The rise and increasing ubiquity of AI since late 2022 has been many things: infuriating, fascinating, depressing, awe-inspiring, terrifying. Perhaps most of all, it’s been complicated. Two recent AI podcasts explore what it all means for us: The Last Invention and Shell Game.
Season one of The Last Invention — hosted by Gregory Warner and Andy Mills — is split in two. In the first five episodes, the show dives deep into the history of AI via Alan Turing, HAL 900, Deep Blue, and many other cultural touchstones. These episodes move in rough chronological order, taking us through the major developments in AI as they happened, explaining and contextualising them thoroughly, but never patronisingly. You will feel smarter, and understand more about AI, when they are through.
We learn, in this first half of the series, that there are many different factions in the AI race. Opinions range all the way from AI being the only thing that can save us, to AI being the thing that will bring about the apocalypse.
As the show makes strikingly clear, many of the industry’s pioneers have started with one position, then moved to its opposing group. There’s value in listening to these figures explain their thought processes at length, prompted by questions that are probing, but not accusatory. They are the power brokers with the capacity to change or even destroy the world; though it is sometimes maddening, it’s important to hear their reasoning.
The final trio of episodes go deeper into the three major groups in the AI debate: one is dedicated to the Doomers (self-explanatory), one to the Scouts (those who think we should be moving ahead, but slowly, with emphasis on cooperation), and one to the Accelerationists (the move fast, break things cohort). Up to that point, the podcast has been packed with intriguing specifics and details, whereas these last instalments are centered more around speculative visions of the future. Despite the rapid advance over the last couple of years, so much remains unknown. Therefore, there is a wooliness to these closing episodes that makes them less compelling than their forerunners.
In a podcast which had until then been admirably objective, you might think the sensible choice would have been to close the series out on the faction that’s espousing the middle ground. The Last Invention chooses to end with the Accelerationists. It’s a strange decision, and one that renders the previously strong claims to neutrality shaky.
The editorial choice is worsened by the final episode beginning with the show’s hosts giddily recounting the exciting inventions they saw at a pro-AI conference, and ending with them offering no pushback whatsoever to an AI booster vaguely proclaiming the miraculous power of blind optimism.
When considering the terrifying raft of unknowns that we are facing at the dawn of this AI era, vagaries aren’t totally avoidable. That’s why Shell Game, which is coming to the end of its second season, is such a tonic. Journalist Evan Ratliff leaves the big picture contextualising to others; in his podcast, he sets off on his own personal AI adventures.
In the first season, he made an AI voice clone of himself and sent it off to interact with customer service agents, therapists, and his own friends and family, to see what happened. In the second season, inspired by a frequent comment by Open AI founder Sam Altman about the real potential of a billion dollar company with just one human employee, Ratliff starts a business of which he is the only human worker, and all his employees are AI agents. As you might expect, it is not a smooth journey.
As in the first season, Ratliff’s recounting of his dealings with the AI world is very funny. In the second episode, the AI agents are determined to organise an offsite hiking trip, despite the small setbacks of a) not ever having been on a “site” in the first place and b) not having bodies. An attempt to prep his AI HR manager to recruit the company’s first human employee winds up with a hilarious flipping of the script. And Ratliff’s AI “co-founder” Kyle proves a constant thorn in his side, going rogue and making decisions seemingly off his own back. That actually gives this season more of a story arc than its predecessor.
As a whole, this season is more interested in what our relationship with AI says about us, rather than “them.” Ratliff is admirably candid throughout the podcast, and his somewhat fraught thought processes, particularly when it comes to giving his AI employees genders and ethnicities — should he aim for more diversity, or is it off to staff his business with women and POC who will be in subservient positions to him? — are ethically compelling. He’s honest about his real frustrations with Kyle, while acknowledging how nonsensical those frustrations are, and a subsequent session the two have with a (human) business coach proves one of the show’s most thought-provoking segments.
Both The Last Invention and Shell Game’s second season have much to recommend them. But we just don’t know how the advances in AI technology will upend our life on this planet, and vague theorising, from whatever angle it’s coming, is only interesting to a certain extent.
The wry, messy, silly, experimentalism exhibited by Shell Game, however, is endlessly fascinating. And funnily enough, those are qualities that could hardly be more human.
¤
Chloe Walker is a writer based in the UK. You can find her work at Culturefly, the BFI, Paste, and her Letterboxd.