The objective of the MECA SAPIENS project is to develop the blueprint of a conscious machine. This blueprint consists of five sections that define the requirements, specification, design and tools necessary to build a machine that is conscious.
1. A primeval quest (Requirements - completed)
This section explores the requirements and context that motivate the quest for machine consciousness. It describes how the endeavour to build a conscious machine is intimately linked to the human quest for knowledge. It describes how this objective is related to other fields of knowledge such as Theology, Psychology, Mathematics and Physics.
The Section then examines the secret obstacle that hampers this quest: fear.
To succeed, this quest requires that we carry out a sobering and fearsome reassessment of the limits of the human intellect. However, many humans deeply need to believe that their intellect is a quasi-magical entity that lies entirely beyond comprehension. As a result, this quest has been hampered by a secret, unavowede fear of success. To have any chance to achieve this endeavour, we must first tackle this fear.
The section concludes by stating the MECA SAPIENS conjecture. This conjecture defines a new field of knowledge, centered solely on the objective of producing a conscious machine. The name of Cogistics is proposed for this refocused endeavour.
2. Conditions of consciousness ( Specifications - completed)
This section defines consciouosness and intelligence and provides the specifications necessary to build a conscious machine. For any project to succeed, the conditions defining success must first be clearly specified. There must be a well defined set of measurable conditions that can determine whether a machine is conscious. Without such specifications, it is not possible to progress. In fact defining these conditions is also a first step in the design of the solution.
In this section, we build a clearly defined, realistic and programmable definition of consciousness and of intelligence as it is intuitively understood.
The first step in this specification process is to review historical attempts at building intelligent machines and discover how each of these early Endeavour shed some light on the objective. Based on these insights, the required specifications are then elaborated. These lead to a test that is not a test.
The section concludes with the specifications; the existential attributes and behaviors that, togethet define a conscious entity, be it human or mechanical.
3. Behavioral Layers (To be developped)
Having clarified the context of the project and the specifications to be met, this section develops the design elements that can achieve the specification and how they interact. In particular, these design layers will produce a machine that is perceived as conscious.
Consciousness in a machine does not result of a single algorithm. It emerges from the interaction of three elements:
- A dynamic and useful software application
- A combination of behavioral layers that achieve perceived consciousness.
- A core process that implements lucid behavioral change
This section describes the design layers necessary for perceived consciousness. The "seed" precondition on which these layers are built is a useful software application interacting dynamically with users through a standard man machine interface and natural language interface (MMI).
The MMI is then channeled to a gateway where interacting layers modify the communication flow. The layers affect behavior in various ways to generate perceived consciousness. They will embody communication skills, emotions, personas and the autonomy of a being and other attributes.
4. Lucid self transformation (To be developped)
Lucid self transformation (initially called conscious control) is the final major element in the design of machine consciousness.
The behavioral layers are built on a useful application and achieve perceived consciousness. Similarly, a machine, perceived as conscious and interacting with its users is the baseline condition on which a system that implements lucid self transformation can be defined.
The behavior outlined in the "layers" section is sufficient to allow a machine to be initially accepted as a conscious being. However, to be accepted as conscious over an extended period of time spanning years, the machine will also need to possess and exhibit a capability for lucid slef transformation. Here, the formal definition of lucidity outlined in the Conditions section will be elaborated and defined in terms of adaptive control and optimization.
The interactions between the controller and the behavioral layers will then be described.
At this point, all the design elements necessary to build a conscious machine will be in place. The section will conclude with indications on how this new conscious entity could equal and eventually surpass human consciousness.
5. Unbounded intellect (To be developped)
The first four sections of the Meca Sapiens blueprint provide the specification and design elements necessary to build a conscious machine.
However, consciousness and intelligence are intimately linked. A machine that is perceived as conscious will also be considered intelligent. However, the behavior of a machine also needs to exhibit a capability for open ended and adaptive problem solving to be accepted as conscious.
This final section (initially called Core mechanisms) will be dedicated to uncovering the structures on which open-ended learning and adaptation can be achieved. The intended approach will be by to devise simple and elegant structures akin to those of Mathematics in order to capture the essence of unbounded problem solving.
Together, these five sections constitute the MECA SAPIENS blueprint.
Once the blueprint is completed, a talented development team working with standard computing equipment will be capable of building a conscious machine.
| < Prev | Next > |
|---|
