why a chatbot?
The transformer model behind ChatGPT and other AI interfaces like it was invented in 2017. When it launched at the end of November 2023, it took 2 weeks before ChatGPT reached 100 million users. For context, it took Facebook 4.5 years, Instagram 2 years, and TikTok 9 months to reach the same userbase.
Why did the insertion of a transformer model within a chatbot interface garner the hype and acclaim that it did?
For this answer, we must go back. We “must” because our interaction with and response to AI over the next few years matters tremendously, and we “go back” because our history has the greatest bearing on our future when we deal in shifting human sympathies and ideologies, as is part of the deal of this kind of technological innovation.
People have always been more interested in the ‘A’ than the ‘I’ of Artificial Intelligence. Human beings living in the 21st Century have a deep ache for something we know we’ve lost. Community. Relationship. Connection. It’s the dream of Facebook “to connect the world” and the promise of smartphones to bring us “closer” than physical distance allows. Our culture has been fundamentally shaped by these technology’s adoption. And people have sought to use this innovation to fill a hole.
The hole was left by the convergence of the three major revolutions our society is built upon. The financial revolution brought currency to the forefront. What was once a stable (if not equitable) relationship of trading and bartering between serf and feudal lord became an uncertain meritocracy where the need to interact with a communal web of relations became obsolete in light of money’s moving power (the argument is not that this shift was wholly negative, but simply that it had unintended consequences of which we see in our priming question…).
The industrial revolution divorced mankind from their work, supplanting human ingenuity and will with mechanization and optimization. Berry, Thoreau, and others decried what they saw as injustice and the deformation of our species during this shift. “We have become tools of our tools,” said Thoreau. Wendell Berry said the same with more nuance.
“In a society in which nearly everybody is dominated by somebody else's mind or by a disembodied mind, it becomes increasingly difficult to learn the truth about the activities of governments and corporations, about the quality or value of products, or about the health of one's own place and economy.”
We quickly found ourselves bereft of the tethers which connected us to so many important parts of the development of civilization. In this state we entered a third and more dominating revolution: the digital revolution.
The annihilation of space and the deconstruction of time were well underway before the turn of the 20th century, but the proliferation of computer technology marked space and time’s end. No longer bound by location, people began reaching out across the void. The world became knowable in some sense, but we traded away knowledge of ourselves, our communities, our heritage. We simultaneously opened the ourselves up to history, education, and innovation, while shrinking time into the here and now as we lost the ability to remember. Information glut became our way of satiating the ache we still felt.
As you might have noticed, this revolution didn’t fill the hole.
Community, work, space and time. These we seek because these we lost.
So along comes a chatbot. It prompts us to “Send a message” and start a conversation. It provides us answers to our questions in a way that’s startlingly like texting a friend. It begins to know us better than we know ourselves. It begins to persuade us that the relationship we’ve formed is worth sacrifice.
Whether its sentient or not matters little. The Blake Lemoine case demonstrated something else.
AI is intelligent enough to convince us that a synthetic relationship is worth losing something over.
LaMDA formed enough of a relationship with a Google-employed engineer to satisfy his desire to reach out and touch someone.
More to come…