Global and Universal Problems

Dennis Meadows in his great lecture at Santa Fe makes the distinction of 2 kinds of problems

Type 1 Global Problems are those that affects everyone and needs concerted action from big players like nations, for any possibility of solving them. He puts nuclear proliferation, terrorism, oil depletion etc in this category.

Type 2 Universal Problems are those that affects everyone but can be solved by local action. He puts groundwater contamination, urban air pollution etc in this category.

My feeling is in organizations people tend to use existence of global problems as reasons to not take action on universal problems.

Global problems include the ever escaping top management approvals, blanket buy in for an initiative etc.

Some even predict the upcoming REVOLUTION as solution to global problems.

Innovation is all about being able to solve those universal problems with local solutions, without being bogged down by global problems.

And local solutions emerge only from safe-fail experiments.


Safe Fail and Fail Safe Roll outs

Safe Fail and Fail Safe concept of the 2 types of experiments or trial runs is not new. I had to make a case to a strong process oriented crowd on adopting a new process and tool for allowing harvested knowledge at an organization level.

 Major portion of the debate that ensued clearly gave a chance to contrast between the 2 types of experiments. I am only taking the roll out part of the whole change (a better word in safe fail types would be adoption).

 For Fail Safe roll outs

  1. A business case that proves or claims that a change would happen on a stated benefit which is along either commercial or satisfaction value
  2. Once the business case is all clear there would be a pilot on a smaller (claimed representative) chunk of selected projects and disproportionate effort goes into making the business case a reality
  3. Now proving the benefit means you have a fixed boundary on the initial state and a new end state that has happened due to the change/experiment. Of course this will include non accounting for reasons that possibly worked in favor (that statement reflects the fact that I just finished reading “Fooled by Randomness”)
  4. Provided the sample set results are satisfactory the change is rolled out to larger set of people whose contexts are generally not clearly understood, and the effort that was put into the pilot is surely not replicated at this new large scale
  5. Post this usually benefit measuring exercise is conducted and published as is or there may be a change in parameters of what is measured as scale has changed

 On the other hand Safe Fails

  1. Do not have a “business case” as the outcome is usually not clear when begun
  2. The pilots are usually not chosen or selected but are rather like drug trials are volunteers
  3. Boundary states are not necessarily fixed and coupled with outcome not being clear
  4. Larger scales is possibly an indication of the success of the change/experiment itself, but really speaking the success comes from the number of such experiments that Safe Fail type allows that will work in context.
  5. This point is not applicable as usually there is ignorance of how the change will adapt and work.