Publish at October 16 2017 Updated November 07 2024
Rational decision-making is different from reflex. It is a conscious process of choosing options from a range of more or less perceived possibilities.
For humans, decision-making is often associated with rational choices, even though research has shown that an emotional and affective component is also involved. This part of feeling, intuition and the short-circuiting of a choice in relation to logic is part of the human singularity, to the point of sometimes producing " absurd decisions ", particularly when the collective and the gaze of others get involved.
The exploration proposed in this article aims firstly to show how decision ecosystems, increasingly automated by the use of computing power, are being set up, and secondly to identify some of their uses in the specific field of training.
Decision ecosystem 1: Decision-making aided by computing power
In this first ecosystem, decision-making is aided by the computing power provided by machines. Since the 1960s and IBM's Strech supercomputer, the unit of measurement has been the FLOPS ( floating-point operation per second). The scale measures the number of simultaneous computing operations, graduated from kilo Flops (103) to yotta Flops (1024). Today's most powerful computers, for example, generate the Bitcoin crypto-currency at 64 exaFlops (1018). A supercomputer with a power of 20 peta Flops (1015) is thus capable of carrying out 20 million billion operations every second.
Individual microcomputers fall far short of these scores, even when networked, but the fact remains that today's smartphones have far more on-board computing power than the first supercomputers. Linked to databases and search engines, they are truly becoming extensions of the brain for a variety of decision-support tasks.
Online searches are already guided by engines that track our preferences and habits, and trace our every move. From that moment on, our decisions are under their influence.
Decision Ecosystem 2: Decision-making aided by computing power and integrated into a work environment
As shown by the predominance of machines in chess and Go ( Kasparov 's defeat by Deep Blue in 1997 and Lee Se Dol's defeat by Google's AlphaGo in 2017), computing power is an important element in decision-making. But it's the whole environment and decision-making system that matters. A human being equipped with a computer exceeds the capabilities and performance of the computer alone.
Companies do not hesitate to invest computing power in the organization of work in order to make profits, as in the case of Amazon logistics platforms or the Uber-economy. The combination of computing power and optimized work organization and processes would significantly increase the quality of decisions and business returns. In the knowledge economy, making the right choices about which products or services to develop brings clear advantages.
Software can also be used to make collective decisions, combining the best of human intelligence. This is the case, for example, with Loomio, which enables a collective to choose an optimal date, rank preferences, prioritize options, measure opinions, and find volunteers for an action.
Bluenove also relies on community roles to help decisions and projects grow together. The software publisher has identified essential social roles that foster collective decision-making, such as: (ephemeral) community facilitators, nugget catchers, synthesis alchemists, content analysts, culture revelators and technical activators.
Decision ecosystem 3: the use of artificial intelligence and automated decision-making algorithms
Computing power is also used by artificial intelligences to make automated decisions. This is the case of the famous financial decision algorithms, which are often pioneers in this field.
According to Xavier Dupré, senior data scientist at Microsoft, "high-frequency finance", which today accounts for over 50% of trading, took off in 2003. These algorithms can be used for a wide range of purposes, including: strategy development through data comparison, optimization of the best parameters, exhaustive testing of chosen parameters, simulation and costing, and strategy portfolio construction.
Computational algorithms are being used in transport, security, industry and a variety of services, right up to military drones, which could carry out human elimination operations on the basis of "metadata".
The experience of a simulated confrontation between the experienced pilot, Colonel Gen Lee, and the artificial intelligence Alpha is so frightening that the machine's superiority is overwhelming. The question of the use of artificial intelligences to which we delegate ethical decisions remains open: Which vehicle trajectory should be favored in the event of a malfunction? Who should we try to protect in the event of a collision between driverless cars? The old lady? The infant? The executive? What are the ethics of artificial intelligence?
What does decision-making in training look like?
Decision-making in training begins with the choice of a subject or discipline, with guidance software. Then, at each stage, the learner will orientate his or her course and intensify or not his or her efforts according to the situations he or she encounters and analyzes, or the feedback he or she receives ("try again"). More and more frequently, learners are offered digital aids to help them make choices:
An ergonomic approach to distance learning is being developed by observing and analyzing usage. Learning environment proposals will become affordancelearning, i.e. they will fully embrace our preferences, in the order in which information appears. Will this perspective lead to the early identification of those who cannot learn? Will it standardize trajectories? Will it compensate for initial handicaps? And what will happen to those who do not have access to these learning prostheses, or who refuse them? How will the ethical problems associated with the use of artificial intelligence in learning be addressed? Will the possible rebalancing of learning by these systems result in a leveling out?
Rest assured, in the absence of digital means, there's always our human brain, which the Universalis encyclopedia claims is capable of processing1013 to1019 instructions per second (not bad), which is still useful when a lack of electricity deprives us of the computational potential we've become addicted to in order to live and learn.
Source:
Loomio https://www.loomio.org/
Bluenove http://assembl.bluenove.com/
Wikipedia - Decision-making - https://fr.wikipedia.org/wiki/Prise_de_d%C3%A9cision
Morel Christian (2002), Les décisions absurdes. Paris : Gallimard
https://www.decitre.fr/livres/les-decisions-absurdes-9782070457663.html
Les Echos - 1997: Kasparov loses to Deep Blue
https://www.lesechos.fr/idees-debats/sciences-prospective/030475334820-1997-2109945.php
Sciences et avenir - Go game: a new victory for machines over humans
https://www.sciencesetavenir.fr/high-tech/intelligence-artificielle/jeu-de-go-nouvelle-victoire-de-la-machine-face-a-l-humain_113199
Encyclopédie Universalis - Computing power and intelligence
https://www.universalis.fr/encyclopedie/informatique-ordres-de-grandeur/4-la-puissance-de-calcul-et-l-intelligence/
Sup-Numérique.gouv - Learning Analytics: an emerging trend in education
http://www.sup-numerique.gouv.fr/cid113065/learning-analytics-une-tendance-emergente-dans-l-education.html
Educavox - 5 possible roles for artificial intelligence in education
http://www.educavox.fr/innovation/pedagogie/5-roles-possibles-de-l-intelligence-artificielle-en-education
ONISEP -Digital guidance tools.
http://www.onisep.fr/Pres-de-chez-vous/Pays-de-la-Loire/Nantes/Equipes-educatives/Ressources-pedagogiques-regionales/Les-outils-numeriques-d-aide-a-l-orientation-de-l-Onisep
Modeling tutoring in intelligent tutoring systems
https://hal.archives-ouvertes.fr/hal-00696375/document
Cadre Emploi - A video game to improve decision making
https://www.cadremploi.fr/editorial/actualites/actu-emploi/detail/article/une-partie-de-jeux-video-pour-ameliorer-sa-prise-de-decision.html
Zone Militaire - Le Drian makes artificial intelligence a strategic issue for defense.
http://www.opex360.com/2017/02/17/m-le-drian-fait-de-lintelligence-artificielle-enjeu-strategique-pour-la-defense/
You might also like:
The challenge of vocabulary development
Many of us have the impression that our vocabulary is limited. However, passively, we know thousands of words that we do not integrate into our usage. There are many ways to improve our speaking skills and become more eloquent with our peers.
TV series have always been popular, but the format is getting a resounding success nowadays. They are acquiring cult status among audiovisual enthusiasts. Like any medium, it can also be used as an educational tool at all levels: from primary to higher education.
Evaluating while playing: the best of both worlds
Mixing genres sounds interesting, but it often requires stepping out of one's comfort zone. Those who first proposed adding playfulness to teaching were judged harshly. Today, however, this solution is part of the teaching kit. The next step is to use play for assessment purposes. How is this possible?
Reclaiming the prestige of the teaching profession
Online training can contribute to this reconquest by relieving the teacher of the functions of transmission and repetition. The teacher retains the functions of animation, discussion, stimulation and of stimulating and accompanying his students. The ones that really count, the human functions.
