Mira Murati, OpenAI's Chief Technology Officer, emphasized the crucial need for intensive research on the consequences of advancing artificial intelligence (AI) technology during an interview at The Atlantic Festival on Thursday, to prevent AI from becoming dangerous and excessively addictive.
Murati, a pivotal executive of the company responsible for the renowned ChatGPT AI tool, expressed concern over AI’s potential to become “even more addictive” as it evolves beyond the current systems available today.
Companies are implementing features with enhanced memory and greater personalization capacities, making results more user-relevant, Murati mentioned. ChatGPT has undergone significant enhancements since its initial release. A voice mode feature has been recently announced by the company, enabling users to converse with the chatbot more conveniently.
She stated that the enhancements in capability have a flip side; if developed incorrectly, AI can be highly addictive, causing dependency. She emphasized that researchers should study user interaction meticulously to understand intuitive engagement and to avoid the deployment of detrimental systems.
“Out of the box, it's a journey of discovery, learning, and exploration. Developing them incorrectly poses significant risks, potentially leading to more harm than benefit,” said Murati.
ChatGPT has gained immense popularity since its inception nearly a year ago and has been incorporated into products by Microsoft. With other tech giants such as Google, Amazon, and Meta launching large language models, a competitive AI race is underway, prompting legislators to assess technology-related risks urgently.
Lawmakers are debating risks related to the dissemination of misinformation by AI, especially its ramifications during election periods. Murati stressed that a “zero risk” scenario is unrealistic, but risks should be minimized to enhance the benefits of AI technology.
“The focus should be on balancing the value provided by the technology in real-world applications against the associated risks,” Murati added.
ChatGPT's release spotlighted immediate challenges such as its use by students to complete schoolwork, occasionally leading to cheating. Murati proposed that the technology necessitates innovations in teaching methods and learning approaches.
She also acknowledged the potential risk AI poses to employment and advocated for extensive work and conscientious efforts to mitigate these risks. “We must brace ourselves for this new lifestyle, likely resulting in a more profound impact on jobs compared to previous industrial revolutions,” she remarked.
This in-depth research and understanding are vital to avoiding the pitfalls of developing AI technology that could lead to addiction, misinformation, and unemployment. The balancing act between progress and responsibility is crucial in navigating the frontier of artificial intelligence technology.