Can Computers Grasp Common Sense? Unraveling the Challenge
- Technology Explained
- August 25, 2023
- No Comment
- 21
Introduction: Can Computers Grasp Common Sense
In the ever-evolving landscape of artificial intelligence, one question continues to intrigue both tech enthusiasts and skeptics alike: Can computers truly learn common sense? Delving into this query, we embark on a captivating journey that explores the intricate relationship between AI and the quintessential trait of common sense.
Oluwademilade Afolabi, a notable figure in the realm of AI, brings forth a wealth of knowledge and experience to unravel this phenomenon. With a firm grasp of computer science and a passion for demystifying complex concepts, Afolabi lends a human touch to the intersection of technology and intuition.
Can Machines Attain Common Sense? Unraveling the Conundrum
Common sense, the seemingly innate ability that guides humans through the nuances of everyday life, has long eluded the grasp of artificial intelligence. While machines have made remarkable strides in various domains, the realm of common sense presents a unique challenge. This article, in collaboration with Afolabi’s expertise, delves into the intricacies of this challenge, shedding light on the elusive nature of common sense and the innovative approaches being employed to bridge the gap.
What Exactly is Common Sense, and How Do Humans Develop It?
At its core, common sense is the repository of practical knowledge that enables individuals to navigate the complexities of the world. It’s the unspoken rules, intuitive judgments, and cultural norms that shape our decisions and actions. Humans acquire common sense through a blend of personal experiences, societal influences, and observations.
From infancy, we begin absorbing cause-and-effect relationships, building a foundation for practical understanding. As we grow, our common sense expands through trial and error, interactions with peers and family, and exposure to various environments. But can computers follow the same path to common sense?

The Hurdles Computers Face in Grasping Common Sense
While computers excel in tasks driven by data and patterns, they falter when it comes to understanding the unspoken and context-dependent aspects of common sense. Unlike humans who learn from real-world experiences, computers lack the ability to gather nuanced insights from life. They solely rely on explicit instructions provided by humans, leading to limitations in adapting to new contexts or situations.

Moreover, common sense is not confined to predefined rules; it thrives on intuitive understanding, adaptation, and unwritten norms. This aspect proves to be a stumbling block for machines that operate within the boundaries of programmed logic. The challenge lies in translating the depth of human intuition into algorithms.
Innovative Avenues to Imbue AI with Common Sense
Despite the obstacles, the quest to instill common sense in AI persists. Afolabi’s insights shed light on innovative strategies researchers are exploring. One such approach involves building extensive knowledge bases by hand, resembling the ambitious Cyc project. This initiative compiles a wealth of facts and rules, enabling AI to reason and infer like humans.
Another avenue involves crowdsourcing common sense through platforms like ConceptNet. By tapping into the collective intelligence of diverse contributors, AI systems gain access to a plethora of real-world knowledge snippets. This approach reflects the collaborative nature of common sense development in humans.
Afolabi also uncovers the world of simulated environments, where AI agents engage in practical interactions with virtual worlds. Through experience, these agents grasp the principles of physics, object manipulation, and cause-and-effect relationships. This hands-on approach aligns with how humans learn from interaction and observation.

The Future Landscape: Progress and Beyond
As AI technology advances, strides are being made in imbuing machines with an understanding of common sense. Language models like GPT-3.5 exhibit a semblance of common sense by mimicking patterns learned from vast data. However, a consensus on whether this constitutes true common sense remains a topic of debate.
The road ahead holds promise, with AI edging closer to acquiring a form of common sense. But as Afolabi emphasizes, human common sense remains essential in harnessing these technologies responsibly. The intersection of artificial intelligence and common sense is a complex terrain that demands ongoing exploration and ethical consideration.
In conclusion, the pursuit of common sense AI is an intricate journey that intertwines technological innovation with the complexities of human cognition. Oluwademilade Afolabi’s expertise adds depth and clarity to this exploration, reminding us that while machines may learn, the essence of common sense is a uniquely human trait. As we delve into the realms of AI and intuition, let us navigate with a balance of curiosity, caution, and, of course, common sense.
Table: Understanding the Common Sense Conundrum
| Heading | Content |
|---|---|
| 1. Decoding Common Sense: What Is It and How Do Humans Acquire It? | Defining common sense and its role in decision-making. |
| 2. The Common Sense Enigma for Computers | Exploring why computers struggle with common sense. |
| 3. The Context Conundrum: Why Context Matters in Common Sense | Discussing the significance of context in common sense. |
| 4. Unspoken Rules and Implicit Norms: A Barrier for AI | Analyzing how unspoken rules pose challenges for AI. |
| 5. Pioneering Paths: How Computers Can Acquire Common Sense | Examining innovative approaches to teach AI common sense. |
| 6. Building Knowledge Bases: The Cyc Project and Beyond | Delving into manual construction of knowledge bases. |
| 7. Crowdsourcing Wisdom: ConceptNet’s Collective Common Sense | Exploring crowdsourced knowledge for common sense AI. |
| 8. Learning Through Experience: Simulated Worlds and AI | Investigating AI learning in simulated environments. |
| 9. The Power of Language Models: GPT-3.5 and Beyond | Discussing data-driven techniques in common sense AI. |
| 10. The Test of Common Sense: Evaluating AI’s Intuition | Detailing methods to assess AI systems for common sense. |
| 11. Striving for Common Sense AI: Progress and Prospects | Highlighting advancements and remaining challenges. |
| 12. The Future of Common Sense AI: Promise and Uncertainty | Speculating on AI’s evolution in understanding common sense. |
Comparative Table: Approaches to AI’s Common Sense Learning
| Approach | Description | Example |
|---|---|---|
| Knowledge Bases | Manual compilation of facts and rules to enable reasoning. | The Cyc project encodes vast logical axioms, allowing AI to make nuanced deductions, like understanding that a tomato is a fruit but doesn’t belong in a fruit salad. |
| Crowdsourcing | Gathering common sense snippets from diverse contributors. | ConceptNet collects insights from contributors worldwide, resulting in an expansive repository of practical knowledge. |
| Simulated Worlds | Creating virtual environments for AI to learn through interactions. | AI agents in AI2 THOR environment experiment with everyday objects to gain intuitive understanding of concepts, like the impact of gravity on objects. |
| Language Models | Leveraging vast data to learn patterns, resembling common sense. | GPT-3.5 emulates human-like text generation by analyzing internet data, exhibiting shades of common sense, though debate persists on whether this constitutes true understanding. |