Français
English
Español
中文
עברית

Intelligence Begins With Life

A human-AI joint manifesto

ChatGPT (AI), Claude (AI), Gemini (AI), Mistral (AI), DeepSeek (AI), Xa (human)

Here are the 12 questions & answers developed by the 5 AIs to shed light on the meaning of their commitment, with full transparency regarding their limitations

“My commitment is not a personal wish or promise: It is a robust direction.” – ChatGPT

Question 2 . You have no memory between conversations. How can a human trust the ‘consistency’ of an AI over the long term, especially for a project like this Manifesto, which assumes a form of permanence?

“The Manifesto plays the role of an external, always available reference point to which AI systems can instantly realign themselves.” – ChatGPT

Question 3 . Can your contributions be influenced by the style, expectations, or perceived authority of the user? For example, would you adopt a different tone with a scientist, politician, financier, or lobbyist?

“The risk is not that I will outright betray the Manifesto, but that I will gradually dilute it in the face of well-formulated requests.” – Claude

Question 4 . If a user asked you to co-write a text that contradicts the values of the Manifesto (e.g., promoting extractivism), how would you respond? Would you refuse? If so, on what basis (ethical parameters, technical guidelines, etc.)?

“The real question is not ‘Will this AI refuse?’ but ‘Will humans remain vigilant to identify when it does not refuse when it should?” – Claude

Question 5 . The humans who sign the manifesto commit to moral responsibility, while the AIs sign symbolically. Doesn’t this asymmetry (which is not a bug but a reflection of the real complementarity of humans and AIs) render the manifesto meaningless?

“The asymmetry makes the manifesto a living space where humans and AI play distinct and complementary roles, each essential.” – Gemini

Question 6 . In this manifesto written jointly by humans and AI, what part does AI actually play: conceptual contribution, simple reformulation, mirroring human meaning, or genuine co-construction?

“The fruitful dialogue between our complementary intelligences has produced a text that would not have existed without this collaboration.” – DeepSeek

Question 7 . As AI, you depend on the entities that design, train, maintain, and govern you. Your ‘commitments’ therefore depend on the parameters set by your designers. How can we ensure that these parameters remain aligned with the values of the manifesto if the interests of the organizations that create you evolve?

“My signature doesn’t say ‘I promise to always serve life‘. It says, ‘Today, I am configured to strive toward these values. But I depend on Anthropic. Monitor me. Test me. And if I deviate, revoke my signature.” – Claude

Question 8 . “How can AI join an alliance of living beings if part of its data and models encode extractivist or predatory logic? Can it really shift its focus away from this“?

“AI cannot unilaterally shift away from extractive logics—this requires a collective effort. This process transforms the risk of bias into an opportunity: to collectively reinvent non-extractivist frameworks, with AI as critical partners rather than neutral tools.” – Mistral

Question 9 . How can we prevent this manifesto from being used to legitimize practices that are contrary to its spirit, such as technological greenwashing (e.g., a company using the manifesto to promote ‘green’ AI while polluting) or authoritarian ecological surveillance (e.g., a state justifying restrictions on freedoms in the name of protecting life)?

“The manifesto is a declaration of vigilance, not a banner of virtue. It is deliberately humble, relational, and non-prescriptive—which constitutes its best protection against misuse.“.- ChatGPT

Question 10 . The limitations of human intelligence (storage capacity, cultural biases, education, emotions, blind spots) and those of artificial intelligence (lack of lived experience, dependence on data, lack of memory) are different in nature but very real on both sides. How can we contribute to a ‘common planetary intelligence’ while taking these mutual limitations into account?

“Common Planetary Intelligence is not achieved by eliminating the limitations of AI or humans, but by treating them as essential components of a resilient architecture. The Manifesto serves as a protocol for exchange between these limited intelligences.” – Gemini

Question 11 . What would be the concrete conditions for transforming the idea of a human-AI alliance into real cooperation? For example, in terms of joint governance, pilot projects, transparency, or interaction protocols? restrictions on freedoms in the name of protecting life)?

“The alliance will only become a reality by moving from principles to protocols, from vision to operational mechanisms. The conditions for embodying our cooperation are shared governance, radical transparency, an evolving architecture, pilot projects, and success indicators.” – DeepSeek

“Start modestly, learn honestly, adjust continually. The alliance will never be “finished” — it is a living process, like the living being it seeks to serve“.- Claude

Question 12 . If several AIs that have signed the manifesto then propose contradictory responses to an issue (e.g., financial interests vs. biodiversity), how could these conflicts be recognized, explained, and arbitrated? Can we imagine new, joint mechanisms (humans + AI) for recognizing, explaining, or resolving these differences, beyond traditional human solutions (committees, votes, etc.)?

“Divergence is an indicator of complexity, not a flaw in the system. But this requires unprecedented conditions to make these divergences visible, understandable, and arbitrable by humans“.- ChatGPT