ChatGPT and Deepfakes: The Rise of AI

By IAN DE MELLO
Staff Writer

If you’ve been keeping up with mainstream news recently, you might have been seeing a couple of particular phrases or words pop up in the headlines. “ChatGPT” and “deepfakes” are some of the most common names thrown around in the tech industry. All of this seems quite sudden and there may be an air of confusion around these terms. It is best to explain what these mean in order to better our understanding of the situation and what truly is at risk here.

ChatGPT is a chatbot that was launched by OpenAI in November 2022 and quickly took the internet by storm. Its features include the ability to write computer programs, compose music, write fairy tales, and essays, and can even answer some test questions at a higher-than-average skill level. Those last two attributes have been quite concerning to some schools, which is why controversy around ChatGPT has been growing. These fears are not unfounded though. ChatGPT, for example, has already been listed as a co-author for several papers. More surprisingly, ChatGPT has passed graduate-level exams at the University of Minnesota at a C+ level and at the Wharton School of the University of Pennsylvania at a B level. Quite an impressive feat for something AI-generated, but there are severe limitations. The program may give plausible-sounding answers but in actuality are ridiculously wrong. ChatGPT even has its own writing style, which gives professors and teachers the edge in finding out if the content is AI-generated. However, the program is improving day by day, being trained on user-given data, meaning that it may soon be impossible to distinguish AI and human work.

The user interface of ChatGPT

Another equally scary possibility is the rise of deepfakes. The term deepfake is used to describe fabricated media in which a person’s voice and face are replaced with someone else’s. This includes images and videos alike. Deepfakes initially began being developed by researchers in the early 1990s, but the rise of the internet has allowed amateurs to throw their hats into the ring. The technology has improved rapidly over the past couple of years with deepfakes moving from just faces, to mimicking an entire body with movement. While it is still in its infancy, realistic deepfakes are becoming increasingly common and it’s more likely than not that you’ve already seen one. Whether it was just a demonstration, a meme, or an application that utilizes it. This technology has mostly been used lightheartedly, but its sinister side is slowly creeping out. Reports of fraud, blackmail, sabotage, and hoaxes have all occurred with deepfakes. One of the more recent events involving deepfake technology came from the Russo-Ukrainian War, in which a deepfaked video shows Ukrainian President Volodomyr Zelenskyy telling his soldiers to lay down their arms and surrender to the Russians. Although this video had no effect on the war, as the shoddy production seriously hurt the possibility of anyone taking it seriously, it is just a mere precursor of what is to come.

Deepfake Example

AI is here to stay and will continue to improve day by day. Due to the majority of AI being improved by user-driven data, we are all in some way contributing to it, whether we like it or not. While there is no way we can stop the development of AI, we must do our best in minimizing the harmful effects that it could bring. The majority of change and regulation will come from legislation, which hopefully comes sooner rather than later. It is not all doom and gloom as AI has the potential to revolutionize certain aspects of our lives. The truth is technology can be a blessing or a curse, it just depends on how we handle it.