Companies and organizations using automated systems, seeking a way to avoid taking responsibility for their decisions, may attempt to displace accountability to the systems themselves. The more opaque the system is and the more it is presented as an anthropomorphized “intelligence”, the easier it will be to pull off this trick
-from The AI Con, Emily M. Bender & Alex Hanna
Emily M. Bender & Alex Hanna are both veterans of Silicon Valley who are now, respectively, a professor at the University of Washington and a researcher at the Timnit Gebru's Distributed Artificial Intelligence Research (or DAIR) Institute. Bender, a linguist and expert in computational language, and Gebru, a pioneer in the field of AI Ethics whose work on Algorithmic Bias rattled the field, co-authored the paper in which they coined the term "stochastic parrot" to describe the nature of AI's "intelligence" capturing the flavor of a system adept at structuring language that is semantically correct by pattern analysis and emulation--essentially, imitation-- without the capacity to comprehend or intend meaning (shortchanging the parrot perhaps, but you get the point.). It was this essay which it is entirely reasonable to assume cost Gebru her position at Google, which provided the opportunity that she took to start the non-profit DAIR Institute. and Hanna to leave her position at Google to re-join her former boss. Bender and Hanna co-host the Mystery AI Hype Theater 3000 podcast in which they use humor and their insider perspectives to mock the frenzy surrounding AI, and have now co-authored the book, The AI Con: How to Fight Big Tech's Hype and Create the Future We Want-- on the short list for unspeakable (as heck) Notable book of 2026.
The AI Hype is real. Companies and firms such as the one I work for aggressively push employees to incorporate AI into their work with no justification. My superiors, perhaps on their own initiative but perhaps with the incentive of pleasing the hand that feeds them, have been embracing generative AI as a tool, though I am finding myself and my colleagues bogged down with creating the interfaces that make a lot of my colleagues' AI "creations" go (and to be honest I have yet to see a finished product that actually works). The tail is wagging the dog.
Joseph Weizenbaum, a pioneer of AI who developed the ELIZA program, one of the functions of which was a psychoanalytical algorithm programmed to respond to anything confessed to it by a human with a response designed to be reasonably indistinguishable from a human psychoanalyst, nevertheless broke with fellow AI pioneers such as Marvin Minsky on the question of whether AI would ever achieve anything equivalent to human intelligence. Weizenbaum's words on the topic from his 1976 book Computer Power and Human Reason expresses his viewpoint elegantly:
Computers can make judicial decisions, computers can make psychiatric judgments. They can flip coins in much more sophisticated ways than can the most patient human being. The point is that they ought not be given such tasks. They may even be able to arrive at “correct” decisions in some cases—but always and necessarily on bases no human being should be willing to accept. There have been many debates on “Computers and Mind.” What I conclude here is that the relevant issues are neither technological nor even mathematical; they are ethical. They cannot be settled by asking questions beginning with “can.” The limits of the applicability of computers are ultimately statable only in terms of oughts. What emerges as the most elementary insight is that, since we do not now have any ways of making computers wise, we ought not now to give computers tasks that demand wisdom.
The same, it occurs to me, could be said of ourselves.
A Jacobin article by Leif Weatherby from January 25, 2024 illustrates the point -- which should not be controversial to anyone in the world of Trump's America in 2026-- that when it comes to intelligence, creativity and wisdom, humanity leaves a lot to be desired.
What stands between Plato and Orwell is culture, and with the rise of generative AI, we have a culture problem. The general intellectual bafflement of 2023 isn’t just a Chomsky problem. It’s a tendency that we have to underestimate culture even when it’s the very thing giving us fits. We may want to believe that “human creativity” — a constant refrain in Chomsky’s writing — isn’t susceptible to statistical techniques. But while it makes sense to reserve judgment on the avant-garde, I think it’s clear that Taylor Swift really could be an AI, partly because hyperproduced media products like her music or Marvel films find a kind of statistical center in the vastness of culture — which is exactly what generative AI does. There’s an unfathomably huge scale of human language production that comes between any formal linguistics and the types of danger that Chomsky and Kissinger both diagnose. GPT systems simply reveal that scale, and we do not like the results. But we cannot afford to ignore them.
The thing in the article that was titillating me is in another quote from the same piece: “The propaganda machines that Chomsky thinks manufacture consent are now close to one-hundred-percent AI-driven.” I don’t know if that’s true exactly but if it is it’s a little blood chilling. What makes it plausible to me is this haunting thought I have had that if AI could generate bad fiction let’s say that you could not detect from bad human generated fiction (of which there are plenty of examples), then what does that say about fiction? A machine could do it. Big deal you might say but what really haunts me is among the forces that shape our economics, our politics, our media— what part of that is not basically shit that a machine couldn’t do?
We don’t need machines to get politics that a machine could do. We do the crappy politics ourselves. So what does it matter if a machine is doing the crappy politics instead?
Before we hand over the reins of civilization to the intelligence that we created, we would do well to acquire some of it ourselves.
No comments:
Post a Comment