Vanessa Nurock once said: “AI has taken root in the form of an experiment that, in a way, cannibalizes ethics and politics by subjecting them to its own logic.”
If that is the case, could AI be considered inherently unethical, imposing its own rules? When we speak of “AI ethics,” are we chasing a chimera? Should we view the coexistence of this philosophical notion and this technological system as impossible ?
Before answering these questions, we first need to define what “ethics” really means.
What is ethics?
Ethics is the branch of philosophy that deals with good and evil. It is essentially a philosophical exercise in moral values, an evaluation of different moral frameworks. Ethics and morality are therefore inseparably linked, one cannot exist without the other.
The works of Luciano Floridi, a renowned Italian philosopher, offer us some insights. According to him, AI represents a form of artificial agency.
This artificial agency, however, is fragmented. It can no longer be considered “intelligence,” since its actions occur without any presence of mind.
The absence of presence of mind leads to the absence of true intelligence, because this lack of awareness results in actions that are unlimited and free of responsibility. Put simply, AI will not restrain its own actions out of respect for virtuous principles.
How can we summarize AI ethics?
AI ethics can be condensed into five fundamental principles, inspired by bioethics:
- Beneficence
- Non-maleficence
- Autonomy
- Justice
- Explicability
The last principle is particularly representative of the intrinsic “black box” problem in artificial intelligence systems. Explicability can be divided into two questions: On the one hand, “How does it work?” and on the other, in terms of responsibility, “Who is accountable for the way it works?”
Explicability therefore also implies transparency, algorithmic transparency.
Is ethics just an empty word, a way to ‘look good’?
According to the AI Ethics Guidelines Global Inventory, there were 160 sets of ethical principles published in 2020. While many overlap, ethics does not mean the same thing for everyone.
The term ”ethics” can easily become a buzzword, especially when used as a form of bluewashing, the practice of making unfounded or misleading claims about the ethical values or benefits of processes, products, services, or digital solutions.
What should we take away from this?
Ignorance is the source of many evils.
Luciano Floridi explains how we might achieve a certain level of ethics in AI: “From a Socratic perspective, where human evil is seen as an epidemic problem, poor practice is often the result of a poorly assessed solution or a misguided opportunity.”
It is therefore crucial to understand that miracle solutions do not exist, especially when they rely on shortcuts. Moreover, humans must not become the means and technology in the end. For this reason, ethics has a vital role to play.
Théo BARTZEN
Sources :
– Luciano Floridi : « L’éthique de l’intelligence artificielle »
– Sabine Marcellin : « Droit de l’intelligence artificielle »
