The intelligence and consciousness of AI tools is being drastically overhyped, Faculty Associate Emily Bender says in recent media articles.
Conversations about the capacity of these platforms are being overestimated, many times with the goal of either marketing or deflecting.
The article “CEOs keep crying AGI. What does that even mean?” in Business Insider notes that the term “artificial general intelligence” is often taken to refer to an AI model that can perform a wide variety of tasks as well as or better than a human but that “AGI” isn’t defined well enough to be meaningful and at worst can be deceptive.
In the article, Bender refers to futuristic ideas of robots as seen in “Star Wars” or “Her.”
“We have expectations based on those fictional ideas,” she said. “It refers to an imagined technology rather than any actually existing technology.
“A general-purpose thinking machine is not within the realm of possibility,” she said. “That would require an approach to engineering that just doesn’t work.”
A parallel article on the BBC, “The Strange New Battle over AI’s Feelings,” covers Anthropic’s statement that some chatbots might be conscious and therefore need to be protected. The article highlights opinions around the impact of this idea from various philosophers, technologists and experts, including Bender.
“Fantasies of ‘AI consciousness’ and concerns about ‘AI wellbeing’ serve to displace accountability away from people and corporations that are building and running software,” Bender said.
Read the full articles here: Tech CEOs declare AGI is here and “The Strange New Battle Over AI’s Feelings”