Default Banner

Stephen Hawking’s vision of future AI

24/04/2018
Stephen Hawking’s vision of future AI

One of the brightest minds of humanity, Steven Hawking, recently passed away at the age of 76. Apart from his legendary insight about black holes and Hawking radiation, Hawking spoke about many different issues during his lifetime - one of which was Artificial Intelligence (AI) and his fear of it overtaking human intelligence and spelling our doom.

AI as a threat to humanity

Today, AI is not at a point of superhuman technology. However, even today, AI can be put to terrible uses in the wrong hands – some scholars have already expressed concerns that self-flying drones may be precursors to lethal autonomous robots. 

In addition, today’s AI raises certain ethical as well as practical problems. AI systems are largely based on opaque algorithms that make decisions which even their designers may be unable to fully explain and the mathematical models they are based on can be biased, and computational errors may occur. As has already been discussed in many previous articles in this column, robotic process automation will go some way towards displacing human skills, resulting in increased global unemployment. It is estimated that 2 billion jobs will disappear to automation by 2030. Limited access to AI might also increase global inequality.

Hawking’s conflicts about AI

Hawking’s views aren’t as gloomy as those described above. He referred to the need to understand and regulate evolving AI technologies. He repeatedly called for more research on the benefits and dangers of AI. One benefit he did describe is that even non-superhuman AI could ultimately help eradicate war, poverty and disease. The contradiction is not unnoticed: On one hand he encourages AI to help solve many of humanities worst problems, and on the other, he speaks about the end of the human race if AI is left uncontrolled.

Hawking knew a thing or two about AI helping human beings. In fact, he eventually had to rely exclusively on AI to interact with the world! At the age of 21, Hawking was diagnosed with a rare and slow-progressing type of motor neurone disease - Amyotrophic Lateral Sclerosis (ALS) - which paralysed him gradually over the decades. Throughout the years, he used a series of different communication systems that helped him to write. Despite eventually losing his ability to speak, he was still able to communicate through use of a hand-held switch linked to a speech generating device, which eventually led to the legendary computer that he operated using just one muscle in his cheek.

The technology that was employed was initially basic AI – a very slow program that was prone to errors. However, as Hawking’s word selection speed improved, the AI within the software analysed Hawking’s words and used the information to assist in expressing his ideas. By processing Hawking’s books, articles and lectures, the system got so good that he did not even have to type whole phrases. When he selected “black”, for example, the word “hole” would automatically be suggested to follow it, since the AI program realised that these words were written in his books or articles multiple times.

Better health with AI

Hawking’s own experience with such a basic form of AI convinced him that it can indeed change people’s lives for the better. He defied predictions that he would only live for a few years at the onset of the disease, overcoming its debilitating effects that left him paralysed. Speech prediction helped him to cope with his neurological disease by allowing his razor sharp mind to express itself. Other similar AI-based systems are now helping to prevent and fight the burden of other diseases. For instance, AI has been used by microbiologists at Harvard University in a microscope that can analyse medical sensors and other health data to predict how likely a patient is to develop a severe blood infection. In studies, results were substantially more accurate and provided earlier warning, compared to other methods.

In another instance, a group of researchers developed an AI program to predict patients’ futures by going through 700,000 electronic health records. ‘Deep Patient’ as the program is known, discovered connections that doctors could not see, detecting new risk patterns for diseases and illnesses such as cancers, diabetes and psychiatric disorders.

AI has even powered a ‘Smart Tissue Autonomous Robot’ that has outperformed human surgeons in a procedure on pigs which is very similar to a type of operation on human patients.

The end of the human race

On the other hand, if AI is left unchecked, it could spell the end of our species, according to Hawking. He argues that humans are limited by slow biological evolution so they cannot compete with superhuman AI, and would therefore naturally be superseded. Once that happens, AI would become its own master and re-design itself at an ever-increasing rate, leaving humanity in the dust.

The good, the bad and….?

The simple truth is that there are many benefits which are readily available today through AI and even more are being developed. However, like science fiction becoming reality, there are also indications that rapid advances in this field would suggest that superhuman AI systems could become extremely powerful, to the point of overtaking human intelligence and ultimately taking over our need to exist. Of course this may never happen at all and regulation may have its part to play as well.

In the meantime, as Hawking knew, there is much to be gained from AI, noting that the technology could help get rid of disease and poverty, and reverse climate change damage (something that he otherwise predicted would turn Earth into a “ball of fire” within six centuries) citing ”We are close to the tipping point where global warming becomes irreversible”. Hawking’s ideas for the survival of the human race has us colonising planets in the future in the closest star system to our, Alpha Centauri.

Although Hawking called for both research and restraint with regards to AI, it ultimately gave him a better and more efficient voice than his body was able to provide and allowed the rest of us to witness the brilliant mind of a genius who dedicated his life to unlocking the secrets of the Universe.

 

Vincent Farrugia is a network and systems manager at Deloitte Malta. For more information please visit www.deloitte.com/mt