The New York Times on collaborative filtering
1

The New York Times on collaborative filtering

The New York Times has a long article on the Netflix prize and collaborative filtering. Very interesting to anyone who’s obsessed with collaborative filtering, as I am. What I found particularly interesting is that customers who pay Netflix but don’t regularly send back movies are bad for the company:

For Netflix, this is doubly important. Customers pay a flat monthly rate, generally $16.99 (although cheaper plans are available), to check out as many movies as they want. The problem with this business model is that new members often have a couple of dozen movies in mind that they want to see, but after that they’re not sure what to check out next, and their requests slow. And a customer paying $17 a month for only one movie every month or two is at risk of canceling his subscription; the plan makes financial sense, from a user’s point of view, only if you rent a lot of movies. (My wife and I once quit Netflix for precisely this reason.) Every time Hastings increases the quality of Cinematch even slightly, it keeps his customers active.

Even though Netflix spends less money shipping movies to them, those customers are at much greater risk of canceling their accounts. Making sure customer queues are full of things they are eager to see is important.

I also thought this was really interesting:

As the teams have grown better at predicting human preferences, the more incomprehensible their computer programs have become, even to their creators. Each team has lined up a gantlet of scores of algorithms, each one analyzing a slightly different correlation between movies and users. The upshot is that while the teams are producing ever-more-accurate recommendations, they cannot precisely explain how they’re doing this. Chris Volinsky admits that his team’s program has become a black box, its internal logic unknowable.

It reminds me of some work I’ve done in the past.

When everything is recorded
5

When everything is recorded

Bruce Schneier’s essay on the fading future of ephemeral conversation is thought-provoking:

This has changed. We chat in e-mail, over SMS and IM, and on social networking websites like Facebook, MySpace, and LiveJournal. We blog and we Twitter. These conversations — with friends, lovers, colleagues, members of our cabinet — are not ephemeral; they leave their own electronic trails.

We know this intellectually, but we haven’t truly internalized it. We type on, engrossed in conversation, forgetting we’re being recorded and those recordings might come back to haunt us later.

I don’t think this is right. Some people haven’t internalized it, but many have. This is, in part, what I was getting at in my levels of candor post.

There are a huge number of things I would never say about a coworker in email. There’s a large number I wouldn’t say over instant messenger. The same goes for texting and voice mail. Some things are only worth saying on the phone or even face to face. I’ve been online in some fashion for over two decades, and I’m a fairly private person to start with, so I am very careful about not saying things that are going to turn up later.

Indeed, these days posting words, photos, and videos online is sort of like getting tattoos. Think ahead, because they’re going to be around forever whether you want them or not.

What I wonder, though, is whether we’re going to see some kind of crest in terms of how harshly people are punished for their previous online behavior. When there are embarrassing photos of everyone online, then by definition their existence will no longer be sensational.

Flat is failure
2

Flat is failure

Anand Rajaraman says that any startup that’s not growing is dying, and that if a startup isn’t going to succeed, it may as well die quickly. It’s outstanding tonic for the Sequoia RIP: Good Times presentation that seems to have been absorbed by everyone in the Internet business. I would guess that in five years, very few of the startups that decided to hunker down and wait for better times will be around.

Google SearchWiki
3

Google SearchWiki

Lauren Weinstein considers the likelihood of success of Google’s new SearchWiki feature. Here’s the official announcement from the Google blog. The basic idea is that users will be able to provide feedback on search results, ranking them up or down in the list and entering comments on them.

It’ll be interesting to see whether spam and other forms of abuse render this feature useless. I can see the black hat SEOs burning the midnight oil this weekend to figure out how to exploit this feature.

Context is everything
2

Context is everything

I don’t trust anyone who doesn’t appreciate context.

I write that sentence after reading Andrew Brown’s post on how the failure to appreciate history on its own terms clouds the thinking of fundamentalists. (In this case, he’s talking about fundamentalist atheists.)

This is the paragraph that grabbed me:

Thinking about the ignorant, angry atheists who infest the Guardian’s comment pages I realised one thing they have in common with scriptural fundamentalists: they have no idea of history. They live in an eternally dazzling present and they can’t imagine that there is anything outside it. Oh, sure, they have legends — the inquisition, the crusades, the middle ages — but within these legends the actors move, as they do in renaissance paintings, entirely in contemporary dress. There is no sense of the strangeness and difficulty of the past; no sense that many things have been tried and failed; no sense that words once meant things entirely different and possibly inexpressible now.

It’s impossible to properly appreciate anything without understanding, to some degree, where it came from. Failure to appreciate things in their own context is a problem I often find when people talk about software development. I read arguments about the superiority of Ruby on Rails to J2EE without any appreciation of the fact that Ruby on Rails is built upon many lessons that were learned the hard way as Java frameworks evolved. Without the 1999 article Understanding JavaServer Pages Model 2 architecture, Struts, and plenty of other lessons learned along the way, there would be no Rails as it exists today. Without Active Server Pages there would have been no Java Server Pages. Without CGI there would have been no ASP. Without Perl and Lisp and Scheme there would have been no Ruby.

Whether the topic is programming, history, politics, or music, attempting to explain or criticize things without judging within their own context is waste of time and energy. The only upside is that when someone persists in doing so, it’s a good signal that their analysis can be safely dismissed without further consideration.

What happens to campaign supplies?
0

What happens to campaign supplies?

I live near one of the Barack Obama campaign offices, and I couldn’t help but notice that it was completely vacant the next morning after election day. It was like the circus leaving town. I wondered what happened to everything in the office, and in the hundreds of other campaign offices around the country. In the case of the Obama campaign, the answer is that they made arrangements to donate all of the computers, furniture, and office supplies to schools.

The new world
0

The new world

Today Barack Obama nominated economist Peter Orszag his to head the Office of Management and Budget. Orszag is currently director of the Congressional Budget Office and writes an official blog in that capacity. Not quite the same as publishing a personal blog, but interesting nonetheless. I wonder if he’ll be blogging from the OMB as well? Marc Ambinder notes that OMB is in charge of implementing Obama’s transparency agenda. Putting that office in the hands of someone who’s comfortable blogging is a good sign.

The Kinkade aesthetic
0

The Kinkade aesthetic

Anyone who loathes Thomas Kinkade as much as I do will appreciate this leaked Kinkade-authored memo offering 16 guidelines for achieving the Kinkade aesthetic in a film. (via Kottke)

You can get an idea of how the guidelines were applied in the trailer for Thomas Kinkade’s Christmas Cottage.

This is the best piece of Kinkade snark since Janelle Brown wrote Ticky-tacky houses from “The Painter of Light™ back in 2002, about a Kinkade-inspired golf course community in California. The prices of those houses have not fared well in the current climate — here’s the 5 year pricing chart for one of the houses in the neighborhood from Zillow.