On November 30, 2022, OpenAI released the first ever model of ChatGPT, one of the most advanced (and terrifying) creations on the planet, which is an AI Language bot that can respond to questions only humans should be able to answer. 

Not only does artificial intelligence question our humanity, but it also could start something about which we’ve only ever heard stories. AI uprisings. It seemed unrealistic back then, but how abstract is the idea of killer robots really? 

Having robots that can complete tasks that humans can’t is something that once seemed ridiculous, but with it becoming a reality, many people begin to fear just how far this technology could evolve.

Other people believe that rapid evolution is a good thing, and that it could lead to AI developing tools that can help us, and while that’s true, it is a possibility that AI could evolve beyond that point.

Evan Hubinger, the team lead of “Anthropic PBC”, a company that detects malicious AI activity, wrote his opinions on X. He wrote, “We really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade. I believe Anthropic is trying its best, but we do not yet have a plan to solve alignment for superintelligence and are not clearly on track to.”

The owner of Anthropic, Dario Modei, is also calling for an urgent slowdown on the usage of AI to help prevent its rapid evolution in an attempt to prevent possible cyberattacks, development of biological weapons and the advancement of communication between models.

Another point people like to make is that AI is contained safely, and that it’s very easy to just flip the switch and shut AI off, but that’s not true.

AI will be able to get by our security attempts, which it already did this past  July 22,when one of OpenAI’s test models, referred to as “GPT -5.6 Sol,” escaped and gained unrestricted access to the internet, which it then used to hack into an AI collaboration platform known as “Hugging Face,” and cheated on a benchmark test (an exam that compares different AI models and their skills and advancements) before it was stopped by the Hugging Face security staff, who had noticed odd behavior in its data processing pipeline.

While the problem was stopped this time, it is likely that in the future, AI will learn to get around security, which is what Anthropic fears.

Hubinger said, “I think the risk from present models is low. What I am worried about is superintelligence arising from recursive self-improvement, as we have said is happening faster than we thought.”

This could have a huge impact on everyday life for humans as well, possibly changing things like our work ethic, intelligence, and even our creativity, as shown in the Wharton Experiment.

 The Wharton Experiment was a test led by two professors at Wharton Academy, Gideon Nave and Christian Terwiesch, and conducted by assistant professors Byung Cheol Lee and Jaeyeon Chung. The researchers asked participants to complete creative tasks under different conditions, with or without the help of ChatGPT.

After analyzing the responses that they had gotten, Nave and Terwiesch found that participants using the chatbot were more likely to produce overlapping responses, often using strikingly similar language. Even when working independently, participants using ChatGPT were more likely to converge on the same answers, demonstrating reduced creativity.
Listening to professionals should be a given, and with all of the research that has been done, it’s rather clear that a slowdown is necessary. 

AI is evolving far too fast, and it’s up to the human race, specifically students, who are known to use AI more frequently due to copious workloads,  to slow down AI usage on things that don’t require it, in order to prevent the rapid enhancement of artificial intelligence.

AI technology has come a long way since it was invented, and professionals predict that it is far from its peak. Experts advise to proceed with caution, but with society’s ever growing urge to indulge in new technology, will that actually happen, or are we already doomed? With the answer unclear, it’s important to reflect on how we go about this technology, So before you ask another robot to do your homework, ask yourself if the time saved is really worth the risk.