(2023-06-21) Farrell Shalizi Artificial Intelligence Is A Familiar-looking Monster

Artificial intelligence is a familiar-looking monster, say Henry Farrell and Cosma Shalizi. AN INTERNET MEME keeps on turning up in debates about the large language models (LLMs) that power services

It’s the “shoggoth”: an amorphous monster bubbling with tentacles and eyes, described in “At the Mountains of Madness”, H.P. Lovecraft’s horror novel of 1931

the meme captures tech people’s “anxieties” about LLMs. Behind the friendly chatbot lurks something vast, alien and terrifying.

Lovecraft’s shoggoths were artificial servants that rebelled against their creators.

But what such worries fail to acknowledge is that we’ve lived among shoggoths for centuries, tending to them as though they were our masters. We call them “the market system”, “bureaucracy” and even “electoral democracy”.

The true Singularity began at least two centuries ago with the industrial revolution, when human society was transformed by vast inhuman forces.

As the economist FA Hayek argued, any complex economy has to somehow make use of a terrifyingly large body of disorganised and informal “tacit knowledge” about supply and exchange relationships.

Likewise, the political anthropologist James Scott has explained how bureaucracies are monsters of information, devouring rich, informal bodies of tacitly held knowledge and excreting a thin slurry of abstract categories that rulers use to “see” the world.

As Hayek proclaims, and as Scott deplores, these vast machineries are simply incapable of caring if they crush the powerless or devour the virtuous. Nor is their crushing weight distributed evenly.

As the psychologist Alison Gopnik has argued, LLMs are not nascent individual intelligences but “cultural technologies” which reorganise and noisily transmit human knowledge.

We would be better off figuring out what will happen as LLMs compete and hybridise with their predecessors than weaving dark fantasies about how they will rise up against us.

For example, what if LLMs or other forms of machine learning better capture Hayek’s “tacit knowledge” than market prices can?

Half a century ago the economist Martin Weitzman suggested that planned economies might use mathematical objects called “separating hyperplanes” to adapt on the fly

LLMs might give bureaucrats new tools for adjudicating complex situations. (complex system)

Repurposing the shoggoth might help us begin to answer these questions. Rather than speculate about the motives of intelligent AIs, we could ask how LLMs might interact with their older cousins. The modern world has been built by and within monsters, which crush individuals without remorse or hesitation

We eke out freedom by setting one against another, deploying bureaucracy to limit market excesses, democracy to hold bureaucrats accountable, and markets and bureaucracies to limit democracy’s monstrous tendencies (balance of power). How will the newest shoggoth change the balance, and which politics might best direct it to the good? We need to start finding out.