Open Intelligence is a very interesting and influential phenomenon. It is especially intriguing to me because I believe that it can play a positive role in developing countries. In order to ensure its long-term sustainability we need to see it "as is" and clearly identify possible pitfalls as well as open intelligence's strong and weak points. Fundamentally, we need a reliable map of the open intelligence environment. As described in the paper, Controlled Intelligent Spaces already exist in the day to day world but to use it in context that targets the needs of the individual is something that needs greater scope in the real space. Context is the real logical model that we as users create and also absorb resulting in a dependency that in-turn creates abundant amount of information that the general user consumes. Building such context driven systems should be the fundamental task of every engineer for the benefit of human future and for preserving valuable information.
Relevance of the Paper
The paper discusses about the potential of building such Context-awareness systems via two main approaches. The Agent-Centered Context-Awareness (ACCA) and the User-Centered Context-Awareness (UCCA). Both of these are such brilliant ways to adopt an environment that is totally open spaced and to bring our own context into it. It talks about the one which revolves around the context that the agent perceives and one that is totally universal out there, to be grabbed and conceptualized; one that is controlled intelligence and the other that is open intelligence.
Such a development demands continuous monitoring and testing to ensure that the context does not fall out of the space and that the boundaries are well in scope of the context. I say, why waste the skills of some talented developers on debugging? In some commercial environments, professional testers provide an important edge. Synergy between developers and testers is so very important. An agent would want to create his own space and work around the context he has chosen for development. In this case the agent is the “self” of the device and hence in this case would even prefer to debug on his own. Representing the tester or the user for instance would be completely secondary to the developer. An inspiring example of such unconventional thinking and development is “Bump Top”; software that applies a 3D metaphor and rough-and-tumble interactivity that delights anyone who sees it in action .
At the same time, because context sometimes is so user centric; there is no harm in placing the common user as the “self” and create a universally agreed-on model. In such a scenario, the system must be dynamic enough to be able to adapt to constantly changing environment. The model must be built such that context is universal and any open context fits into the model of the “self” and the environment. The best and the most common example of this is the Windows Operating System; a brilliant software that fits into every possible corporate environment and is reusable in every sense of software development.
Cultivating a user base is very important for any system. The context that is so “self” reliant, which is the case for both ACCA & UCCA, needs to be complete and the best available. For this, and to manage context dependencies, user relationship needs to be established and maintained. Context quality needs to be assessed by each agent and conflicts resolved within the boundaries of open intelligence using the user base.
Also, the context driven system built under such environments largely rely on the resources of the controlled intelligent space for governing the basic security principles of Confidentiality, Integrity and Availability. This is a limitation that needs to be overcome by writing applications that fit into open intelligent space.