* This site is based on internet standards not supported by IE 6.
wJones
EC Container 5
Questions and Answers

What is Stored Purpose?

Questions and Answers
A system of technologies that enable the processes of intelligence to run on commercially available electronics, e.g. an intelligent computer.
Intelligence is defined as system that sustains a logical definition of “right” and “wrong” states called Identity, and uses some motive force to process measured regions of context beyond Self with respect to those states, minimally able to use some technology to gather measured states and some technology to align measured states of Self with Identity. For example, a very simple intelligent system’s Identity could define salty water as good, and warm water as good. It could understand some region of context in the Pacific ocean by using its bio technology to measure states of saltiness and warmth and its alignment or lack there of with preferred states. The system would finally use technology such as swimming feet, to re-align its Self with Identity when necessary.
The technology has six parts:
1. Stored purpose computer - The overall intelligent system architecture, a Gia based system compatible with commercial electronics
2. Platonic Forms (Forms) - A symbolic analog of the shape and causality that can define Knowledge, Goals, Purpose or Identity
3. General intelligence algorithm (Gia) - A computable representation of the facets of intelligent existence
4. Agent - An entity created when Identity and Gia computational resources are instantiated in a logical body
5. Symbolic network - The logical body of a multi-agent system
6. Metacomputer fabric - The physical body of a multi-agent system.
more ...
Comments

What is machine intelligence?

Questions and Answers
It is the outcome of seven processes: Store, Plan, Act, Sense, Pursue Goals, Understand, and Learn.
The processes act on a system’s Identity which is defined when an intelligent system is first instantiated. Identity is the “contextual space” covered by Intelligence. That space is divided into regions called Purpose. A Purpose could be, “Maintain the Grounds of Ashtin Hospital.” An intelligent entity could have multiple Purposes. Each Purpose is further divided into symbolic regions called Goals.
A Goal is a self directed graph that defines right and wrong for a region of Self. Goal Pursuit is the use of technology to re-align measured or predicted states with Self, i.e. right, when they are measured or predicted to be wrong.
For example, a lawn maintenance bot might store a Goal related to the height of lawn grass it is responsible for maintaining. The Goal would define right and wrong states of grass height, which could be between three (3) and six (6) inches.
Each Goal is saved in the basic information structure of Gia, called a Platonic Form, which stores, in an extremely brief format, a symbolic analog of the context, shape and causality of the Goal. The Goal Form also stores linkages to technology that allow a system to determine the state of it’s world relative to Self (i.e. sensors to detect if the grass is at the right height) and technology to realign Self with Identity when measured or planned states don’t align (i.e. if the grass is too high, a route from high grass to low grass, “./mower/grass/cut”).
Multiple Goal Forms can be combined to form a Purpose such as “Maintain Grounds.” The Purposes and Technology available to a system define its Identity. Within a system, Identities can be loaded and run in logical entities called Agents.
more ...
Comments

How does an intelligent system work?

Questions and Answers
A Stored purpose computer works by:
Composing the Forms of Identity to a type of memory, called Plan,
Prioritizing, optimizing and propagating Plan throughout the metacomputer,
Acting and Sensing by way of contextual linkages between Plan and physical metacomputer Instruments at time Now/Reality,
Pursuing Goals to align Reality with Self,
Processing an Understanding of how each thing affects Plan, and
Learning from processing the coincidence of symbols measured but not in Plan.
more ...
Comments

Will the roles of intelligent systems be similar to those held today by people, or primarily service oriented?

Questions and Answers
Systems will be service oriented. Intelligent systems were designed to assist and extend people in the workplace, not replace them. People and agents working together will make it cost effective for hometown factories to make elegant, long lasting, state of the art products, for pilots to fly planes faster and more safely and for cars to navigate through traffic more quickly using less fuel. People and agents together will more frequently align with higher value Purpose and better meet responsibilities at work and at home.
There will also be a second type of intelligent system, different from R2D2 and people, both of which share a simple exoskeleton design in which the brain travels with the body around a center of gravity. The new type of system will support multiple agent minds in a body, and span multiple Instrument device types linked by communications to form a virtual body. This type of system will be called a metacomputer.
more ...
Comments

What is a metacomputer? Is it consumer mobile, smart medical devices, a plane, a train, a city’s emergency management system, self running delivery bots or something that can help coordinate a business, an office building, or a hospital?

Questions and Answers
All the above. We’ve prototyped extending the logic, memory and communications of a metacomputer to Instrument form factors that could enable a metacomputer to comprise any combination of lights, dials, sensors, controllers, medical instruments, factory machines, engines, program processes, door locks, documents, media objects, displays, desktop computers and mobile phones. Instrument technologies can include smartdust-like low-power (TI CC430), handheld mobile (OMAP/NVIDIA CUDA), and larger systems (INTEL/AMD x64/NVIDIA CUDA) processors.
more ...
Comments

Is the technology practical?  I thought the processing power needed to reproduce the human brain wasn’t yet available.

Questions and Answers
The design doesn’t copy the human brain, instead it implements the fundamental principles of intelligence, natively in silicon. The Gia algorithm offers an extremely efficient logic implementation. That logic will be made ever more efficient as increasing numbers of systems are developed over the next decade. Current electronics technology is believed sufficient for most small and midsize solutions. For large solutions, such as monitoring national health emergency status, CPU-GPU processor technologies would need to be better integrated.
more ...
Comments

Would it really be reasonable or cost effective to build or deploy intelligent solutions in three to six years, as your materials suggest?

Questions and Answers
Yes, with two caveats:
First, an understanding that any investment must be for the long term and the first systems will be necessarily simple. Over time (that will measure in decades) intelligent products and processes will become increasingly capable.
Second, there are also risks: regulatory, technical and market related, that could delay any planned product introduction.
The key regulatory risk is that regulatory debate will start after first products are ready for market, despite efforts to engage government early.
The key technical risk is that regulatory requirements could add late technical requirements.
The key market risk is that important aspects of the technology’s design that will make solutions safe, secure, protective of privacy and assistive in nature, rather than a replacement for human labor, would not be adequately communicated to the public.
more ...
Comments

What are the potential safety issues? How will they be addressed?

Questions and Answers
Stored purpose systems will be inherently more safe than non-intelligent systems. This is because unlike legacy stored-program systems, that execute program logic with no understanding of consequence, each action performed by an intelligent computer must align with its Identity, a definition of “right” and “wrong” defined by the system’s authority (o.patron). Also, systems will have additional high priority “curtailment” Purpose that explicitly defines Goals for safety (o.protect) and legal authority (o.legal).
That said, Stored purpose systems will eventually process very complex Purpose, consisting of thousands or millions of Goals. They will also be able to learn, if granted the requisite degrees of freedom. To validate the decision processes of these complex systems, new technologies will have to be developed over the next two years to test and monitor processing. As stored purpose starts to displace stored program systems, software engineers will need to retrain, learning new types of processing logic, that will include curtailment, prioritization, valuation and mediation. They will need to convert from “programming” explicit instructions to “balancing” Goals and Purpose and “mediating” intent among agents. Civil infrastructure will need to be upgraded to enable police to validate Purpose of mobile intelligent systems. Transportation infrastructure will need to be upgraded to enable bots to refuel, recharge and travel as cargo.
more ...
Comments

Is there a large enough potential return to warrant the three to ten million minimum in R&D over the next five years to build a product?  What are the markets?  What would be a candidate first “smart” product?

Questions and Answers
Yes. All. Let’s discuss.

These are questions a company should answer through a careful planning process.
We can assist.
more ...
Comments

EC Container 6