The concept of AI as an ancient technology is intriguingly presented in Venkatesh Rao’s essay, “Superhistory, Not Superintelligence.” This idea reshapes our understanding of AI and ourselves, positioning us as ‘centaurs’ – beings augmented by artificial time.
AI, in essence, is Artificial Time (AT) rather than just Artificial Intelligence. It embodies superhistory by absorbing vast quantities of data, encompassing centuries of human knowledge. This ‘data aging’ concept means that AI experiences years’ worth of information in a drastically shorter period, aging centuries in mere days. It surpasses the limitations of human information consumption, which is bound by our physical capabilities.
Understanding this, we can redefine human wisdom and experience not in terms of chronological age but in data age. A person’s true age could be a blend of their biological years and their exposure to information or data years. This is where the concept of a ‘centaur’ comes into play: a human augmented by artificial time, whose wisdom and capabilities extend far beyond their physical years.
Take the example of chess grandmaster Magnus Carlsen, who, at 22, wasn’t just a young chess champion but an individual who had accumulated centuries of chess knowledge through his training with AI. His true ‘player age’ was a combination of his biological age and the immense data age he had gained through AI.
This idea also applies to our everyday lives. The internet and AI tools like Google and ChatGPT have accelerated our data aging. They allow us to access and learn from a vast array of information far beyond what we could physically experience or learn in our lifetime. As a result, someone who actively engages with these tools can possess knowledge and wisdom that far exceed what their chronological age would suggest.
However, this also highlights a disparity between generations. Those who grew up with limited digital augmentation, like some older generations, have a significantly different data age compared to digitally-native younger generations, leading to a gap in knowledge and understanding.
The concept of superhistory takes an even more fascinating turn when considering that AI can generate its own history. For instance, AlphaGo, after learning the basics of Go, improved its skills by playing against itself, thereby creating a unique dataset and history beyond human experience.
“Superhistory, Not Superintelligence” provides a profound perspective shift on how we perceive AI, human intelligence, and the concept of aging. It encourages us to consider not just the passage of time but the depth and breadth of information we have experienced and absorbed. This understanding can significantly impact how we value and assess wisdom, experience, and intelligence in the modern world.
Enjoyed the article?
Subscribe for more and follow us on Twitter at @Blastfutures