Understanding Intelligence in AI: From Narrow AI to Artificial General Intelligence
As we continue to develop and integrate AI into various aspects of our lives, it's crucial to recognize the distinctions between the two primary forms of intelligence: Artificial General Intelligence (AGI) and Narrow AI.
As the world of artificial intelligence (AI) continues to evolve, it's essential to understand the concept of intelligence and its different manifestations within AI systems. Intelligence can be described as the ability to acquire, process, and apply knowledge and skills. In the context of AI, we generally categorise this into two main types: Artificial General Intelligence (AGI) and Narrow AI.
Artificial General Intelligence (AGI) refers to AI systems that can understand, learn, and apply knowledge across a wide range of tasks and areas, similar to human intelligence. AGI can adapt to new situations, transfer knowledge from one area to another, and display creativity and problem-solving skills that aren't limited to a specific domain. In essence, AGI can perform any intellectual task that a human can do, making it a versatile and powerful form of intelligence. The ultimate aim of artificial general intelligence is to build an agent/ algorithm, that can learn to solve any problem from scratch without being taught how.
Oxford philosopher Nick Bostrom defines superintelligence as
“any intellect that greatly exceeds the cognitive performance of humans in virtually all domains of interest”
In contrast, Narrow AI, also known as Weak AI or Artificial Narrow Intelligence (ANI), is designed to handle specific tasks or solve particular problems within a limited area. These AI systems are highly specialised and optimised for their designated functions but lack the ability to generalise their knowledge or skills to other areas. Examples of Narrow AI include image recognition algorithms, speech recognition systems, and recommendation engines.
Any AI you can think of that is currently in use is an example of narrow AI; these are not the robots taking over the world as seen on the silver screen or in the pages of science fiction novels, though. Here are eight real-world instances:
Digital Voice Assistants
Digital voice assistants like Siri and Alexa are often regarded as the best instances of narrow AI and are examples of narrow AI that we use on a daily basis. The AI must quickly classify data and reply to queries in order to function properly.
These recommendation systems are useful whether Netflix suggests what movie you should watch next or Amazon or other retail websites provide you helpful suggestions about what else you might be interested in buying.
Image and speech recognition
Narrow AI is behind image recognition in many industries. Through image recognition, narrow AI is having a significant impact on the healthcare industry by assisting radiologists detect disease from the scans of their patients. It is also behind speech recognition and translation services like Google Translate.
Search engines like Google and others are another instance of a narrow AI. When you type in your query, the algorithm goes to work categorising it and returning replies by searching through its enormous database.
If you've ever interacted with a company via chat, whether it was your bank, internet service provider, or preferred online retailer, you were probably speaking to artificial intelligence. The majority of the time, chat features are an AI algorithm that takes care of responding to frequently asked queries in order to free up the humans who formerly performed this work to finish more difficult duties.
Narrow AI is the technology that enables automobiles to drive without a human driver. The algorithms carry out preprogrammed tasks. Since an AI doesn't have complete cognitive capacities like a human brain, it is difficult to programme and train it to recognise any potential hazards on the road or situations that the vehicle might face.
Robots don’t have a mind of their own. Manufacturing robots and drones have limited artificial intelligence (AI) and can only execute the limited range of tasks that they have been programmed to do. Delivery robots were highly helpful during the epidemic as disinfecting robots and for following social distance requests.
As we continue to develop and integrate AI into various aspects of our lives, it's crucial to recognise the differences between AGI and Narrow AI. While AGI has the potential to revolutionise many industries and lead to groundbreaking discoveries, Narrow AI serves as a valuable tool for addressing specific challenges and streamlining processes within its designated domain.
It's important to note that despite the exciting potential of Artificial General Intelligence (AGI), no AGI systems have been successfully developed so far. All current AI applications and systems fall under the category of Narrow AI, focusing on specific tasks and areas. The development of AGI remains an ambitious and complex goal within the field of AI research, as it requires building systems that can understand, learn, and adapt across multiple domains, closely emulating human intelligence.
The current state of AI research and development is primarily centred around improving and refining Narrow AI systems, making them increasingly efficient and effective in their respective areas. As our understanding of AI and its underlying technologies advances, researchers continue to strive towards the realisation of AGI, which could significantly transform our world and unlock countless possibilities. However, until AGI becomes a reality, Narrow AI will continue to play a key role in enhancing various aspects of our lives, from healthcare and transportation to entertainment and communication.