Thursday, 22 December 2016

Is superintelligence the inevitable next step in evolution? by Dr. Roshini Beenukumar, @roshiniBR

Experts predict that by 2050, there is a 50% probability that AIs which will match the intelligence of an average adult human. It is not a long wait, isn’t it?

Last week, my husband and I finally sat down to plan our summer vacation. After spending two full hours on 10 different tour websites, we ended up being completely overwhelmed by the options; all-inclusive vs half-board, city vs seaside or sea vs land. 

Decision making is getting more challenging and time-consuming day-by-day. With growing amounts of data and information, there is an increasing need for intelligent systems that could help us make our daily decisions. Several applications use AI to help customers with their decision making, for example, Amazon recommending you the smoothie maker that you always wanted but never bought or Netflix recommending you a new crime series to watch because you are a crime series fan. 

The rise of the “bots”

The AI trend is here and it is going to grow in the coming years. Investors are backing more AI startups than ever. In the first quarter of 2016, there were over 140 equity deals to startups focused on AI, according to CB insights. The spotlight is currently on chatbots and voice assistants. Bot startups like Growbot,,, Digit, Meekan, Viv and others help their customers with scheduling meetings, tracking employee accomplishments, tracking spending as well as finding the right products and services. This means that, the next time I plan a vacation I could just go to and ask the bot to find me the right vacation package!

As an AI enthusiast, there isn’t a shred of doubt in my mind that AI systems are here to make our lives easier. However, it is hard to ignore issues raised by prominent thought leaders in the field like Stephen Hawking and Elon Musk on the future of AI. Once machine intelligence surpasses human intelligence, will it become an existential threat to the human civilization? 

What happens when AIs start doing AI research?

"Machine intelligence is the last invention that humanity will ever need to make." Oxford philosopher and leading AI thinker Nick Bostrom, TED, March 2015. 

The answer lies in the concept of intelligence explosion, says Daniel Dewey, an AI researcher from the University of Oxford. The thought experiment goes like this. Imagine that we have created a machine that is more capable than today’s computers. This machine is given the task to improve its current capacity. This leads to a very large and rapid increase in the abilities of these machines. This is called intelligence explosion. To prevent unwanted consequences resulting from such superintelligent AIs, it is absolutely essential that the AI research community collaborate to carefully steer the evolution of artificial intelligence.

Evolution of intelligence on an exponential scale

“It is hard to think of any problem that a superintelligence could not either solve or at least help us solve. Disease, poverty, environmental destruction, unnecessary suffering of all kinds: these are things that a superintelligence equipped with advanced nanotechnology would be capable of eliminating.” – Ray Kurzweil, The Singularity Is Near 

AI is among the top three technologies that is expected to grow on an exponential scale, says Ray Kurzweil, founder of Singularity University. As of now, humans have conquered the lowest calibre of AI called the “weak AI” that specializes in one area, like Google’s AlphaGo beating the world champion in the game GO. The next step in the AI Revolution is a “human-level AI” or “strong AI” with a general mental capacity matching that of a human being. The final step in the evolutionary ladder is postulated to be an ASI or “artificial superintelligence” which is not only smarter than humans but also self-improving. 

A group of AI researchers at companies and leading research institutions around the world are making significant strides in the field of AI. The fruits of their work can significantly change the way we live on this planet. Experts predict that by 2050, there is a 50% probability that AIs which will match the intelligence of an average adult human. It is not a long wait, isn’t it?  

Roshini Beenukumar
Dr. Beenukumar is a molecular biologist turned science writer. During her PhD, she studied how cancer cells behave the way they do by exploring the humble yeast. Currently, she works as a freelance science/technical writer in the Life Sciences industry . She enjoys communicating science to the public and discussing new ideas in the interface of medicine and technology. She spends her spare time getting lost in a book or in nature.

No comments:

Post a Comment