5
minute read
"You don't want to examine the basis of your computer's morality any more than you want to see sausage being made,” computer scientist John McCarthy warned after he coined the term ‘Artificial Intelligence’ at the 1956 Dartmouth Conference in New Hampshire.
McCarthy, along with Marvin Minsky, Nathaniel Rochester, and Claude Shannon, are considered the founding fathers of AI but little did we know then that AI had long-standing ties to the intelligence community. Join us on our wild ride through the history of AI and espionage.
Giant bronze robot Talos
Thousands of years before Elon Musk began dreaming about self-driving cars, there was the giant bronze robot Talos, the artificial woman Pandora, and their ancient Greek creator god, Hephaestus, according to Stanford researcher Adrienne Mayor. It seems the idea of automata can be traced to the Middle Ages, and the concept of artificial, lifelike creatures dates back at least 2,700 years, foreshadowing our obsession with creating intelligent machines.
AI emerged through the collective efforts of countless scientists, some notable figures include Alan Turing, Herbert Simon, Arthur Samuel, Geoffrey Hinton, Yann LeCun, and many others.
The Turing Test
Bletchley Park codebreaker Alan Turing, probably best known for cracking the German Enigma code, proposed the Turing Test (1950). Known as ‘The Imitation Game’, it introduced the concept of evaluating a machine's ability to exhibit intelligent behavior indistinguishable from a human being by using the replies to questions put to both.
Eliza ‘Doolittle’ AI
AI still hadn’t passed The Imitation Game by 2023, but there have been contenders. MIT computer scientist Joseph Weizenbaum developed a ‘chatbot’ in 1966 called Eliza (named after Eliza Doolittle from the play Pygmalion) programmed to search for keywords in questions and issue responses. The program allowed a crude type of conversation between humans and machines: Eliza rephrased whatever speech input ‘she’ received in the form of a question. If you said you were feeling anxious, Eliza might ask, “Why do you feel anxious?” Things were about to get a whole lot more intense though, as foreshadowed in Stanley Kubrick’s 2001: A Space Odyssey (1968).
Expert systems (1960-1980)
Expert systems - aka knowledge-based systems - such as MYCIN for medical diagnosis and DENDRAL for chemical analysis were developed from the '60s to '80s. They utilized rules and knowledge bases to mimic human expertise in specific domains. Dendral (short for Dendritic Algorithm) was an expert system for chemical analysis that demonstrated the potential of AI to assist in complex problem-solving tasks.
The emergence of neural networks and deep learning (1980 - )
Beginning in the 1980s, researchers Geoff Hinton, Yann LeCun, and Yoshua Bengio made significant contributions to the development of neural networks and deep-learning algorithms which have revolutionized AI applications including computer vision and natural language processing. One huge breakthrough was the advent of deep convolutional neural networks (CNNs) and their application in image recognition. The introduction of AlexNet in 2012 paved the way for advancements in computer vision, leading to applications like facial recognition, object detection, and autonomous driving.
Deep Blue, 1996-97
World Chess Champion Garry Kasparov faced off against supercomputer Deep Blue in a six-game series in 1997. Only 35 percent of US households had a computer at the time and Kasparov did not expect to lose. Why would he? A year earlier, they’d squared off and Deep Blue won only one game in six. But that one game showed IBM they were on the right track. In the 1997 rematch, the upgraded Deep Blue could evaluate 200m chess positions per second. Kasparov could evaluate about three per second. Deep Blue defeated Kasparov 3½-2½ - the first artificial intelligence construct to ever win a chess match against a reigning world champion.
