The Myth of a Superhuman AI
Kevin Kelly

When I looked at the large number of reactions already given on this article I decided not to digest them all before putting a reaction of my own.

What intelligence exactly is seems hard to define. To me it seems intelligence of an entity is about the ability of that entity to perform tasks and the minimum amount of time in which each of these tasks can be performed by that entity at best.

But given the vast amount of different tasks one can think of measuring intelligence for all of these tasks combined is impossible. Furthermore, when the time of the performance of a certain task by an entity was measured, how do we determine whether this was (within certain limits of measuring) its best possible performance? And how can the changed condition of that entity after that measurement be determined to be able to predict with an acceptable amount of uncertainty what best possible performance for that task will be in the near future? An approach can be to do more measurements in time to get a better picture of the ability of an entity to perform a certain task in a certain amount of time and how that ability possibly evolves over time. And this is yet only about the performance of one certain task of a certain entity.

So when considering intelligence of an entity one must somehow limit the (kinds of) tasks to be considered to stay on the practical side.

Anyhow, all one can do is measuring the amount of time in which an entity performs a task. So maybe instead of talking about intelligence one should talk about an ability to perform certain (kinds of) tasks within a certain time.

To be continued later …..

Like what you read? Give Erik van Zomeren a round of applause.

From a quick cheer to a standing ovation, clap to show how much you enjoyed this story.