I'm Jonathan and I built Ask Hacker Search (https://hackersearch.net/ask), an LLM-powered version of Hacker News' Ask HN.
Unlike Ask HN, Ask Hacker Search doesn't solicit new contributions from HN readers. Instead, it leverages Hacker News' historical data to answer questions, and offers LLM-generated summaries of those. I've used it for questions like "Should I use Drizzle or Prisma?" or "What is a good screen capture that allows easy zooming effects on Mac?".
It is particularly useful when you're interested in understanding HN readers' sentiment about a topic, or when looking for expert insights on topics of interest to HN readers. I've been using it continually while building it, and have found it particularly useful to find software libraries recommended by HN or get quick vibe checks on hot topics.
This builds on my release of Hacker Search two weeks ago (https://news.ycombinator.com/item?id=40238509), which offered a semantic search engine over top HN submissions. It's not just a small upgrade: covering comments was the #1 requested feature after that launch, so I rebuilt the near entirety of the product to support that.
Please try it out and let me know what you think of it! I have to limit the number of LLM summaries each person can get for free, as this is entirely self-funded. If you hit the limit, you can subscribe for more summaries generated by a better model ($8/month), or bring your own compute by running inference on Ollama on your machine!
One thing I've been wanting for with search in general, especially LLM-powered, is having some kind of date relevance - especially with the fast moving world of technology.
For example I want to know how ProductX and ProductY compare. Last year ProductY didn't have FeatureZ, but they implemented and announced it last month. There might be several comments lamenting the lack of FeatureZ from 2 months ago, but they shouldn't be considered with the same weight now that it does exist.
I don't have any ideas for how this should be done but it's something I'd like to see tackled in RAG systems in general, and wanted to put it out there.
I could definitely use this as an alternative to hn algolia for some things but it’s going to be hard for me to remember. I’d recommend doing this as a browser extension or something so it’s not 100% lost in the void.
I tested it a bit and it seems pretty decent, although for some really niche theoretical questions it wasn't successful in retrieving the answers I wanted even if alot of the results were really good in other aspects. It could simply be because the answer is not available anywhere in hackernews.
I'm wondering if someone were to build a similar project but for other sites, what would your advice be? For instance what technical difficulties did you stumble on that you think would be good to be aware of?
Thanks in advance and once again congratulations on the project!
Personally I like that it doesn’t look like a cookie cutter 2024 templates site and has a more raw feel but I’m not sure that’s the best way to gain mass users these days.
10/10 for me
I’m impressed and most of the time I’m disappointed in the show hn section
Can I ask what the tech stack is? Is it open source?
>Except as expressly authorized by Y Combinator, you agree not to modify, copy, frame, scrape, rent, lease, loan, sell, distribute or create derivative works based on the Site or the Site Content, in whole or in part,
>In connection with your use of the Site you will not engage in or use any data mining, robots, scraping or similar data gathering or extraction methods.
Who owns the posts that you are trying to profit from?