Empire of AI: "Prologue", "Divine Right", and "A Civilizing Mission" (Patreon)
Content
note: these are just my first impressions of the reading---in the comments, feel free to engage with this or start completely unrelated threads~
"Over the years, I've found only one metaphor that encapsulates the nature of what these AI power players are: empires. During the long era of European colonialism, empires seized and extracted resources that were not their own and exploited the labor of the people they subjugated to mine, cultivate, and refine those resources for the empires' enrichment. They projected racist, dehumanizing ideas of their own superiority and modernity to justify---and even entice the conquered into accepting---the invasion of sovereignty, the theft, and the subjugation. They justified their quest for power by the need to compete with other empires: in an arms race all bets are off...In the simplest terms, empires amassed extraordinary riches across space and time, through imposing a colonial world order, at great expense to everyone else."
My first awareness of this book had been hearing Karen Hao make this comparison of tech companies to empires in an interview sometime last year, and I've been vying to read it ever since. I remember thinking it almost sounded to extreme at first, but the more she said and the more I thought about it, the more I feel like she's absolutely right: there is nothing else that compares in scope or scale of power.
Something I'm already enjoying a lot about this book is how much it reads like an actual narrative, almost like a novel (for anyone who's not participated in our other book clubs, I think this one is a first in that respect). And maybe because of the story-telling structure of it, my biggest impressions from these first few chapters are less about the tech or the culture---which are things I'm already pretty familiar with---but the personal stories. Like, you know, everything we know about Altman and what he's responsible for at this point aside, he's a fascinating character. I don't anticipate that my takeaway by the end of the book is gonna be 'huh Altman's just a normal dude', but I was surprised by how much Chapter 1 humanized him to me. Something about the line describing him as sitting "with a leg popped up or perched like a bird atop his chair" really had me feeling he's justlikemeforreal. But like, a far more ambitious and socially skillful me.
Kind of keeping in the theme of seeing more of the human side of these people for whom I usually...don't.... Chapter 2 had me curious about the earnestness of supposed good initial intentions. Things like OpenAI initially being intended as a legitimately open source resource or the Anthropic folks going their own way to try to be meaningfully better. On the one hand, I've been skeptical/outright disbelieving of say, Elon Musk's early stated concerns about AI safety as his motivation for pushing so hard to be involved in its development (it's something of a trope to have tech people express concern for possible harms in order to position themselves as the ones who will "handle it", as opposed to regulation, sometimes called "ethics washing"). On the other hand, I (maybe naively?) want to believe that when Google was still just a cool search engine, they really meant it when they said their overarching rule was "don't be evil". I suppose in the grand scheme of things, it doesn't matter much whether the initial good intentions were sincere or not if they're eventually deviated from, but idk, I find myself wondering how much was just fraud from jump versus the system pushing well intentioned people to abandon their values.
Besides all the humanizing, something that stood out to me was the uncanniness of how people not even a decade ago were talking about what would become our reality. To end on one such unsettling quote AI acountability researcher Deborah Raji wrote addressing the existential fearmongering around possible rogue super-intellgient AGI:
"It is not just the actions of an AI agent that can produce side effects...In real life, basic design choices involved in mode l creation and deployment processes also have consequences that reach far beyond the impact that a single model's decision can have. In reality, for AI systems to even be built, there is very often a hidden human cost."
yeah, pretty much.