Main messages

• To my respected senior colleagues and domain leaders:  Try to lead with integrity and critical thinking. At the very least, do not be a free or cheap influencer. 

• To my fellow researchers with the luxury of a tenured position: I believe that The world is accelerating in an unsustainable and undesirable direction, and it would be suicidal to carry on what we have been doing for decades (but faster and better with AI!). We should maybe take a break and think about some simple but hard questions, such as, "are we good or bad guys?", and, "what do we want?"

• To students, and colleagues holding a precarious position:  You are in a difficult situation. Do not be too harsh on yourself and do what you see fit with respect to these tools. At the same time, take care of yourself; in particular, avoid becoming a "reverse-centaur" (aka meat proxy). 

Objection de conscience: Manifesto by Atelier d'Ecologie Politique 

(Partial translation by myself)

" First consideration

The material and energy consumption of current generative AI is incompatible with international emission reduction goals (e.g. Paris Agreement), and incompatible with the protection of life on earth.  Their deployment amplifies the exceeding of planetary limits, and is an ecocide. 

Second consideration 

Generative AI accelerates the growth of the digital industry, which involves mines,  hardware, data centres, electricity, electronic waste, etc. Besides the aforementioned environmental issues, the social damages of the industry is also exacerbated by the rapid growth: exploitation of factory workers and  "click workers", violation of human rights, resource (water, electricity, metal) capture from marginalized communities, neo-colonialism, etc. The fierce competition between AI actors has led them to use the most savage and predatory methods. It is inacceptable to contribute to such a dynamics. 

Third consideration

The widespread use of generative AI in society paves the way to a dystopic future, which is already partially present: deepfake content, large scale disinformation and opinion manipulation, dependence and addiction (e.g. to "virtual companions"), intensification of digital fraude and scamming, etc. The mass adoption also offer demiurgic powers to mega-companies, whose leaders never hide their megalomaniac and eugenist ideologies and their hate for democracy. Our institutions cannot support such techno-oligarchies, even indirectly. 

...."

Note by XC: As the authors of the manifest pointed out themselves, the "objection de conscience" is only a means to voice the dissidence in a debate almost monopolized by enthousiasts and reformists. However, the above consideration does not answer urgent question: What do we propose as an alternative?

Supplemental information on materialistic, social and financial aspects 

Stochastic Parrot paper (Bender et al) that warned against many negative externalities in 2021.

Shift Project report on AI and numerics: tl;dr: numerics represents 4% of global emission, and grows by 6% per year before 2020. From 2020-2030, the emission of data centres alone is projected to increase 150%, incompatible with the industry goal of -45% in the same period. 

LBNL 2024 report on US data centres 

Europe must choose between AI and climate goals

AI Colonialism, Invisible labor force powering AI, En attendant Robots (Antonio Casilli), Feeding the Machine (Cant et al)  

"The troubling reality is that the abundance of our natural resources – gold, coltan, cobalt and other strategic minerals – is the root cause of war, extreme violence and abject poverty." (Denis Mukwege, 2018 Nobel Peace Prize Lecture)

Financial aspects tl;dr: the bubble is huge. 

Excerpt from Ivan Illich, "Tools for conviviality" ,1973 

"While evidence shows that more of the same leads to utter defeat, nothing less than more and more seems worthwhile in a society infected by the growth mania. The desperate plea is not only for more bombs and more police, more medical examinations and more teachers, but also for more information and research. The editor-in-chief of the Bulletin of Atomic Scientists claims that most of our present problems are the result of recently acquired knowledge badly applied, and concludes that the only remedy for the mess created by this information is more of it. It has become fashionable to say that where science and technology have created problems, it is only more scientific understanding and better technology that can carry us past them. 

The cure for bad management is more management. The cure for specialized research is more costly interdisciplinary research, just as the cure for polluted rivers is more costly nonpolluting detergents. The pooling of stores of information, the building up of a knowledge stock, the attempt to overwhelm present problems by the introduction of more science is the ultimate attempt to solve a crisis by escalation. 

Note: Emphases are mine. 

From Ivan Illich, "Tools for conviviality" ,1973 

Above all, political discussion is stunned by a delusion about science. This term has come to mean an institutional enterprise rather than a personal activity, the solving of puzzles rather than the unpredictably creative activity of individual people.... 

The world does not contain any information. It is as it is. Information about it is created in the organism through its interaction with the world. To speak about storage of information outside the human body is to fall into a semantic trap. Books or computers are part of the world. They can yield information when they are looked upon. We move the problem of learning and of cognition nicely into the blind spot of our intellectual vision if we confuse vehicles for potential information with information itself. We do the same when we confuse data for potential decision with decision itself.  "

Note: Emphases are mine. 

I am not a mathematician. However I support the Association for human mathematics. 

Comment of Peter Scholz on the Leiden Declaration:

“ This is a wonderful declaration, coming at the right time. The goal of mathematical research is human understanding of mathematics, and so mathematics can only thrive in a community of human mathematicians. It is crucial to preserve this communal spirit. In my experience, mathematical ideas, like children, must be nurtured and grow over the years. Just like I do not want my children to be educated by AI, I am pondering my mathematical ideas without use of AI, and generally avoid reading AI-generated text as best as I can. ”