For years, artificial intelligence (AI) research has chased the dream of Artificial General Intelligence (AGI) – a machine that can think and learn like a human. But what exactly makes us intelligent? In the quest for AGI, researcher Joscha Bach proposes a novel framework: The Big Nine.
Beyond Language Processing: Unveiling the “Big Nine”
Large Language Models (LLMs) like GPT-3 have captured headlines with their ability to generate human-quality text. However, Bach argues that LLMs represent a narrow slice of a much larger pie. True intelligence, he suggests, encompasses a broader spectrum of capabilities. These nine fundamental abilities, the “Big Nine,” form the building blocks for comprehensive intelligence:
- Embodiment: Having a physical body allows us to interact with the world, not just process information about it.
- Perception: Sight, touch, hearing, and other senses provide the raw data for understanding our surroundings.
- Social Learning: We learn a tremendous amount from observing and interacting with others.
- Motor Skills: The ability to move and manipulate objects is crucial for navigating the world.
- Motivation and Drive: The internal forces that push us to act and achieve goals.
- Common Sense Reasoning: Understanding the everyday cause-and-effect relationships that govern the world.
- Planning and Problem-Solving: The ability to think ahead, strategize, and overcome challenges.
- Creativity and Imagination: The capacity to generate new ideas and concepts.
- Metacognition: The ability to think about our own thinking – to be aware of our own strengths and weaknesses.
The LLM: A Piece of the Puzzle, Not the Answer
While LLMs excel at language processing, they fall short in most of the other “Big Nine” categories. They lack a physical body, have limited perception, and struggle with tasks that require common sense reasoning or motor skills. This suggests that AGI may not emerge from a single, all-encompassing invention, but rather from a collaborative ecosystem of AI systems, each with its own strengths aligned with different aspects of the “Big Nine.”
The Road Ahead: Diverse AI
By focusing on developing AI systems that excel in different areas of the “Big Nine,” we can pave the way for a more robust and well-rounded form of intelligence. Imagine an AI that can not only generate text but also navigate the physical world, learn from its interactions, and solve problems creatively. This diverse “orchestra” of AI has the potential to unlock advancements far beyond the capabilities of any single system.
The “Big Nine” framework offers a compelling alternative to the singular focus on LLMs. It reminds us that true intelligence is a multifaceted phenomenon, and the key to AGI may lie in fostering a rich tapestry of AI capabilities working together.