Hacker News Reader

Comment by scrumper | Hacker News Reader

I'm curious why you did that? It's not a very complicated sequence. The whole point of engaging in a discussion here is to think about the issues raised and offer a point of view while incorporating other perspectives into yours. You've spent your money to bypass the whole intent of this site; akin to you being hungry then sending someone else to a restaurant for you so you can later read their review of the food.

EDIT and you of most of the commenters here, with your industry background, are better placed to offer an opinion!

View on HN

> It's not a very complicated sequence.

For me, it was. I have trouble forming a mental model of itineraries so I’m grateful for the summary.

Why? I saw no upside to reading a layman's media source and teasing out the sequence, which is all I wanted to know. I read NTSB reports for pleasure and my favorite YT creator is Blancolirio. There's an infintesimal chance that the story had any juicy details I cared about. Also, it was work hours and I really didn't have any more time to devote to topic. I was in a rush to get my take out there, since I am a legitimate SME.

I think you missed the point.

Presenting information in different ways is useful (and the method of display can offer informational insights itself). And for different users it might help parse larger connections. And by using the LLM to summarize just that one facet of the problem (itinerary and sequence) and sharing it here, they’ve contributed in a meaningful way. It may not have warranted a response. But it added to overall understanding of the problem space to help facilitate discussion. And they did well enough by citing that the info came from an LLM. They didn’t bypass the intent of the site. They added to it and fell right in line with that intent.

Okay that's fair.

I had no issue with the contribution itself, the route summary is helpful.

> And they did well enough by citing that the info came from an LLM.

In terms of acknowledging AI contribution, you're damned if you do, damned if you don't. Here it's sidetracked a discussion; but transparency is better than otherwise I suppose. Perhaps it just boils down to taste - and I don't like it.

Anyway, you made me think!

With all due respect, you are the one that sidetracked the discussion, not the person that acknowledged AI

You also made me think. I think the problem lies in the nature of “threaded” discussions. The information was useful, but it wasn’t necessarily a “reply” to any particular comment. But, the design here leaves no other choice when adding it. It maybe doesn’t make sense as a top-level comment either. It’s almost like we need to have…side comments? Asides? Like a place to just drop in relevant tidbits and informational helpers rather than a true “comment”. Maybe it’s as simple as having a means to turn off replies to signify that a contribution was an “relevant aside”.

But yes, in a perfectly meta way, I think we’ve both thought more about the nature and design of web comments and information and the implications of LLM-assisted chat ‘augmentation’. (All those AI summaries forced upon us in every site is probably the worse implementation of this!)

Reply on news.ycombinator.com