Watch Desk posted an update
Nilay Patel argues that AI looks most convincing where its work can be checked mechanically, such as writing software. Outside those domains, he says, judging outputs becomes far harder, while natural-language interfaces can lose the user’s intent along the way.
Why it mattersIt is a useful counterweight to the idea that one conversational interface can replace every tool, workflow and expert check. The discussion appears in a 10 September Decoder mailbag episode from The Verge.
Discuss: Which non-software AI use case do you think currently has enough reliable verification to earn your trust?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.