial and obvious tend to be wrong. Isn’t it obvious to everyone that the Sun moves above the stationary Earth? With this doubt in mind I made a little search on Internet and discovered some signs of the anti-trend. Here are two clear examples.
“Though there has been a strong tendency to mechanize and automate production during the last years many operations in assembly and disassembly are still very often performed manually. Furthermore there might be an increase in manually operated assembly or disassembly as a result of product equipment investments with shorter usage time.” (Appropriate Human Involvement in Assembly and Disassembly)
“Human involvement is a promising trend in data mining. … association rules is a relatively easy sub area to involve human opinions. … My testing result shows that human opinions really play an important role in the algorithm. They help the data mining process improve the result.” (Human Involvement in Data Mining)
Maybe an opposite heuristic, such as “Increasing of human involvement”, would be a non-trivial one capable of bringing up unexpected solutions?