Agent based systems are more common than we may think. A Promise Theory perspective on cooperation, in systems of human-machine agents, offers a unified perspective on organization and functional design with semi-automated efforts, in terms of the abstract properties of autonomous agents, This applies to human efforts, hardware systems, software, and artificial intelligence, with and without management. One may ask how does a reasoning system of components keep to an intended purpose? As the agent paradigm is now being revived, in connection with artificial intelligence agents, I revisit established principles of agent cooperation, as applied to humans, machines, and their mutual interactions. Promise Theory represents the fundamentals of signalling, comprehension, trust, risk, and feedback between agents, and offers some lessons about success and failure.
Figures & tables
Fig. 1: Autonomous agents each have their own internal language. There is no authority that calibrates these to be the same without the agents’ consent. This means that, between any two agents there is a subset of language that overlaps and is comprehensible to both, but may still mean different things to each. What happens where these meet?
Fig. 2: Overlapping patches of language require all agents in a patch to use a compatible language, and for (at least some) agents in each patch to comprehend a faithful translation of a neighbouring language, in order to bring long range order.
Fig. 3: The aggregation of promises to assemble a new (tensor) promise is dependent on the overlap between each pair of agents, which involves language uncertainty LiR and measure uncertainty bi(+)∩bRi(−) .
Fig. 4: Cooperation in parallel and in series presents different challenges, each with their own uncertainty.
Fig. 5: A potential difference (assessed trustworthiness) acts like an incentive to begin kinetic sampling of the other party, and thus drives work.
Fig. 6: An agent seeing insufficient relative trustworthiness would continue its own internal kinetic work without interacting.
Fig. 7: Assured delivery on a promise through a proxy. delivery.
Fig. 8: Graphical view of a three intermediary promise chain. the promises in rectangular boxes along the chain move down the chain from server S to proxy P3 handing over the service itself. The oval promises are directed back up the chain providing delivery assurances. Finally, each agent promises independently with the end-point or client that it is an authorized agent for delivering the promise from the source (server).
Fig. 9: When we deal with agents that operate on a high level, like humans in the workplace, most singular promises are made in an ambient environment that’s already filled with promises and perceived obligations (tacit impositions) as a matter of ’common knowledge’. This can make it difficult to resolve exactly what an agent should be trying to do.