NOTABLE MILESTONES IN THE DEVELOPMENT OF AI – PART 1
In our last post I wrote about how AI as we know it today has been decades in the making and how the modern era of AI can be traced to the Dartmouth College meeting in the summer of 1956 and was organized by John McCarthy who coined the term “Artificial Intelligence.”
Within a very short time, Allen Newell, J.C. Shaw and Herbert A. Simon developed what is considered to be the first AI program to solve mathematical theorems. They then developed a more generalized problem-solving program. This laid the foundation for “Perception” which was developed in 1957 by Frank Rosenblatt. Perception was an initial “neural network” that could learn simple tasks.
Over the next 20 years we had the development of “Expert Systems.” These used knowledge bases encompassing rules, heuristics and facts. Heuristics are problem solving techniques that are geared to find a solution quickly even though it may not provide perfect results. Expert Systems were used for guidance in the fields of Engineering, Troubleshooting, and Medicine and they saw significant adoption in the 1980s.
In 1986 an algorithm called “Backpropagation” gained traction and it dramatically improved the training of neural networks. Then in the 1990s there were machine learning advances and improvements in speech recognition. Notably in 1997 IBM’s Deep Blue defeated the world chess champion.
In the 1990s with the adoption of the internet methods were developed to access, retrieve and organize the massive amounts of online data. Having this huge data set available aided AI which requires data for learning and validation. Tangentially, internet adoption also assisted researchers around the world in their collaboration and knowledge sharing which significantly sped up the development of AI.
In the next blog post I will fast forward through the rest of the big milestones in AI development to the present time…