By Ashoke Agarrwal
It is conventional today to blame many of modern society’s ills on social media and its tendency to nurture and amplify echo chambers. Echo chambers are fertile grounds for bigots and cranks for all hues to gather and indoctrinate the impressionable and idle among us.
The more technologically sophisticated among the commentariat identify the algorithms that social media companies employ to be at the core of the echo chamber phenomenon. Social media platforms design these algorithms to maximise the time an individual spends with them and thus maximise advertising revenue. To many, these algorithms are simple forms of Artificial Intelligence (AI) – a relentless machine that, one day in the near future, will control aspects of life, robbing the ordinary individual of many degrees of freedom.
To my mind, the debate and apprehension over AI are overblown and misplaced. A crucial input into the algorithms that bring about social media’s echo chambers is the interaction of the individual with the content presented to him. The algorithms respond to the time he spends on a particular piece of content and the reaction button he presses (Facebook currently has six options – Like, Love, Haha, Wow, Sad and Angry). In that sense, social media platforms’ algorithms are a form of Augmented Intelligence (AuI) – albeit what I call Weak Augmented Intelligence (WAuI).
A strategic shift from WAuM to Strong Augmented Intelligence (SAuI), whether voluntary or forced through regulation, can address the ills of social media. While WAuI implies near-involuntary participation by the human in the human-machine interaction, SAuI requires the human component to be conscious and informed. Imagine a social media platform genuinely sharing and seeking feedback. For example, say Facebook shares with its users at periodic intervals the user’s profile built by its algorithm and allows them to input suggestions. Furthermore, imagine Facebook periodically seeking their personal growth goals from its users and making these a part of the algorithm’s input. That would be the SAuI system, and I contend that, sooner or later, persuaded either by a maverick competitor or by regulators, the social media world would move to SAuI.
I also contend that SAuI will be the dominant form of AI across domains over the next few decades.
Take the example of AI applications in the daily lives of people that have been around for some time now – recommendation engines on e-commerce and streaming platforms and chatbots in the customer service arena. A white paper in a recent issue of the Journal of Marketing titled “Artificial Intelligence in Utilitarian vs Hedonic Contexts: The “Word-of- Machine” Effect” by Chiara Longoni and Luca Cian provides some interesting insights. People generally vary with inputs and advice from those they perceive as “Word of Machine” – inputs coming from AI with little or no human inputs. They are a little less vary when the “Word of Machine” is about a product or issue with well-defined rational dimensions at play – that is, they are “utilitarian” in nature. However, when the product or issue has emotional dimensions at play – that is, they are “hedonistic” at heart, the scepticism against “Word of Machines” is high. Some proponents of AI believe that as Natural Language Processing (NLP) engines cross the Turing threshold, pure AI-driven recommendation engines and chatbots will get wider acceptance and usage. The Turing threshold, to be clear, is when Natural Language Processing (NLP) based AI becomes so sophisticated that a human at the other end of a conversation cannot make out that it is a computer. They cite the emergence of Google’s LaMDA chatbot as that end is neigh. I think advanced AI technology like LaMDA can improve and be more valuable if it employs SAuI instead of pure AI. While the quest driving pure AI and GI pushes the boundaries, it is by incorporating SAuI that tools based on these advanced technologies can increase utility and find broad societal acceptance.
A review paper titled “Artificial Intelligence and Management: The Automation-Augmentation Paradox” by Sebastian Raisch and Sebastian Krakowski for the Academy of Management Review validated this insight in business and management.
As for the broader fields of arts and sciences and the broader human quest for knowledge, let us consider the predictions of Ray Kurzweil, the world’s leading authority on artificial intelligence and pre-eminent forecaster. In his 1999 book, “The Age of Spiritual Machines”, Mr Kurzweil made pithy milestone-type predictions for the next 100 years. Mr Kurzweil”s forecast for 2099 is “There is no longer any clear distinction between humans and computers. Most conscious entities do not have a permanent physical presence…the goal of education, and of intelligent beings, is discovering new knowledge to learn…Life expectancy is no longer a viable term in relation to intelligent beings.”
That is a breathtaking and bold prediction, but Mr Kurzweil is taken seriously in most quarters because his forecasts for 2009 and 2019 have, by and large, come true, and the one he made for 2029 looks to be heading that way.
In the context of pure AI and SAuI debate, it is essential to take note of two key phrases in Mr Kurzwell’s 2009 prediction – “no longer a clear distinction between human and computers” and “the goal of education and intelligent beings is discovering new knowledge and learn..”.
The culmination of AI is a symbiosis of humans and computers, and the ultimate human goal is innovation and creativity. It, therefore, stands to reason to accept that somewhere early in the development of AI, the focus will shift from pure AI to what I term Strong Augmented Intelligence (SAuI). SAuI will be a jugalbandi between the extraordinary powers of computers married to the essential human quest for knowledge and creativity.
PS: I have explored these themes in my earlier MxMIndia columns – two of them being:
“The Coming Post Digital Age”, published on Jan 6, 2022,
and “From Machine Learning to Machine Creativity”, published on Jan 20, 2022
