Inductive reasoning can be dangerous. For many phenomena, several failed beginnings precede a successful take-off. Examples are: 1. Initial sputterings of a fluorescent lamp finally give rise to a steady bright light. 2. Early Earth life did not immediately take off; the first successful blob was probably preceded by many failures. 3. The first few sparks may not start a sustained fire. 4. All previous civilizations (about a dozen) collapsed and failed; need we despair of ours?

In each of these cases, inductive reasoning would lead us to believe that light, life, and fire could never be sustained. Likewise, we could predict the downfall of our civilization. Yet light, life, and fire have finally been sustained.

The simple saying “if at once you don’t succeed, try and try again” is productive of hope and courage in the face of adversity. Yet it is “counter-inductive”; by the rules of inductive reasoning, we should rationally conclude that we must abandon the project. So induction can be counter-productive. Hope and courage are DEFINED as NOT yielding to inductive reasoning.

The opposite of these negative induction processes are also seen, for example in the “gambler’s trap”: just because you won the last few bets, you will not necessarily win the next one.

These inductive series, bad or good, do not have to continue. So why do we believe so firmly in inductively established natural laws? Is it simply the large number of repetitions? Sometimes they are not all that large; results are reproduced in experiments 3 or 4 times. So perhaps there even could be perpetual motion machines, and costless entropy decreases – a self-organization of order in closed systems.