The AI Hype?!
Much has been said about the future involving Artificial Intelligence (AI). Till recently the discussion on AI was directed into two paths. There were people like Ray Kurzweil who wholeheartedly embraced AI technologies, advocated the view that AI would be a new blessing for humanity and predicted a tipping point (singularity - when machines would take over) coming in 2045. Please mind that Kurzweil’s 147 predictions since the 1990s have turned out to be accurate in 86 percent. Therefore, his views ought to be taken seriously. There were also those people who voiced concern about the development of AI and pointed to life-threatening consequences of the unregulated AI boom (e.g. Nick Bostrom, the author of “Superintelligence: Paths, Dangers, Strategies”, Stephen Hawking who claimed that AI could spell the end of the human race, or even Elon Musk who is afraid that one of the side effects of advancements in AI may be genetically modified humans).
We live in the world which takes for granted that AI will revolutionize our everyday lives, will reshape our understanding of “work”, and wrought significant social and economic reforms. We fret over the future of our children in the environment of total digital control and automation and we get overexcited about all the joys, new experiences and healthier existence that AI promises to accomplish. However, now there are critical voices that begin to question all those AI powers. Some of us start to reconsider the true implication of artificial intelligence.
There is no doubt that since 1956 when John McCarthy coined the term “artificial intelligence” research in AI has had spectacular achievements and its practical applications have been felt by many. It is worth remembering, however, that, for instance, in the end of 70s and the end of 80s of the previous century AI was not a buzz word at all. We call these period of “AI winters”. The recent successes have come mainly because of a tremendous increase in available data and a not less gigantic increase in computing power. Suddenly algorithms have started to work effectively. This way we have managed to teach computers how to identify items in pictures and recognize voices by using machine learning and deep learning. The former uses structured data to produce the output. The latter may not have labeled data, it produces the output going through various layers of decision making. Yet ultimately both machine and deep learning depend on the quality of data. The quality of data is key.
It looks like AI is good at discerning patters (e.g. pinpointing faces from camera footages) and making approximations (what might happen next given tons of data regarding the preceding stages of the system). Any assertions about reaching the human level of intelligence by robots like Sophia are utterly exaggerated. It is worth noting that representatives of Hanson Robotics do not deny that Sophia is just a computer program which cannot think for itself. They just let the popular misconception prevail.
If AI is not about intelligence (let along consciousness) then why is it so eagerly promoted? It is not an easy question to answer and it may have various layers. Definitely it is financially advantageous for some big companies. It is also profitable for scientists who could ride the wave and capitalize on publicity applying for grants, etc. It is worth considering, however, wider implications of AI like making human mind more artificial and less creative. To Lord Kelvin everything that could not be measured did not exist (he also argued that plains would never be able to fly). To us nowadays everything what cannot be digitalized in a stream of data does not exist either. It seems that we lost the sense of transcendence a long time ago, then we parted with many virtues. We live in the world which becomes more and more artificial and we tend to become robots faster than robots are rendered humans.
It is essential to work out our own judgement and do not yield do current fashions. It is true that at the beginning of the XX century there were hardly any car in New York. In 12 years, however, there were hardly any horse. Cars won over truly fast. On the other hand, we do remember glorious promises of cloning, stem cells, nanotechnology, growing human organs in animals, making new neurons grow in human brain, etc. We still have to wait for serious advances in these fields. Artificial Intelligence has been around for about 60 years. In a few last years it has successfully been implemented in some businesses. Whether it would revolutionize our life as a washing machine did remains to be seen.
More about critical view on AI you may read here.