Straight to the point

The spotlight beams down on the stage, illuminating the faces of the evening’s celebrities, who are seated in a semicircle like ancient oracles of comedy. At the centre of it all is the roastee — the guest of honour — who will soon be subjected to a brutal barrage of jokes. “James Franco…” started Natasha Leggero’s demolition job 10 years ago. “Acting, teaching, directing, writing, producing, photography, soundtracks, editing — is there anything you can do?” Then there was Gilbert Gottfried’s audacious pivot at the Hugh Hefner Roast, held just two weeks after 9/11, which descended into the “filthiest joke ever told”.

But as much as these moments underscore the art form’s past audacity, they also highlight the pallor that has settled over it. Once a platform for such biting wits as Don Rickles and Joan Rivers, the comedy roast has become a sanitised ritual, a showcase of quips that hardly go beyond the skin. Perhaps the last time we heard a roast joke that truly shocked was in 2019, when Blake Griffin took the mic to thank Caitlyn Jenner “on behalf of black men everywhere” for giving her daughters “daddy issues”. Since then, roasts have transitioned into an assembly line of safe, formulaic jokes that don’t even scratch the surface.

Yet looking at the roast’s decline, perhaps it’s understandable that nowhere in the entertainment world is the existential crisis over the rise of AI more palpable than in comedy. After all, there is no field of creative endeavour that’s become more dependent on cliches, groupthink and repetition. But while the malaise is widespread, the roast, in particular, looks set to be an early casualty.

Consider the recent “roast battle” between a human comedian and an AI counterpart. Earlier this year, Matt Maran, a comic from Queens, faced off against a ChatGPT-powered version of Sarah Silverman. Maran faltered early, attempting several jokes that clearly misfired. In contrast, the AI dished out jokes without so much as stopping for a sip of water. Though the AI was unflappably corny — “you’re as edgy as a butterknife,” it told Maran — it won the roast by ceaselessly throwing cliched barbs that were more agile and responsive to Maran’s hit-or-miss material. (Perhaps the AI’s biggest hit: “You’re from Long Island and you lost your virginity to a prostitute… talk about starting from the bottom and staying there.”)

Yet rather than serve as another warning of a Skynet-like robot takeover, this episode was more an indictment of human comedians’ inability to excite. Look past the hype, and ChatGPT stands as a glaring example of AI’s limitations: its output may be rapid and cost-effective, but its algorithms still can’t replicate the spark of human wit or the depth of human emotion. It certainly can’t capture the true essence of laughter: what French philosopher Henri Bergson described as “a slight revolt on the surface of social life”. In his view, comedy is not just about robotically eliciting a laugh; it’s a social endeavour. The emotional connection between the human comedian and his human audience isn’t just beneficial — it’s essential for the success of the joke.

Ironically, then, it’s almost as if AI were precisely calibrated for churning out the type of lame, uninspiring content that now defines roasts as well as late-night comedy — a genre that has long suffered from a lack of originality; jokes about SNL being “Saturday Night Dead” have made the rounds almost since the show premiered in the late Seventies. Given one or two fast-working hack writers to oversee its output, ChatGPT could easily pump out entire seasons of it.

And perhaps it already is. Joe Toplyn, for instance, a former writer for David Letterman, has ventured into this uncharted territory with his bot, Witscript, which takes topical headlines or themes and generates what it believes to be the “best” joke, leveraging algorithms to optimise for elements such as timing and punchline effectiveness. The output is beyond cornball: in response to a user prompting it to make a joke about the Bidens’ aggressive German Shepherd, it wrote: “Sounds like he was more of a BITE-en than a Biden.” Yet for all their banality, Witscript’s jokes are merely a reflection of the modern comedy circuit.

By contrast, genuinely artistic comedy will always thrive on a certain rebellious spirit — a willingness to stake out and work on that edge. This is the comedy of David Letterman in his early years at NBC, of Bob Odenkirk and David Cross at HBO’s Mr. Show in the Nineties, of contemporary comedians such as Sam Hyde and Nick Rochefort of Million Dollar Extreme. This is the comedy that doesn’t just elicit laughter, but also provokes thought.

Recall that last show’s most unforgettable sketch, in which Rochefort portrays a burnt-out teacher so convincingly that it is almost uncomfortable to watch. His character wrestles with self-loathing and moral apathy, manifesting a bleak internal monologue that touches on career failures, his complex sexuality and an eagerness to escape his dismal reality. This form of humour is deeply unsettling — it ends with his imagined suicide set to rock music — and yet groundbreaking, the likes of which AI can’t come close to replicating. Here, then, is an important clue for saving the traditional comedy roast from automation: re-embracing the raw, unfiltered humour that made roasts a staple of American comedy in the first place.

Yet, if we are to be brutally honest — in the true spirit of the roast — the odds of such a revival seem extraordinarily slim. From economic pressures that favour risk-averse humour to a polarised comedic landscape that leaves little room for edgy jokes, the variables are stacked against a revival. The recent Writers Guild of America strike exemplifies the trend, with the industry’s creatives clamouring for a bigger slice of an increasingly bland pie — temporary protection from the bots — rather than attempting to surpass them by creating great art. As a result, any hope of reigniting the fire in the roast seems less like a potential future and more like the fading light of a bygone era.

view comments

Source: UnHerd Read the original article here: https://unherd.com/