The Maturing of AI: from Science to Technology and Back

Artificial intelligence is a science or a branch of computer science. But it is now also becoming a technology for the real world. This event is happening at the right time, when the complexity of today's business environment is increasing faster than human skill. In all industries and government bodies, events happen at a pace requiring the information infrastructure to quickly adapt (figure 1). This adaptation cannot occur using conventional techniques. Until recently, modern information systems could withstand the demand of the user by somehow bringing more power than could be used, more languages and tools than could be learned. The remarkably rapid adoption of these new techniques, from computer-aided design to databases and personal computing software, has generated complex environments that require a new level of computing. The in-frastructures need to be managed, the software needs to be glued, the productivity of application development needs to be increased, and the development cycles need to be reduced. It is the combination of the systems' complexity and the fluidity of today's business contingencies (that these systems help establish) that requires a new generation of software. Human skills simply cannot evolve fast enough and need to be better used. It is clear that the environment we create evolves far faster than these limited resources. Undoubtedly, we need to either manage this environment or have it manage itself. This feat can only be achieved by manipulating knowledge. It is not just a research perspective for some future world; computing with knowledge has become a natural need, arising from the complexity of the world we built. Many examples illustrate that this knowledge computation is becoming a reality. AI has emerged as the technology that allows us to bridge the gap, adapt our existing environment to the rapidly changing conditions , and create new software applications designed to be an extension of not only the physical body but also the mind. As shown in figure 2, the place of this new technology is on top of the existing one. One result of this interdependence between conventional and AI technology is the increased attention being paid to integration of AI and conventional technology. Just as the mind without the body is meaningless , there also cannot be an " AI-centric " view of the world. Some examples of this evolving complexity follow. If these global x RAPPAPORT & SMITH complexity human skills business environment time 1990 …