Covering Disruptive Technology Powering Business in The Digital Age

Written by: Rogelio Legaspi, Journalist, AOPG

When a technology arises, particularly one that is cutting-edge, people tend to use its name nonchalantly – either because of the lack of better terms or to ride on the wave of interest that it carries. One example is Artificial Intelligence (AI). Today, the term AI is used in various areas but are people using the word correctly?

Simply put, AI is a broad and vague concept that refers to any machine (whether virtual or physical) that is able to perform tasks that require some sort of intelligence of a human. However, there is also a notion that something can only be considered as AI when it already possesses human-level intelligence.

Non-experts or typical consumers may think that the simple voice assistant on their phones can be considered AI. There are also algorithms on various apps, which seem intelligent as they can recommend products or services accordingly. Whereas some experts in the field might create an intelligent system (such as those that provide automation and analysis) and just throw the word AI at it.

While this is not exactly wrong, it is also not entirely true. We should think of AI as more than just an intelligent machine because after all, AI is inspired to imitate the intelligence of humans. But since the influence of AI has already reached a wide range of applications today, we can just distinguish AI in its level of intelligence compared to humans.

AI Stages: Narrow, General and Super

Today, it can be said that every AI in existence falls under the category of Artificial Narrow Intelligence (ANI) or ‘weak’ AI. Why? For starters, AI systems in the current landscape only solve the problems of their particular and assigned application.

Say virtual assistants that use Natural Language Processing (NLP) to detect voice and execute commands accordingly, like tell the weather or open a specific app. There are also automated systems within organisations, using machine-learning, that aim to hasten the process within the IT infrastructure or provide data analysis for insights.

Quite simply, ‘narrow’ AI only uses a portion of human intelligence and it only does what a human can do, albeit in a faster and more efficient way. It is also far from replacing humans in the workspace, as humans can still do such tasks.

Moving forward, these narrow AI systems can integrate with each other and evolve in the future, essentially forming ‘true intelligence’ or ‘strong’ AI, also called Artificial General Intelligence (AGI). General in a sense that narrow AI can only do one task, whereas AGI can do several things a human can do.

In this stage, AGI is at the level of human-like intelligence: able to reason, make judgments, be creative, argue, err, understand semantics or anything a human can do. Think of the AI depicted in the movie ‘Her’, where the AI can converse with a human much like a real person.

In essence, AGI could also (potentially) pass the Turing test, as it can act and talk like a human – indistinguishable from “real” people. Up to this day, no AI has ever fully passed this test.

Now, what if AI surpasses human intelligence in every imaginable way? This is where Artificial Super Intelligence comes to play – the AI that takes the credit for humans worrying about the technology, in the fear of probable world domination by sentient robots or AI or even human extinction. However, we are still far from this stage, as we have not yet developed AGI.

What is True Intelligence?

These three stages or forms of AI can be referred to if we are talking about where we are right now in the progress of the technology. As for the definition of true intelligence, it depends on what you think about the capability of humans, as AI works on the basis of human intelligence.

Some think that we are already in the age of AI, either augmenting the potential of people or altogether replacing them. Others imagine AI as a system that could work on a planetary scale. We may also think that humans are not intelligent enough to be the basis of such a degree and that AI already surpasses human intelligence.

Whichever it is, AI remains in the hands of humans, from the developers to the databases used to create such systems (which involves many other people) and that AI is a reflection of what humans want to achieve.

(0)(0)