AI or Machine-Learning?

Since the creation of computers, the world of technology kept advancing with an astonishing speed. Today, the world rely on intelligent automated systems, AI, which performs a big percentage of our daily tasks from creating grocery lists to performing surgeries. Artificial intelligence or technically speaking machine learning has demonstrated capacities that goes beyond the simple “tell me to do something” of robots; but what is AI and where can we find it?

Have you ever taken a picture with a smartphone? Have you ever had to label the names of the people in the picture after prompted to? If yes, then you already played apart in Machine Learning. AI are intelligent systems that are able to learn and adapt in function of the challenges bought to them. The better the adaptation the closer they approach full autonomy; but they need us to make that happen through machine learning. By teaching a system to recognize people faces, for instance, after going through thousands of pictures representing humans; we implement a pattern in the computed neural systems that continues growing and getting better at recognizing human faces. However, even if machine learning is actually one of the major reasons AI is developing; is intelligence the only thing that make automated system valuable?

The answer to this question is no. As true as having a blender which can program itself and come up with the “perfect” smoothie may come handy; it’s to note that automated systems are famous for well…the fact that they are automated primarily. Now it is to clarify that automation does not always imply intelligence. To develop, let's look at the example of a robot's arm designed to cut a 10 cm*10 cm square shape on fabrics after they are sensed on a tabletop. This robot's arm possesses an automated system; in programming a very basic representation of the algorithm is that if the sensors give True (the presence of something on the tabletop) the arm will start cutting (True as well). But what happens if instead of fabric someone accidentally put their laptop on the table-top? Will the machine recognize it and not cut or will it do as programmed? In this case, if the program and sensors are only set to cut after anything is put on the tabletop; yes the computer will get damaged. That’s where machine automation and machine intelligence show their boundaries. In automation the system does not really “think” by itself; and by thinking is meant estimating which actions are appropriate after constantly analyzing and balancing the data provided in real time.

In the field of business information systems; companies rely heavily on automated and intelligent systems in order to keep up with an ever-increasing demand. Pants for instances, because of goods’ storage, shipment and procurement, will heavily rely on both automated systems for (movements) and intelligent business systems. But where is the place of AI in this framework ? Well, it’s everywhere. In fact AI offer the best of all worlds; they may be robots not only able to lift heavy loads and accurately input data, but able to speed up the chain of production therefore increase benefit (if the demand is high). AI do not require health insurance coverage, they don’t need many breaks or paid vacations; they are therefore a promising future for companies always wanting and expecting more. AI are taking away many positions that could have been filled by humans; a downside? If everyone has their jobs taken by AI, how will people be able to afford the services provided by companies?