AI Scientist

A scientist. Not a chatbot with a lab coat.

The Eidosoma AI Scientist is part of the lab's research infrastructure. It reads across the literature, proposes hypotheses, writes and runs computational experiments, and returns its results to human scientific review — expanding the space we can investigate without surrendering scientific direction, taste, or judgement.

Principles

Five principles for a scientific collaborator.

01Expands the range of questions a small lab can investigate
02Remains directed and reviewed by human scientists
03Keeps code, parameters, reasoning, and negative results connected
04Links literature, hypotheses, experiments, and synthesis in one research loop
05Revisits results as new evidence changes the context

Modules

Built from composable research modules.

The lab's AI Scientist draws on a set of composable research modules. Below are three of the core ones. Tap to unfold.

Collective Intelligence Processor

Continuous literature and web review across active research questions.

The CIP follows preprints, patents, code, conference schedules, registries, and lab notes, then filters and cross-references them against the lab's active questions. Novelty, surprise, and contradiction are prioritised over volume.

Experiment Coding Manager

Runs a team of AI coders that build and execute computational experiments.

The ECM decomposes a research question into runnable experiments, spawns a team of specialised AI coders, reviews their pull requests, and orchestrates execution on the cluster. It keeps a full experiment graph — every run, every parameter, every negative result — so that nothing is ever re-discovered by accident.

Evolution Engine

Optimisation through genetic algorithms and open-endedness — MAP-Elites, NSLC, quality-diversity.

Good science rarely comes from climbing the steepest gradient. The Evolution Engine maintains a living archive of diverse candidates — models, protocols, organisms, hypotheses — and keeps exploring the space of what has not yet been tried. Underneath: MAP-Elites, novelty search, and open-ended divergence.

Research cycle

A continuous loop between evidence and experiment.

Literature intake — follows new evidence and connects it to the lab's active questions.

Hypothesis formation — proposes explanations and identifies assumptions that can be tested.

Computational experiments — turns selected questions into reproducible runs.

Adversarial review — challenges methods, interpretations, and apparent positive results.

Human review — returns evidence, uncertainty, and open questions to the research team.

Next cycle — updates the experiment graph and revisits earlier results in their new context.

Explore a question with us in living science.

We welcome conversations with researchers and labs working across bioelectricity, regeneration, artificial life, and AI-enabled science.