What I have learned from Hacker News


Hacker News Logo.


Hacker News is run by Y combinator- a startup incubator that aggregates content from the Internet with a simple premise, “anything that gratifies one’s intellectual curiosity“.

Here’s an interesting blurb from Wikipedia:

The intention was to recreate a community similar to the early days of Reddit.[2][8] However, unlike Reddit where new users can immediately both upvote and downvote content, Hacker News does not allow users to downvote content until they have accumulated 501 “karma” points. Karma points are calculated as the number of upvotes a given user’s content has received minus the number of downvotes.[2] “Flagging” comments, likewise, is not permitted until a user has 30 karma points.

I am very keen to see a similar model adopted for sourcing scientific literature; especially in Oncology. In an ideal world, all content would be freely accessible with the actual value debated/hashed and assessed by the scientific community. While SciHub has made some progress to make scientific content available, it comes with a moral censure. Users across the world could submit and rate the literature (and key takeaways) based on the model above (or work through an iteration that suits the scientific community).

Here’s a bit more technical nuance sourced from Wikipedia:

The site has a proactive attitude in moderating content, including automated flame and spam detectors. It also practices stealth banning in which user posts stop appearing for others to see, unbeknownst to the user.[10] Additional software is employed to detect “voting rings to purposefully vote up stories”

As is the case- when communities grow, end-users take advantage of anonymity and tend to harass others. I have seen flame wars erupting (on opinions!) and it can get worse on other social media networks like Reddit or Twitter.

Submitted articles can quickly get overwhelming; as such, there are several services which allow to filter out strong signals from the noise. The same can be easily adapted for Oncology. Long time back, I had started Twitter account pushing the updates with specific hashtags and collect “twitter likes” so that could attempt to glean relevance. However, I realised that Twitter has reduced rates of engagement, and that was the end of my “experiment”.

Personally, I don’t expressly use specific applications to “track literature” or “keywords” but instead rely on “old-world” RSS.

Numerous engineering resources escape my comprehension, but it exposes me to a different world on what’s out there. A considerable source of articles linked here is usually the ones upvoted the most or have had most comments/discussions on HN forums. I don’t expressly claim to understand machine learning (or artificial intelligence) in Oncology, but have domain expertise of one to understand the application of other.

It is a continuous learning curve, and through this blog post, I thank the forum moderators and all the other contributors for keeping it alive. I am usually lurking in the background to make an attempt to understand what is being discussed and follow-through links that expose me to a deep dive of various topical ideas.