What is artificial intelligence and what is it not?

What is the difference between Artificial Intelligence (AI) and Machine Learning (ML) and Data Science (SS)?

(in English: Artifical intelligence (AI), Machine learning (ML), Data scinece (DS))

These terms are often confused by people, unfortunately also by those who are supposed to be doing this. Many people say they are the same or, better, overlap. Other times they are simply used for marketing purposes because it is more sellable. But let’s get this sorted out.

Machine Learning in Agriculture (GT)

This is about the ability of a computer or system to learn things without human intervention. That is, it doesn’t have to be specifically programmed to do this – this or do that. Typical examples in agriculture are the detection of diseases in plants. The system gets a lot of pictures, gets some samples of healthy and diseased plants, and starts to analyze them and slowly starts to learn which are healthy and which are diseased leaves.

You don’t know at the beginning, about. guessing, but by the end of the process it can tell with great accuracy whether a leaf is healthy or not.

Similar to machine learning, learning to recognize different growths. Or what weather is doing to a particular plantation, or programs that detect the relationship between soil composition and plant development.

Artificial Intelligence in Agriculture (AI)

Artificial intelligence in agriculture means that a system tries to appear intelligent and make decisions. Such a system could be a self-driving tractor or a self-driving drone. In other words, these systems make decisions, hold the line, but if they see an obstacle, they can go around it or even turn around and take a different route to the desired area.

It is also artificial intelligence if the system detects certain needles in a plantation and based on that it orders some soil tests, weather data and determines by itself what is causing the problem and sends out the self-driving tractor to spray with a certain agent.

If you send out a drone to map the area along a certain route, it is not artificial intelligence. If there’s a tree in between, it’ll go in without thinking. But if,to the drone, it’s enough to say, there’s that big cornfield, map it. And the drone can fly out, determine its path, detect the edges of the board and avoid obstacles, then it’s an intelligent drone.

In short, it’s intelligent – it makes decisions by analyzing the environment.

It seems that AI uses a lot of Machine Learning subsystems.

Data Science in Agriculture (AT)

And finally, let’s look at data science. Data science also relies heavily on Machine Learning, but it does not make decisions. Its task is to discover the relationships in the data and organize it in a way that is understandable to non-programming decision makers.

Data Science is a system that analyzes satellite images to create a weed map and then uses that to create a nice color infestation map. It does not make a decision, but based on that, the experts can decide what intervention is needed.

A complete report visualizing the soil analysis and the environment and the soil analysis and crop averages for the last 5 years is also Data Science. If based on this the system recommends things like fertilization or tells you what to grow in the field next year instead of the current one, that is already Artificial Intelligence.

And one or two more things

In the above cases we are only talking about Narrow Artificial Intelligences. Which means that it can only exhibit intelligent behavior in a particular domain.

Narrow AIs are what we find in the Science Fiction literature. We do not have any at present.