Skip to content

Computers have limitations

Thoughts about what computers can, what they can’t and whether or not the robots might throw us humans out with the bathing water.

A computer is a piece of technology for “using scientific knowledge and technical aids for practical purposes, eg for solving certain production and work tasks” faster, easier or with fewer errors than if people had to solve the tasks themselves.

When humans and computers solve tasks in the same way

it is called “computational thinking” what we do. It goes like this:

  1. a complex task is ‘dissolved’ into many small sub-tasks (decomposition)
  2. the small sub-tasks are compared to each other to find common features in them (pattern recognition / data representation)
  3. the smallest of the sub-problems and the outliers are put aside, so it is easier to get a clear vision of the common features of the majority of sub-problems (generalization / abstraction) and finally
  4. an action plan is made, including order, for solving the small sub-tasks which together are the solution of the complex task (algorithm).

Computers are unmatched in the power of computing because all the computing power available is put into solving the designated task.

Computers cannot practically compute everything

because even the most well-defined task may need more computing power than what the computers have available.

For example, a bird can build a nest with the materials that are now available and achieve that the nest is stuck on the branch in rain and wind and is warm enough for eggs and chicks. A computer can’t build a nest because there are too many computations on all the materials available and the framework of the task.

Another example, we humans can quickly identify which of the six photos show a bridge. A computer can not pass the CAPTCHA test, i.e. the “I am not a robot” test, because computers can not ‘recognize’ bridges, but must analyze all the pixels and compare with what the computer ‘knows’ about how things look.

Computers need people to ‘understand’ people

It is, of course, captivating to mention Heisenberg’s uncertainty principle, Gödel’s incompleteness theorem and “messy”, “wicket problems” as reasons why not everything can be computed. But I will tame myself, because I have no clue what that is about, other than that it’s supposedly some of the good reasons why computers can neither understand nor solve human problems without common sense. So it’s actually quite simple: When we humans act, it is because we are motivated to act. When computers act, it is because the programmer and the user have asked for it and power has been turned on.

  • We humans are blessed with common sense and understand why Chomsky’s phrase “colorless green ideas sleep furiously” is meaningless. Computers does not ‘understand’ meaningfulness.
  • If we humans do not get the information we think we need in relation to solving a task or in relation to action, then we seek them out ourselves. This is because we understand the motivation behind a given task and can analyze the planning from that perspective. Computers cannot do this, they only receive the data that the programmer and users give them to work with.
  • We humans reflect on the goal of an action (we have agency) and can therefore act wisely. Sometimes we can even see that not doing something is not meaningfull or against the intention. Agency gives us initiative and independent action because we can understand the consequences and take responsibility
There is more than motivation and responsibility that separates people and computers

Above all, it is us humans who experience and define what problems are, who the problems are problems for and how we ‘best’ solve them. We do this on the foundation of our values from the intersection between virtue, duty and utility. We define which problems have value for us to solve and which tasks are needed to solve the problems sensibly and appropriately.

In addition to being ethical herd animals, we have some strengths where computers can not yet match us in solving tasks:

  • Finger dexterity
  • Manual dexterity
  • Cramped Work Space
  • Assisting and caring for others
  • Persuasion
  • Negotiation
  • Social perceptiveness
  • Fine arts
  • Originality

Computers are technology and technology are human tools. Like tractors, pencils, and fire are tools for solving very specific tasks. We know the fire to be “a good servant, but a cruel lord“. This is also the case with all other tools we use to solve tasks for us and even though the usefulness of tractors / pencils / fire is extremely limited, we probably would rather not do without them.

Personally, I think it will take a while before we become obsolete because of our own technology 🙂

What are your thoughts?