Vous êtes sur la page 1sur 5

Il convient ici d’étudier un article rédigé par Caroline Davies le 4 avril 2022 sur le journal

« The Guardian » dans la rubrique « Tech » et « Robots ». Cet article, illustré de photos, nous permet
de faire la connaissance de AI-DA, un robot peignant des œuvres d’art. Ce robot, pensé et construit à
Oxford construit par Aidan Meller et son équipe de programmeurs, roboticiens, experts en art et
psychologue, lie pour la première fois au monde l’AI et l’art de la peinture. Cet article, tant descriptif
qu’interrogatif, nous questionne, non pas sur le fait que la peinture par un robot est possible, mais sur
l’attractivité d’une telle pratique. Cet article sous-tend plusieurs question qui semblent pertinentes
d’abord sur l’apparence physique et les propos tenus par d’AI-DA mais aussi sur l’attractivité de cet
art et enfin sur le caractère éthique de ce projet.
D’abord, on peut s’interroger sur l’apparence physique donnée à AI-DA, si c’est l’art d’un
robot que l’on souhaite observer, pourquoi lui donner une apparence humaine ? Les traits humains du
visage du robot ainsi que sa perruque interroge sur l’intention et l’objectif de la pratique. On pourrait
se demander s’il n’aurait pas été opportun que le peinture robot n’ait pas une apparence typiquement
robotique. Son prénom lui aussi rappelle un prénom humain « Aida », il également possible que ce
prénom soit une référence au nom du créateur Aidan.
De plus, nous pouvons nous interroger sur les paroles rapportés de Ai-Da « j’aime peinture ce
que je vois », celle-ci n’ayant pas d’imagination. Cela interroge sur l’utilisation des données
photographiques. L’expression artistique du robot doit se cantonner à ce que celle-ci est autorisée à
voir, c’est-à-dire des natures mortes ou des personnes ayant données leur autorisation. Si ce cadre
n’était pas respecté, se poserait alors d’important problèmes juridiques sur l’utilisation des données,
cela faisait écho notamment au Règlement Général des données, régies par l’Union européen et
adoptée le 14 avril 2016. Ce problème juridique est renforcée par le fait que l’AI, comme le dit AI-DA
elle-même, utilise des données que nous donnons « librement sur nous-même et en parlant à nos
téléphones, ordinateurs, voitures et même appareils de cuisine ». Nous savons déjà que l’utilisation de
donnée contre le gré de l’utilisateur ou sans explicitation de leur usage a déjà posé beaucoup de
problèmes juridiques dans le passé.
Ensuite, cet article interroge sur l’attractivité de ce type d’œuvre, l’art étant en premier lieu la
représentation abstraite ou fidèle de la vie vu par l’humain au cours de l’Histoire. Indépendamment du
fait que c’est un bras bionique qui manie le pinceau et non pas un humain bien réel, la réalisation de la
peinture par un robot semble dénué de conscience, c’est l’apprentissage automatique qui donne à AI-
DA ce talent et non pas une observation critique sur le monde. L’art, qui semblait pourtant être l’un
des domaines les plus préservés de la croissance exponentielle de l’utilisation de l’AI, est aujourd’hui
divisé en deux, distinguant les œuvres humaines ou non humaines. On peut penser que les amateurs de
nouvelles technologies créeront un marché parallèle avec ce type d’art, on se souvient des FNT.
Néanmoins, on peut facilement imaginer que beaucoup de puristes resteront attacher à l’art humain et
des œuvres humainement authentiques resteront prisées par certains collectionneurs.
Enfin, on peut s’interroger profondément l’intention du créateur sur ce projet, pour Aidan
Meller, « c’est un projet éthique ». Ce projet vise-t-il a sensibiliser les consommateurs d’art contre la
pratique de l’art par un robot ? Promeut-il l’art humain ? Selon le concepteur, c’est « un
commentaire et une critique » sur l’évolution technologique rapide et automatique des IA. En réalité,
la lecture de l’article et l’observation des images apparaissent plutôt comme une publicité qu’une mise
en garde d’ordre publique. Son exposition en avant-première mondiale à la Biennale de Venise en
2022 ne fait que confirmer cette hypothèse, on peut imaginer que l’entrée à l’exposition n’était pas
gratuite. Il est difficile de déterminer si cette opération est maladroite ou un important coup de buzz.
Peut-être aurait-il intéressant de lui faire réaliser des œuvres qui mettent réellement l’observateur mal
à l’aise et qui rend inconfortable face au fait que c’est un robot qui a réaliser cela et pas un autoportrait
ou des fleurs.
En conclusion, cet article et surtout le projet étudié ne peuvent que créer une ambiguïté entre
le discours tenu et les faits observés. Dès lors, on peut réellement s’interroger sur le bien fondé de la
bonne volonté des équipes ayant conçu ce robot, leurs intentions ne sont surement pas mauvaises, mais
on peut douter du fait que ceux-ci n’ont agi que par prévention et pure bienveillance. Ceux-ci sont
peut-être, comme beaucoup, en conquête du monde par l’IA et passer par l’univers de l’art (monde qui
n’a rien de plus humain) permet une entrée en matière forte et garanties reconnaissance parmi leurs
homologues.
Cet article interroge sur la façon dont l’AI intervient dans des milieux qui étaient pourtant
réservés aux humains il y a peu : ceux faisant intervenir le libre arbitre. Il conduit à s’interroger sur les
dangers de l’AI. Il est tout à fait possible que nous nous inquiétons de façon primitive du simple fait
que nous n’avons pas encore de connaissance et de recul nécessaire à la bon appréciation de la
situation. Néanmoins, il est également possible que cette inquiétude, du fait de l’importance que prend
l’AI et des vastes domaines qu’elle concerne soit justifiée. Indifféremment de l’effet que nous produit
cet article, il démontre une nouvelle fois l’omniprésence grandissante de l’AI. Comme le suggère
Aidan Meller, nous devrons être de plus en plus prudents avec les données que nous donnons à l’AI.
Dès lors, si l’art, qui est un milieu qui fait intervenir le libre arbitre mais qui n’a aucune conséquence
sur notre monde est désormais accessible aux robots, qu’en sera-t-il des métiers de demain qui font
intervenir cette même compétence telle que les métiers de la médecine et du droit ?
Affaire à suivre…
----

Here we take a look at an article written by Caroline Davies on 4 April 2022 in the 'Tech' and
'Robots' section of The Guardian newspaper. This article, illustrated with photos, allows us to get to
know AI-DA, a robot that is creating works of art. This robot, conceived and constructed in Oxford by
Aidan Meller and his team of programmers, roboticists, experts in art and psychologists, is the first in
the world to combine AI and the art of painting. This article, which is both descriptive and
questioning, asks us not whether painting by a robot is possible, but whether such a practice is
attractive. The article raises a number of questions that seem relevant, firstly about the physical
appearance of AI-DA and what it can say, but also about the attractiveness of this art form and, finally,
about the ethical nature of this project.
Firstly, the physical appearance given to AI-DA is questionable: if it is the art of a robot that
we want to observe, why give it a human appearance? The human features of the robot's face and its
wig raise questions about the intention and purpose of the practice. We might wonder whether it would
not have been appropriate for the robot painting not to have a typically robotic appearance. Its first
name is also reminiscent of a human name, "Aida", and it is also possible that this is a reference to the
name of its creator, Aidan.
In addition, we can question Ai-Da's reported words "I like to paint what I see", as she has no
imagination. This raises questions about the use of photographic data. The robot's artistic expression
must be limited to what it is authorized to see, i.e. still lifes or people who have given their permission.
If this framework were not respected, major legal problems would arise over the use of the data,
echoing in particular the General Data Regulation, governed by the European Union and adopted on
14 April 2016. This legal problem is reinforced by the fact that AI, as AI-DA itself puts it, uses data
that we give "freely about ourselves and by talking to our phones, computers, cars and even kitchen
appliances". We already know that the use of data against the will of the user or without explicitly
stating its purpose has already caused many legal problems in the past.
Secondly, this article raises questions about the attractiveness of this type of work, art being
first and foremost the abstract or faithful representation of life as seen by humans over the course of
history. Regardless of the fact that it is a bionic arm that wields the paintbrush and not a real human
being, the creation of the painting by a robot seems devoid of any conscience; it is machine learning
that gives AI-DA this talent and not any critical observation of the world. Art, which seemed to be one
of the fields best protected from the exponential growth in the use of AI, is now divided in two,
distinguishing between human and non-human works. It is conceivable that fans of new technologies
will create a parallel market with this type of art, as we remember from the FNT. Nevertheless, we can
easily imagine that many purists will remain attached to human art, and works that are authentically
human will continue to be prized by certain collectors.
Finally, we can ask ourselves what the creator's intention is for this project. For Aidan Meller,
"it's an ethical project". Does this project aim to educate art consumers against the practice of art by a
robot? Does it promote human art? According to the designer, it is "a commentary and a critique" on
the rapid and automatic technological evolution of AIs. In reality, reading the article and observing the
images seems more like an advertisement than a public warning. Its world premiere exhibition at the
Venice Biennale in 2022 only confirms this hypothesis, and we can imagine that admission to the
exhibition was not free. It's hard to say whether this was a clumsy operation or a major buzzkill.
Perhaps it would have been interesting to have him create works that really make the viewer feel
uncomfortable and uncomfortable about the fact that it was a robot that did it and not a self-portrait or
flowers.
In conclusion, this article, and above all the project studied, can only create an ambiguity
between what is said and what is actually observed. Their intentions are certainly not bad, but there are
doubts as to whether they only acted out of prevention and pure benevolence. Like many of their
counterparts, they may be in the process of conquering the world through AI, and using the world of
art (a world that has nothing more human about it) provides a strong entry point and guarantees
recognition among their counterparts.
This article looks at the way in which AI is intervening in areas that were, until recently,
reserved for humans: those involving free will. It raises questions about the dangers of AI. It is quite
possible that we are worried in a primitive way, simply because we do not yet have the knowledge and
hindsight necessary to properly assess the situation. However, it is also possible that this concern,
given the importance of IV and the vast areas it covers, is justified. Regardless of the effect this article
has on us, it demonstrates once again the growing ubiquity of AI. As Aidan Meller suggests, we will
have to be increasingly careful with the data we give to AI. So if art, which is a medium that involves
free will but has no impact on our world, is now accessible to robots, what will become of tomorrow's
professions that involve this same skill, such as medicine and law?
To be continued...
---
Caroline Davies wrote an article on 4 April 2022 in The Guardian newspaper's 'Tech' and
'Robots' section. The article features AI-DA, a robot that creates works of art. Aidan Meller and his
team of programmers, roboticists, art experts, and psychologists conceived and constructed the robot
in Oxford. AI-DA is the world's first robot to combine AI and painting. This article questions the
attractiveness of robot-created paintings, rather than their possibility. It raises relevant questions about
AI-DA's physical appearance, speech capabilities, the appeal of this art form, and the ethics of the
project.
Why give AI-DA a human appearance if we want to observe the art of a robot? The physical
appearance of AI-DA is questionable. The robot's human-like face and wig raise questions about the
purpose of the practice. It may have been more appropriate for the robot painting to have a non-human
appearance. The robot's name, 'Aida', may be a reference to its creator, Aidan, as it shares similarities
with a human name.
Furthermore, there may be doubts about Ai-Da's statement 'I like to paint what I see' due to her
lack of imagination. This brings up concerns regarding the use of photographic data. The robot's
artistic expression should be restricted to what it is permitted to observe, such as still lifes or
individuals who have granted their consent. Not respecting this framework could lead to significant
legal issues regarding the use of data. This echoes the General Data Regulation, which is governed by
the European Union and was adopted on 14 April 2016. AI-DA itself acknowledges that the use of data
that we provide freely about ourselves, by talking to our phones, computers, cars, and even kitchen
appliances, poses a legal problem. The use of data without the user's consent or without explicitly
stating its purpose has caused legal problems in the past.
Additionally, this article questions the appeal of using data in art, which should primarily
represent life as seen by humans throughout history. Although a bionic arm wields the paintbrush
instead of a human, the creation of a painting by a robot appears to lack conscience. AI-DA's talent is a
result of machine learning rather than critical observation of the world. The field of art, once thought
to be protected from the exponential growth of AI, is now divided into human and non-human works.
Fans of new technologies may create a parallel market for this type of art, as seen in the FNT.
However, many purists will likely remain attached to human art, and certain collectors will continue to
value works that are authentically human.
Finally, we should consider the creator's intention for this project. According to Aidan Meller,
it is an ethical project. Is the project intended to educate art consumers about the practice of art by a
robot or to promote human art? The designer states that it is a commentary and critique on the rapid
and automatic technological evolution of AIs. However, upon reading the article and observing the
images, it appears more like an advertisement than a public warning. The world premiere exhibition of
this work was held at the Venice Biennale in 2022. It is unclear whether admission to the exhibition
was free. The artwork was created by a robot, which may have been intentional to provoke
discomfort in the viewer.
In conclusion, this article and the project it discusses may create ambiguity between what is
stated and what is observed. While their intentions are not necessarily bad, there are doubts as to
whether they acted solely out of prevention and benevolence. Like many other AI systems, they may
be attempting to conquer the world. The world of art, which has nothing inherently human about it,
provides a strong entry point and guarantees recognition among their peers.
This article examines how AI is intervening in areas that were previously reserved for humans,
particularly those involving free will, and raises concerns about the dangers of AI. We may be worried
due to lack of knowledge and hindsight to assess the situation properly. However, given the
importance of IV and the vast areas it covers, this concern may also be justified. This article
demonstrates the growing ubiquity of AI. As Aidan Meller suggests, we must be increasingly careful
with the data we give to AI. If robots can now access art, a medium that involves free will but has no
impact on our world, what will happen to professions that require the same skill, such as medicine and
law?
The answer will be continued...

Vous aimerez peut-être aussi