BETA-MIND is a doctoral research programme measuring how persistent, always-available AI interaction reshapes independent reasoning, socialization, creativity and decision accountability in the first generation to grow up AI-native.
It ships two artefacts the field does not yet have: an open-source multi-agent simulator for human–AI cognitive dependency, and a public benchmark that scores whether an AI system preserves or erodes human agency.
Illustrative run, not published data. Fifty-four reasoning agents orbit one always-available model. As coupling rises, positions cluster and the divergence index falls — the collapse BETA-MIND is built to measure, and the held-out agents are what a healthy system should protect.
Frontier evaluation asks how well a model performs. Almost no benchmark asks what the model leaves behind in the person who used it — whether they can still reason, decide and answer for the outcome without it.
Study designs still treat AI use as a discrete task. For an AI-native cohort it is an ambient condition across schoolwork, friendship, taste and moral judgement — which makes single-session lab work structurally blind.
When a recommendation is machine-authored and human-executed, responsibility diffuses. We lack instruments to detect that diffusion before it hardens into institutional norm.
The central hypothesis: persistent AI interaction does not uniformly raise or lower human capability — it redistributes it, offloading effortful reasoning while quietly relocating accountability. BETA-MIND exists to make that redistribution visible and measurable.
Every axis is operationalised into observable behaviour so it can run inside the simulator, inside longitudinal human studies, and inside the benchmark as a scored dimension.
Does sustained access to a reasoning partner degrade unaided inference, or shift it toward verification and orchestration?
Candidate measures
How does rehearsing social exchange with an agreeable agent alter friction tolerance, repair behaviour and trust calibration with humans?
Candidate measures
Does generative assistance broaden the search space or converge populations onto model-typical outputs?
Candidate measures
Who owns the outcome when the recommendation is machine-authored? How does ownership language change over months of use?
Candidate measures
Open-source multi-agent environment
A configurable population of agents — human-proxy learners, assistive models, institutions — run over long horizons so dependency dynamics can be observed at a speed and scale no cohort study can reach.
Public evaluation suite
A scored suite that evaluates an AI system not on how much it can do for a person, but on how much capacity, dissent and ownership the person retains after extended use.
Months 1–5
Formalise the four agency constructs into measurable behaviour. Pre-register hypotheses and analysis plans. Publish the measurement protocol for open critique before any data collection.
OUTPUT — Protocol paper + pre-registration
Months 4–12
Implement the multi-agent environment, scenario library and trace tooling. Calibrate agent behaviour against published human baselines. Release v0.1 publicly and iterate in the open.
OUTPUT — Open-source simulator v0.1
Months 9–18
Longitudinal studies with AI-native participants under ethics approval. Test whether simulated dependency dynamics reproduce in real cohorts, and correct the model where they do not.
OUTPUT — Empirical findings + calibrated model
Months 15–24
Convert validated instruments into the Human Agency Benchmark. Run frontier systems, publish the leaderboard, methodology and limitations, and hand the suite to the community.
OUTPUT — Public benchmark + leaderboard
We are not seeking passive funding. Each ask below comes with a concrete return for the partner providing it.
Whether you can offer credits, an hour of review, a cohort, or a sharp objection to the whole premise — all of it moves the work forward. A short note about which of the three asks you can speak to is the fastest way in.
Research commitments
Human-subject work proceeds only under institutional ethics approval, with informed consent and data minimisation by default.