The second part of the Ethika AI development is all about disinformation. But before I launch a whole new chapter about disinformation, let me break this down into sub chapters. The future of AI is not built on sophisticated coding routines, the future of AI is inclusive. Inclusive as in a symphony between computer components, advanced coding routines, arranged and perfected through human ingenuity. At THEMATIC AI we define our AI, Ethika AI, as an Inclusive AI system that seamlessly combines safety, security, and versatility. I made the introduction about our design decision to focus on safety and security, effectively bypassing public concern about AI behavior without human oversight. Let me explain versatility as the architects of Ethika AI and why I use the term Inclusive AI.
To fully appreciate and make use of AI, such a system would factor in the capabilities of computer components, the digital pathways to align and optimize the stream of data between components and computations with reverse feedback of combined data output and so on.
With the recent ChatGPT v4 release to the public, its data collection is from large databases stored in cloud server pools. AI data research is very much in the stage, and continues rightly, to improve accuracy. This refinement, even with excellent optimizing coding structures, is a huge and time-consuming task. Industry groups with a focus on research, communications, data investigations or narrow data retrieval requirements do not benefit from this large data pool. The bulk of cloud data might be stored somewhere but is difficult to pinpoint because of the complexity and interconnection with too many data access points. In other words: the term making sense out of raw data is applied here. In our design approach to make Ethika AI the most accurate, safe, and versatile system, the AI capabilities are extended by introducing sensors and circuitry with a specific purpose to collect data.
I name the examples of advanced CMOS photocells, environmental sensors, motor components, and semi-conductors to improve accuracy feedback to the AI system. Ethika AI is from start ready to accept industrial data exchange, the same coding-based on Apache's MXNET is easily expanded to cater different industries: from environmental detection to support better modelling and tracking of climate changes, rapid and efficient sampling of disease specimen, monitoring of hazardous environments to conduct and support investigations of (war) crime sites or residential environmental improvements.
The word is versatile, as the architects designed an AI system, Ethika AI, capable of real-time fact-checking its environment to enhance the feedback from Human Experts. An Inclusive AI system capable of matching demanding customer profiles from different industries.
PART IIb: Educating the Ethika AI system about Disinformation.