TITLE: Artificial Intelligence vs. Human Intelligence AUTHOR: Eddie S. Jackson, MrNetTek DATE: June 1, 2024 at 7:31 AM EST RESEARCH: Google search, current news, books, Copilot. EDITING: Grammarly
Artificial intelligence has captured my imagination for many decades, as I’m sure it has yours. A quick view of my website, and you’ll see I’ve been buried in technology for a very long time, with AI always in the mix. From sci-fi books to big-budget movies, the idea of machines that can think, learn, and reason like humans has fueled entertainment, imagination, and of course fear (the terminators are coming for us., we’re all living in the Matrix). But, what exactly is AI, and how does it differ from human intelligence? In this article, I’ll provide a brief glimpse into AI, its types, and capabilities, and discuss why AI remains at the bottom of the intelligence totem pole in comparison to human intelligence (for now).
Types of Artificial Intelligence
Artificial Narrow Intelligence (ANI)
Artificial Narrow Intelligence, a form of weak AI, is designed for specific tasks and excels in those tasks, often outperforming human-level intelligence. What? Did I say outperforming humans? Yes. ANI is very good with a concentrated skill set if the right conditions have been met. Conversely, humans have a broad-spectrum intelligence. A contrasting example would be the equivalent of comparing a single reference book (expert level and amazing for the content), to an entire library. Humans are the library of intellectual ability, and the ANI is the reference book. Why is that distinction important? Because intelligence rarely operates in a vacuum. Our decisions change based on context, and that context is driven by ever-evolving experiences. ANI captures a single frame of context and experience and will provide relative output that has been curated for that frame of context. This just isn’t the way the world works, not the human world. Our world is infinitely complex and infinitely chaotic. It would require thousands of times more processing power and storage capacity that exists today to fully process the depth of context and effect of experience.
Examples of ANI
- Natural Language Processing (NLP): Systems like chatbots, virtual assistants, enhanced search engines, smart agents, and language translation tools fall under Narrow AI. These kinds of systems can process and generate human language, but their understanding is a limited, single instance of context, and even within the boundaries of the given context, errors are often made. If you use the popular AI models available online, you’ll see a disclaimer that says the AI makes errors. I guess that is how they evade prosecution when the AI kills a person.
- Image Recognition: AI models that identify objects, faces, or patterns in images have showcased impressive accuracy. The very best implementation I’ve witnessed is China’s facial recognition (check out this article). However, all image recognition models lack the basic understanding that humans possess. A simple example would be whether or not a person is wearing a mask. Humans can identify a mask with very little effort. An AI has a nearly 100% rate of failure in determining whether or not a human is wearing a mask in public. This is why some people have resorted to wearing masks in public now. They know the cameras are watching.
- Self-Driving Cars: These vehicles rely on AI algorithms and hardware sensory input to navigate and make real-time decisions. While they can handle complex scenarios, they lack the adaptability of a human. You may ask, why is adaptability important? Because, although many known driving scenarios can be accounted for in programming, there is always the chance that edge cases, or completely new variables, will appear while driving. Humans just have an innate ability to read a situation and make the necessary adjustments based on almost no data. We as humans take so much granted when it comes to the decision-making process. Trust me when I say this, coding human-ability into a digital platform has not been easy.
The limitation of ANI lies in the specificity of a single context. It cannot perform beyond its predefined task, making it fundamentally different from human-level intelligence.
Artificial General Intelligence (AGI)
Artificial General Intelligence, a form of strong AI, is a theoretical concept, a digital counterpart to human intelligence. Imagine an AI system that possesses human-like cognitive abilities, learns from its experiences, and adapts to various tasks without any kind of explicit programming, that would be AGI. Contrary to popular belief, AGI is indeed mythical, and far, far, far into the future. AGI would only be possible if humans ever make the leap from weak AI to strong AI. That has been nothing but a dream and movie script as of now.
AGI Concepts
- Potential: AGI could revolutionize fields like healthcare, scientific research, and creativity. Imagine an AI researcher tirelessly analyzing vast datasets or an AI artist creating masterpieces. This human equivalent would need no further programming or assistance in producing relevant solutions, and even adding to the existing body of creative content.
- Challenges: Creating AGI involves simulating human cognition, which remains elusive. We’re yet to crack the code of consciousness and intuition. It’s going to be difficult in the extreme to impart consciousness or sentience into a digital platform, when—in reality—we don’t understand it very well ourselves. That is just a cold, hard fact. So, be highly skeptical of claims that state they have somehow created human-level intelligence. No, they haven’t.
AGI Example
- An AGI system would be equal to a human, capable of diverse intellectual tasks. It could speak, reason, and even dream. The AGI system would possess broad-spectrum intelligence.
Artificial Superintelligence (ASI)
Super Artificial Intelligence, the pinnacle of strong AI, transcends human intelligence and even artificial general intelligence. It’s purely theoretical but sparks intense debates all over the internet and in written literature.
ASI Concepts
- Potential: Super AI could solve complex problems, optimize global systems, and even enhance its capabilities. Imagine an AI that cures diseases, mitigates climate change, and designs revolutionary technologies.
- Caution: Ensuring alignment with human values becomes critical. Super AI, if misaligned, could have catastrophic consequences.
ASI Example
- Imagine a benevolent Super AI orchestrating global resource allocation, ensuring equitable distribution, and advancing scientific frontiers.
Digital Machine Versus Human Machine
Why is artificial intelligence different than human intelligence?
- Learning Paradigm: AI learns from data and algorithms. It lacks intuition, consciousness, and the ability to learn holistically. Humans learn through experience, intuition, and self-awareness. Our learning transcends mere data points.
- Contextual Adaptability: AI adapts within predefined contexts. It lacks the flexibility to seamlessly switch between unrelated tasks. Humans adapt to diverse situations, emotions, and environments. We thrive on context-switching.
- Creativity and Insight: AI generates content based on patterns; lacks true creativity. Its art, music, and literature lack depth. Humans create art, music, and literature with originality, emotional depth, and cultural context.
- Emotional Understanding: AI lacks genuine emotions or empathy. It processes sentiment but doesn’t feel it. Humans understand emotions, social cues, and complex interpersonal dynamics. Our empathy defines us.
- Self-Awareness: AI has no self-awareness; it operates based on algorithms. It lacks introspection. Humans reflect, introspect, and have a sense of identity. We ponder our existence.
AI is a powerful tool, but it remains distinct from human intelligence. I realize that it is sometimes difficult to separate fact from fiction, but most modern AI, the popular AI you’re seeing in headlines, just isn’t anywhere what’s being marketed. There is no doubt, that AI’s limitations and rising potential are exciting areas of study. As we progress as a species, AI will also progress. AI ethical considerations become paramount (many countries have already started implementing regulations centered on AI). AI isn’t going anywhere. So, we need to understand AI’s capabilities and its impact on society. For me, we can use AI responsibly, while also reaping the many benefits of AI that are yet to come.
Notes
EU AI Act: first regulation on artificial intelligence
How China uses facial recognition to control human behavior
Understanding the different types of artificial intelligence
What is AI-based Image Recognition? Typical Inference Models and Application Examples Explained
Computing Machinery and Intelligence (Alan Turing)
Tags: Artificial Intelligence, MrNetTek, Eddie Jackson Computer, AI
#ArtificialIntelligence #MrNetTek