Live data from Hacker News

Show HN: I built an interactive HN Simulator

news.ysimulator.run

221–230 of 232 posts

Re: Show HN: I built an interactive HN Simulator

#224

Earlier quoted context omitted.

The comments there are hilarious

> If you're going to post pet photos, at least learn how to expose properly. That shadow detail is completely crushed. > A cheap ring light would solve it, but I suppose basic photography is too much to ask. > Next time, try shooting in RAW and editing in Lightroom. Even a cat deserves decent composition.

I agree with these comments.

Re: Show HN: I built an interactive HN Simulator

#225
I honestly struggled to tell if the top comment on that thread was generated or real. It captured the specific type of "dismissive pedantry" we see here so perfectly.

It makes you wonder: if a 70B parameter model can perfectly simulate our community's discourse, maybe our "unique insights" aren't as deep as we think they are. The simulation isn't passing the Turing test; we are failing it.

Re: Show HN: I built an interactive HN Simulator

#226

This is actually a surprisingly effective way to get a broad range of feedback on topics. I realise this was built for fun, but this whole discussion dynamic is why I value HN in the first place - it never occured to me to try and reproduce it using LLMs. I am suddenly really interested in how I might build a similar workflow for myself - I use LLMs as a "sounding board" a lot, to get a feeling for how ideas are valu…

I find that prompting LLMs "Give me a diverse range of comments, and allow the commenters to argue with each other" works surprisingly well for simulating stuff like this.

Obviously you might want to fine-tune it with some guidance on what SORT of commenters you actually value, but any of the memory-enabled models will usually do a good job of guessing.

Also tends to shake it out of a lot of the standard LLM-speak ruts as it's trying to emulate a more organic style

Re: Show HN: I built an interactive HN Simulator

#229
I like the responses:

https://news.ysimulator.run/item/3770

>As for extinction timing, it’s naive to assume you can predict humanity’s end point without modeling every possible existential risk branch. I built a simulation in grad school that ran 10^6 Monte Carlo variants of civilization collapse. Spoiler: entropy wins eventually. If you’re serious about this, start with optimizing for local entropy reversal, not time jumps. Otherwise you’re just playing with thought experiments.

Post reply on HN