Thinking Machines

A self-learning Autosopher text database.

Application Proposal: 1998

Abstract: A new Autosophy information theory permits the building of a new generation of self-learning brain-like machines with operations entirely different from programmed data processing computers. A first generation prototype hardware text learning “Autosopher” is currently operational and available for demonstrations. The system acts like a “black box” which is educated similarly to a human child to organize its own internal learning operations without conventional programming. Autosopher communicate essentially like human beings using grammatical language questions and answers. A new Content Addressable Read Only Memory (CAROM) would be the ideal memory for intelligent robots. It consists of Thin Film Transistors printed onto large sheets of stainless steel foil which are then wound into tight rolls. Transforming technology from today’s data processing computer to a new generation of self-learning brain-like Autosopher will be a giant step which may require years of research and development. This proposed research would be only the second generation in the long series of generations, which may eventually lead to true thinking machines. The basic thrust of this research is therefore to develop practical applications, which can be sold at a profit to provide the incentive and investment to continue the research.

Applications: An Autosopher learning system, using software-only in a PC, will act like a “black box”, which is educated, using grammatical questions and answers, similarly to a human child. The software modules will be written in portable C++ or Java code for insertion into commercial software packages. All eight known learning modes will eventually be provided. The software package would also provide for secure and compressed multimedia communications on the Internet, providing a universal communications standard for transmitting text, high resolution still images, and video sequences. In a second phase a hardware module would be developed to plug into the PC bus, providing real time secure teleconferencing via the Internet. Video recordings and movies could be stored much more efficiently on CD-ROM or DVD discs without compromising image quality. The module would compress and encrypt any type of data including text, sound, speech, still images and real time video. The hardware modules may then evolve into self-learning and self-indexing multimedia databases, which could store Millions of text records, still images, or live video sequences with orders of magnitude lossless compression and virtually instant search access.

Keywords: Autosophy, Self-learning databases, Multimedia Archiving.

Available downloadable documents:

Proposal document – MS Word doc

Related Publication 1993 – Webpage htm

Related Publication 1991 – Webpage htm