Artificial Intelligence (AI) has seen a resurgence recently with breakthroughs in machine learning and the application of deep neural networks. This article explains what has caused this resurgence, what challenges and opportunities AI offers, and a view on where this technology may go in the future. AI is not new. The US Dartmouth College Conference in 1956 established many of the founding principles. Since then successive generations of AI technologies, each embodying growing levels of intelligence enabling a new set of AI applications, have emerged.The 1970s and 1980s started with simple chess or checker games, natural language parsers and expert systems. These were hand-crafted using symbolic techniques such as heuristic or rule-based technologies. By the 1990s, a growing number of large-scale AI applications emerged. Soft computing techniques, such as fuzzy logic and evolutionary computing, started to bridge the world of symbolic AI and numerical technologies with applications in elevator and washing machine control systems, autofocus cameras and early assisted-driving systems. The first telecoms applications of AI also appeared where constraint logic programming and heuristic search techniques were applied to field force management.
展开▼
