3 ms·
This points to a more general problem with all online discussion forums: The signal to noise varies (and may on occasionally be zero) but in the end all the sig
by college_physics 4y ago
This points to a more general problem with all online discussion forums: The signal to noise varies (and may on occasionally be zero) but in the end all the signal is lost / buried deep into some archive (like tears in rain if you want to be melodramatic about it).
Think about those millions of people, interacting for billions of hours online over almost two decades now. What is there to show for it at the end of the day?
If even 1% of that exchange its somehow "valuable", it means there is a lost opportunity. And it will be kept thrown away in the future.
Wouldn't it be useful if we somehow could use technology to persist the "better" bits across this ever growing digital ocean? Something like wikipedia but autogenerated from diverse sources, with no claim to "truth", but rather a concise, searchable repository of whatever people are interested and are discussing online.
- beckingz 4y agoPeople use technology to manually curate the better bits into things called books, and sometimes implement them on physical biodegradable materials to serve as offline backup and as a low bandwidth distribution method.
- college_physics 4y agoA book is probably near the other end of the spectrum in terms of signal to noise and requires very serious investment of time. It is also not yet very typical to continuously update them to reflect new information
- prox 4y agoThe thing is, the curator of any topic X, also needs to be at an expert level of topic X. On HN I often gain more insights, so I add to the signal, not noise (I hope) But it’s sometimes hard to gauge who is an expert if it’s a topic you are unfamiliar with. It helps if someone mentions (“I have been doing this for the last 5 years”) or wrote a literal book on the subject.
- gus_massa 4y agoI used to post a link to the previous thread and sometimes copy a short version of the best[1] comment if I remember it debunk or enhance the article. Lately [2] I had no enough time to look up, and usually dang post a link before I see the repost. [1] Best in my opinion. Usually the top one, but sometimes the second one, or a mix of a comment and a reply. [2] In the last few years...
- 1123581321 4y agoGoogle (and now Open AI, I suppose) are excellent at surfacing those old discussions. Forum software like Discourse also reliably brings them up in an effort to avoid deduplication of discussion. The best discussions are also springboards for articles, books, and sometimes whole companies.
- college_physics 4y agoSearch is indeed the only tool we have now to salvage all this body of content, but it is limited: you need to have a 'prompt' (using modern terminology :-). It is closer to a book index than a table of contents (to use older information technology tools) but one where the whole index is not accessible.