Does AI Now Mean We're Living in Pluribus?
Centralized knowledge or decentralized, which has more value?
Like many, there was a single article last year that I could point to that was a much needed lens to understand the big bang of AI. If you haven’t already, I highly recommend reading first Death of a Knowledge System by Nicolas Michaelsen which is some of the best writing and analysis I’ve seen in years.
This article really got me thinking that the deployment of AI and Large Language Models are not much different than the Pluribus virus from the hit Apple TV show. It is analogous in that LLMs ingests the collective corpus of all human knowledge (the entire internet, books, movies, social media, transcripts, and software codebases) and unifies it into a single, accessible interface as the LLM. It connects the disparate nodes of human thought and knowledge, completely abstracted from geography, degrees, race, nationality, and class into a “Digital Hive Mind.”
Will the knowledge economy transition into a Pluribus reality? The “Knowledge Broker” represents the old world of friction—the gatekeepers between immune individuals. Previously my innovation teams would rely on very expensive subject matter experts rented through GLG or AlphaSites to learn about an industry or technology quickly. However, AI replacing the knowledge broker signifies the “Joining”, just as in the show Pluribus there are no intermediaries. You don’t ask a grocer to find stock; any person in the system simply is the stock.
Just as the hive mind restocks the shelves instantly in the grocery store episode or a waitress can now pilot an airplane, AI provides instant answers to everyone at all times. There is no scarcity to knowledge or thinking. A toddler who can only speak can be a software developer to vide code a custom video game they want to play and a grandparent can do deep technical research creating a pitch deck with Gamma in minutes. Similarly, we are moving to a world where we don’t ask a broker to find an expert; we query the “Shelf Service” of the AI.
Everyone now has equal access to the same “technical information”, a democratization event that impacts both the digital divide and digital equity more than the invention of the internet. A junior analyst with an AI agent has the same technical reach as a veteran expert, just as any member of the Pluribus hive can access the collective’s skills and data sets. AI is now the great equalizer of specialized knowledge.
So I ask you, does centralized knowledge or decentralized have more value?