Roger Hanson: Don't fear the new AI revolution

Tuesday, 24 July 2018

Artificial intelligence is not something that will happen in the future. It's happening right now and you're already using it.

OPINION: In 1955, American computer scientist John McCarthy coined the term 'artificial intelligence' and in 1958 documented his ideas in a scientific paper, 'Programmes with Common Sense', in which he proposed using logic to represent information in a computer.

Traditionally, computing has involved writing a set of instructions called the computer programme and installing that program in the computer.

The operator then feeds data into the computer and the programme slavishly applies the line by line instructions to the input data and produces an answer.

Today, artificial intelligence (AI) is software that writes itself. Computing has now reached the stage where the speed at which computer hardware can process data is many thousands of times faster than the human brain.

* Never mind the AI apocalypse, AI can be useful too

* It's not coming for your job, yet**

The combination of software able to write itself and super fast processing speeds has opened the door to the AI revolution.

This revolution has two aspects, the first is the ability to manage prodigious amounts of data and the second is called machine learning.

An example of the power of the data processing ability of AI is cameras in public places.

These can be configured not just to look at one person in the frame, but to observe thousands of people and using the AI software, simultaneously track each individual.

Each person can be identified using recognition software and a vast data base interrogated to report on when each individual was last seen in the location.

Machine learning is still in its infancy but this uses the ability of machines to retain information to improve their knowledge with experience.

In a medical application, at first a computer might have difficulty in telling the difference between healthy tissue and malignant tissue, but in a remarkably short time, equipped with its trial and error software, it can learn and become expert in diagnosing malignant tumours.

IBM's Watson computer, famously demonstrated its potential in 2016 when its performance was acknowledged by a team of top oncologists.

The computer began with no knowledge of cancer but its software searched the internet and very quickly learned.

We are already using and interfacing with computers underpinned by AI. The chances are that when you book a hotel or flight, the price of that booking is decided by an AI machine which, every second, is receiving a huge number of requests from dozens of customers for hotels and flight bookings.

This computer is, within milliseconds, deciding based on a list of criteria, what the price should be.

Should we fear AI? No, it should be embraced, it is our friend, it will open a world of new opportunities.

AI will do what every technological revolution has done before, substantially contribute to the progress of human development and improve the quality of many lives.

Like all new technologies there is a downside; defence experts warn of drones that conduct missions with no human intervention apart from the 'kill decision' which will be made by a human (maybe).

However we have lived with the dreadful consequences of new technologies for years, from the bow and arrow to the machine gun to the nuclear bomb.

Will AI take all our jobs? This is a much debated question and there are compelling arguments on both sides. Many jobs will go, but many will be generated.

If history is anything to go by, it has shown that humans adapt supremely well to technological change and reorganise their lives accordingly – to include new jobs and activities.

Sport is now a huge international industry and has grown exponentially in the last 50 years because millions of people have time to watch and participate in sport.

These people don't have to spend every waking minute working in places like mines hand-digging coal.

The lives of billions of people have been enriched by the computer revolution – today, for most people, life without a smart phone would be unimaginable.

In the 1970s scientists were warning that computers would be the cause of mass unemployment, yet here we are in 2018 with some of the lowest unemployment figures since the 1950s.

It is true that no one can accurately predict the future, but there is every chance the AI revolution, like the computer revolution before it, will improve human lives.