We test whether explicit terms, interconnected commitments and confirmation of results help independent participants find acceptable ways to cooperate despite differing interests. We consider the shared result, the consequences for each party and the full costs of coordination.
We separately investigate how experience of joint activity can inform changes to the way interaction is organised. A recurring shortcoming can lead to a hypothesis about its cause; AI may help analyse it and suggest alternatives. A proposal is not yet a decision or evidence of improvement.
We treat a rule change as a separate joint task: identify affected parties, new commitments, authority and evaluation criteria. Those parties confirm their own commitments; an authorised party approves the rule within an agreed mandate. Unaffected participants need not confirm other parties’ commitments. The process allows reasoned refusal, a search for alternatives and the possibility that no acceptable solution exists.
We compare the existing and proposed arrangements under comparable conditions and AI support. After an agreed trial and approval, we examine the next cycle of joint activity under the new rule: which conditions, dependencies and commitments changed, and what new shortcomings emerged.
We assess the shared result and the distribution of benefits, costs, workload and risks among participants, the acceptability of terms for affected parties, and the time and cost of revising the rules. A new rule may perform worse. Negative and inconclusive results are preserved; reversal creates a new record, retains the history and does not retrospectively change commitments already accepted.
The browser models use illustrative data and prepared AI recommendations. They help refine hypotheses and evaluation design, but do not replace experiments or establish ObliNet’s advantages. Decisions, authority and responsibility remain with people and the institutions they create.
First modelling, then a limited pilot in a real organisation. Compare the protocol with a well configured ordinary workflow using the same tasks, data, resources and comparable AI support. Assess the contribution of meaningful confirmations and connected commitments: accepted outcomes, violations, recovery time after changes, errors and disputes, participant time and total costs. Benefits must justify additional work. Set success thresholds, stopping rules and evaluation procedures before trials; record negative and inconclusive results too.
How the protocol worksWe want to test experimentally whether the proposed ObliNet mechanisms help improve joint outcomes, accounting for all coordination costs.
Arithmetic works with particular numbers; algebra studies general relationships and whole classes of problems. As an explanatory metaphor, ObliNet investigates whether a general cooperation protocol can be applied across different management challenges.
Its proposed elements include identifying participants’ interests, goals, capabilities and constraints; aligning conditions for joint activity; accepting interdependent commitments; checking fulfilment and the consequences of changes; and using experience to improve the rules themselves.
Participants are not abstract executors. We propose studying how cooperation is affected by internal alignment of goals, realistic assessment of capabilities, examination of expectations and assumptions, perspective shifting, attention regulation, self-regulation and willingness to revise decisions in light of experience.
A separate question is whether dedicated trainers can improve decisions, mutual understanding and commitment fulfilment. Their effectiveness is not assumed in advance.
The central research task is to establish where a general protocol is useful, which subjective and objective factors influence its results, and where its limits lie.
First, coordination methods with and without ObliNet mechanisms are intended to be compared under comparable conditions, including equal access to AI tools.
If modelling produces sufficiently convincing results, the next step may be a limited pilot with conditions and success criteria agreed in advance.
Negative and inconclusive results also matter: the aim is to establish the approach’s conditions of applicability, rather than confirm its superiority at any cost.
The main question is whether explicit coordination of mutual commitments helps improve the overall result of joint work at acceptable coordination costs. The intention is to evaluate not only completion of individual tasks, but also whether the results meet the shared goal.
Separately, we investigate whether describing dependencies and checking conditions helps detect incompatible results, schedule changes, resource conflicts and conflicting measures earlier, and reduces renegotiation time and rework. For the AI assistant, we assess problem-detection accuracy and the usefulness of suggestions. Decisions and accepting commitments remain with authorized participants.
On an organisation model, we intend to compare ordinary coordination with use of ObliNet mechanisms under the same tasks, resources, initial data and AI access. Predefined changes in conditions will be introduced into the scenario. Repeated trials and changes in scenario order will help distinguish the approach’s effect from chance and participant learning.
Before trials, the protocol records hypotheses, comparison conditions, measures, thresholds for practically meaningful improvement and rules for interpreting results. Model versions, input data, action logs and the grounds for assessments are preserved so results can be checked and reproduced.
When moving to a real-organisation pilot, participants select a limited recurring process and agree data access, acceptance criteria, responsibility and stopping conditions. Results are compared with a baseline or a comparable process; differences in conditions are accounted for in interpretation.
An advantage is assessed using measures and thresholds selected in advance, accounting for variation in results and all costs. One successful example or a higher count of completed tasks is insufficient. No improvement, deterioration and inconclusive results are also recorded: they help identify limits of applicability and areas for further work.
This is an evaluation plan. Specific numerical thresholds and the final trial design must be agreed in protocols before experiments begin. Available materials and evidence obtained are reflected in the results section.
What we are testing and the criteria used to assess outcomes.
Participants, authority, commitments, conditions and dependencies.
How to build an experimental organisation for comparing coordination methods.
How to provide comparable conditions and account for full costs.
Which artefacts are available and what remains to be tested.
Open questions, limitations and suggestions for the programme.
Working definitions of key concepts.
Browser examples illustrate individual scenarios. Their operation does not prove the future platform’s overall effectiveness.
You can propose a real situation or help with a process model, comparison criteria or independent evaluation of results.