Been hearing the term ChatGPT a little often in your everyday conversations? Well, you’re not alone. This whirlwind of a model that crossed one million users in less than a week since it was officially made available to the public, has taken the world by storm for its herculean abilities. The Generative Pre-trained Transformer artificial intelligence developed by OpenAI, is the world’s fastest cutting-edge language model that can be used to perform an array of tasks, including language translation, text summarisation, and creative writing of several formats.
The advent of this wonder writer made jaws drop and has the digital world gasping at its efficiency levels. Many say that GPT has the potential to change the world in many ways- for both better and for worse. On one side, though it has helped people from all walks of life and prompted easier writing facilitation, in hindsight has also posed a threat by impacting jobs and transforming industries.
This is where we get familiar with a 22-year-old senior from Princeton University- Edward Tian. Over the last couple of years, he’s been researching how to detect text written by AI while working at Princeton's Natural Language Processing Lab.
Thus, when ChatGPT entered the chat, it was a no-brainer that it caused a stir, with students all over the world quickly cottoning onto the idea of getting AI to write their essays, dissertations and other scholarly assignments. Tian was quick to identify the AI as a mechanism that went beyond merely being an aid, to a threat to human skill. The question of Autobots and their malicious ability to kill jobs, affecting the education system holistically and denouncing human ability must have felt like an episode of the dystopian sci-fi series Black Mirror coming to life for Tian.
This concern of his turned to action rather quickly, which enabled him to create GPTZero, the application that has the ability to identify prose that has been written by a machine. Tian’s inspiration for creating his brainchild was to not defeat the power of AI but remove the deceit that came from it. "Humans deserve to know when something is written by a human or written by a machine," he says.
There is no stopping AI, that’s for sure. But it does present some interesting quandaries. If children don’t have to develop critical thinking skills - and leave it to AI - will we become a dumber species? A matter more worrisome- will we become so reliant on AI, that we begin to lose sight of what our species is truly capable of?
GPTZero is looking for patterns a human likely wouldn’t notice, or cannot make sense of. But an AI model trained to provide responses with high perplexity and burstiness can most likely evade, if not completely destroy detection.
Films like The Matrix and Terminator were considered science fiction. But could we be moving one step closer to a world where those ideas become a reality? Let us all sit back & collectively observe how interesting it is going to see this arms race play out!