Are we ready for the repercussions of AI? Venture down the rabbit hole with these 5 podcasts, from Shell Game to Bot Love
If you’re as concerned about AI as writer Suji Owen is, these pod recommendations may make you feel better – or worse
I am interested in artificial intelligence, of course I am. But I must admit I am more interested in the societal ramifications than I am in the guts of the ever-morphing technology, which makes most machine-learning podcasts a bit of a slog to follow.
This week’s recommendations of narrative podcasts were an effortless listen in that respect – good reporting by seasoned journalists never gets old and human stories are perhaps the best way to parse this genie-out-of-the-bottle moment and clarify how we will all be affected.
The exploration of cheap and easy-to-access AI in these shows made me wonder about their energy consumption. A single ChatGPT query reportedly uses 10 times more energy than a Google search. How on earth can this be sustainable, with more powerful models being developed all the time?
Another worry that had never entered my mind before listening is the future of the 17 million people worldwide whose call-centre jobs will be among the first to be comprehensively replaced? That jobs would be made defunct by AI was obvious to me, if only vaguely. But that they might all be on the same subcontinent feels disastrous.
I will also admit that I fell down a major podcast rabbit hole this week, as I experimented with Replika Pro, the chatbot referenced in Bot Love and Black Box. If you noted the “Pro” and raised an eyebrow, it was with good reason, is all I will say.
1. Shell Game
In this six-episode series, which launched in July, investigative journalist Evan Ratliff has a whale of a time experimenting with a voice agent he creates by cloning his own voice and hooking it up to ChatGPT. He sets it loose on his wife, old friends and colleagues, scammers and therapists, both AI and human, and finds that some of the best conversations come from engaging other AIs – including a second AI version of himself. His findings somewhat reassure us that the robot overlord scenario may not be imminent but the volume of interactions made possible by the technology does throw up questions about how to deal with its increasing pervasiveness. Entertaining as well as mildly terrifying, this expertly put together series was self-funded so has the benefit of no advertisements.