Hacker News new | past | comments | ask | show | jobs | submit | treefarmer's comments login

Is there distributed server support? I see it on the list of new features with (currently PoC) next to it, but is the code for the PoC available anywhere?

Also, would there be any potential issues if the index was mounted on shared storage between multiple instances?


The code for the distributed search cluster is not yet stable enough to be published, but it will be released as open-source as well.

As for shared storage, do you mean something like NAS or, rather Amazon S3? Cloud-native support of object storage and separating storage and compute is on our roadmap. Challenges will be maintaining latency and the need for more sophisticated caching.


S3 support would be absolutely killer.


The content and idea of the article are interesting but the writing is so terrible that I couldn't bring myself to finish it. The article is clearly written/augmented poorly by AI because of how many meaningless paragraphs (that convey absolutely zero information) there are.


I thought I was super clear, I will take your comments into consideration next time it is the first time I've heard someone say that. Out of all the comments. I really appreciate the feedback this is the only way someone gets better by getting feedback.


Yeah, this is the main issue with the suggestion. Embeddings can only be compared to each other if they are in the same space (e.g., generated by the same model). Providing embeddings of a specific kind would require users to use the same model, which can quickly become problematic if you're using a closed-source embedding model (like OpenAI's or Cohere's).


Had the same thought - it's also annoying to update the PDF once links die, so I doubt that'll happen often. I guess it might be helpful if you want it as a coffee table book...


Would love it if it was available and open source so people could use it in their own projects (or on their own hardware), instead of only being available on Intel's AI Cloud. But cool idea and execution nevertheless!


Yeah, would love to built-in support for this in PyTorch or TF


For those interested in hearing more about the criticism behind Freakonomics, there is an If Books Could Kill podcast episode going over it: https://open.spotify.com/episode/5wHpooGMRsSBrUHhQZbOZp

I don't agree with all of their criticism but it contains many valid points


Funny podcast in general and great if you are looking for one that is highly critical of whatever book they are reviewing.


Yeah, I'm still surprised they haven't added a list of unsupported extensions (that you can add but they're not responsible for the performance of).


Amazingly their own aws_s3 extension isn't even supported in their multi-az cluster configuration.


But then you have a paid "featured" tier, along with a leaderboard showing the most upvoted ones. How is this different from ProductHunt in the end? Isn't it still pay to win?


I understand your concern, but our paid tier simply boosts visibility, not rankings. It’s not “pay to win.” The underdog feature gives lesser-known projects a fair shot, chosen randomly from those with fewer upvotes. This ensures that every product, regardless of budget, has a chance to shine without favoritism. also, we have a plan to change all voting system for the future too.


Products with higher visibility will automatically get more upvotes, so even if you don't directly pay for upvotes, you indirectly do pay for upvotes.

If you let people pay to be featured, you'll always end up favouring those who pay. And if you didn't favour those who pay, why would people pay in the first place?

This is the big problem with all recommendation sites: The easiest way to monetize them is by charging vendors for visibility; so sooner or later all recommendation sites start recommending the most profitable products. High quality fair priced products don't have a chance, since they will always be outbid by someone who makes a cheaper product or charges more.


I don't disagree with the point that visibility = upvotes = leaderboards. It's possible it biases rankings as you outlined. That said, given the problem: "How do I monetize my rankings/listings website without introducing pay to win or relying on third-party ads?"

I think the author implemented a tasteful and respectful solution. Other sites inject sponsors atop search results, every nth result, block parts of pages, force extra navs to link-spammed pages to get the product site, or sends sponsors to the top of rankings regardless of votes. You never have to glance at that portion if you're not interested, and can find it if you are. It's good design and advertises to the user in a respectful and convenient way imo. This is a small fixture that's static, and not even in the common paths followed by users ("F" eye track)

@Author: Perhaps a voting freeze during the sponsored period or a vote expiration (votes older than a certain amount of time fall off the total or such) would address the visibility = rank boost concern. Two random examples for moderating it with zero future thought there, but I think you get the idea.


Thank you for the thoughtful feedback. I agree that visibility can affect upvotes and rankings. Given the challenge of monetizing without pay-to-win or intrusive ads. Your suggestions—like a voting freeze during sponsorships or vote expiration—are good, interesting ways to further address this concern.


There is no way to square this circle. If you take payments from the vendors you recommend, you will never be fair. Either you take their payments and don't promote them, then they'll stop paying. Or you take their payments and promote them, then you'll end up doing paid promotion.

If you want to make a "fair" recommendation site, you need to find some other way to monetize. One possibility would be to charge potential customers. I believe there are people out there who want fair recommendations, I'm just have no idea how to get them to pay.


And god forbid you use a VPN and try to do anything on a Cloudflare site


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: