link

Godseye is a project that allows its users to attach pictures and stories to homes, trees, shops, and other parts of their neighborhoods and cities. You begin by gradually zooming in on high-quality, color aerial photographs, then selecting a building or other feature. When you focus on a specific feature, you can view and contribute storefront photographs, histories, or personal stories.

The potential uses of this data are vast and compelling. We spend lots of time playing around in online communities now, but there can be a mutually beneficial relationship between physical and online neighborhoods, and godseye proves it beautifully.

Unusual but interesting meeting tonight. I gave a demo of frassle; it was a little nerve-wracking at first because frassle's basic functionality has lots and lots of rough edges, but by the time I got to the cool stuff about learning category relationships it was well received.

It's clear at this point that I should put together a guided tour of frassle. It will be demoed at least once in the next few days, and if people are coming to it without a real guided tour from me, they will need some explanation. The interface does not, at present, explain itself.

I also brought my roommate Josh. He seemed to really enjoy the meeting and I'm sure we'll be chatting about it during the ride to work tomorrow morning. Another new visitor was Sarah, who is a smart and charming postdoc researcher blending social science and medicine. Many good conversations before, during, and after dinner, and I'm still riding high from my little demo.

One other note. I don't know if anyone noticed, but famous Harvard (ex MIT) linguist Stephen Pinker was sitting at a nearby table in the Bombay Club. On my way back from the restroom, I saw someone sitting at a table and thought, geeez, that guy has hair just like Stephen Pinker. I walked around, and there he was!

link

The CEO of Overstock writes a letter to stockholders, frankly explaining his company's achievements and mistakes.

Conclusion

A man crosses a desert by shooting an arrow from his bow and retrieving it as he walks. He must cross the desert with the fewest shots possible. At times he strains the bow with all his strength and lets his arrow fly, but sacrifices accuracy, at times shooting the arrow wildly off course. Other times he aims carefully and draws timidly, attempting little but guaranteeing himself small, solid progress. In time he finds the balance of draw and aim that covers the most ground in the fewest shots.

I have not yet found such balance. Yet in 2003 Q4 we drew the bow deeply, and our shot went far and fairly straight. On such a deep draw I could have blundered the shot (as I have before), and only the excellent work of my colleagues prevented mishap. I do not enjoy chancing so much on each shot. Yet I saw an opportunity to let one fly, and we still have much ground to cover.

Your humble servant,

Patrick M. Byrne
President

 

link

Andrew suggests I should check out these docs to see if there's an easy way to siphon the feeds.scripting.com data out of his oracle DB into a DB of some sort that I can play with. It may be easier to use the feeds.scripting SDK, but we'll see.

The next major feature for frassle will be an aggregator I won't mind using myself. Currently the "aggregator" functionality means that I can have feeds come in from RSS rather than be written on frassle itself. This enables the basic inter-category relationship statistics, and the essential technical stuff, but the interface blows. Aggregators should, well, aggregate.

The updated design will show recent postings from any feeds below the current category. So, for example, activating the aggregator inside my incoming category will show you the latest posts in all my feeds. I will probably break down incoming into subcategories for people I know personally, tech news, normal news, etc., and so you'll be able to narrow down your aggregated feeds that way too.

SharpReader, an RSS aggregator for Windows written in C#, has been influential here. SharpReader does a similar thing with categories, although it is a three-paned reader. Another cool feature it has is automatic threading together of related posts; when posts link to each other or both link to the same external URLs, they are visibly tied together. This is enormously convenient and I intend to copy and extend this idea for frassle. Extend because the careful threading together of related ideas is not something that should be left solely to computer programs. It's an intelligent activity.

A busy day at work. A few people who are not me are hitting my frassle site, which is very cool. Hopefully there'll be time for me to give a short demo at berkman tonight.

One nice thing about putting your software out there for other people to use is that its shortcomings keep showing themselves to you, over and over and over again. Some people (not our industry's finest) might mistake this for a bad thing.

During my freshman year of college, while walking through the path in front of the Williams College Museum of Art, I invented a piece of personal doctrine. The principle I invented was small and easy to understand but with profound implications, just like a good computer program. It was that

I should never do anything I'd be ashamed to have appear in a personal journal on a public website.

Naturally, there are some obvious exceptions. For example, if a friend confides in me and I listen to his secrets, those do not go up on the web journal of my thought-experiment. Details of my sex life are also not the public's business. But the point was to give myself a tool for evaluating certain long-term consequences of my decisions as I decided them.

In particular, this got me into the habit of asking myself how my actions could be scrutinized. I didn't want to please everyone or get everyone to agree with me. What I wanted to do was avoid any situations where I might have to defend something I felt was indefensible.

As a result, I always wanted to find and reconcile small personal conflicts before they became big ones. I got into the habit of reflecting on my actions and trying to understand what motivated them. I'd take a devil's advocate position, pretending I was someone out to prove that Shimon was stupid or evil. Real Shimon and Devil Shimon would argue (politely) in my head. The goal was to be able to either honestly refute or honestly take responsiblity for any shortcoming Devil Shimon could find.

Did this prevent me from making any mistakes? Hardly. But it made me realize that I can make mistakes, even mistakes that hurt people or hurt me, and that it's OK. I can learn from them, and I can be careful and avoid them in the future. But the doctrine of honesty gave me a powerful tool for dealing with my mistakes: it enabled me to consider them scientifically.

If you have not been in a good writing workshop, you may not know the value of confronting your own mistakes scientifically. The reason is that by nature, human beings have a certain level of pride in and expectations for thier work. We are usually not conscious of our pride and expectations until they're injured or unmet, hence the popular truism you don't know what you've got till it's gone.

The problem with having pride and expectations is that you subconsciously support your pride, or your expectations. These concerns become hidden assumptions; the things you can't say. Well, at most crucial junctures in thinking, there is good sense to saying those things you can't say. In science, the major breakthroughs must not, by definition, fit into current paradigms. In therapy, the fears that a patient is unable to talk about are the most crippling.

Unsurprisingly, this scales to one's day-to-day life. Those actions you take that you're least inclined to examine, those motivations you're least inclined to explore for fear of what it might admit, are exactly your greatest opportunities to improve yourself as a human being. Just like those problems in a poem you bring to a workshop that you're most afraid will expose your lack of skill will in fact do that and show you, conspicuously, how to improve both your current poem and your overall ability.

Maybe that's why I blog— to keep myself honest, and to remind myself that I can make mistakes. In writing. Several times a day.

link

link

A sensible take on offshoring from both sides of the debate/ocean.

But the rest of us, like it or not, will have to adjust. The hints about how to make this adjustment are evident at Patni. As I meet programmers and executives, I hear lots of talk about quality and focus and ISO and CMM certifications and getting the details right. But never - not once - does anybody mention innovation, creativity, or changing the world. Again, it reminds me of Japan in the '80s - dedicated to continuous improvement but often at the expense of bolder leaps of possibility.

And therein lies the opportunity for Americans. It's inevitable that certain things - fabrication, maintenance, testing, upgrades, and other routine knowledge work - will be done overseas. But that leaves plenty for us to do. After all, before these Indian programmers have something to fabricate, maintain, test, or upgrade, that something first must be imagined and invented. And these creations must be explained to customers and marketed to suppliers and entered into the swirl of commerce in a fashion that people notice, all of which require aptitudes that are more difficult to outsource - imagination, empathy, and the ability to forge relationships. After a week in India, it seems clear that the white-collar jobs with any lasting potential in the US won't be classically high tech. Instead, they'll be high concept and high touch.

link

Basic overview article on blogs in the corporate world.