Wednesday, December 1, 2010

Current Projects: Twitter Newspaper & Automated Refactoring

tl;dr: I'm building Twitter Newspaper, a sequel to Hacker Newspaper. It's extremely private beta status at the moment, but you can sign up to hear more when more is ready.

Sign up for beta invite
* indicates required

Powered by MailChimp


Since it appears the Freakonomics guys have conclusively discredited Derek Sivers's claims that publicly stating your goals diminishes their probability of success, I'm no longer of two minds about announcing my projects. So: I'm working on putting together some iOS apps, and I spent most of the holiday building a successor to Towelie, the code repetition and similarity detector I wrote in 2008. The new version is closed source and intended to serve as the foundation for an automated refactoring business - maybe a service, maybe a web app, I'm not sure yet. What I already have is much faster and more effective than Towelie was, and covers one additional language beyond Ruby; the new angles of attack I'm taking are even more interesting. Unfortunately, I can't go into more detail about that yet.

I'm also working on a Twitter client, both to prevent people from reaching me with exasperating tweets, and as a way to replace Hacker Newspaper, since Hacker News got less interesting. Twitter is a much more effective link aggregator than Hacker News, but it's also much noisier. Its "follow" model means the range of available sources is much more finely-tuned to your personal interests than any site like Hacker News could ever be. For instance, in addition to the usual tech industry and open source suspects, I get tweets daily about dance music and the film industry. The follow model also makes Twitter completely immune to sudden population shifts, like the one which appears to have caused HN's most recent dip in interestingness. Much more importantly, link aggregation via Twitter operates within the conditions necessary to leverage the wisdom of crowds - namely, independent agents working without coordination, in competition, to achieve disparate goals.

No link aggregator site in the (by now very old) Slashdot/Digg/Reddit/Hacker News model can harness those conditions. After I explained what I believe to be the economic naivete of Hacker News's design assumptions, Peter Cooper asked me what a better system would look like; I've been thinking about the question off and on ever since, and I became convinced that passive/implicit link aggregation was the answer long before paper.li came along with things like The #graffiti Daily. I think Peter's own project coder.io follows a similar line of reasoning.

I prefer Twitter to Hacker News because there's something very artificial about attempting to build a community around gamified link aggregation. I also find Hacker News annoying because I'm banned on Hacker News; not only can I not comment on anything, but if you attempt to post something from my blog there, it'll get automatically blocked. I never received any warnings or communications of any kind from HN about this, and I never post my own stuff on any site of that nature; I only know because people tried to post my stuff, failed, and told me about it. I don't know why it happened or how to fix it, and finding out is more trouble than it's worth. Like a gated community, Hacker News is fundamentally suburban, in a very artificial way. Twitter, to paraphrase William Gibson, is like a city; anyone can come and go, so you have a living, vital community with no intentional theme or all-powerful "benevolent" overlord.

However, cities are noisier than gated communities, and that can be a problem. Going back to the wisdom of crowds, if anybody has anything to say about the wisdom of crowds that does not actually stem from having read The Wisdom of Crowds by James Surowiecki, not only do I not want to hear it, but that is exactly the kind of lazy, not-doing-your-homework horseshit which I am building this Twitter client to keep people from wasting my time with. The sad truth is that opinions are like assholes; everybody has them, and most of them are not worthy of detailed investigation. (This is especially true when it comes to the wisdom of crowds, or any concept which comes from a book which many people have heard of and only a few took the time to actually read.) The Twitter filter I'm working on is partly a cure for that problem. You can think of it as a pair of dark sunglasses to rock on your dome while you go about your business in the big city.

In the past I've gotten angry with people for talking to me on Twitter about things like the wisdom of crowds without doing any research first, but to be fair, it's really almost impossible to convey in 140 characters whether I happen to be in the mood for random, pointless banter or serious analysis driven by research. I read a lot; half the time I'm reading doctoral-level research on postmodern story structure and half the time I'm reading a comic book written by a teenager. There's really no way to predict via Twitter which of these two extremes I happen to be in the mood for at any given time, if you happen to be outside of my brain. On the inside-my-brain side of the equation, however, it's trivial to just choose to use an awesome new Twitter client with built-in filtering and aggregation when you're in the mood for a research-driven, grown-up approach, and fall back to the usual inane gabber at other times, as appropriate. Since there does not yet exist an awesome new Twitter client with built-in filtering and aggregation, and since its existence will make the choice of using it an easier choice to make, I'm building it.

I'm hoping to make it a business as well, but I haven't done enough work on it yet to say anything further on that.

I'm also working on some new musical projects. More news about that soon. And if you're wondering what happened to the SEO-loving, direct-marketing Giles Bowkett of 2010, don't worry. He has plans for 2011 too.



Oh yeah, I'm also working on a book.