Epistemic stewardship
How should authority, accountability and authorship work when machine cognition can produce artefacts more sophisticated than their human curators can fully defend?
Open research toward abundance
The goal is not more ideas. It is greater capacity to reach worthwhile futures.
Agalmic Research studies how knowledge, energy, computation, institutions and cognitive tools can expand the set of worthwhile futures people can actually reach. Candidate abundance is an input. Frontier-expanding capacity is the aim.
Why this exists
Abundance is not a prediction. It is a direction.
Scarcity should be studied not only as something to allocate around, but as something human knowledge and institutions can sometimes transform.
Agalmic Research asks which constraints can be made less binding, which new bottlenecks appear when they are, and how knowledge, technology and institutions can increase broadly useful human capability. It does not assume that technological progress, artificial intelligence or economic growth automatically produce abundance.
The manifesto states the direction. The research programme is responsible for finding out which claims survive contact with prior art, logic, evidence, experiment and criticism.
Meta aim & lineage
Expand the effective adjacent possible, without claiming the phrase.
The adjacent possible and the effective adjacent possible are established research ideas.
Stuart Kauffman's adjacent possible describes possibilities reachable from a present state. Josef Taalbi's recent innovation research explicitly distinguishes an effective adjacent possible constrained by search complexity, absorptive/search capacity and resources. Agalmic Research uses that lineage rather than reinventing it.
The project asks what cognitive tools, provenance, validation and realization infrastructure do to the set of possibilities a human-machine system can actually reach.
Personal and programme discipline
Let the possibility space be abundant. Keep commitment scarce.
Ideas are options, not obligations.
The public Possibility Portfolio preserves worthwhile branches without silently converting them into projects. The Active Frontier is capped at three substantial bets by default, each with a discriminating next action.
Capture promiscuously. Commit selectively. Realize deliberately.
Prior art first
This programme is not starting on empty intellectual ground.
Many of the intuitions explored here have important antecedents, some very close.
Robert Levin used agalmics for the study of non-scarce goods decades ago. Herbert Simon tied information abundance to attention scarcity. Nelson and Winter, Weitzman, Kauffman and later innovation scholars treat innovation as search and recombination. Cohen and Levinthal formalized absorptive capacity. Grandori already used the term epistemic economics. Computational discovery, idea-management, innovation-portfolio and knowledge-brokerage literatures also substantially predate present generative AI.
Novelty is a conclusion of search, not a tone of voice.
The project maintains a living lineage and novelty register. Where prior art exists, we intend to use it rather than rename it; where a narrower extension remains, we will say exactly what that extension is.
Epistemic status
The programme begins by stating its own constraint.
Machine-assisted performance can outrun the human curator’s understanding, but that underlying problem is not ours alone.
Empirical and conceptual work on metacognitive calibration, verification gaps and claim accountability already develops closely related concerns. Agalmic Research uses Authorial Capacity Constraint as a local organizing label and asks how the problem connects to contribution provenance, handoff, uptake and the economics of abundant candidate knowledge.
Epistemic authority should not exceed defensibility.
Epistemic handoff
A good idea can outrun the expertise of the person who found it.
The right response is neither false authority nor dismissal.
The contribution/authority distinction builds on mature contributorship, epistemic-dependence, absorptive-capacity, knowledge-brokerage and intermediary traditions. The narrower research question is whether a claim-sensitive, provenance-preserving handoff can improve routing to missing expertise under machine-generated candidate abundance.
Origin deserves credit. Authority requires warrant.
Working thesis · Displaced Scarcity Hypothesis
When a previously binding constraint becomes sufficiently abundant, system performance and value become increasingly determined by complementary constraints that remain scarce.
This is an organizing synthesis, not a demonstrated law. Fixed-proportions bottlenecks, Simon's attention scarcity, Teece's complementary assets, Weitzman's idea-processing constraint, strong-complementarity models and recent AI-abundance economics all supply antecedents. The research task is to derive conditions and predictions that are not merely standard production theory in new clothing.
Distributed discovery
Machine generation and discovery credit are not the same question.
Computational scientific discovery and collective-centred accounts of AI-assisted discovery already exist.
Agalmic Research therefore does not claim to have discovered the distinction between machine generation and human significance recognition. The prospective contribution is operational: preserve problem framing, candidate generation, selection/recognition, validation, integration and realization as provenance events, while keeping novelty, priority, contribution and epistemic authority separate.
Novelty is a relation between a result and prior knowledge. Discovery attribution is a relation between a discovery episode and its contributors.
Research programme
A programme built from antecedents, extensions and open questions.
Existing theory is an input, not an obstacle. The goal is to reuse mature ideas, identify what machine cognition actually changes, and turn the remaining questions into testable work.
Epistemic handoff & uptake
Building on brokerage and absorptive-capacity research, how can low-authority contributions be routed toward missing expertise without confusing originator credit with epistemic authority?
Agalmic economics
A contemporary extension of earlier agalmics: what happens to value and allocation when useful knowledge, computation and other non-rival goods become dramatically cheaper to create?
Economics of epistemic abundance
When candidate knowledge becomes cheap, how should scarce attention, validation, trust, interpretation, institutional capacity and realization be allocated?
Distributed discovery
Building on computational and collective-discovery scholarship, how should novelty, priority, contribution and authority be represented when humans and machines split the work of discovery?
Innovation search under AI
Building on established search and adjacent-possible literatures, how does machine cognition alter search cost, branching, recombination, selection, diversity and realization?
Representation as discovery
How can systems compress, transform and embody information more cheaply while preserving the structure humans need to perceive what matters?
Knowledge, property and rents
How should publication, defensive disclosure, patents and incentives change when the cost of generating potentially valuable ideas collapses?
Applied research
Nemosyne as a testbed.
The arguments should survive contact with software.
nemosyne.world explores how representations can make meaningful structure perceptible without spending more computation, complexity or human attention than necessary. Broad claims about representation and discovery have substantial prior art; any defensible novelty must live in concrete Moneta/Nemosyne mechanisms and evidence.
Operating model
Search lineage. Preserve possibilities. Limit commitments. Build capacity.
Research begins as a canonical, versioned object. Ideas can be preserved without promotion. Before a claim is presented as a contribution, the project searches its intellectual lineage. Before an idea becomes an active commitment or publication, the project asks what new capacity it creates and what worthwhile future becomes more reachable.
Finding prior art, parking an idea, handing it off or closing a branch can all be successful research outcomes.