Estimated reading time: 6 minutes

SMT Perspectives and Prospects: Artificial Intelligence, Part 5: Brain, Mind, Intelligence

Filmmaker James Cameron, who directed “The Terminator,” “Avatar,” “Titanic,” and other award-winning movies, equates generative artificial intelligence (AI) with human dreams. Can we explain dreams’ origins, content formation, and links to real events, emotions, and memories? Are dreams the result of the brain interpreting neural signals during sleep?

Although it is generally understood that dreams are the confluence of neurological, psychological, and physical processes during sleep, can we answer the above questions through experiences or science?

Experts widely agree that early childhood is the most sensitive period for forming strong neural connections and establishing foundational neural architecture, making it the most important period for brain development. Early childhood lays the foundation for future learning, behavior, and emotional well-being. The brain has all the power in connections, wiring, storage, memory, and processing to function as a human being. The human brain also requires a lot of fuel; it reportedly reaches its peak with approximately 100 billion nerve cells or neurons, and it takes 20W to power a brain1.

If transmitting and receiving electrochemical signals via neurons are essentially the thoughts, emotions, actions, and automatic functions of the human body, then the neurons’ knowledge controls how to use the combined power of the conscious and unconscious mind to think in a healthier, more flexible, resilient, and goal-supporting way. Is thinking linked to electrical signals zooming inside our heads, forming a complex code carried by our neurons? Can we say the brain is the hardware, the mind is the software, and the operating system gathers, stores, and manages information by using our brain's massive processing resources and capacities as the basis of human intelligence?

What is artificial intelligence, and what is human intelligence?

Artificial Intelligence vs. Human Intelligence

Currently, there is no single established test that can authoritatively measure artificial or human intelligence. Humans generate data, acquire information, and translate it into knowledge. Cumulative knowledge builds intelligence. As such, it is plausible to define human intelligence as the capacity to acquire knowledge and the ability to apply it to achieve desired outcomes.

Computer science defines artificial intelligence as any device that perceives its environment and takes actions that maximize its chance of successfully achieving its goals. It also includes a system's ability to interpret correctly external data, learn from such data, and use what it has learned to achieve specific goals and tasks through flexible adaptation.

At its broadest level, intelligence is the ability to achieve a range of goals in different and unpredictable environments. Higher intelligent systems can fulfill a wider range of goals in a wider range of predictable and unpredictable environments.

A marvel of the human brain is that it can deal with the unexpected. Today, we need to advance the understanding of the brain to advance AI. Meanwhile, AI is fostering brain research.

Brain Research
AI enables new kinds of research2. The capacity of modern computer systems to process more data than in the past opens immense possibilities. Mathematics makes many kinds of AI possible, such as cluster models. One example of this methodology is to amass a pool of data that groups people into different clusters and uses artificial neuron networks to interpret the electrical signals of hundreds of neurons in the brain. The research’s main finding was that the actual substance of thought and the patterns that constitute the mind we use to read is dynamic electrical activity in our brains, rather than something physically anchored to neurons. This is an important finding. It perhaps points to the immense complexity, nuance, and intricacy of brain dynamics—our mind, thoughts, and the mechanism of thinking and reasoning.

While neural networks in the brain are vastly more complicated, the result of this simulation is a model system that is both close enough to its biological equivalent and simple enough to offer hints about how the brain works.

AI in Brain Research
Brain study and neuroscience continue to advance. Two notable clinical research trials are being conducted at Neuralink and AI at Meta labs. Reportedly, a second human (as of this writing) has received a Neuralink brain implant, which could lead to a potential milestone in developing brain-computer interface technology. The implant device is one-fifth the thickness of a human hair and is designed to sit on top of the brain and detect neuron spikes by detecting signals from individual neurons inside the brain—a potential advance that could decode higher-quality brain signals. A brain-computer interface, such as a brain implant, allows humans to have direct control of a computer or external device solely using human thoughts. It is a set of tiny electrodes (e.g., platinum) embedded in a thin film that conforms to the surface of the brain. Each electrode listens to the electrical activity underneath the brain and takes an electrical video in real-time of the thoughts taking place on the brain’s surface to record, amplify, digitize, and then transmit them using AI to compute the vast number of signals in real-time to aid studies.

Separately, the Meta AI lab studies how to read mind and brain activity using a self-supervised learning model that can extract meaning from giant pools of data without human instruction. The goal is to create a “speech decoder” that can directly transform our brain activity (our thoughts) into words. 

Present and Future
How neurons in our brains communicate and explore the nature of cognition is still an enigma and human intellect is still intriguing. Take the stock market as an example. One can use a computerized analysis of market data to detect hidden patterns and write AI algorithms to pick stocks3. However, at present, no AI model can consistently and reliably predict the stock market. This, in part, is a result of the stock market data being “noisier” than language and other data, making it harder to explain or predict how the market moves4.

In his book The Transcendent Brain: Spirituality in the Age of Science, author Alan Lightman writes, “Some human experiences are simply not reducible to zeros and ones.” This reflects the current and future challenges of AI to reach the capacity and capability of the human brain, dubbed Artificial General Intelligence (AGI).

Is the concept of and the technology behind the large language models (LLMs) that mimic the way humans think, act, read, write, and reason a promising path to achieve AGI and beyond? 

Can we turn the human brain into a machine or, when can we turn the human brain into a machine?

This is an opportune moment to quote Albert Einstein, who noted, “Computers are incredibly fast, accurate, and stupid. Humans are incredibly slow, inaccurate, and brilliant. Together, they are powerful beyond imagination.” He was correct then, and he’s still correct today.

References

1.  “The blood-brain barrier: an engineering perspective,” by A.D. Wong, Frontiers of Neuroengineering 6, e7, 2013.
2.  “A new era in cognitive neuroscience: the tidal wave of artificial intelligence (AI),” by Z. Chen and A. Yadollahpour, BMC Neurosci 25, 23 (2024).
3. The Man Who Solved the Market: How Jim Simons Launched the Quant Revolution, by Gregory Zucherman, Portfolio, 2023.
4.  “AI Can Write a Song, but It Can’t Beat the Market,” Wall Street Journal, April 12, 2023.

Appearances
Dr. Jennie Hwang will present two webinar courses for IPC: “Reliability of Electronics—Solder Joint Voids—All You Should Know” May 13 and 15; and “Artificial Intelligence—A Primer and Essentials,” June 17 and 19.

This column originally appeared in the April 2025 issue of SMT007 Magazine.