I recently refactored my Hacker News Amplifier 1.6 (Chrome extension) and added some new features:

  1. support for Reddit, so a name change was in order: the new version is Hype-Amplifier 1.7
  2. a little animation for shifting to the right news with less points and comments
  3. a separation of concerns, with all amplification code based on callbacks specific to each news site
  4. upload all code to https://github.com/aercolino/Hype-Amplifier, ready to be forked


See also: Reddit, […] and Hacker News Algorithms Exposed!


About Software

A program is (a) the expression of an idea (b) such that a machine can run it and produce a change (c) compatible with that idea. (b is what makes software unique)

difficulty(a) >> difficulty(b) > difficulty(c)

Not only software is not physical, but any program is like many alphabets (coding languages/environments) + many dictionaries (names for blocks/data) + craftsmanship (using information flows and domain knowledge for conveying the idea into the program).




Ant Colony Optimization

I found this interesting article in Wikipedia about Ant Colony Optimization Algorithms. The basic idea is that collective intelligence can result from a big number of repetitions of very simple efforts. In the case about ants, the paradoxically best features of their simple effort (leave a trace back to food while taking some of it back to the nest) is that ants look for food by means of random walks (this guarantees they’ll find anything in their area) and the trace they leave vanishes after some time (this guarantees they’ll concentrate on shortest paths from nest to food).

Many crowd based ideas somehow relate to this algorithm, but they are much more complicated than needed. The hope is that if we as humans apply our intelligence to imagine a possibly optimal result we’ll save time, also collectively. What bothers me is that ants do not need this individual intelligence to find a solution that collectively is optimal. So I’m wondering if we are not really losing time paying attention to unneeded details.