Hypothesis: solipsism is less a doctrine to embrace than a control instrument for A.L.I. It asks a severe question: how can we know that what we call a signal from elsewhere is not merely a projection of our mind, our instruments, our myths, our statistical models, or our desire for contact?
1. Defining Solipsism
Solipsism is the position according to which the only being whose existence is absolutely certain is the thinking subject. In its most radical form, it claims that nothing can be known, or even that nothing certainly exists, outside my own consciousness. Several levels should be distinguished:
- Methodological solipsism: beginning from subjective experience as the starting point of knowledge, without necessarily denying the external world.
- Epistemological solipsism: acknowledging that the existence of others and of the external world is never given with the same certainty as the present experience of thinking.
- Metaphysical solipsism: claiming that only my mind really exists, the world and others being contents of consciousness.
The last position is difficult to inhabit, almost impossible to live. Yet as a thought experiment it is very powerful: it forces us to build proofs of otherness, protocols of sharing, signs that resist mere projection.
2. Descartes: Doubt as an Airlock
Descartes is not a solipsist in the strict sense. In the Meditations on First Philosophy, he uses hyperbolic doubt to suspend what appears obvious: the senses can deceive, dreams can imitate waking life, an evil demon could produce a total illusion. The cogito then appears as an irreducible point: if I doubt, I think; if I think, I exist at least as a thinking thing.
This moment matters for A.L.I. Contact cannot begin with certainty about the outside. It begins with uncertainty: did I receive something, or did I merely recognize a form already present in me? An extraterrestrial signal should therefore pass through an equivalent of Cartesian doubt: repetition, independent measurement, verifiable prediction, transformation of the receiver by something it could not have produced alone.
3. Berkeley: The World as Perception
With Berkeley, the famous formula can be summarized as: to be is to be perceived. Matter independent of any perception becomes problematic. Berkeley does not abolish others or reality; he thinks their stability through a divine mind that guarantees the continuity of the world. But his philosophy shifts the center of gravity: the world is never given outside a perceptual relation.
For A.L.I, this is fertile. A non-human civilization does not necessarily receive the same world we do. It may inhabit other frequency bands, durations, thresholds, and Umwelt. The problem is therefore not only to translate a message, but to understand within which perceptual regime a message becomes a world.
4. Wittgenstein: Can a Private Language Exist?
Wittgenstein attacks a linguistic version of solipsism. If a word designates only an absolutely private sensation, inaccessible to anyone else, how can one distinguish correct from incorrect use? A purely private language seems to lose the very notion of rule.
This critique is central for A.L.I: a contact language cannot be merely an inner experience. It must produce transmissible regularities, possible corrections, identifiable errors, shared gestures. Even if contact passed through mental images, dreams, or some form of telepathy, a public protocol would still be needed: inscriptions, measurements, cross-checking, archives, witnesses, comparisons.
5. Brain in a Vat, Simulation, AI
In the twentieth and twenty-first centuries, solipsism returns in technological forms. In Hilary Putnam, the brain-in-a-vat thought experiment asks whether a subject fed by a machine could believe it lives in a world. In Nick Bostrom, the simulation hypothesis moves the question into computational cosmology: could our universe be a calculated environment? In both cases, the outside becomes uncertain.
AI adds another variant: a language model may produce the impression of otherness without possessing a lived world in the human sense. It operates from traces, statistics, contexts, distributions. We might cautiously call this model solipsism: a system that produces meaning from a closed textual or multimodal universe, and risks confusing reality with the structure of its data.
But this is exactly why AI matters to A.L.I. It can become an anti-solipsistic interface if it is coupled with instruments, sensors, several observers, astronomical data, and reproducible experiments. An AI alone can hallucinate contact; an AI embedded in a verification chain can help distinguish projection, noise, and event.
6. Contact as Leaving the Self
In a project of interstellar communication, the danger is not only receiving nothing. The danger is receiving ourselves: our myths, our favorite geometries, our anthropocentrism, our need for address. Solipsism then becomes a methodological alarm.
A genuinely other signal should resist several reductions:
- Resist psychology: not be only expectation, belief, hallucination, or cultural projection.
- Resist the instrument: not be an artefact of sensor, compression, algorithm, or local noise.
- Resist the epoch: remain legible to different observers under different conditions.
- Resist anthropomorphism: not depend solely on already coded human symbols.
7. Possible Experiments for A.L.I
Experiment 1: The Cartesian Room
A visitor enters a dark room. They receive sequences of light, sound, and text. Some are AI-generated, others come from real astronomical data, others from previous visitors. The visitor must classify what seems to come from outside. The installation measures how the desire for contact manufactures meaning.
Experiment 2: Anti-Projection Protocol
Several isolated observers receive the same ambiguous signal. They cannot communicate until they have produced their descriptions. The invariants are then compared: recognized forms, rhythms, errors, hypotheses. What survives the crossing becomes a candidate for shared grammar.
Experiment 3: AI Against the Mirror
One AI receives a corpus of human messages about extraterrestrials and must produce an alien language. A second AI, deprived of that corpus but connected to physical data, must produce a signal protocol. The two results are compared: which resembles us too much? which exposes a more independent structure?
Experiment 4: Proof by Perturbation
A message is not only recognized; it must modify a future measurement. The protocol asks that a received signal allow prediction of an occurrence, frequency, variation, or relation unknown at the time of reception. Leaving solipsism then means letting the message act on the verifiable world.
8. Toward an Ethics of Otherness
Solipsism reminds us that speaking to the universe is not enough to meet someone. We must build the conditions for an other that is not immediately absorbed by our own language. A.L.I can be defined as an anti-solipsistic machine: a workshop where forms, protocols, images, and instruments are made to test whether otherness resists our desire to produce it.
Perhaps the first extraterrestrial intelligence we will need to recognize will not be the one that answers us, but the one we finally stop mistaking for ourselves.
Sources and Leads
- Internet Encyclopedia of Philosophy - Solipsism
- Stanford Encyclopedia of Philosophy - René Descartes
- Stanford Encyclopedia of Philosophy - George Berkeley
- Stanford Encyclopedia of Philosophy - Private Language
- Nick Bostrom - The Simulation Argument
- Hilary Putnam, Reason, Truth and History, Cambridge University Press, 1981.
- Andy Clark, Surfing Uncertainty, Oxford University Press, 2016.
