Optimizing human under uncertainty

Gill Eapen
2 min readJan 1, 2023

It appears that humans did not cultivate skills of optimization under uncertainty. This is likely because for most of their time on the blue planet, their horizons were very limited with most decisions made with low uncertainty. In the last hundred years, it dramatically changed, many sporting lifespans well in excess of three times the original humans. Now, human decisions have to systematically consider future uncertainty and flexibility to optimize life.

If happiness is the singular parameter in the human objective function, optimization under uncertainty is a hard math problem. Humans are exposed to sequential decisions with options characteristics, and the decision options tree one needs to consider is complex. As future decisions are intricately connected with present decisions, any human decision to maximize happiness will have to consider all future possibilities. And, this differs from person to person and is likely not generalizable.

Although it is a difficult problem to solve, one could create a few heuristics that may be true for most.

  1. Committing suicide is likely suboptimal unless future uncertainty substantially narrows and the aggregate remaining disutility is high
  2. Accumulating wealth is likely suboptimal as the marginal utility of wealth is low after a threshold. The threshold can be derived from the required satisfaction of needs and desires within an expected lifespan
  3. Committing a crime is likely suboptimal, regardless of the objectives as the disutility associated with the expected outcomes far exceed any presumed benefits
  4. Being sad is likely suboptimal as such a state does not optimize utility within a limited life horizon that show uncertainty
  5. Helping others is likely optimal as such an act is shown to have multiplicative utility for the helper and the resources utilized is marginal
  6. Praying to God or partaking in religious activities are likely suboptimal as the uncertainty of outcomes on these activities are very narrow
  7. Having hard goals are likely suboptimal as this will lead to decisions that will not consider uncertainty and emerging changes to regimes
  8. Staying in one place is likely suboptimal as such a human will lack flexibility to tackle future uncertainty
  9. Education in a specialized field is likely suboptimal as future uncertainty may change professions and requirements
  10. Living long is likely suboptimal as the uncertainty in the ratio of disutility to utility predictably narrows as time passes

The time is limited, uncertainty high and optimization not well understood.

--

--

Gill Eapen

Gill Eapen is the founder and CEO of Decision Options ®, Mr. Eapen has over 30 years of experience in strategy, finance, engineering, and general management