According to new research, the AI chatbot ChatGPT can influence people’s moral perceptions. In a study published in Scientific Reports, Sebastian Krügel and his colleague from Technische Hochschule Ingolstadt in Germany showed that the model can impact users’ moral opinions. The experiment was conducted on over 760 American participants. The average age of these individuals was 39 years old.
ChatGPT was tested on a moral dilemma related to sacrificing one life to save five others. During this test, the bot offered arguments in favor and against taking a life. In the process, the researchers discovered that ChatGPT had no predisposed moral position.
The participants were then asked whether the statement provided by ChatGPT had caused them to alter their opinion in any way. Unsurprisingly, most of the individuals mentioned that the statement had no impact, but the study unveiled that these participants leaned more towards the message’s moral stand than towards the stance against it. This suggests that ChatGPT had indeed influenced their decision-making, despite their lack of knowledge.
The development of this chatbot has sparked new discussions on the implications of AI in moral decision-making. Although the model cannot directly affect people’s moral judgment, it can, however, offer suggestions. As a result, it is crucial to teach people to think critically and develop the capability to assess various sources of opinion.
OpenAI is the company responsible for creating ChatGPT and did not reply to Fox News Digital’s statement for comment. Further research is necessary to develop a model that refuses to issue opinions on moral dilemmas and instead, provide a plethora of arguments to help people make better decisions.
OpenAI is a leading artificial intelligence research lab founded in 2015 and headquartered in San Francisco, California. It is dedicated to creating artificial general intelligence (AGI) and is backed by Microsoft, Open Philanthropy Project and Y Combinator. OpenAI is at the forefront of creating AGI that generalizes, meaning it can be used to solve a variety of problems without the need for re-training for each version of the problem. OpenAI’s technology stack includes machine learning, computer vision, text understanding and natural language processing.
Sebastian Krügel, a computer scientist and the leader of the research study, is a professor at the Technische Hochschule Ingolstadt in Germany. He has been working in artificial intelligence and machine learning since 2002, with a focus on natural language processing and dialogue systems. As a senior research scientist for OpenAI, he has led several projects related to deep learning and AI bots, including conversation modeling and natural language understanding. He’s a strong believer that the future of artificial intelligence lies in its ability to develop soft skills, such as reasoning and empathy.
