New Sapience has achieved a breakthrough in computing: Modelled Intelligence, a software technology that endows machines with Computational Knowledge: human knowledge that has been transformed into a computable structure in a computer program. It is the result of a multi-disciplinary approach combining state-of-the-art computing techniques, practical experience in modeling complex real-world systems, and an original theory about the underlying structure of human knowledge.
Endowing computers with the Computational Knowledge in addition to their already excellent capacities to process data and information is the key to achieving Artificial General Intelligence (AGI). Considered the “holy grail of computing,” its possibility has been a subject of fascination since the dawn of computing, hyped as the savior of mankind and sometimes dreaded as its doom. But one thing is certain. Our greatest power, the one that most distinguishes us from all other known species, is our mind’s capacity to create knowledge. Combining that with computer speed, connectivity and perfect memory will result in the greatest amplification of human power ever.
Modelled Intelligence contains, extends, and applies knowledge. The looked for revolution from Artificial Intelligence is not decades away – it is happening now.
Today, Artificial Intelligence and Machine Learning are being used to solve a number of distinct problems in specific verticals. Verticals are narrow and deep, which is why these various distinct techniques are often referred to collectively as “Narrow AI.”
Modelled Intelligence is general as opposed to narrow. It can do something very similar to what humans do when exercising their intelligence: acquire and create knowledge of the world useful for doing whatever needs to be done. Potentially, the problems we solve are the kind that people, when using their intelligence, solve. However, while we think it is fair to say our technology is in essence – generally intelligent, our initial products will not be nearly as intelligent as a human in degree.
Computers are already intelligent in the sense of reasoning, pattern matching and memory management, but up until now they have been ignorant. Computational Knowledge is game changing. As we increase the quality and quantity of computers’ world knowledge, those superb reasoning capabilities can finally be harnessed to solve real world problems with human level generality and flexibility. The day when computers must be accounted as having true AGI in degree as well as in essence is not far off.
In the near term, even modest degrees of general intelligence and knowledge can solve a problem that has solidly eluded narrow AI techniques. Being able to know what words mean is the key to genuine comprehension of natural human language. Combining a human-like capacity to understand language with the perfect memory, speed and connectivity of computers opens up a vast realm of new possibilities, even with our first-generation products where applications’ level of intelligence will still be modest compared to humans. Future generations of products will ultimately be able to augment/replace knowledge workers.