Faced with the frenzy of the AI, the Hiatus collective offers a counter-discourse

With our partner Salesforce, unify sales, marketing and customer service. Accele your growth!

At the top of the AI ​​action on AI, rhetoric around the unavoidable revolution does not convince everyone. In a platform published in The worlda coalition of organizations engaged in the defense of human and environmental rights denounces a dynamic which it deems dangerous, both on the social, ecological and democratic level. Under the name of Hiatusthis collective, which brings together in particular The quadrature of the net, the human rights league, Attac and act for the environmentalert on the consequences of an uncontrolled technological race and requires stricter regulation.

Far from the speeches that praise AI as an inevitable progress, Hiatus offers a radical counterpoint. For its members, this technology, in its current form, “Already have disastrous consequences”because she “Access the ecological disaster, strengthens injustices and aggravates the concentration of powers”. Far from a simple technological evolution, its deployment is above all a political and economic project shaped by large companies, under the guise of modernization.



An extractivist and monopolistic model

Far from being intangible, artificial intelligence is based on an infrastructure of considerable magnitude. The data centers that host the most advanced models consume astronomical quantities of electricity and water, while the manufacture of semiconductors and processors necessary for their functioning is based on the extraction of rare metals in often conditions disastrous.

“” To be carried out (these innovations) require, in particular, to multiply the power of graphic chips and data centers, with an intensification of the extraction of raw materials, the use of water and energy resources. »»

But the stake is not limited to natural resources. For Hiatus, AI is also an instrument of economic and political concentration. Its development “Extend neocolonial dynamics”with increased exploitation of underpayed workers in the countries of the Global South, responsible for the annotation of data and content moderation.

Upstream, it consolidates the stranglehold of large technological companies on strategic infrastructure. “It is the whole of society that is summoned to adapt to put itself on the page of this new industrial and technocratic watchword. »»

Industrialization at the service of private interests

For the collective, governments do not seek to slow down this dynamic, but actively support it by financially supporting the sector. In Europe, AI Act, supposed to supervise the potential drifts of AI, is perceived by hiatus as a text “Designed to consolidate the forward flight of artificial intelligence”.

Far from being an obstacle to abuse, this regulation is analyzed as a tool aimed at promoting a competitive market, to the detriment of real risk control. “To justify this blindness and silence criticism, it is the argument of geopolitical competition which is most often mobilized. »»

This competition logic pushes Europe to invest massively to promote the emergence of national champions in AI. But for the collective, this argument of technological sovereignty is a dead end. “Everything suggests that the delay of Europe in this area cannot be caught, and that this race is therefore lost in advance. »»

AI, a social control automation tool

Beyond economic and environmental aspects, Hiatus warns against “The growing delegation of crucial social functions to AI systems” in administration and public services. The automation of decisions via predictive algorithms leads to loss of democratic control and increased monitoring of populations.

“” That in public action, it acts in symbiosis with the austerity policies which undermine socio-economic justice? Asked the signatories, who denounce the growing appeal to algorithmic tools in the management of social aid, justice or recruitment.

This approach, which promises greater administrative efficiency, risks turning against the most vulnerable populations, due to the lack of real consideration of biases and errors in automated systems.

Is an alternative still possible?

Faced with a regulation deemed insufficient and a dominant discourse which presents AI as inevitable, Hiatus claims a “Democratic control of this technology and a drastic limitation of its uses”.

According to them, it is imperative to question in depth the real AI purposes and the interests it really serves. “The proliferation of AI may be presented as inevitable, we do not want to resign ourselves. »»

In a world where technology is too often considered an autonomous and unstoppable force, Hiatus recalls that each advance is a political choice that can and must be discussed democratically. Far from being a technophobic movement, this coalition raises an essential question: “Against the multiple unthoughts which impose and legitimize its deployment, should we not collectively rethink the use of these technologies before it is too late? »»

Among the first signatories of this forum:

Judith Allenbach, president of the magistracy union, Judith Krivine, president of the French lawyer union (SAF), Julien Lefèvre, member of scientists in rebellion, Nathalie Tehio, president of the LDH (League for Human Rights), Raquel Radaut, spokesperson for the quadding of the net, Sandra Cossart, director of Sherpa, Sophie Venetitay, secretary general of SNES-FSU, Stéphen Kerckhove, Director General of Act for the Environment, Vincent Drezet, Porte d’Atac France…