Discussion about this post

User's avatar
Nirav Bhatt's avatar

As always, an intensely thought provoking post, Terry!

As someone who interacts with GenAI for the purposes of programming problems, I am accustomed to the fact that text proximity is the backbone of GenAI's effectiveness.

In other words: One's likelihood of producing human-like text soley depends upon the vocabulary diversity.

To give an example: A novice writer is more likely to describe a dream using adjectives: "vivid" and "lucid". Because of their vast availability in public content, GenAI is likely to use them generously, too.

The novice writer's content has a higher chances of being flagged as AI-generated. More so, if she relies upon GenAI in doing her writing exercises.

To drive home the point: While it will be intensely valuable to write about one's unique conscious experiences using an authentic voice, it is equally important to keep in mind the tactical details to evade the gaping mouths of the word-proximity bots.

This is, until popular AI levels up to focus on improved comprehension rather than lifeless text. It's called inference based AI. Verdict on this isn't out yet.

Expand full comment
Nick Potkalitsky's avatar

This is it, Terry!!! Sign me up!!!

Expand full comment
1 more comment...

No posts