The Malta Independent 20 October 2019, Sunday

Artificial intelligence: Friend or nemesis?

George M Mangion Sunday, 3 July 2016, 09:11 Last update: about 4 years ago

The result is that the massive computer power so harnessed helps us to analyse what has happened in the past and, with the use of predictive analytics techniques, opens a window leading to accurate predictions.

Undoubtedly, artificial intelligence is fast becoming a major technology for prescriptive analytics, the step beyond predictive analytics that helps us determine how to implement and/or optimise optimal decisions. In business applications, it can assess future risks and quantify probabilities, giving us insights into how to improve market penetration, customer satisfaction, security analysis, trade execution and fraud detection and prevention, while proving indispensable in land and air-traffic control, national security and defence, not to mention a host of healthcare applications such as patientspecific treatments for diseases and illnesses.

Typically, the giant search engine firm Google is a pioneer in the field of artificial intelligence, developing self-driving automobiles, smartphone assistants and other examples of machine learning, while it is no secret that Facebook founder Mark Zuckerberg and actor Ashton Kutcher recently invested $40 million in a project focusing on developing artificial brains.

In science fiction films such as Matrix, we have seen how futuristic devices will facilitate facial recognition, interpret human comments and perform complex language translations. Some of these devices are no longer in the realms of fantasy but have become real tools as a result of advances in science and social media. Readers will appreciate that, through social media links, this technology is helping various countries improve crop yields and speed up the progress in complex human Genome classification. Delivery drones, both wheeled and airborne, may in the near future compete with couriers while supermarket robots silently stack food items on shelves and move merchandise in warehouses.

Let us stop for a moment and take a look at major stages of human advancement over the centuries, starting with the ancient Greeks who laid the foundations of democracy. Plato was so particular in his dialogues with Socrates about the importance of human values which, of course, are alien to robots and faceless machines powered by artificial intelligence.

The next big advance in knowledge was during the Renaissance, which heralded a flowering of the arts, general knowledge and the invention of the printing press. This was followed by a quantum leap in technology during the industrial revolution which saw the invention of combustion engines, increasing productivity in factories, providing families with cheap electricity and the discovery of the germ theory of disease that heralded vast improvements in medicine. Einstein was born and later on the Internet age was developed and continues to expand apace.

The way forward beyond 2016 is awe-inspiring because soon there will be computers that can run faster than the human brain. It is unstoppable - computing power will continue to grow at a faster rate. It could be said that this is a blessing and shows how mankind has developed the ability to programme machines using computers running at incredible speed - effortlessly solving practical problems. The dilemma is whether all this computing power (now housed in Cloud) can be harnessed to be our servant. The million dollar question is: can we succeed in instilling values in brains of such behemoths? Not so fast, as scientists warn us that such unbridled power threatens the very existence of mankind.

Last year, famed astrophysicist Professor Stephen Hawking warned of the dangers posed by developing human-like intelligence in computer systems. Just consider how computer researcher Alan Turing has succeeded in designing one computer that can successfully imitate a human in conversation, although few doubt that more research is still needed in order to optimise neural design to make it function like the short-term memory in human brains. Simply put, it consists of a special network programmed to learn as it stores memories, empowering the device to mutate beyond its original code.

Artificial intelligence in machines can even replicate human judgements previously considered to be too complex. Devoid of emotions, it can supervise hundreds of skilled factory operatives. As if by magic, complex algorithms that 'learn' from past examples can relieve engineers of the need to write out every command.

It comes as no surprise that cash-rich Google has invested $400 million in this project which may, in the near future, turn out to be the building block for functional robots. The investment in robotics points to a race to produce superior machines with the ability to recognise objects, people and pets and to determine their own actions. For example, one finds robots at home using artificial intelligence to tell us to refill the cat's food dish when they sense it is empty.

Quoting Ray Kurzweil, a futurist and inventor recently hired by Google as director of engineering, he believes that computers will soon become more intelligent than human beings. He believes that this event, which he refers to as 'the singularity', will take place by the year 2045. Equally ominous is the prediction of Stephen Hawking, who exclaimed that "The development of full artificial intelligence could spell the end of the human race." Professor Hawking says the primitive forms of artificial intelligence developed so far have already proved very useful, but he fears the consequences of creating something capable of matching or surpassing humans that would take off on its own and re-design itself at an ever-increasing rate. He fears that humans, limited by slow biological evolution, could not compete with such superior machines and would be superseded.

In fact, Intel already develop smart chips that could pave the way for sensor networks that mimic the brain's capacity for perception, action and thought, so we could ask what is there to stop computer scientists developing a machine with a brain that is even more intelligent than that of humans? And yet there is a contrary view to the claim that we have lost our 'man versus machine' fight. Some computer scientists believe that artificial intelligence can be harnessed to be a servant not a master. Charlie Ortiz, heading the software company Nuance Communication, feels that increased AI might be useful in solving current problems that human intelligence cannot fathom. In his opinion, there is no reason to assume that AI will harm us, even if it becomes more complex than human intelligence, but it is in our own interest to set up safeguards that will protect our own intelligence when AI (a new god) reaches its acme and possibly mutates to try to become our master.

To conclude: not everyone is so negative with regard to the destiny of humanity. So let's cheer up and for a moment forget the scary predictions of robots ruling over mankind and pray for the day when the Internet will dutifully connect every gadget in the home: not all fridges warn us when we're out of milk.

Mr Mangion is a senior partner of PKF an audit and consultancy firm, and has over 30 years experience in accounting, taxation, financial and consultancy services. He can be contacted at [email protected] or on +356 21493041. 

  • don't miss