You are viewing a read-only archive of the Blogs.Harvard network. Learn more.

The Longest Now

Getting the word out
Sunday November 16th 2003, 6:18 pm
Filed under: metrics

Broadcasting timely / breaking information to those who care about it should be easy

The absolute cost of broadcasting from one computer to another is under $1/GB, even after amortizing the cost of the sending machine/hardware and assuming moderate economies of scale.  This is on the order of a millicent per memo(sorry, Millicent)

Clustering on each end of this process can further reduce cost by another few orders of magnitude.  Making use of something simple like a public electronic bulletin-board would allow each such titbit to be broadcast to thousands; similarly, gathering together similar sources of information to reduce {time, topic, audience flag} redundancy would allow a five-fold reduction in the number of broadcast titbits. 

Now, however, you must pay for hardware; half of your $60 bulletin board is replaced with a ten thousand dollar screen.  If a hundred clustered titbits are broadcast a day, a hundred thousand over 3 years, this means an extra twenty cents per tb, or an extra centicent per memo[= 5 tb] if two thousand people pause long enough during the day to scan the board.

Oops.  Looks like we’ve actually upped the cost of transmission by a magnitude.  But now we’ve also taken care of contextualizing (clustering increases relevance, helps recipients compare similar events/happenings/announcements), dissemination (how did you get your list of interested parties in the first place?  were they able to tell they shouldn’t filter it into the trash? [this happens to me with a few messages a day that I honestly care about] etc…), and have left the vagaries of displaying information in the central hands of experts.  Before, you were limited to the lowest common denominator of the display terminals of your recipients; but now you can access just about any high-resolution display format, if the display boards are well-designed (which extras you have on the order of $1k per display to pay for).

Good displays can provide another half magnitude of information density; good use of familiar logos can greatly increase scanning[reading] speed; more subtle interleaving and refreshing can improve on the transmission rate of 100 tb a day.  Each of these improvements makes the use of the system enjoyable, a more rewarding daily ritual — and we’re also back down to the microcent memo.  This means that reminding five hundred people twenty times each about a timely bit of information costs you around $1… after paying off middle-men, and accounting for the many people who won’t care about it in the first place. 

[of course the sources are now spending a bit more time dealing with clustering, but they will  welcome the excuse to stop padding one-fact messages.  When we have a true information-glut problem, our world will already have become a much richer place.]

That’s about as much as it costs to have your secretary spend five minutes proofreading it and sending it off.  As for acquiring timely information and splitting it up into useful bits… I’m not yet sure how that works  I’d have to ask a news channel, calendar maintainer, news librarian, or maybe even a high-speed current events commentator

Once I do, have no fear, you’ll be the first to hear about it. 

Okay, you’ve asked. Let me mull over a response and I’ll answer.

Comment by j 11.16.03 @ 9:38 pm

Speaking of news librarians…. that’s what I like to hear. For instance, why is news propagation about university news so poor at Harvard? Half of campus reads the Crimson, half reads the Gazette, and some read nothing at all.

It’s shameful that there are people who
a) visit major campus gathering places like dining halls, Mem Church, Johnston Gate, the Science Center, and the Quad shuttle drop-off point, but
b) don’t find out about major campus news items [for instance: many undergrads in the phys department don’t know that Xiaowei Huang won a genius grant… or even that she’s *in* the phys department].

These places should have pretty, visible news-centers for official and strictly-approved notices, which have some standards for information-density(not too high ot too low), subject relevance, and subject classification (so it can be put in the right section of the news center; for fast skimming).

The university should be willing to spend at least 10% of their news-printing budget on creation and upkeep of public centers like this; the result would affect far more than 10% of their audience.

Comment by sj 11.16.03 @ 11:30 pm

My response to your blog post is at:

Comment by j 11.17.03 @ 9:03 pm

Why do I feel targeted and trapped? You knew I would come back here to post a URL and you’ve written a loaded comment that seems to be directed at me. I cannot speak, especially publicly, about this issue on behalf of the office where I happen to work. And I don’t want to post anything in a place where I’ll lose control over what I’ve typed. I hope you can understand that. <grumble grumble, but smiling while grumbling>

Yes, news dissemination at Harvard is a problem. Many people have brought this to my attention recently, too. My thoughts:

Comment by j 11.17.03 @ 10:14 pm

Why trapped? I was responding to your first comment by expanding on how my post would apply to Harvard. Nothing you do in your daily job would change the parts of university operations I mention. You are tangentially related to the Gazette editors, managers, etc. and I mention the Gazette as one of two major sources of campus news… any targetting is wholly incidental. Thanks for posting your thoughts! Don’t forget that as soon as you publish something, you’ve already lost control over what you’ve typed… you can spend much of your life trying to regain control over a school of thought that you founded.

Comment by sj 11.18.03 @ 7:46 pm

“Nothing you do in your daily job would change the parts of university operations I mention.” Boy, do I feel valued. ; )

Now that I know you’re using the Gazette and Harvard as an example, I’m a little relieved. I interpreted your comment as something you were asking me directly, like you wanted *me* to explain why news distribution at Harvard is so flawed. Many people confront me about things on campus because they think I have some kind of inside knowledge I can share with them. Sometimes, I get sick of that. Sometimes, I’m pleased to have the opportunity to dispel myths about Harvard.

I’m not angry or upset about you for asking about it. Re-reading my comment about feeling trapped, I noticed that it sounds far more pointed than I intended it to.

You’re very intelligent and have a perspective I lack and appreciate. Thank you for sharing your thoughts with me.

Comment by j 11.20.03 @ 6:51 pm

Bad Behavior has blocked 160 access attempts in the last 7 days.