Error message

Deprecated function: The each() function is deprecated. This message will be suppressed on further calls in menu_set_active_trail() (line 2394 of /home/khalifac/mbh/eng/includes/menu.inc).

2017: Artificial Intelligence on the Hype !

Artificial intelligence, or AI, is classified by the Gartner hype Cycle amongst most visible emerging technologies for the year 2017. Mike J. Walker, research director at Gartner Inc. says in this respect: "Consider the potential impact of AI-enabled autonomous vehicles. They could reduce accidents, improve traffic, and even slow urbanization as people can use travel time and won’t need to live near city centers. “When autonomous vehicles, AI, IoT and other emerging technologies are combined with economic trends like the sharing economy, we truly see different business designs that profoundly disrupt the market” (1).

AI combines several technologies that stimulate human cognitive processes. Its purpose is to make machines perform tasks that humans perform using their intelligence. The AI ​​intervenes in very diverse fields like the voice recognition, the automatic translation, the robotics, the video games. Many digital tools, software and gadgets are already the product. With evolving robotics, are we ready for the era of symbiotic man and cyborg?

The term Artificial Intelligence (AI) was coined in 1956 by John McCarthy, an American computer scientist at the Dartmouth conference where the discipline was born. Today, it is a general term that encompasses everything from speech recognition, writing and face recognition, video games and conversational agents to automation and automation of robotic processes. It has recently become more important because of the large volume of data on networks.

AI is the simulation of human intelligence processes by machines, especially computer systems, which can now perform value-added tasks such as identifying customer models and consumption trends in data from sale of businesses. The Larousse Dictionary defines AI as a "set of theories and techniques used to create machines capable of simulating human intelligence". These techniques include learning (acquiring information and rules for the use of information), reasoning (using the rules to arrive at rough or final conclusions), and self-correction (carrying out varying degrees, analyzing the system for errors, and offering alternatives for replacement).

Computer approaches to artificial intelligence

Originally, AI is in the field of computer science where it has gone through three phases of evolution, first based on algorithms, then based on expert systems and finally based on neural networks.

As a reminder, the notion of algorithm is now linked to computers. It consists of a "finite sequence of rules to be applied in a definite order to a finite number of data to arrive with certainty in a finite number of steps to a certain result" (2). The algorithms developed with computers and the need to automate calculations. A program is an algorithm written in a digestible language for the computer. It should also be remembered that expert systems are smarter than algorithms in the sense that they are able to reproduce the cognitive mechanisms of a human expert in a particular field. This is one of the paths trying to lead to artificial intelligence. An expert system can be broken down into two main components: the knowledge base and the inference engine whose role is to reason from the data in the knowledge base.

Today, we are moving to neural systems. "The main idea of the neuron approach of artificial intelligence is to model the basic entity of the human brain, that is to say the neuron, and to assemble several to get closer to human reasoning We then build a network consisting of several layers of neurons, the number of which is determined by the complexity of the problem to be solved For example, for linear problems, so simple, one to two layers are sufficient and the resolution of the problem is very fast ".

Types of Artificial Intelligence

Arend Hintze, professor of Integrative Biology and Computer Engineering at Michigan State University, distinguishes four types of artificial intelligence:

1. Type I: responsiveness:
This is the most basic stage of artificial intelligence, and the most developed at the moment: a machine capable of perceiving the surrounding world and acting on these perceptions - the best illustration being Deep Blue, the computer of IBM, which defeated chess champion Garry Kasparov in 1996. According to some thinkers of artificial intelligence, research should also focus on the development of this type of machines.
 

2. Type II: limited memory:
Unlike reactive machines, memory-limited machines have the ability to rely on representations of the world to make a decision. Autonomous cars are a good example: equipped with a repertory of pre-programmed world representations, they are able to adjust their speed or their trajectory according to several factors, such as the traffic or the curves of the road.
 

3. Type III: Theory of Mind:
The third type of artificial intelligence marks a point of rupture between the machines that exist and those that we will build in the future. Through the theory of mind, robots will not only be able to apprehend and classify representations of the world, but also understand and prioritize the emotions that can influence human behavior, and adapt their behavior accordingly.
 

4. Type IV: self-consciousness:
Self-awareness is the final stage in the development of artificial intelligence. Self-awareness is about giving machines the ability to form representations of themselves. As an extension of the artificial intelligence type III, self-conscious machines not only understand human emotions, but manage to apply them, while at the same time predicting those of others.

On basis of these evolution indicators, do we dare predict that Man, formerly master and ruler of nature, would be about to be supplanted and subjugated by the machines he created? Anxiety for some, prophetic exaltations for the promoters of posthumanism, this vision finds roots in science fiction since the twentieth century, in literature and especially in movies and television : Robocop, Terminator, etc. How far are we from the Cyborgs and posthumans myth?

Cyborgs and posthumans, so close, so far?

A cyborg, or Cybernetic Organism, is a hybrid creature composed of organic and cybernetic components. We argue that, unlike the caricatural vision of a kind of robot made up of a metal endoskeleton covered with living tissue, the post-human will be a human being like you and me, but probably "improved" thanks to joint advances in cybertechnologies, biotechnologies and nanotechnologies. In particular, devices allowing permanent connection to information networks will not be implanted in the body, but rather integrated into clothes close to the body.

Defining the tansision process from human to Cyborg, “The transition from human to posthuman can be defined physically or memetically. Physically, we will have become posthuman only when we have made such fundamental and sweeping modifications to our inherited genetics, physiology, neurophysiology and neurochemistry that we can no longer be classified with homo-sapiens” (More, 1994).

___________

Notes:

(1) : https://www.gartner.com/smarterwithgartner/top-trends-in-the-gartner-hype-cycle-for-emerging-technol...

(2) : une application informatique de commande vocale qui comprend les instructions verbales données par les utilisateurs et répond à leurs requêtes.