# What is the ideal TRIZ ?

## Y. B. Karasik

### Thought Guiding Systems Inc. Ottawa, Canada e-mail: karasik@sympatico.ca

TRIZ is a technology, the technology of innovation. And as such, it is subject to the same laws of evolution as any other technology. Hence, it makes sense to speak, for example, about the ideal ARIZ, the ideal contradiction matrix, and the ideal TRIZ on the whole.

What is the ideal ARIZ ? It is such an algorithm which contains no steps but still works. Surprisingly enough, the evolution of ARIZ went in the opposite direction. The first version contained just a few steps. But in the course of time it swelled significantly.

Similarly, the ideal contradiction matrix is no matrix at all. However, some researchers are pushing for expanding the matrix.

By the way, there is an interesting technical contradiction associated with such an approach. Altshuller managed to create a matrix, which is more or less accurate. By the latter I do not mean that it is very helpful. However, when you eventually find a solution, you can retroactively recognize that it is based on those principles which are suggested by the matrix. Thus, the matrix is accurate from the "explanatory" point of view rather than an effective tool of problem solving.

The attempts at increasing its effectiveness give rise to the swelling of the matrix by adding more contradictory parameters on the top of the initial 39 and by adding more principles on the top of the initial 40. However, this results in loss of its accuracy. The chances that solution has nothing to do with the principles suggested by the matrix, increase.

How to get rid of this contradiction ? Of course, by getting rid of the matrix itself !

But may be we should measure the ideality of TRIZ tools not by their sheer complexity. May be the correct measure is as follows:

 The ideality of a TRIZ tool = effectiveness of the tool complexity of the tool

For example,
 The ideality of ARIZ = effectiveness of ARIZ complexity of ARIZ

Then it might appear that ARIZ-85 is more ideal than ARIZ-59. At least one can argue that increase in the number of steps was compensated by a manyfold increase in the effectiveness of the algorithm. The critics disagree. They claim that effectiveness remained the same. The experts in fuzzy systems say that fuzziness cannot be eliminated by increase in the number of words !

On the other hand, nobody knows how to correctly measure the effectiveness of various versions of ARIZ. By conducting the comparative tests ? Have anybody done this ? And how to factor the difference in people participating in the tests ? (you cannot try the same people on ARIZ-59 and ARIZ-77 !)

I met inventors who got acquainted with ARIZ back in the 1960s and they were very happy with those versions to this day ! They did not want to learn anything else ! By the way, some of them were most prolific inventors of all familiar with TRIZ. So, has TRIZ been evolving towards the ideal TRIZ ?

(TO BE CONTINUED)