Several groundbreaking scientific discoveries have happened by accident, for example the effect of X-rays, penicillin, the microwave oven, plastic and Viagra. I have experienced an accidental scientific discovery myself, where a programming error in the implementation of an entropy-based metric turned out to be advantageous in my research, something that subsequently also could be verified theoretically.
Have any of you experienced anything similar? Can such accidental scientific discoveries, which may be groundbreaking as the examples show, be utilised in any structured way? For example by using some sort of process for performing controlled accidental experiments?
Accidental discovery - serendipity - is very common in science.
The phenomenon was first explored by Koestler in his book The Act of Creation, in which he drew an analogy with humour. Both involve taking an idea and putting it in a different context, where it may be absurd (humour) or enlightening (discovery).
I explained the phenomenon in my book The Serendipity Machine (Springer 2004) and argued that many areas of modern computing use serendipity systematically. In Data mining, for instance, Knowledge Discovery exploits what I call the serendipity effect: combine enough different data "layers" and the number of possible combinations grows exponentially so that "interesting" relationships are bound to emerge. Swarm intelligence and evolutionary computing use an effect known as Wisdom of the Crowd (James Surowiecki 2004). Ants for instance, find food by sending off scouts in many different directions; most find nothing but a few are bound to hit the jackpot.
More recently (Of Ants and Men, Springer 2014), I've argued that serendipity is related to a more general process by which side effects of activities lead to unintended consequences in different contexts - responsible for accidents, as well as many socio-economic trends. I'm currently exploring the basis for all this in terms of underlying networks.
Accidental discovery - serendipity - is very common in science.
The phenomenon was first explored by Koestler in his book The Act of Creation, in which he drew an analogy with humour. Both involve taking an idea and putting it in a different context, where it may be absurd (humour) or enlightening (discovery).
I explained the phenomenon in my book The Serendipity Machine (Springer 2004) and argued that many areas of modern computing use serendipity systematically. In Data mining, for instance, Knowledge Discovery exploits what I call the serendipity effect: combine enough different data "layers" and the number of possible combinations grows exponentially so that "interesting" relationships are bound to emerge. Swarm intelligence and evolutionary computing use an effect known as Wisdom of the Crowd (James Surowiecki 2004). Ants for instance, find food by sending off scouts in many different directions; most find nothing but a few are bound to hit the jackpot.
More recently (Of Ants and Men, Springer 2014), I've argued that serendipity is related to a more general process by which side effects of activities lead to unintended consequences in different contexts - responsible for accidents, as well as many socio-economic trends. I'm currently exploring the basis for all this in terms of underlying networks.
Thank you, David.
Very interesting links about the serendipity effect, and also interesting books you have written yourself. Evolutionary computing is clearly one way to try to break something in a random way to see if the end result is better than the original.
Simply by keywords and Google.
Also, Intellectual Ventures owns a huge percentage of the world's Intellectual Property, so just searching against their database would yield much fruit.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Artificial Intelligence