We're nearing the precipice of more natural human-computer interaction that will need to rethink the interfaces and conventions.
Alexa and Siri seem like Model T Fords when there's a jet aircraft flying overhead. I'm thinking these agents need to be replaced by more natural agents who can co-create a language with their human counterparts rather than relying on fixed, awkward, and sometimes unhelpful commands. It would behoove us to expose APIs and permissions delegation in a more consistent and self-describing (OpenAPI + OAuth / SAML possibly) manner for all possible services one would wish to grant to an agent. If a natural language agent is uncertain, it should ask for clarification. And on results, it is necessary to capture ever-more-precise feedback from users because positive and negative prompts aren't good enough.
> “The way it functioned was very interesting. When the Drink button was pressed it made an instant but highly detailed examination of the subject's taste buds, a spectroscopic analysis of the subject's metabolism and then sent tiny experimental signals down the neural pathways to the taste centers of the subject's brain to see what was likely to go down well. However, no one knew quite why it did this because it invariably delivered a cupful of liquid that was almost, but not quite, entirely unlike tea.”
I think Douglas Adams was closer to the truth on the subject of AIs and tea. I don't want to be overly cynical but I suspect we'll just get used to saying "OK, close enough" when dealing with LLMs.
We're nearing the precipice of more natural human-computer interaction that will need to rethink the interfaces and conventions.
Alexa and Siri seem like Model T Fords when there's a jet aircraft flying overhead. I'm thinking these agents need to be replaced by more natural agents who can co-create a language with their human counterparts rather than relying on fixed, awkward, and sometimes unhelpful commands. It would behoove us to expose APIs and permissions delegation in a more consistent and self-describing (OpenAPI + OAuth / SAML possibly) manner for all possible services one would wish to grant to an agent. If a natural language agent is uncertain, it should ask for clarification. And on results, it is necessary to capture ever-more-precise feedback from users because positive and negative prompts aren't good enough.