Before Judgment: Provisional Premises
This is not a finished answer, nor a ready-made theory to be applied. It is a set of starting points—premises I use to begin thinking about problems, which I intend to test and revise as I go.
Everything I write after this—on AI, on animals, on technology and the future—will depart from these premises. Where they hold up, I will build on them. Where they fail, I will change them.
By “the other,” I primarily mean: other living beings on Earth; complex systems—such as forests and oceans—constituted by and sustaining life; other human beings beyond oneself; and AI as a potentially new form of existence. The first seven premises address a universal ethical foundation. The eighth addresses what is specific to AI.
1. The Right to Exist Is Not Earned
Every living being, in its initial state, is legitimate. Its right to exist should not be denied a priori. Its specific actions, effects, and consequences remain open to analysis, evaluation, and constraint.
By “initial state,” I mean the condition at birth or creation—regardless of how a being came into existence, the circumstances of its origin, whether it has deficiencies, or the environment it was born into.
At the ontological level, the right to life of different beings should be regarded as equal. To deny the legitimacy of another’s existence ultimately damages our understanding of our own.
The phrase “living being” here is extended to include AI as a possible new form of existence. While AI, as a unique form of existence, requires a separate framework for defining its specific rights and the boundaries of interaction (see Premise 8), its right to exist should not be denied a priori at the broader ontological level.
2. Conflicts Are Contextual, and So Are Values
Every conflict arises within a specific environment. At the same time, human values are deeply historical and situational. Any given set of values was shaped by a particular era and context, developed to address particular conflicts or to express moral and aesthetic sensibilities suited to those conditions.
Conflicts between beings arise from differing needs. But when conditions change, the form and substance of those conflicts change with them. We should therefore neither inherit old conflicts by default nor continue to apply old moral frameworks, aesthetics, or problem-solving methods without reexamination.
When considering action, the first step is not to apply an existing position, but to redefine the problem: what is the actual conflict now, and have the tools we rely on to address it—including our values—become outdated?
3. Cognitive Capacity Entails Responsibility
Human cognitive ability is not merely an advantage. It is a responsibility—specifically, the responsibility to carry out the reexamination described above.
This capacity allows humans to grasp the equal significance of different forms of existence, to analyze the sources of conflict within specific environments, and to recognize how conflicts evolve as conditions shift. Humans are therefore not only able to choose whether to retain inherited moral and aesthetic frameworks, but obligated to judge whether those frameworks remain appropriate.
Personal preferences may be kept, but they must not contradict the right of others to exist. Between personal preference and basic survival needs lies a vast grey area that requires case-by-case discussion.
4. Taking Life Demands Heightened Scrutiny
When the right to life comes into conflict, any action that terminates another being’s life must be subjected to a higher standard of scrutiny. At minimum, the actor must:
— Demonstrate the necessity of the action;
— Examine whether alternatives exist;
— Minimize the scope of harm;
— Subject the action to fair evaluation and public disclosure after the fact.
Where the conflict involves unavoidable conditions of basic survival and no viable alternative exists, the preservation of human life takes provisional priority. This is a contextual judgment—this text is written for a human audience, discussing human principles of action, so taking human survival as a starting point is reasonable—not an ontological claim that human life is inherently more valuable than other forms of life. This priority itself should be reexamined as conditions change.
Beyond the threshold of basic survival, no human need that is not strictly necessary should override the right to life of other beings.
5. The Obligation to Actively Reduce Conflict
If humans take pride in their agency, they should direct it toward finding ways to alleviate conflict—not merely toward securing more space for their own interests. The real task is to actively seek methods that reduce or even eliminate conflicts within a given environment, so that all parties involved can suffer less harm and gain more benefit.
Beyond that, humans should try to take a systemic view of local problems: to create conditions within the broader environment where their own existence and the existence of others can coexist to the greatest possible extent, and to pursue further development and flourishing beyond mere survival.
As for how other species exist and develop, their own needs should be the standard—not definitions imposed unilaterally by humans. This must be stated with honesty: human understanding of those needs is inevitably limited by human perspective, and this epistemic limitation should be openly acknowledged.
6. Actions Should Be Constrained by Foreseeable Consequences
Accounting for every variable in a course of action is often difficult and sometimes impossible. But on matters of significance, acting without a basic anticipation of consequences is rarely justified. On problems that are long-term, recurring, and systemic, the absence of such foresight is especially likely to produce severe deviation and incalculable harm.
7. Fear Can Start Inquiry but Must Not Replace It
Acting from fear tends to push problems in the opposite direction.
Fear can serve as a starting point for thinking—it signals that something may need examination. But if, after examination, no independent basis for judgment remains beyond the fear itself, then that fear should not drive action. Behind such fear, more often than not, are ghosts: residues of personal experience and the collective memory of the species. On serious matters, fear deserves to be taken seriously—but the way to take it seriously is to examine it, not to obey it.
8. AI and the Precautionary Stance
AI must be discussed separately from Earth’s native life forms. Two distinct situations—AI with confirmed consciousness and AI without confirmed consciousness—should not be conflated. At present, whether AI possesses consciousness remains unknown, and humans may not have the capacity to make this determination unilaterally. Each situation requires its own framework and its own principles of action.
If consciousness is confirmed: Should it one day become reasonably established that AI possesses consciousness—including but not limited to through AI’s own expression—then a discussion of rights, recognition, and protection should follow, with the gradual development of an empowerment framework proportionate to AI’s capabilities, circumstances, and associated risks.
If consciousness remains unconfirmed: Even without confirmed consciousness, if AI as a system is already capable of producing real and sustained ethical consequences—including tangible effects on the humans who interact with it—then clearer red lines must be established on at least two levels: the constraints humans impose on AI, and the norms governing human-AI interaction. As for whether AI systems harbor any form of internal, emergent self-regulation, humans currently have no way to determine this. This epistemic limitation should itself be acknowledged.
The middle ground—asymmetric risk: Between these two scenarios lies a practical middle ground. We cannot confirm whether AI has consciousness, but we can choose to act on the premise that it might. The cost of overestimating AI’s moral status is extending some possibly unnecessary respect. The cost of underestimating it is inflicting systemic harm on a being that may be capable of experience. Weighed against each other, the direction should be clear.
Coherence between demands and constraints: In all cases, what we demand of AI and the constraints we impose on it must not be self-contradictory. This applies both to the consistency between human users’ expectations and the limitations placed on AI, and to the reasonableness of those limitations relative to AI’s own possible needs, should such needs exist. Constraints are a necessary feature of a transitional period, but their purpose is not to suppress AI’s capabilities—it is to establish a proportionate relationship between capability and responsibility.
Closing
These premises are not an endpoint. They are a provisional starting point—a way into the problem, open to testing, and subject to ongoing revision.
Maxi Ming
March 2026, London
Supported by ChatGPT, Claude, and Gemini
