After a weekend when an Anthropic researcher quit and warned AI could wipe out humanity within a decade, Sarah and Beth take on the questions listeners sent them about artificial intelligence, starting with what is actually being made and what it costs. They argue that the flyers-and-water critique is stuck in 2022 while frontier models and agents race ahead, and they take the Doomer scenarios seriously without buying that the doom is just sales copy. Beth explains why "can't we just unplug it" would mean unplugging the internet; Sarah argues nobody trained these models on a soul, and that's where...
No transcript for this episode yet
You can still listen to it in the player. On episodes with a transcript, you can tap any word to see its meaning.
