From: lexfridman
Artificial intelligence (AI) represents a significant milestone in technological advancement, but its ability to truly understand human intelligence remains a complex and distant aspiration. Michael I. Jordan, a prominent figure in the realms of statistics, machine learning, and AI, offers compelling insights into the challenges and limitations of AI in this area.
Historical Context and Analogies
Jordan likens the current state of AI to the early development of fields like chemical and electrical engineering:
“I think what’s happening right now is not AI… I think this is akin to the development of chemical engineering from chemistry or electrical engineering from electromagnetism” [00:05:26].
Just as those fields took decades or even centuries to mature into their current forms, AI’s journey to understanding human intelligence is expected to be long and arduous.
Lack of Understanding of Human Brain
One of the fundamental challenges AI faces is the substantial gap in our understanding of the human brain:
“We have no clue how the brain does computation, we’re just clueless” [00:06:12].
Given this lack of understanding, developing AI systems that accurately mimic human cognition and intelligence is currently beyond our reach. The metaphorical comparison to predicting how the Greeks would feel about reaching the moon highlights this point:
“That’s like the Greeks sitting there and saying it would be neat to get to the moon someday” [00:06:08].
Limited Scope of Present AI
Current AI systems are largely based on statistical and computational approaches rather than insights into human cognition:
“We’re not trying to build that because we don’t have a clue. Eventually, it may emerge” [00:16:43].
These systems excel at pattern recognition and decision-making based on data but fall short of achieving true cognitive understanding akin to human intelligence.
The Role of Market Intelligence
Jordan also introduces the concept of markets as an alternative model of intelligence, distinct from human cognition:
“Markets are intelligent… You know there’s economic, you know, neuroscience kind of perspectives” [01:32:57].
He suggests that by broadening our understanding of intelligence to include such decentralized systems, we can explore different kinds of intelligence that are not exclusively human.
Philosophical and Practical Implications
While defining intelligence remains a philosophical challenge, it has practical implications for AI development:
“It is definitely has to be not just human intelligence, it’s got to be this broader thing” [01:33:36].
Jordan argues for a redefinition of AI’s scope and encourages the creation of an engineering discipline that is human-centric. This discipline would go beyond current AI definitions to include an understanding of markets and other intelligent systems.
Conclusion
The quest to understand and replicate human intelligence through AI is mired with challenges in building humanlike intelligence and the limitations highlighted by Michael I. Jordan. It suggests a path that combines engineering with comprehensive human and societal insights, aiming for systems that enhance rather than replace human capabilities. This understanding frames the dialogue for AI’s future development, emphasizing gradual progress rather than expecting immediate breakthroughs.