Can AI Become Self-Aware?

Technology becoming self-aware has been a popular plotline in sci-fi books and movies, and now that AI technology has become more commonplace, that has led to the question, "Can AI become self-aware, or is that just something made up to sell media?" That question isn't easy to answer because it depends on the definition of self-awareness, a topic philosophers have debated for centuries.

In the past, the Turing Test was considered to be the best test for determining self-awareness in machines. The Turing Test was designed in 1950 by Alan Turing, and it's based on the idea that if a computer is truly capable of intelligent thought, its responses to questions should be indistinguishable from those of a human when examined by a human. Current AI language models are able to pass the Turing Test with ease, which has led some to believe that self-awareness could be on the horizon, if it hasn't already been achieved. However, experts are now debating whether the Turing Test can still be considered a reliable test of self-awareness, given that Turing himself could not have foreseen what today's technology would be capable of in the 1950s.

Not everyone considers the Turing Test to be the ultimate measure of self-awareness in machines, especially since there's no universally accepted definition of self-awareness. If self-awareness is the ability to experience feelings, then no current AI is self-aware. Today's AI models were programmed to perform specific tasks and solve defined problems. The conversational capabilities of chatbots are not due to their ability to truly understand the conversation and form opinions. Rather, they're trained on human communication and understand the patterns in how humans speak and write well enough to be able to respond the way a human would. This is why tools like ChatGPT can easily pass the Turing Test: They're capable of mimicking human speech because they were trained on human-written text on the Internet.

Whether AI could become self-aware in the future depends on how technology evolves and on our definition of self-awareness. It's hard to predict what technology will look like 70 years from now, just as it was hard to predict today's technology in the 1950s. It's possible that AI could become self-aware in the future, but as there's no objective standard to evaluate whether something is sentient, it may be difficult to determine when AI becomes self-aware or if it ever does. But today's experts suggest that while it may seem like AI has human-like qualities in the way it responds, the odds of it achieving self-awareness remain slim.