Skip main navigation

The AI Effect

The idea of artificial intelligence is hard to define, but has been around for a long time. Read more about the history of AI here.

Modern artificial intelligence (AI) can appear daunting to newcomers, but it all started from a very simple aim: to automate tasks that require intelligence. The current complexity is in part due to a phenomena known as “the AI effect”.

The AI Effect

Let’s start with a simple definition of what AI is considered to be:

A computer system able to simulate an intellectual task
This definition rules out mechanical processes, such as lifting or motion and focuses instead on an intelligent outcome. The problem here is how you define “intelligence”.
AI research has followed this pattern for many years:
  1. Define a problem that the researchers think requires “intelligence”
  2. Work towards and achieve (either in full or close enough) a solution to the problem
  3. Change the definition of “intelligent”
  4. Repeat
The result is that the definition of artificial intelligence is constantly changing as technology becomes more advanced. As soon as something is achieved, the process is no longer deemed “true artificial intelligence”.
For an example of the AI effect, you can look at satellite navigation systems. Navigation is a complex process that definitely is AI but most people would not consider Google Maps (or whichever map service you use) as an example of modern AI.
There is a long-running and impossible to properly source, joke definition of AI that captures this idea perfectly…
AI is cool stuff computers can’t do yet

The Birth of Artificial Intelligence

The original seed of artificial intelligence did not come from scientists and researchers but rather from the minds of artists and writers. Some of the earliest references to intelligent machines can be found in Ancient Greek mythology. There are multiple references to intelligent tripods that serve the gods.
Tripods, gifted with wills of their own, attend the banquets of the gods…
Philostratus, Life of Apollonius of Tyana 6

The idea of intelligent creations continued to pop up through the centuries, and the influence can even be seen in more modern texts. Frankenstein by Mary Shelly follows the creation of a “monster” that is able to think and feel for itself after being awoken with a lightning bolt. More direct comparisons can be found in the 1927 film Metropolis from director Fritz Lang, which shows a mechanical woman impersonating a rebellious teacher.

The Digital Revolution

When computers first appeared, the scientists who worked on them had long been exposed to fictional depictions of artificial intelligence. They saw that computers were able to perform complex calculations and so the AI effect took hold and they redefined intelligence.

One such researcher was Alan Turing, who considered the idea of “thinking machines” at great length. Turing proposed a game to test an AI, called the imitation game, where one adjudicator would ask two participants a series of questions. The twist is that one of the participants is a computer and the adjudicator needs to guess which one is which.

I will explain the game further in the next steps, but I want to highlight one important point here.

The focus of the imitation game for Turing was on the questions. He thought an artificially intelligent machine would be able to answer questions on a wide variety of topics with the same accuracy as a human. In his paper Computing Machinery and Intelligence, he suggested asking the participants to write poetry, select a chess move and calculate some complex arithmetic.

Generalised AI

Taking the working definition developed by Turing, researchers in the 1950s began programming “logic machines” that would be able to deduce the answer to any question; this type of AI is called generalised.

To achieve this lofty goal, the computer scientists borrowed learnings from other fields. The Bayesian model of statistics had been around since the 18th and 19th centuries, but it found a new use in helping machines deal with probability. The work of biologists also influenced the development of neural networks that mimic the cells in a brain. Logic and philosophy helped developers deal with decision making under uncertainty.

The AI Winter

The promises of this golden age of AI could not last. Companies that were started in the 1950s had secured investment by promising all-powerful thinking machines. When they were unable to deliver, money for AI began to dry up. This led to a period of time between 1970 and 1995 known as “the AI winter”.

The Return of AI

In the 1980s, some researchers began to look back over the AI research and they found some new applications.

Rather than a whole system that was “intelligent”, the techniques the researchers had developed in the 1950s could be used in a more specialised context to add a pinch of intelligence. The final push AI needed came with the development of the world wide web in the 1990s. As the decade progressed, the real product of the web became clear – data. Both humans and machines had access to more data than was ever possible before. Machines could process and analyse this data to make better decisions, to answer more questions and to be more “intelligent”.

In modern times, the hunt for generalised intelligence has not stopped, but it is no longer the main focus.

AI Effect Examples

I challenge you to find an example of a technology you use regularly that has fallen victim to the AI effect. Alternatively, search the internet for the AI effect to find another example to share with the other learners.

This article is from the free online

Introduction to Machine Learning and AI

Created by
FutureLearn - Learning For Life

Our purpose is to transform access to education.

We offer a diverse selection of courses from leading universities and cultural institutions from around the world. These are delivered one step at a time, and are accessible on mobile, tablet and desktop, so you can fit learning around your life.

We believe learning should be an enjoyable, social experience, so our courses offer the opportunity to discuss what you’re learning with others as you go, helping you make fresh discoveries and form new ideas.
You can unlock new opportunities with unlimited access to hundreds of online short courses for a year by subscribing to our Unlimited package. Build your knowledge with top universities and organisations.

Learn more about how FutureLearn is transforming access to education