Yes, the provocation is the point, but still — it’s hard not to be lured by the bait. Quite the kerfuffle was whipped up over the past month after Dudesy, a comedy podcast and YouTube show hosted by Will Sasso and Chad Kultgen, published an hour-long special titled George Carlin: I’m Glad I’m Dead!, which the show purports to have created purely from generative AI technology that’s been trained on the late comedian’s body of work.
The stunt hit the news cycle after Carlin’s daughter, Kelly, criticized the effort on social media. The Carlin estate later filed a lawsuit against the podcast, which seeks to prohibit the Dudesy duo from using Carlin’s copyright material to create future “specials†and to compel them to “destroy any video or audio copies†of I’m Glad I’m Dead!
This section of the lawsuit, which moves to cut off a free speech or parody defense, stands out: “The Dudesy special has no comedic or creative value absent its self-proclaimed connection with George Carlin. It does not, for example, satirize Carlin as a performer or offer an independent critique of society.†It’s kind of a funny burn via legalese, though one I’m afraid I can’t fully vet for myself. The video, which was released on YouTube, was made private after news of the legal action went public. Reflecting the generally porous and wanting nature of platform safeguards, you can still find rips of the thing on YouTube if you look around enough, but I’m reticent to do so. As such, my exposure to the actual work is limited. I sampled the video out of curiosity when it initially dropped to get a sense of what the hubbub was about. The video was composed of the purportedly AI-generated Carlin audio track — which sounds like a plausible copy in brief spurts — overlaid with a succession of images meant to accompany the topics of the routine; they all possessed the strange, curvy, and hell-house style that defines everything a visual generative AI system like DALL-E will churn out in its current public iteration. After a few minutes, I checked out. I got bored.
This isn’t the first time Dudesy has pulled this particular stunt. Last April, Tom Brady threatened to sue the show after it similarly published an hour-long fake Brady comedy special that’s framed as a creation of artificial intelligence trained on recordings, footage, and data produced by the NFL megastar, but that threat didn’t appear to amount to anything. It’s also worth noting that garish attention-grabbing stunts like these are par for the course for Kultgen in particular, who, in 2015, conjured an elaborate hoax where he digitally posed as a woman threatening to have an abortion unless she could raise $1 million from pro-life supporters as means to promote his novel about a woman who commits a similar scheme.
Anyway, all of these antics are broadly in line with the overarching Dudesy schtick. The show launched about two years ago with a superficially grabby premise, billing itself as what you get when two dudes allow an AI to access wide swathes of their digital data for the purposes of generating episode prompts and conceits. It’s a funny premise, even if it isn’t completely legit. It’s never made explicitly clear to what extent artificial intelligence actually factors into the design of the podcast and to what extent it’s just a bit; what’s supposed to be the Dudesy AI is incorporated in the form of a disembodied voice that’s clearly some guy reading out a script with a half-assed AI voice. Of course, that hand-waving ambiguousness is the joke of the podcast, which ultimately conforms to the podcast ür-shape of two dudes shooting the shit (see also: How Long Gone, Hollywood Handbook). (The Dudesy team declined to comment for this article.)
For the most part, Sasso and Kultgen do commit to the kayfabe of it all. It shouldn’t surprise you, then, that both hosts are noted wrestling heads, Sasso in particular. You can sometimes find him on the YouTube version of the show cradling a Dudesy-branded championship belt, as he does in an episode of the show released after the I’m Glad I’m Dead! drop where, presumably underlining the intent of the stunt, the two men discuss the finer points of generative AI and creative production within the context of the pseudo-Carlin special. Kultgen is the more AI-pilled of the two, usually assuming the technobro stance of how this technology is gloriously inevitable. Sasso not so much, who ruminates on the surreal lifelessness of the performance. “Parlor trick,†he called it. “Fuckin’ parlor trick.â€
Amidst this particular Carlin brouhaha, the show finally made a minor public concession earlier this week. In a development that should shock no one, a rep for Sasso told the New York Times that the AI central to Dudesy’s conceit is not actually an AI, but a “fictional podcast character created by two human beings.†The rep also noted that the Carlin special was in fact written by Kultgen himself, though they have apparently yet to clarify whether the synthetic Carlin voice heard on the special was produced by AI technology. (But I mean, come on.) Responding to the concession, the Carlin estate says it’s insisting on plowing ahead with the suit nevertheless on the grounds that they “don’t know what they’re saying to be true†until they’re deposed in court. This legal sortie may or may not go anywhere, but if it does persist it could set up a possible situation where the joke may very well backfire on the Dudesy duo — or simply spirals to grow even bigger and stranger. Frankly, both outcomes seem equally plausible.
Again, the provocation is the point, and while the actual artificial intelligence involved is somewhere between fully and partly simulated, the feelings generated by this entire ordeal are very much real. As gaudy as the stunt is, it felt especially agitating because it explicitly seized on a sacred cow: Carlin, a comedian whose deal was largely defined by a fierce advocacy for lived authenticity and who’s probably among the most painful subjects one could inflict with digital necromancy. There’s a reason this feels unambiguously disconcerting compared to, say, Who? Weekly’s dupes of the very much still-alive Sydney Sweeney and Glen Powell, done to illustrate the uncanniness of AI creations.
The Dudesy-Carlin affair fits neatly into a cluster of ongoing stories concerning public figures and deep fakes. Right now, there is the horrible stuff that’s happening with Taylor Swift and the mass YouTube celebrity deep-fake scams, an unsettling phenomenon first reported on by 404 Media. More broadly, there are the big-picture questions surrounding the use and ownership of digital likenesses in the media and entertainment space. Beyond that, there are growing concerns over AI-generated misinformation within the context of the upcoming presidential election. But these stories are just the tip of the iceberg.
Part of what I’ve found so frustrating about mainstream coverage of artificial intelligence is how it seems predominantly driven by transgressions experienced by public figures — celebrities, politicians, and so on. These are instances where the conflict between people and artificial intelligence do get identified and litigated because they’re loud, obnoxious, and splashy, where the vast majority of that conflict will play out will actually be invisible: quieter, subtle, and structural changes that affect the rest of us, people who nobody seems to want to read about. I suppose you have to start this public conversation somehow. If dumb Dudesy stunts are how that can happen, I’ll begrudgingly take the bait.
Another view …
I asked our resident comedy reporter, Herschel Pandya, what he thinks about the matter. He writes:
“There’s a tendency on the part of some commenters to reflexively dismiss the merits of any media generated by AI. From where I stand, this is wrongheaded. For all its flaws, the technology is only going to get better. The question we should be asking isn’t whether it’s good; it’s whether this is what we want from art. And with that in mind, the first time I watched Dudesy’s ‘special’ George Carlin: I’m Glad I’m Dead!, I was not entirely unimpressed. It struggles, as you’d expect, to tie together disparate ideas in a surprising manner befitting of good punch lines, but it’s worth noting that Carlin’s later specials (while better than this) were similarly better at eliciting applause than laughter. Its timing is poor — a dual byproduct of the current limitations of AI-voice generation technology, and not having had the opportunity to of workshop this material in front of an audience to refine delivery — but it’s also unreasonable to expect a crudely illustrated YouTube video to possess the same gravitas as living George Carlin. What it is is a starting point — a script waiting to be punched up by a skilled comedian to bring it up to task. It’s not a revelation, but it’s also not nothing.
But to be clear, these were my thoughts before the big reveal that the ‘special’ was actually written in full by Kultgen. I had my suspicions, as I did when Dudesy put out its ‘special’ written from the perspective of Tom Brady in 2023, but now that the show’s creators have confirmed this, I no longer feel obligated to grade it on a curve. This sucks. If Kultgen wrote it hoping people would be blown away by the fidelity of his fan-fiction, he underestimated how much work it needed to be ready for consumption. If he intentionally wrote something of middling quality to deepen the debate of how much AI actually factored into its creation — a debate the podcast’s listeners have been engaging in since the show’s inception — he succeeded, but gravely miscalculated the context collapse it would undergo when consumed by the rest of the internet. If it was just a provocation to stir up anger, he may as well just have lifted it from ChatGPT. It’s all a bit like a stunt Nathan Fielder might concoct, if he were rendered by AI.â€
News and Notes
➽ Valley Heat, the very good and very funny fiction podcast about a freelance insurance agent who scopes out his neighborhood in L.A., has joined Maximum Fun and will now be producing consistent episodes. Sweet.
➽ Sizzle sizzle: Sohla and Ham El-Waylly have a new podcast, Deep Dish, that’s releasing through The Sporkful feed where they plumb into the history of a specific dish.
➽ Planning to spend some time this week checking out the new season of Blindspot, “The Plague in the Shadows,†that digs into the less-documented aspects of the HIV and AIDS epidemic.
➽ I’m doing my usual thing this time of year where I catch up on a few podcast archives for the Oscars: The Big Picture and Little Gold Men, of course, and I need to queue up the “Class of 2023†ep of This Had Oscar Buzz at some point.
➽ Speaking of Oscar news: The Ankler is partnering up with LAist/Southern California Public Radio to build a brand new entertainment-industry show for the radio. A sort of answer to KCRW’s The Business, I suppose.
➽ Quick correction to last week’s blurb on In the Dark: “The Runaway Princesses†is not the long-awaited third season of the podcast, but a miniseries meant to tide us over for the time being. No word yet on when that season will actually drop, sadly.
➽ Smartless is moving from Amazon to SiriusXM under a $100 million multiyear deal, according to Bloomberg. Previously a cornerstone of Amazon Music, I imagine Smartless will now fit perfectly into the SiriusXM machine. The latter is cleaning up.
➽ NPR has a new CEO: Katherine Maher, formerly the chief executive of the Wikimedia Foundation and, most recently, Web Summit. She had actually just taken over leadership of the latter in October after the previous Web Summit CEO, Paddy Cosgrave, stepped down amid controversy; she’ll step away from that position on March 1. A surprising choice, frankly, given Maher’s lack of news and journalism experience — but these are particularly strange and fraught times for the media business. Perhaps an unconventional choice is needed.