[ExI] Speed Reading ChatGPT Output

[Adrian Tymes]

On Sat, Jan 6, 2024 at 11:21 PM Kelly Anderson via extropy-chat <
extropy-chat at lists.extropy.org> wrote:

> Has anyone besides me noticed that it is cognitively easier to speed
> read AND STILL comprehend the output of ChatGPT than it is to read and
> comprehend human generated text? Because it is generated by a
> predictive algorithm, would this be predictable? Or, is it just my
> imagination?
>

I think you are on to something here.  It's not just because it's generated
by a predictive algorithm.  The output is more often summarized by
paragraph: one can read the first few words or sentence of each paragraph
in the middle and have a good understanding of the main point without
parsing the rest of that paragraph, especially when it bolds or capitalizes
them as headers.  This style can be imitated by humans, but ChatGPT more
often actually does it, akin to how a factory robot has more consistent
output than a human factory worker.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20240107/b140ad59/attachment.htm>