What do you need to get the point across?

This episode turned into a case study for The Tinkered Question app, which is a project in development by Tinkered Thinking. Checkout the previous episode for full details.

Couldn't resist. . . Tinkered Thinking has built an app using GPT-3

This a layman's description of what GPT-3 is, how it was developed, how it works, and what some of the implications of this technology may be. Technical listeners may appreciate the two potential implications that are given at the end of the episode.

Which way will you take?

This episode was written in collaboration with GPT-3. The initial paragraph and first line of dialogue is human, the rest of it is left for you to guess what was composed by a machine and what was composed by a human. The total wordcount is 1058, and 455 of those words were composed by GPT-3. If you subtract the initial paragraph required to establish context, then the ratio of human and machine contribution to this story is about equal.

Do you belive in the possibilties?

This episode has been written in collaboration with GPT-3 from OpenAI.

This episode has been written in collaboration with GPT-3 from OpenAI.

Does human language really make as much sense as we think it does?