Today, people know AI as one of the most powerful tools available to humans. With the technology’s extreme complexity, it’s not hard to believe that it was nothing more than a theory not too long ago. However, AI actually has quite a detailed history that many are unaware of.
So, where did AI start? Here are 10 key moments in the rise of what many consider to be the technology of the future.
1950
Computing Machinery and Intelligence - Alan Turing
Often considered the “father of AI,” Alan Turing published one of the most influential works to the field of artificial intelligence: “Computing Machinery and Intelligence.” This paper explored whether or not it was possible for computers to think. He addresses this topic by describing the “Turing Test.” Turing suggests that if a human can converse with a machine without realizing that they are not speaking to a human, then the machine should be considered intelligent. He refutes objections to the existence of machine intelligence, while not outright claiming that machines can think in the same way humans can. All in all, this paper and Alan Turing as a whole were foundational in the philosophy of artificial intelligence and shaped the way scientists and researchers approached the field.
1956
Dartmouth Conference
Organized by John McCarthy, Marvin Minsky, Nathaniel Rochester, and Claude Shannon, the conference brought together a small group of researchers to discuss the possibility of creating machines that could simulate human intelligence. This conference was the birth of AI as a field, and was proposed by John McCarthy, a highly respected computer scientist who coined the term “artificial intelligence.” The goal was to precisely describe the meaning of intelligence and how a computer could be made to possess it, setting the agenda for AI development for years to come.
1966
ELIZA
Generally considered the first chatbot, ELIZA was developed by Joseph Weizenbaum as the first program to simulate a human-like text conversation. It operated on several templates, most famously being the “DOCTOR” script. While it lacked real comprehension, it gave the impression of understanding the patient’s feelings and expressed sympathy, care, and intelligence.
1986
Introduction of Backpropagation
The publication of the seminal paper “Learning representations by back-propagating errors” by David E. Rumelhart, Geoffrey E. Hinton, and Ronald J. Williams solved a standing issue with neural networks. While the idea of training neural networks with multiple layers was already established, it was unclear how to effectively update the weights of neurons. Backpropagation solved this issue by providing a way to make these computation and allow networks to learn patterns. This breakthrough was foundational in the development of deep learning.
1987-1993
Second AI Winter
The second AI Winter was a period of reduced interest, funding, and optimism for the development of AI. One cause of this stagnancy was the heavy marketing for AI in the 1980s, which overhyped the technology and led to unrealistically high expectations that were ultimately not met. This issue was exacerbated by the limited hardware and software capabilities as well as the lack of large datasets available. As a result, many AI researchers moved to other fields and were unable to find support for their work.
1997
Deep Blue vs Gary Kasparov
A series of Chess show matches were held between the IBM supercomputer Deep Blue, and Gary Kasparov, arguably the greatest Chess players ever in his prime. Winning 3.5-2.5 against Kasparov, Deep Blue became the first computer to win a match against a World Chess Champion. This match marked a significant point in AI history, spurring increased interest and funding while showcasing the technology’s potential.
2011
IBM Watson on Jeopardy!
Similarly to Deep Blue vs Gary Kasparov, IBM Watson’s appearance on Jeopardy! was a landmark event in demonstrating the capabilities of AI. The computer competed against two champions of the show, ultimately coming out on top with $77,147 which was donated to charity. This demonstration of skilled natural language processing helped prove that AI could efficiently comprehend language while effectively analyzing and retrieving information, boosting public awareness and confidence in the technology.
2012
AlexNet
AlexNet, developed by Alex Krizhevsky, Ilya Sutskever, and Geoffrey Hinton, transformed the field of computer vision by winning the 2012 ImageNet competition with a breakthrough accuracy. The model’s innovative design, which included eight layers and the use of GPUs for faster training, made deep learning more practical and accessible. To enhance performance, AlexNet used techniques like data augmentation and dropout to improve its ability to generalize and avoid overfitting. This success sparked the deep learning revolution, showing the world the potential of AI in understanding and recognizing images on a large scale.
2016
AlphaGo vs Lee Sedol
AlphaGo, an AI developed by Google’s DeepMind, was the first computer to beat a professional human Go player. Go is an ancient strategic board game that is known to be difficult for AI to master. Lee Sedol, the computer’s opponent, was widely regarded as one of the best players at the time. The AI used deep learning to develop strategies new to even expert Go players. The defeat was shocking to many, and just like Deep Blue and IBM Watson, has a legacy that speaks to the advancement of technology.
2022
Release of ChatGPT
ChatGPT continues to revolutionize the world by making advanced large language models available to anyone. It has influenced the public’s perception of AI, shifting it from a niche technology to a mainstream tool with practical applications. For the AI community, ChatGPT has set new benchmarks in natural language processing and has driven research into more ethical and responsible AI development. Today, many would consider ChatGPT the face of AI to the general public.
See more like this:
