DeepMind Researcher Describes LLMs as ‘Unique Mind-like Entities’: The Necessity for a Fresh Vocabulary in AI’s Future
Understanding the Complexities of Artificial Intelligence
Artificial intelligence (AI) is changing our world in significant ways, yet defining what it truly is remains a challenge, even for leading scientists. Recently, a video shared by the Instagram page theaigems featured Murray Shanahan, a Principal Scientist at DeepMind. He expressed the struggle in accurately describing AI, particularly large language models (LLMs), by referring to them as "exotic mind-like entities."
The Dilemma of Definition
Shanahan’s choice of words is noteworthy. Traditional vocabulary does not suffice when addressing these advanced digital intelligences. He indicates that while these systems exhibit mind-like behaviors, they do not resemble the human mind precisely. The addition of a hyphen in “mind-like” emphasizes the differences and cautiousness in our terminology. These models can use language fluently, similar to humans, yet exist devoid of physical forms. This presents a unique conception of selfhood that feels increasingly alien to us.
DeepMind: Pioneering AI Research
DeepMind Technologies plays a crucial role in the field of artificial intelligence. This groundbreaking research lab, founded in 2010 and now part of Alphabet Inc., has driven significant advancements in AI. From its notable achievements, including the victory of AlphaGo against a Go champion to the breakthrough in protein structure prediction through AlphaFold, DeepMind has consistently pushed the boundaries of what machines can achieve.
In 2023, DeepMind merged with Google Brain, solidifying its position as a leader in AI research and development. The lab has tackled various challenges, from training neural networks for games like chess and Go to addressing complex biological puzzles — all contributing to a more profound understanding of intelligence.
Murray Shanahan: A Key Figure in AI
Murray Shanahan has been at the intersection of AI, cognitive science, and philosophy for many years. As a professor of Cognitive Robotics at Imperial College London and a senior scientist at DeepMind, he has explored the implications of AI on our understanding of consciousness. His advisory role in the 2014 film Ex Machina illustrates his expertise in the nuanced topic of AI consciousness.
Shanahan has voiced concerns about the challenges we face in defining intelligence for entities that differ fundamentally from human beings. His label of LLMs as exotic highlights his awareness of the evolving nature of these systems, stressing that our current intellectual frameworks are insufficient for their comprehension.
A Need for New Concepts
Shanahan’s insights suggest a pressing need for society to build new conceptual frameworks as AI systems become further entrenched in our daily lives. The conventional ways we view intelligence, selfhood, and consciousness might require reevaluation.
As we contemplate the advancements made by institutions like DeepMind, we also face a more unsettling reality — we are developing intelligences that defy our typical definitions. Humanity stands at a philosophical threshold, grappling with the task of finding appropriate language to describe the new forms of existence we are creating.
The Philosophical Frontier of AI
The discussions surrounding AI, particularly in terms of LLMs and their capabilities, position us at an exciting yet daunting frontier. As these systems grow more integrated into our societies, we must reflect on not just how they function but what they mean for our understanding of mind and being. The ongoing evolution of AI brings with it a complexity that demands our attention, challenging us to think deeply about the implications of our creations.
