The Future of AI: Understanding Language Through the Lens of Neuroscience
Do new AI tools like ChatGPT actually understand language the same way that humans do? This question is currently sparking intense debate among the inventors and researchers behind these large language models (LLMs). The implications of their answers could profoundly affect education and various aspects of society, particularly as we edge closer to achieving Artificial General Intelligence (AGI).
The Origins of ChatGPT: Bridging AI and Brain Science
A pioneering new book titled “ChatGPT and the Future of AI,” authored by Terrence Sejnowski, delves into the intersection of brain research and AI development. Sejnowski, a professor of biology at the University of California, San Diego, co-directs the Institute for Neural Computation and is affiliated with the Salk Institute for Biological Studies.
Get EdSurge journalism delivered free to your inbox. Sign up for our newsletters.
From Physics to Neuroscience: A Journey into the Brain
Sejnowski’s unique background as a physicist began with exploring the origins of black holes. However, he soon realized that it might take decades for new instruments to adequately measure the gravitational waves he was studying. This realization prompted him to shift his focus to neuroscience, wanting to “pop the hood” on the human brain and understand its complexities.
“It seemed to me that the brain was just as mysterious as the cosmos,” Sejnowski tells EdSurge. “And the advantage is you can do experiments in your own lab, and you don’t have to have a satellite.”
The Challenge of “Understanding” in AI
For decades, Sejnowski has concentrated on applying insights from brain science to constructing computer models, often collaborating with Nobel laureates John Hopfield and Geoffrey Hinton. The advancements in computing power and algorithms today are reshaping how neuroscience and AI inform each other, challenging conventional notions of thinking.
“What has really been revealed is that we don’t understand what ‘understanding’ is,” Sejnowski states. “We use the word, and we think we understand what it means, but we don’t know how the brain comprehends something. We can record from neurons, but that doesn’t really tell you how it functions and what’s really going on when you’re thinking.”
Potential for Educational Revolution
Sejnowski believes that new chatbots could revolutionize learning by acting as personal tutors for students. However, he cautions that current LLMs focus primarily on one aspect of how the human brain organizes information, neglecting many other crucial areas involved in survival and awareness.
According to Sejnowski, “There are a hundred brain parts that are left out that are important for survival, autonomy, and maintaining activity and awareness.” He suggests that simulating these other components may be essential for effective tutoring.
Societal Implications and Regulatory Needs
The researcher also raises concerns about potential unintended negative consequences of technologies like ChatGPT, reminiscent of how social media has led to misinformation and other societal challenges. He emphasizes that clear regulations will be necessary, but acknowledges that we may not completely understand how to regulate until these technologies are widely used and their impacts are observed.
A Shift in Human-Computer Interaction
Sejnowski predicts a future where traditional keyboard interactions become obsolete. Instead, most people will rely on voice commands to communicate with their devices. “You’ll be able to talk to your car and say, ‘How are you feeling today?’ [and it might respond], ‘Well, we’re running low on gas.’”
Deepening the Dialogue on AI and Human Learning
For more insights, listen to our conversation with Sejnowski on this week’s EdSurge Podcast, where he elaborates on his research aimed at more accurately simulating human cognitive functions. He also discusses his previous educational initiative, a widely popular free online course titled “Learning How to Learn,” which has attracted more than 4 million students over the past decade.
Conclusion
The ongoing discussion around AI tools and their understanding of language underscores the complexities shared between human cognition and machine learning. As researchers like Sejnowski navigate these frontiers, society must remain vigilant about the implications these technologies carry, particularly in the realms of education, information, and personal interactions.
Questions and Answers
- What does Terrence Sejnowski’s book focus on?
Sejnowski’s book explores the origins of ChatGPT and the intersection between neuroscience and AI, aiming to unravel the complexities of understanding in both fields. - Why did Sejnowski switch from physics to neuroscience?
He switched after realizing that developing new instruments to study gravitational waves would take decades. He wanted to explore and experiment with the human brain instead. - What potential does Sejnowski see for chatbots in education?
He envisions chatbots acting as personal tutors for students, revolutionizing learning by providing tailored support. - What concerns does Sejnowski raise about AI technologies?
He warns about possible unintended consequences such as misinformation and emphasizes the need for regulation, which will become clearer as these technologies impact society. - How does Sejnowski predict human-computer interaction will evolve?
He predicts that voice commands will replace keyboards, allowing for more natural dialogue between humans and devices.
