# Editorials

Hors Piste

What if in 2 years from now 90% of the internet content was generated by Artificial Intelligence? Only one year later, Artificial Intelligence is already starting to self-train, as ChatGPT does by training on Wikipedia – itself nourished by AI. Considering this phenomenon creating a potentially toxic self-feeding cycle, it seems crucial to me to mention the principle of Garbage In/Garbage Out which reminds us that the quality of the output data from a system is directly linked to the quality of the input data. This situation raises an interesting question: if an AI trains itself with its own creations, can we expect new artistic movements resulting from it, or do we risk a kind of degeneration, or even another 'AI winter' ? The human creative process still stands out clearly : it is defined by basics such as freedom, finitude and the ability to navigate the unknown. According to Marcel Duchamp, artists act like a medium, guided by intuition more than by a full and complete aesthetic consciousness. On the contrary, AI, with its simplified and predictable approach, struggles to emulate this depth of innovation and surprise.  Can we then imagine an AI capable of surprising or rebelling against its own algorithms? It seems to me essential and urgent to preserve a place for non-AI creation, in order to protect the human one, but also to extend the synergy between AI and humans. Human creativity, driven by an innate understanding of the sense of vague and lack, goes beyond any algorithms' limits by offering an inexhaustible source of innovation. Hartmut Rosa and Edgar Morin remind us of the importance of balance and complexity when it comes to technological acceleration and algorithmic simplification. History shows us a cultural resistance to new technologies, highlighting the dangers of losing our "human measure". In this era of rapid innovation, #ResponsibleIA, #HumanCreativity, and #SustainableInnovation must guide our integration of AI.  As Éric Sadin points out, our unique and free relation with language and creation, sharply contrasts with the standardized approach that has AI.I am not seeking to speculate or predict anything, but to encourage a reflection on how to promote a human-AI interdependence and interdisciplinarity, to maintain a long-term constant enrichment of the AI . We will then be able to avoid the impoverishment of creativity and ensure a future where innovation is both technologically advanced and deeply human. 

- Dimitri Daniloff

 © Joris Bacquet

 

#ArtificialIntelligence #ArtAndAI

 

# 3 Reinventing AI: Towards a successful collaboration with Humans / © Joris Bacquet - Hors Piste

1/2

9 rue Sainte-Apolline - 75003 Paris +33 9 78 80 08 56

© All rights reserved / Credits