How intelligent is Artificial Intelligence really?

16th April 2020

‘I propose that we need to develop a deeper understanding of natural intelligence so that we can really advance artificial intelligence,’ says Claude Calleja, Executive, eSkills Malta Foundation

Artificial Intelligence (AI) is a broad term with different definitions, especially from the perspective of general intelligence. AI was first used by John McCarthy in 1955 and referred to a machine that mimicked human detection and problem solving.

Machine learning is a branch of AI that was already shaped by Arthur Samuel in 1959 as a field of study that enables computers to learn without being explicitly programmed. A newer and more common definition of machine learning is the scientific study of computer algorithms that automatically extract patterns and create models from existing data and make predictions and conclusions about new data.

AI is an attempt to have computers perform services for which humans need intelligence. It is therefore advisable to be careful with the term artificial intelligence, since not everything that is currently sold to us as such can certainly be called AI. The dictionary defines intelligence as the ability of man to think abstractly and rationally and to derive useful actions from such thoughts.

“I propose that we need to develop a deeper understanding of natural intelligence so that we can really advance artificial intelligence.

“We need to better understand human intelligence and understand how a conscious person acts, what goes through their mind, what drives the decision-making process and how emotions matter to say the least.

“Then we can try to implement this newly gained understanding in the form of algorithms, machines, robots and software applications such as chatbots and virtual assistants,” says Claude Calleja, Executive, eSkills Malta Foundation.

In short, there has been tremendous progress in the ability of AI systems to incorporate intentionality, intelligence, and adaptability into their algorithms in recent years.

“With the right security measures in place, countries can make progress and take advantage of artificial intelligence and emerging technologies without sacrificing the important characteristics that make up humanity.

“There is no easy answer to this question, but system developers need to incorporate key ethical values into algorithms to ensure that they meet human needs, learn and adapt in a way that is compatible with community values.”

It can more broadly be described as the ability to perceive or derive information and retain it as knowledge to be applied to adaptive behaviours within an environment or context.

Common to both is the idea that intelligence is about using knowledge and applying it in different ways to solve a new problem. Professor Jon Crowcroft explains in his blog that most things that we consider artificial intelligence are certainly artificial.

The earliest affect detection work began almost three decades ago when physiological sensors, cameras and microphones were used to detect a variety of affective responses.

The ability to take appropriate action based on a person's inner cognitive state is essential for an emotionally intelligent agent. Applications such as automatic tutoring systems, support for mental and physical health and applications to improve productivity are at the forefront of the current measures.

“We are in the middle of a new wave of excitement about artificial intelligence. In some areas, due to recent advances in computing power, artificial intelligence is being touted as finally realistic.

“Once again, the public is invigorated with visions of how our lives will change completely and how millions will be replaced by intelligent machines. The immediate availability of low-cost graphics processing units has made Convolutional Neural Networks commercially viable for some applications such as image recognition.

“We now have more strength in our pockets than we had at home 10 years ago. Artificial intelligence has been a favourite topic in novels and films for decades. But now it is an important part of our technology in different ways.

“However, the limits of deep learning are still a cause for concern. We are still quite a long way away from Skynet!”

This article was prepared by collating various publicly available online sources.

Claude Calleja, Executive, eSkills Malta Foundation

17th April 2020

Almost 17,000 businesses applied for the Government wage supplement

Rebecca Anastasi - 15th April 2020

The figures for the year’s first quarter show that only 212,341 passengers made the trip last month,...

14th April 2020

Pursuant to the legal notice, the Central Bank of Malta issued a directive to define the eligibility...

13th April 2020

In a statement, the department urged the public to get in touch online and by telephone.

How intelligent is Artificial Intelligence really?