Where are all the AI converations headed?
Currently, like cars on an unmarked highway five lanes wide, conversations involving AI tend to drift between lanes. At any point, the conversational thread weaves into the various lanes: scientific, technical, #AI_Ethics, political, and the aspirational futures of humankind.
At this point, technologists’ claim the wheel. They steer the conversation because, well “It’s my car!”
With one hand rewiring the navigation system and the other thumbing through a college prep handbook on utilitarianism, technologists own all five lanes, even those where they have no specialized expertise in traversing.
Meandering out of your lane while driving is… fine-ish. If people know what they’re doing and the speeds and stakes aren’t too high. Live and let live I say. Afterall, a front left panel with scraped paint isn’t going to kill anybody.
Speeds and stakes: Accelerating and existential
In the case of AI, however, few are arguing the speeds and stakes of AI are low. To the contrary, the operative watchwords in that context are acceleration and existential. At their nexus is the fuzzy, magical idea of Accelerationism, a grouping of intense phenomena – technological change, peak capitalism, racial conflict, and the singularity – not simply destabilizing the world-as-we-know-it, but deconstructing it and wiping the slate clean. With that in mind, the use of the word existential as table stakes to enter the AI conversation clearly makes sense:
“Howdy pardner, we’re talkin’ AI? Deal me in, here’s my table stakes …AI is existential.”
The word existential is always tossed onto the table in conversations about AI’s stakes. From what I can recall, the word has gotten more mileage in the past 18 months than it did even in the bong-fueled dorm room sessions about nuclear annihilation cascading from 1983 to late 1984, during the height of the Cold War. (But don’t hold me to it. Knowledge and memory tend to get anchored to a bonged state of mind, aka ‘state-bound knowledge’. But, I digress.)
Back to our analogy about AI conversations as careening cars, which we can now clearly see are accelerating wildly in a life-or-death race driven by the tech giants. The tech giants who say “Obviously, we’re in a life or death race with each other, and btw with China and Russia. But, no worries, you’re spectators. You are exempt from the existential stakes (especially if we win the race).”
The tech giants then point to regulators who are frantically building a chain link crash fence meant for race cars even as they notice the racing vehicles look increasingly like an Air Force general’s wet dream. These all purpose jets have horizontal lift, cruise at Mach 3 (2151 mph), and are covered in racks for attaching specialized weapons and gear. No worries, the tech giants whisper, sit back, grab a beer and enjoy the race!”
Everything is not fine.
The lanes are there, but aren’t well marked. Moreover, the drivers who originated in the technical lane are supported by drivers from the scientific lane and are speeding up quickly.
Getting home to the spouse and younguns, no doubt. Isn’t that nice? Yes, but.. they are weaving into others where they have no experience or expertise.
In fact, they’re acting like they own the road – they’re hogging it. They’re veering into conversational lanes originally devoted to drivers whose dispositions – expertise, experience, worldview etc. – primarily stem from ethical, political, and aspirational knowledge and expertise. Lawyers, doctors, priests, shamans, church volunteers, truck drivers, caddies…everyone.
Not because they should – clearly the technologists’ interest in the other lanes is based on their technologist-based dispositions and in service to the interests ($$$) of their own lane – but because they’re being allowed to.
In essence, the current AI conversational corollary to a rule-of-the-road looks like ‘He who builds the vehicle owns the lanes in the road also’.
The AI conversation is disguised as a vehicle in a shared conversation on a shared road.
Whether they know it or not, it is actually more of a monologue and a declaration by technological triumphalists building the accelerationists’ Mad Max-ian jet. Like the chrome paint huffing from the Mad Max movies, the news about Nvidia computational increases and scaling of LLMs adds further to AI technologists’ disassociation from fellow travelers (cue the dysopian heavy metal music).
Beep! Beep! (Honda 2005 CRV honking noise).
The AI conversation, in short, needs other voices from other lanes.
Unfortunately, AI technological triumphalists don’t have chrome painted teeth and aren’t identifiable in that way. However, when the AI pontification floats into lanes in which the speakers have scant dispositional expertise consider it a sign – the paint fumes are rising and its time to get off at the next exit.
At minimum, AI panel discussions about AI should include experts from at least one of the other lanes.
