How AI has changed over the years?

During the Covid-19 pandemic, if someone had told me that in a matter of years an AI program that adapts to your prompts and gives a unique response to each person would be developed, I would laugh in their face and tell them they’ve watched too many science fiction movies. But now, all you need to do is look up “Chat GPT” in the search bar and you immediately get a website that does everything described earlier, and possibly more. It contains many other programs and apps having similar features, even being able to create images. What breakthroughs and discoveries about AI have been found leading up to this point.

First of all, in 1950, the concept of “artificial intelligence” was coined, with Alan Turing publishing “Computer Machinery and intelligence” which proposed a test of machine intelligence called The Imitation Game, eventually becoming the well-known Turing test.

From 1950-1956, AI began development, already managing to reach huge milestones in these few years, such as Arthur Samuel developing an AI program that knew how to play checkers and respond to moves and workshops on AI popularising the concept further. 

AI then started to mature in 1957-1979, with the term “machine learning” being coined, idealising the concept of a program eventually becoming smarter than it’s creator through independent learning and adaptation to its environment, and even bigger milestones being achieved, such as the first industrial robot being created and the first autonomous vehicle being created (albeit only being used in a controlled environment).

In 1980-1987, the concept of AI started to boom, being a period of rapid interest and growth as well as breakthroughs in AI research. Researchers were being funded millions of dollars and coming out with expert results, with a driverless vehicle able to drive properly on a road unlike the previous vehicle, a system that automatically picks computer components based on a customer’s needs, and an autonomous drawing program known as AARON.

After an “AI Winter” from 1987 to 1993, in which AI got less funds from the government and less breakthroughs, AI finally started to get some strides forward in the time period of 1994-2011, with a robot simulating human emotions using different physical features such as eyebrows and a mouth in 2000, companies such as Twitter, Facebook, and Netflix utilising AI algorithms as a part of their advertising and user experience as early as 2006, and Apple managing to launch their famous virtual assistant Siri.

Finally, from 2012 to the present, AI has become more mainstream than ever, with more companies such as Google and Amazon making their own virtual assistants, Facebook teaching two chatbots negotiation, and finally, OpenAI starting testing of the groundbreaking ChatGPT in 2020, a chatbot generating extremely humanlike response to almost any prompt you give it. Nowadays, ChatGPT is a widely accessible website that anyone can take a couple of seconds to search up and immediately have an AI companion to ask questions and get answers, essentially a more developed version of the version of ChatGPT tested in 2020.

Overall, it’s very interesting to see all of the breakthroughs and discoveries of AI slowly becoming greater over the years, from simply conceptualising a test aimed at differentiating humans and computers to managing to create a machine that can learn like a living creature. AI has come a long way since 1950.