By Ali Asgar Sanwarwalla, VP- Social Insights and Analytics AGR
A few years ago, the Dead Internet Theory whose broad tenets were that the internet was algorithm-driven and bots and automated content were widely deployed by vested interests to control perception and opinion was considered a conspiracy theory and thus not a part of mainstream discourse. Over the last few years, from the statements of the co-founder of Reddit, Sam Altman of OpenAI to data from Cloudflare and other cyber security firms has shown that this theory has gained mainstream traction and is no longer a fringe theory floating aroud in cyberspace. From a third to a half of the internet traffic and interactions originating from non-human sources is a shocking statistic that should alarm everyone.
Social media pages, conversations forums and review sites are already seeing the consequences of the bot-based and non-human source interactions. For example, let us take the case of Reddit. The challenge that platforms like Reddit have faced (and continue to face) is the complex relationship they have with AI. Reddit is an important source for data training sets for AI and has data licensing agreements with OpenAI and Google and is leveraging AI for its own answering service as well as for its advertising tools which have helped boost its revenue. At the same time, Reddit understands that its core appeal to users lies in the deep and varied human interactions and engagements among human beings that the platform is able to provide. Reddit has recently sued AI and data scraping companies for what they believe to be unlawful usage of their data since no partnership or licensing was sought. Reddit is also deeply watchful of AI bots infiltrating chats which are between human beings. A recent experiment on Reddit by researchers at the University of Zurich came under a lot of flak and raised ethical concerns. The study used artificial intelligence generated content to engage in discussions in order to test the efficacy of AI in changing people’s perceptions when compared to human interaction. Facebook and others have long been accused of creating feeds which serve as self-reinforcing bubbles that only serve to deepen people’s biases and beliefs with the sole aim of creating stickiness on the platform. Influencers, celebrities and brands too have sometimes been accused of “buying” engagement which has become very easy with the advent of bots and AI.
The role of AI, whether artificial general intelligence is a myth or a reality, is undeniable. The effect it has and will continue to have can range from moderate to transformational across many industries and does definitely indicate a technological leap. As providers of social media insights and services, our business too uses AI and benefits from it. We strive to marry technological prowess with human intelligence in order to create value for our clients. However, it is the latter which is the focus of this article. Customer service firms, research firms and other businesses are often asked if they foresee AI replacing all of their human resources. A common question asked for social media engagement is that with the evolution of chat bots and LLMs, will there be a need for human agents at all?
In order to answer the question above, let me provide some examples of the value of human interactions and not just for platforms like Reddit but for all brand engagement. Every time, a human engaging with an irate customer on behalf of a brand has provided a canned response in recent times, the customer’s responses to that often include phrases like “Can you please train your bot a little better” or “Can I speak to a human please”. Sometimes, these responses are sarcastic but often, customers are genuinely concerned about speaking to other human beings. Further, when a brand receives a lot of positive engagement, whether genuine or paid for, irate complainants and even observers wonder how much of it is genuine and how much is fake. When moments of truth or real experience when it comes to interactions with brands and businesses arise, and these do not align with online reviews, individuals express opinions like the “reviews are manipulated” and “fake”. Recent surveys have indicated what intuitively we know to be true that humans prefer to engage with other humans.
A survey of online conversations itself reveals that human opinions vary regarding AI with some people denouncing it as another nail in the coffin of what used to be a genuine place of human engagement- the internet and social media- while others are either neutral or on the absolute opposite side of the spectrum where they want to embrace the new technology and see it as a leap in human prowess.
The challenge – as the author sees it – is not whether brands and businesses should or should not embrace AI. Big tech and large organizations are highly enthusiastic in using and implementing AI although they are still to find a firm footing when it comes to large scale deployment. Working professionals and other individuals are also embracing AI in some way or the other. The real challenge is the need to be verifiably human and to have human-to-human interactions in the age of “the dead internet”. What has powered the growth and survival of Reddit and other genuine conversational platforms is what will continue to power the growth of brands and businesses specially in their engagement online. Authentic content, human touch and genuineness are not naïve values but form the bedrock and foundation of communication. Automation without a human heart and engagement without human elements is not a matter of how “humanized” AI can get but about how we as embodied beings understand and feel and experience life. In 1972, the widely regarded Hubert Dreyfus wrote “What Computers Can’t do” which was a critique of artificial reason. The world has come a long way since then and although Dreyfus is widely attacked, he is also studied because of his arguments on what it is that makes human beings unique. A casual stroll through the conversations of human beings on Reddit or on social media indicates how important humans regard “real” opinions or “real” humans and the need to be verifiably human with the emergence of AI is going to be a pressing way in which humanity asserts itself. Brands and businesses that seek to build engagement online will have to evaluate automated and AI engagement versus genuine human experience and interactions. Far beyond blue ticks, verifiable humanness will assert and reassert itself in one form or another.
This article was written by a verifiable human life form called Ali Asgar Sanwarwalla who heads the social insights and analytics practice at AGR.
