Hacker News new | ask | show | jobs
by furyofantares 103 days ago
The aesethetic tells are excessive headings for small sections, many of which are largely just a list. There's the numbered list but then the sections that are a sequence of "Bold Text: Single sentence." Plenty of em-dash too, but I don't really look at that much.

It's basically a long list of factual statements all with the same weight, very little opinion expressed about the experience. I don't actually mind an infodump from a human, you usually can still work out some of what they care about, and you also can be reasonably sure they didn't fill in the gaps. Most likely I would have enjoyed whatever the author fed into the LLM.

2 comments

Speaking as someone who has used em-dashes and semicolons for years, it’s a shame what the AI companies have done with punctuation.
Yeah we’re in a weird place where the prompt would still have to contain most of the content and then be “prettied up”. Since generic AI “prettied up” style is so same-y and dry it would be refreshing to just see whatever the original was, warts and all. Just blog-post the prompt!