Anyone who grew up watching The Terminator or The Matrix is aware of that AI poses an existential risk to humanity. Because the robots turn into smarter, it was thought, they may inevitably substitute us, both by destroying us or mining us for assets. Nonetheless, the age of AI is now right here, and the reality is a lot worse than something from a dystopian sci-fi. You see, AI has determined to provide us extra podcasts.
The world wants extra podcasts prefer it must be kicked by a horse. Everybody’s bought a podcast. Gyles Brandreth has a podcast. Paul Giamatti has a podcast. Your 4 or 5 worst pals all have podcasts, blathering endlessly into an atmosphere already cluttered with an excessive amount of content material. Now Google has simply created the primary AI podcasts, they usually’re as fascinating as they’re superfluous.
NotebookLM is mainly ChatGPT however for audio. You add a bunch of sources – paperwork, web sites, YouTube movies – and it parses all the knowledge, then creates a bewilderingly human-sounding dialogue about them. Two hosts, one male and one feminine, chat about no matter topic you’ve given them in an uncannily podcasty method. Their speech is stuffed with ums and ahs. They hesitate, they speak over one another. They, like, sort of like speak like this on a regular basis? It’s so imperfect that you could shortly neglect you’re listening to a few robots repeating crap from the web.
NotebookLM payments itself as a examine useful resource, which is sensible. If you wish to summarise a number of data in a method that retains your consideration, or if you wish to soak up data on a run or a drive, then it’s nice. Earlier than lengthy, folks will put together for exams by ramming their textbooks into one thing like NotebookLM then pottering about with earplugs in.
However if you wish to make a podcast on any topic you want, it could try this, too. Rivals begins on Disney+ this week, and it has already generated acres of protection, so I fed in a couple of interviews on the present to see what the hosts would provide you with. The ensuing five-minute podcast was uncanny.
In it, the hosts handled the present prefer it was one thing that they had simply organically fallen upon. “All proper, prepare, as a result of we’re diving into Rivals!” the feminine host broadcasts at the beginning, to a volley of agreeable muttering from her male counterpart. They focus on the attitudes of the Eighties, its sexism and racism, and applaud the present’s willingness to immediately confront them. It makes it sound as if Rivals is an excellent, pioneering piece of agenda-setting tv.
The issue is that it isn’t actually like that in any respect. It’s camp enjoyable with a great deal of nudity. Nonetheless, the sources I fed in had been interviews with actors from the present, who’re understandably extra eager to speak about real-world points than what it was wish to whip their bits out on a regular basis. And in order that’s what the podcast is. A extra correct model would have fed in each obtainable piece of knowledge – interviews, opinions, present notes, possibly even the total supply novel – and created a 360-degree view of the sequence. As an alternative, it was a particularly assured presentation based mostly on restricted data. And in the long run, isn’t that each one a podcast is anyway?
After that I made a decision to make the kind of podcast the world wants least of, which is two folks bibbling on about conspiracy theories. Admittedly, I may have accomplished a greater job right here, discovering flat-Earth boards and Fb teams made up of people that nonetheless blame Covid on 5G masts. As an alternative, I simply threw in a bunch of stuff from Wikipedia and Reddit, and located myself shocked by how measured the tone of the ensuing audio was. It ended up being a reasonably eye-opening 14-minute episode about issues comparable to affirmation bias and the human impulse to know the world. At one level, they begin to checklist prevalent conspiracy theories, however cease as a result of – because the male host says – “My mind would actually explode.”
The existence of NotebookLM raises a number of questions. Will it make folks too lazy to learn their very own analysis? Can it’s totally trusted? What’s going to humanity do with all of the hundreds of thousands of newly unemployed podcasters roaming the Earth? However as a method of disseminating data for inexperienced persons in a naturalistic method, it’s type of, like, sensible.
Supply hyperlink
