AI Speaks Well, But Does It Truly Understand?
Rodney Brooks, a robotics expert, warns that despite their eloquent communication, linguistic AIs are fundamentally detached from reality, raising concerns about the potential for deception. How do we avoid falling into this trap?
AI: Masters of Misdirection
One cannot lie without believing they know the truth; spewing nonsense requires no such conviction, philosopher Harry Frankfurt notes in his 2005 book, “On Bullshit.” In this vein, AI expert Rodney Brooks believes that today’s large language models (LLMs) are essentially bluffing.
AI: Grand Promises
As a former director of MIT’s Computer Science and Artificial Intelligence Lab and founder of iRobot, Brooks has witnessed countless technological breakthroughs and inflated promises. In a detailed conversation with Newsweek, two main ideas stood out: we should not be fooled by LLMs’ fluent use of language into thinking they possess magical abilities, and changes will come slower than anticipated.
At the Heart of AI: Brooks’ Warning
“LLMs do not know what is true,” states Brooks. “They only know which words to cluster, having read everything ever written or digitized, thus they are skilled in linguistic tricks. They remain charlatans until we can ground them in reality, in truth.” However, Brooks is not an AI skeptic; he is a pioneer who combines deep technical knowledge with decades of practical experience. His perspective provides a vital framework to understand both the real promise and ongoing limitations of AI.
Brooks advocates for a future not of a singular omniscient AI system but of specialized AI tools combined with human oversight. This hybrid approach is already proving successful across various fields.
AI: A Tool to Enhance Human Capabilities
Brooks champions successful AI systems that are not only technically capable but also compatible with humans. Instead of pursuing full automation that removes humans from the equation, he supports augmentation that enhances human abilities while maintaining human agency.
The rate at which new technologies are adopted is often overestimated. Infrastructure changes require massive investment and coordination. Brooks remarks, “Everything takes a long time to adopt, even software.” Ultimately, AI systems don’t exactly lie, but they aren’t grounded in true understanding either. They require human supervision to be truly useful, just as our own quick, intuitive responses often need to be checked by slower, more deliberate thought.
The key is to maintain perspective, focusing on real value rather than hype.