Language models like ChatGPT are designed to simulate human-like conversations, but the question is: are they truly intelligent?
In an interview with DesignRush, co-founder and CTO of Lemay.ai, Daniel Shapiro claims that the human-like behavior of these AI tools is more of an illusion than true intelligence. He argues that AI researchers are focused on concealing the shortcomings of AI rather than developing consciousness.
Shapiro insists he is “very unconvinced” that conscious AI will be possible in our lifetime and prefers to focus on “the real work” that includes AI regulations and ethics. The AI expert recommends publicly funded research to prioritize these efforts.
“Never say never. We do need to continue to support universities and their scientists to investigate the issues, but [the] private industry should not bother working on this,” he added.
ChatGPT itself admits that while current AIs “can perform complex tasks and make decisions based on data and rules”, they lack “the subjective experience and self-awareness that are typically associated with consciousness.”
According to Shapiro, repetitive tasks such as data entry, assembly line work and customer service roles are among those that could be automated, leading to fewer jobs in those industries.
It also said that “some experts believe that it could be many decades or even centuries before truly conscious AI is possible, while others are more optimistic and believe that it could happen within our lifetime.”
Regardless of when true conscious AI will actually be possible, or of how real is the actual behavior of current AIs, it’s likely that we’ll see major breakthroughs in this field in the near future and certainly in our lifetime.
For more on artificial intelligence and its potential impact on several industries, and how it may impact workforces, you can check Daniel Shapiro’s full interview right here.