Fungible Convictions

I tried KGB for the first time tonight, and it looks like the whole thing is automated using a semantic language program (similar to how Ask.com worked).

Here in Boston there’s an commercial running where an auto dealer will pay the first year of a lease if the temperature at Logan Airport reaches 96 degrees this Labor Day. I wanted to see what the chances of that happening are, so I texted KGB “What’s the hottest Labor Day on record at Boston’s Logan Airport?” KGB’s reply was, “The highest temperature ever recorded in Boston, MA was 107 degrees Fahrenheit on Aug 2 1975.”

So KGB ignored two key parts of the question that a human would see—that I’m asking specifically about Logan Airport and specifically about Labor Day—leaving me to think a computer is doing the answering, at least initially. (There’s a third part, “on record,” that’s more or less redundant.)

When I replied that they didn’t answer the question, they followed up with an acknowledgment that they couldn’t find the answer and they were issuing me a credit for the $0.99 charge per answer. It’s a little disappointing overall, because there is an answer—KGB staff would simply have to click 122 times (the first official Labor Day in Boston was in 1887) through a page like this one at Weather Undergound. KGB just wasn’t interested in spending the time it takes to look it up.

I went ahead and did it. The answer to “What’s the hottest Labor Day on record at Boston’s Logan Airport?” is 94 degrees in 1928*. So if you’re thinking of leasing a car with Pride Motors of Lynn, Massachusetts, don’t do it just because you think you might get a year free.

* Temperature records at Logan go back to 1920, and the airport itself opened in 1923, making ‘23 the latest possible year applicable to the question.

It’s tough to describe the awesomeness of the conference we just ran at MIT. It was exhausting, yes. But I designed/printed the conference program, helped set the schedule, managed 200 attendees, kept an eye on an intern, and got to work with some incredible colleagues.

Based on the syntax of that last sentence, you can tell I’m exhausted. But I got to meet some folks that I’ve admired for a long time, such as Dan Gillmor, and got to promote the 2009 Knight News Challenge winners.

I’m conflicted. This conference was the last big set of tasks from now until the fall, so I’m glad I can rest a bit. But it was why I wanted to work with MIT’s Center for Future Civic Media—a chance to rock out with media innovators and meet a few of my long-time heroes.

I’m glad there’s a full year until the next conference, but I hope I get to convince all of these folks to come hang out at MIT before then.

I was recently interviewed by blogger Jillian C. York, one of those handful of people with whom I have an oddly enjoyable entirely Twitter-based relationship. The interview was part of a set she’s doing on people’s use of Tor, a web anonymity tool. (It’s run within a program called Vidalia, like the onion, an apt metaphor for how Tor anonymizes your web surfing by passing your data through layers of other users).

The interview was the first time I’d had a chance to think through my use of Tor and other electronic security tools. It comes down to: while I don’t really have anything to hide; while I’m not a security master; and while I’m not a paranoid, it still feels like an obvious best practice, like locking up your bike. It’s easy, and it’s free, so why not take that extra step? (And sometimes you get props, or suspicion, or both, like when Chris Csikszentmihalyi walked by my laptop and said conspiratorially, “You’re running Vidalia?”)

One thing I mentioned in the interview but largely glossed over was my use of TrueCrypt, a harddrive encryption program.

TrueCrypt is freaking awesome. It would take thousands of years to decrypt your data if someone ever got a hold of it. Except there’s a weak link: you have to remember your password. There’s absolutely no password-recovery option. When I went out sick in ‘07 with the memory problems, the person Tufts brought in to cover my work didn’t know the password. And neither did I, anymore. I had to go into the office a couple weeks after my surgery, and luckily, amazingly, my fingers had enough muscle memory that they typed out the password on the first try. (But I uninstalled TrueCrypt on both my computers after that. I wrote to Bruce Schneier sometime afterward and asked him what you’re supposed to do about a TrueCrypt password if you have a crappy memory. His three-word reply: “Write it down.”)

Electronic security tools today are dead-simple to use, free, and open source (therefore verifiably safe). They don’t get a lot of attention, but each one of them—Vidalia, TrueCrypt, or a password-organizer like KeePassX—are all worth the 5 minutes to set up.

Bain News Service,, publisher.

Fenway Park exterior

[between 1910 and 1915

1 negative : glass ; 5 x 7 in. or smaller.

Notes:
Date based on research by the Pictorial History Committee, Society for American Baseball Research, 2006.
Forms part of: George Grantham Bain Collection (Library of Congress).

Format: Glass negatives.

Rights Info: No known restrictions on publication.

Repository: Library of Congress, Prints and Photographs Division, Washington, D.C. 20540 USA, hdl.loc.gov/loc.pnp/pp.print

General information about the Bain Collection is available at hdl.loc.gov/loc.pnp/pp.ggbain

Persistent URL: hdl.loc.gov/loc.pnp/ggbain.11857

Call Number: LC-B2- 2554-7

I have to admit I’m biased in favor of Google. I have friends who work in both the Cambridge and Mountain View offices. I’ve tried, and provided feedback on, every beta Google has produced. I worked for a group trying to get funding from its philanthropic arm, Google.org. And every time I hear CEO Eric Schmidt speak at a conference, he strikes me as one of the most intelligent, well-versed, sober, geektastic corporate leaders I can think of. (If you have an hour, this interview with the New Yorker’s Ken Auletta is definitely worth watching:

.)

So perhaps I’m biased when I don’t see a problem with Google archiving so-called orphan works, publications that have been abandoned by both author and publisher, are out of print, and are effectively if not technically out of copyright. I don’t see a problem with making available works that no one can easily see/acquire, that no one is promoting, and that no one is making money from—but that may, and often do, still have great value.

I’m also biased, however, in favor of one of the great archival minds of our age, Robert Darnton:

Critics say that without the orphan books, no competitor will ever be able to compile the comprehensive online library Google aims to create, giving the company more control than ever over the realm of digital information. And without competition, they say, Google will be able to charge universities and others high prices for access to its database.

The settlement, “takes the vast bulk of books that are in research libraries and makes them into a single database that is the property of Google,” said Robert Darnton, head of the Harvard University library system. “Google will be a monopoly.”

The question for Darnton and others, though, is: is this a bad thing? Google does not somehow become the exclusive copyright holder to orphan works. Other groups and companies are welcome to do the same thing and to also make money from it. And this particular monopoly is, contradictorily, limited and temporary. There will be well-funded competitors. There’s no indication that Google wishes to charge for access—it’s fair to assume Google will monetize the collection through targeted advertising as it does with search results and within Gmail. The original orphan works don’t disappear.

So I don’t begrudge Google its ambition. While experience shows that powerful groups try to control archives as a way of shaping history, experience also shows that seemingly dominant businesses, such as General Motors and Microsoft, are inevitably outflanked. And most important, as Schmidt explains in the Auletta interview, Google thrives only in so far as it is trusted. It’s a business that deals in user data, and that demands trust. Trust broken once is trust lost, so it’s in Google interest to welcome competing ideas, to accept criticism, and to be, above all, open.

Though I’m still waiting for video to be posted, CBS Sunday Morning just aired a segment by Nancy Giles that was on par with an Andy Rooney rant in terms of its technological ignorance.

The technology in question was Twitter, which Giles knocked as yet another distraction. She ultimately derided it as something that splits our attention, leaving people a puddle of half-thoughts.

Though Twitter is indeed pitched to its users as a way to “Answer a single question: What are you doing?”, Giles ignores the ways Twitter is actually used (hence the Rooney level of non-research). She doesn’t mention that people can reply to one another. She doesn’t describe the use of hash tags to organize erstwhile strangers around a single topic. She’s oblivious to the potential for viewers of CBS Sunday Morning to immediately, publicly respond to Giles’ segment.

And though it’s a little more technical, she never once explained what every general-interest piece on Twitter has to explain: why a message is limited to 140 characters. The character limit makes sure every message can be sent as a text message. Though most messages are sent from desktops, the uses of Twitter from a mobile device are profound.

An example of how all this comes together, all of which Giles is oblivious to. Let’s say it’s November 2010. A election-monitoring group in northern Virginia has set up a “#nova10″ hashtag to track tweets about voter problems. (A hashtag is any arbitrary term preceded by “#” and is a Twitter feature that organizes all messages sharing that term.) Two hours after polls open in Fairfax Co., Virginia, voters notice irregularities with some voting machines—the machines have frozen, poll workers are rebooting them, but a Republican poll watcher is yelling that rebooting is deleting the first two hours worth of votes.

So voters waiting in line use their cell phones to send messages to Twitter using the #nova10 hashtag. What happens? Factual information is shared and replied to among voters. A political science professor and lawyer down the road at George Mason University see the messages and zip over to the polling station to help smooth out the legal issues. And, if problems escalate, local reporters tracking the hashtag can cover the dispute.

In all, Twitter helped announce a major problem and coordinated its response and solution.

If you think this is all pie-in-the-sky, consider the fact that this not only already happened here during the 2008 Presidential election but also in Pakistan, leading to circumvention of marshal law and ultimately the ouster of Pervez Musharraf.

If only CBS’s Nancy Giles would have bothered to do some research on how Twitter is actually used, she could have put together a story that didn’t leave her sounding like Andy Rooney.

For months—months—I’d had trouble submitting the Comparative Media Studies Colloquium podcast feed to iTunes.

Today I finally had a free morning to really sort through these issues.

1. When I tried to submit Feedburner’s podcast feed, I would get the message “It appears the feed has already been submitted.”
2. That was a problem because the podcast was not listed in the iTunes podcast directory
3. iTunes is notorious for not responding to requests for help, though I can’t blame them

(There were also problems with converting our podcast feed to XML and discovering that Feedburner now requires a Google account, but that’s not for now.)

After rooting around a dozen different help forums, one thing was clear: the only way to resubmit a podcast feed so that iTunes doesn’t think it’s a duplicate is to change certain key XML data. Though no one in these forums knew so, it turns out that you can change this data within Feedburner itself:

1. Log in to your feed
2. Click the Optimize tab
3. In the sidebar under Services, click Title/Description Burner
4. And within that, tweak the description of your podcast

I went an extra step and changed more, since all of that “Optimization data” is actually what shows up in your podcast feed’s XML. So I added geotag info, a new image, and selected the SmartCast and SmartFeed options….all the more to make it appear different from the previous version already submitted to iTunes.

The podcast should be on iTunes sometime next week. And though I hate to wait for it, this email I just got was so very sweet:

Dear Podcast Owner

Your podcast feed, [ http://feeds2.feedburner.com/MITCMSColloquium ] was successfully added and is now under review.

Sincerely,

The iTunes Store Team

The wife and I wanted some tips on how to comfortably file our dog’s nails. Naturally Google is a first destination for “how to” searches, but I’m always a little weirded out at what Google’s autocomplete feature—Google Suggest—presents as the most common searches starting with “how to…”:

And once you type some more, “how to file…” isn’t much more reassuring:

What about you guys? Ever run across creepy suggestions while searching for something totally different?

The battery for my Macbook Pro completely died—I can only get a 10-minute charge of out it—just days after my warranty expired. It was a frequent problem with batteries from a particular batch.

So this morning I called up Apple customer service and found out that a new battery would cost $129. So I did what Consumerist always suggests and asked flat out what I could do to get it for free. The friendly rep put me on hold for a minute to talk to a superior and came back with my full customer history and said, “It looks like you’ve been a customer with us for about four years, so we’re going to go ahead and wave that charge.” All I have to do is make sure to send the old battery to Apple, presumably so they can confirm it’s defective and make sure it doesn’t end up in a landfill.

Me = happy customer.