Visualizing Social Networks -- PieSpy
This is a newly released tool for visualizing social networks over IRC chat. What I find most useful about this site is the detailed discussion of both SNS representation and visualization algorithms, at a high enough level to be comprehensible to the non-programmer. If you want to know how the many visual tools such as Touchgraph work, check this out.
Tech-related musings. Occasional rogue war pieces. Hosted and led astray by Ethan Stock, founder and CEO of Zvents. This blog reflects my own views, not the position of Zvents.
Monday, January 26, 2004
Thursday, January 22, 2004
Touchgraph and Amazon
Alex Shapiro over at Touchgraph (proof that Open Source can be truly innovative!) has added an Amazon browser to the way-cool Google browser. For all you teenage girls out there, there's a LiveJournal browser as well. The Amazon implementation is based on 'customers who bought also bought' and the discussion includes a link to a good IEEE article on recommendation algorithms, including collaborative filtering, cluster algorithms, and search-based methods. The article serves as a handy collaborative filtering primer.
Alex Shapiro over at Touchgraph (proof that Open Source can be truly innovative!) has added an Amazon browser to the way-cool Google browser. For all you teenage girls out there, there's a LiveJournal browser as well. The Amazon implementation is based on 'customers who bought also bought' and the discussion includes a link to a good IEEE article on recommendation algorithms, including collaborative filtering, cluster algorithms, and search-based methods. The article serves as a handy collaborative filtering primer.
Wednesday, January 21, 2004
Social Networks, Information Economics, and Search
SNS has become a big enough field with sufficient internal dialogue that it's easy to forget that the promise of the whole shebang can be reduced to a single sentence:
Reducing informational exchange costs, particularly transactional search costs, through leveraging existing information exchange pathways such as personal relationships.
That's it.
We're reminded of that singular purpose by a new SNS-driven search engine, Eurekaster. Danny Sullivan over at Search Engine Watch has a good article on this new twist on a major field.
And yes, Eurekaster, you should offer your service as a layer built on top of the Google API. That API is a fantastic tool for developers and experimenters; the fact that it acts as an innovation magnet for Google, by which the company can observe --> copy or observe --> acquire promising new technologies is just the price of the piper.
SNS has become a big enough field with sufficient internal dialogue that it's easy to forget that the promise of the whole shebang can be reduced to a single sentence:
Reducing informational exchange costs, particularly transactional search costs, through leveraging existing information exchange pathways such as personal relationships.
That's it.
We're reminded of that singular purpose by a new SNS-driven search engine, Eurekaster. Danny Sullivan over at Search Engine Watch has a good article on this new twist on a major field.
And yes, Eurekaster, you should offer your service as a layer built on top of the Google API. That API is a fantastic tool for developers and experimenters; the fact that it acts as an innovation magnet for Google, by which the company can observe --> copy or observe --> acquire promising new technologies is just the price of the piper.
Monday, January 19, 2004
Interest cost and real value
While I am talking about property, I'll throw out another thought. It's conventional wisdom that home-buyers are most sensitive to their monthly payment, and declines in interest rates drive up selling prices in a near-perfect inverse relationship, as lower interest cost is swapped for higher sticker price, leaving the buyer with the same total cost. Call me contrarian, but I think that buying a house at times of historically low interest rates is madness. Why? Because you're fully leveraged. Let's assume that there is a nominal value to a house, and an actual value. The nominal value is the sticker price -- say, $400K for that two-bedroom apartment in Berkeley. Since the vast majority of buyers are on 15- or 30-year mortgages or their functional equivalent, a significant interest cost drives the actual value of the place -- the contracted financial commitment which the buyer undertakes in order to acquire title -- up to something like $700K. In times of low interest rates, this $700K (which is based, conventional wisdom tells us, on the buyer's ability to make 30 years x 12 months x a given monthly payment) might be made up of a nominal/sticker cost of $550K, and an implied interest cost of only $150K. In times of *high* interest rates, that $700K number might be made up of a nominal cost of only $320K, and the same overall total due to a vastly increased interest cost of $380K.
If I buy at a time of high interest rates and low sticker costs, it seems to me that I have nothing but capital gains upside from interest rates, which historically have varied cyclically (and which are due for a massive rise, if our trade deficits and the plummeting value of the dollar are any indication -- but that's another story as well). In addition to my capital gains upside (the nominal--> real value of my place going from $320K to $550K) I have the potential to refinance, getting a low low interest rate on a new mortgage to go with my low nominal cost, giving me a genuine total-cost bargain.
But what if I bought in times of low rates? The nominal value of my property can only fall on a cyclical basis, even if overall rising prices mask this somewhat; and if I've taken a variable-rate mortgage, my interest cost can only go up. In the best case, if I am forced to sell my house for any reason during a high-interest period, I'll lose out on its declined nominal value and avoid (through a fixed rate mortgage) any increase in interest costs. In the worst case, I am faced with a rising monthly payment on an asset whose value is declining precipitously, in an economy which is tanking because interest rates kill business investment. Ouch.
Have I mentioned that I'm feeling very secure being in cash?
While I am talking about property, I'll throw out another thought. It's conventional wisdom that home-buyers are most sensitive to their monthly payment, and declines in interest rates drive up selling prices in a near-perfect inverse relationship, as lower interest cost is swapped for higher sticker price, leaving the buyer with the same total cost. Call me contrarian, but I think that buying a house at times of historically low interest rates is madness. Why? Because you're fully leveraged. Let's assume that there is a nominal value to a house, and an actual value. The nominal value is the sticker price -- say, $400K for that two-bedroom apartment in Berkeley. Since the vast majority of buyers are on 15- or 30-year mortgages or their functional equivalent, a significant interest cost drives the actual value of the place -- the contracted financial commitment which the buyer undertakes in order to acquire title -- up to something like $700K. In times of low interest rates, this $700K (which is based, conventional wisdom tells us, on the buyer's ability to make 30 years x 12 months x a given monthly payment) might be made up of a nominal/sticker cost of $550K, and an implied interest cost of only $150K. In times of *high* interest rates, that $700K number might be made up of a nominal cost of only $320K, and the same overall total due to a vastly increased interest cost of $380K.
If I buy at a time of high interest rates and low sticker costs, it seems to me that I have nothing but capital gains upside from interest rates, which historically have varied cyclically (and which are due for a massive rise, if our trade deficits and the plummeting value of the dollar are any indication -- but that's another story as well). In addition to my capital gains upside (the nominal--> real value of my place going from $320K to $550K) I have the potential to refinance, getting a low low interest rate on a new mortgage to go with my low nominal cost, giving me a genuine total-cost bargain.
But what if I bought in times of low rates? The nominal value of my property can only fall on a cyclical basis, even if overall rising prices mask this somewhat; and if I've taken a variable-rate mortgage, my interest cost can only go up. In the best case, if I am forced to sell my house for any reason during a high-interest period, I'll lose out on its declined nominal value and avoid (through a fixed rate mortgage) any increase in interest costs. In the worst case, I am faced with a rising monthly payment on an asset whose value is declining precipitously, in an economy which is tanking because interest rates kill business investment. Ouch.
Have I mentioned that I'm feeling very secure being in cash?
Why are Bay Area housing prices still stable or rising?
The New York Times has an article today, the damn-with-faint-praise title of which is Job Losses Slow in Silicon Valley. Seems that we only lost 5% of local jobs in the past year, which is relative cause for celebration after the bloodbath of 2000 - 2003. What is staggering to me is the raw numbers. Peak employment in 2Q 2001 was 1.38 million; now it's 1.18 million, after a loss of 202,000 jobs. Average pay (meaning, I presume, wages; not income, which would have been further inflated by capital gains) was $81.7K in Y2K; it's now $62.4K. A little algegra shows that there was about $112 billion of annual salary in SV in 2000-2001; now there's about $74 billion.
For those scoring at home, that's a decline of more than a third -- 34% -- in local salary from 2000 -- 2003.
Since we are talking wage-earners here, it's not the $2 million options pads in Portola Valley that should be affected, but rather the $500K shoeboxes in Palo Alto and Mountain View. Instead, fueled by historically low interest rates, prices for homes have remained surprisingly robust during the past three years of nuclear winter -- an employment freeze that, if fear-mongering about job outsourcing overseas is true, may never actually end.
Six or nine months ago, you still regularly saw news articles about the 'housing bubble' in America, which laid out a case that people had pulled their money from the stock market and put it into real estate, taking advantage of low rates and upset at the declines in their portfolios. You don't see those articles any more, and yet prices continue to be surprisingly robust. What's going on here?
And yes, I am a frustrated wannabe-homeowner. But boy, does being in cash feel nice! Don't even get me started on the dollar and the 'which currency to be liquid in' question...
The New York Times has an article today, the damn-with-faint-praise title of which is Job Losses Slow in Silicon Valley. Seems that we only lost 5% of local jobs in the past year, which is relative cause for celebration after the bloodbath of 2000 - 2003. What is staggering to me is the raw numbers. Peak employment in 2Q 2001 was 1.38 million; now it's 1.18 million, after a loss of 202,000 jobs. Average pay (meaning, I presume, wages; not income, which would have been further inflated by capital gains) was $81.7K in Y2K; it's now $62.4K. A little algegra shows that there was about $112 billion of annual salary in SV in 2000-2001; now there's about $74 billion.
For those scoring at home, that's a decline of more than a third -- 34% -- in local salary from 2000 -- 2003.
Since we are talking wage-earners here, it's not the $2 million options pads in Portola Valley that should be affected, but rather the $500K shoeboxes in Palo Alto and Mountain View. Instead, fueled by historically low interest rates, prices for homes have remained surprisingly robust during the past three years of nuclear winter -- an employment freeze that, if fear-mongering about job outsourcing overseas is true, may never actually end.
Six or nine months ago, you still regularly saw news articles about the 'housing bubble' in America, which laid out a case that people had pulled their money from the stock market and put it into real estate, taking advantage of low rates and upset at the declines in their portfolios. You don't see those articles any more, and yet prices continue to be surprisingly robust. What's going on here?
And yes, I am a frustrated wannabe-homeowner. But boy, does being in cash feel nice! Don't even get me started on the dollar and the 'which currency to be liquid in' question...
Sunday, January 18, 2004
Parties, Food, and Anti-Food
I just threw a dinner party last weekend, and in the post-mortem with my girlfriend, happened upon an interesting concept. The 'dinner party' as formally constructed probably originated around 1900; when the bourgeousie acquired dining tables and formal dining rooms and all the tat necessary to impress the neighbors. The key point of differentiation here is that a 'dinner party' is not for a special occasion like a wedding, funeral, holiday, etc; includes more than just family; and is thrown by people not of the nobility or aristocracy.
Food was then expensive, and manners were formal, so throwing a mini-feast for no particular reason was an impressive thing to do.
Today, dinner parties are all-but-dead; the closest analogue being a BBQ in the back yard during summer. I would suppose that this has a lot to do with cooking skills being on the decline, but I think that in fact the key difference is that food is now cheap, so having a party centered around a meal (when half the people there are likely on a diet) makes very little sense. In fact, most parties I attend these days center around activities -- volleyball in the park and a burrito afterwards; a hike or run in the hills and burgers afterwards; etc. It's as if anti-food has replaced food as the centerpiece of our socializing.
This needs a bit more thought, but it's a dichotomy worth recording in its present inchoate form.
I just threw a dinner party last weekend, and in the post-mortem with my girlfriend, happened upon an interesting concept. The 'dinner party' as formally constructed probably originated around 1900; when the bourgeousie acquired dining tables and formal dining rooms and all the tat necessary to impress the neighbors. The key point of differentiation here is that a 'dinner party' is not for a special occasion like a wedding, funeral, holiday, etc; includes more than just family; and is thrown by people not of the nobility or aristocracy.
Food was then expensive, and manners were formal, so throwing a mini-feast for no particular reason was an impressive thing to do.
Today, dinner parties are all-but-dead; the closest analogue being a BBQ in the back yard during summer. I would suppose that this has a lot to do with cooking skills being on the decline, but I think that in fact the key difference is that food is now cheap, so having a party centered around a meal (when half the people there are likely on a diet) makes very little sense. In fact, most parties I attend these days center around activities -- volleyball in the park and a burrito afterwards; a hike or run in the hills and burgers afterwards; etc. It's as if anti-food has replaced food as the centerpiece of our socializing.
This needs a bit more thought, but it's a dichotomy worth recording in its present inchoate form.
Thursday, January 15, 2004
Visual Representation of Social Networks
Great overview of the visual representation of social networks over the past 30 years. Note to self: Read this in detail!
Great overview of the visual representation of social networks over the past 30 years. Note to self: Read this in detail!
Friday, December 19, 2003
Hyperdictionary - designed to become a web service?
In my opinion, Hyperdictionary is the best dictionary on the web. Fast, clean, easy to use. And it's dead-easy to build the service into other sites and scripts, since each definition page is simply the word appended to the URL http://www.hyperdictionary.com/search.aspx?define= like this: http://www.hyperdictionary.com/search.aspx?define=petard
Very cool. This seems to be run by a group of Canadians known as Webnox - I don't know much about them, but they've got a great service.
In my opinion, Hyperdictionary is the best dictionary on the web. Fast, clean, easy to use. And it's dead-easy to build the service into other sites and scripts, since each definition page is simply the word appended to the URL http://www.hyperdictionary.com/search.aspx?define= like this: http://www.hyperdictionary.com/search.aspx?define=petard
Very cool. This seems to be run by a group of Canadians known as Webnox - I don't know much about them, but they've got a great service.
ENUM and SIP
I've decided that I need to understand this in detail, very soon. I don't yet get all the implications, especially in combination with IPv6. ENUM is more or less a registry translation service done by an interesting Quango (Handy UK term - what's that mean?) called Neustar. Here is their whitepaper on ENUM -- I need to find similar discussions of SIP, process, and blog it here. To be continued.
I've decided that I need to understand this in detail, very soon. I don't yet get all the implications, especially in combination with IPv6. ENUM is more or less a registry translation service done by an interesting Quango (Handy UK term - what's that mean?) called Neustar. Here is their whitepaper on ENUM -- I need to find similar discussions of SIP, process, and blog it here. To be continued.
John Battelle's Searchblog
This is an excellent site. On a quick read, I ended up at a really thoughtful discussion of Google. I highly recommend this in conjunction with Tim Bray's Search Series.
This is an excellent site. On a quick read, I ended up at a really thoughtful discussion of Google. I highly recommend this in conjunction with Tim Bray's Search Series.
Searching gets smart filtering
This is damn cool. Google-enabled 'search my trusted universe' tool. When you start cascading it, you will *really* have something: "Search my trusted universe, and the trusted parties of my trusted universe, unto three degrees of separation. Go forth, my spider."
How to add it to your site:
http://www.alpern.org/weblog/php/blogsearch/mysite.php
How it works:
http://www.alpern.org/weblog/php/blogsearch/writeup.html
I found it on Kevin Werbach's site (he's implemented it) and followed it back to the source. Excellent.
This is damn cool. Google-enabled 'search my trusted universe' tool. When you start cascading it, you will *really* have something: "Search my trusted universe, and the trusted parties of my trusted universe, unto three degrees of separation. Go forth, my spider."
How to add it to your site:
http://www.alpern.org/weblog/php/blogsearch/mysite.php
How it works:
http://www.alpern.org/weblog/php/blogsearch/writeup.html
I found it on Kevin Werbach's site (he's implemented it) and followed it back to the source. Excellent.
Tuesday, December 02, 2003
Commodity hardware + open source = asterisk PBX
This in and of itself is not a game-changer. This is the equivalent of the "garage hosting" model of ISPs that was prevalent through the early 90s, with Mr. Modem Bank selling to local customers. However, it presages what larger players can do with the same technological building blocks -- after all, those little ISPs led to AOL and Earthlink.
"Asterisk is a complete PBX in software. It runs on Linux and provides all of the features you would expect from a PBX and more. Asterisk does voice over IP in three protocols, and can interoperate with almost all standards-based telephony equipment using relatively inexpensive hardware.
Asterisk provides Voicemail services with Directory, Call Conferencing, Interactive Voice Response, Call Queuing. It has support for three-way calling, caller ID services, ADSI, SIP and H.323 (as both client and gateway). Check the Features section for a more complete list.
Asterisk needs no additional hardware for Voice over IP. For interconnection with digital and analog telephony equipment, Asterisk supports a number of hardware devices, most notably all of the hardware manufactured by Asterisk's sponsors, Digium. "
Software: http://www.asterisk.org/
Hardware: http://www.digium.com/
This in and of itself is not a game-changer. This is the equivalent of the "garage hosting" model of ISPs that was prevalent through the early 90s, with Mr. Modem Bank selling to local customers. However, it presages what larger players can do with the same technological building blocks -- after all, those little ISPs led to AOL and Earthlink.
"Asterisk is a complete PBX in software. It runs on Linux and provides all of the features you would expect from a PBX and more. Asterisk does voice over IP in three protocols, and can interoperate with almost all standards-based telephony equipment using relatively inexpensive hardware.
Asterisk provides Voicemail services with Directory, Call Conferencing, Interactive Voice Response, Call Queuing. It has support for three-way calling, caller ID services, ADSI, SIP and H.323 (as both client and gateway). Check the Features section for a more complete list.
Asterisk needs no additional hardware for Voice over IP. For interconnection with digital and analog telephony equipment, Asterisk supports a number of hardware devices, most notably all of the hardware manufactured by Asterisk's sponsors, Digium. "
Software: http://www.asterisk.org/
Hardware: http://www.digium.com/
Tuesday, November 04, 2003
Ray Ozzie and Power to the Edge
Ray says:
Power to the Edge: A new book by Dave Alberts and Richard Hayes - open sourced in its entirety by CCRP.
This book is truly a must-read for anyone interested in decentralization and the social and organizational relevance of shifting power to the edge, whether in a commercial or a defense context. As you read about the technology enablers of the edge, it'll become clear why products such as Groove - as COTS enablers of the fully-networked collaborative environment - have such immediate relevance to the defense community.
Note to self, and world: Read this thing. Yes, all 303 pages. It looks really interesting, and the DoD is funding most of the fundamental research in America right now.
Ray says:
Power to the Edge: A new book by Dave Alberts and Richard Hayes - open sourced in its entirety by CCRP.
This book is truly a must-read for anyone interested in decentralization and the social and organizational relevance of shifting power to the edge, whether in a commercial or a defense context. As you read about the technology enablers of the edge, it'll become clear why products such as Groove - as COTS enablers of the fully-networked collaborative environment - have such immediate relevance to the defense community.
Note to self, and world: Read this thing. Yes, all 303 pages. It looks really interesting, and the DoD is funding most of the fundamental research in America right now.
Tuesday, October 28, 2003
Sony: Dot-Bomb Stock?
As Silicon Valley creakily starts to turn around and begin to innovate again, I have been required to fend off the doubts and dismissals of non-local naysayers who came late to the last tech party, and cannot believe that real value is created through the venture process. One of the favorite gambits is to point out the utter collapse of equity valuations even among the "real" companies in the tech sector -- those who had products and revenues, like Cisco or Sun or Inktomi.
This quote in today's NYT article on the Sony layoffs was thus a revelation: "Before today's announcement, Sony's stock closed at 3,960 yen, about 89 percent off its lifetime high of 33,900 yen in March 2000." Hmmm. 89 percent for a device company? 89 percent for a 50-year-old industrial bellwether? 89 percent for a company based 5000 miles from San Jose? Maybe, just maybe, the fault of the bubble was something called the stock market, not someplace called Silicon Valley.
Innovation is happening. These are good times for those who create.
As Silicon Valley creakily starts to turn around and begin to innovate again, I have been required to fend off the doubts and dismissals of non-local naysayers who came late to the last tech party, and cannot believe that real value is created through the venture process. One of the favorite gambits is to point out the utter collapse of equity valuations even among the "real" companies in the tech sector -- those who had products and revenues, like Cisco or Sun or Inktomi.
This quote in today's NYT article on the Sony layoffs was thus a revelation: "Before today's announcement, Sony's stock closed at 3,960 yen, about 89 percent off its lifetime high of 33,900 yen in March 2000." Hmmm. 89 percent for a device company? 89 percent for a 50-year-old industrial bellwether? 89 percent for a company based 5000 miles from San Jose? Maybe, just maybe, the fault of the bubble was something called the stock market, not someplace called Silicon Valley.
Innovation is happening. These are good times for those who create.
Friday, October 03, 2003
The Metaphor and Reality of Silos
Every time someone starts talking about legacy enterprise applications, the term 'silos' jumps into the conversation. Silos are bad. We need to break away from all these old vertical silos to something horizontal, and dynamic, and cool. It's worth remembering a few things about silos:
1) They are designed to hold grain in and keep the rain outside
2) They also keep rats away from the grain
I believe that a new generation of dynamic software will improve the way enterprises operate immeasurably, within the next decade. And I think that whatever solution emerges will make considerable use of silos.
Every time someone starts talking about legacy enterprise applications, the term 'silos' jumps into the conversation. Silos are bad. We need to break away from all these old vertical silos to something horizontal, and dynamic, and cool. It's worth remembering a few things about silos:
1) They are designed to hold grain in and keep the rain outside
2) They also keep rats away from the grain
I believe that a new generation of dynamic software will improve the way enterprises operate immeasurably, within the next decade. And I think that whatever solution emerges will make considerable use of silos.
Wednesday, October 01, 2003
Thursday, August 07, 2003
Last single person, turn off the Match.com?
It has just dawned on me that if forums like Match really are fundamentally more efficient at putting couples together than were traditional dating methods, we should see a measurable decline in the number of single people in America over the next 10 years as a percentage of the population, and possibly an absolute decline. Aside from some hard core of want-to-be-alone people, most single people are "waiting for the right one" and if you can find the right one faster, their numbers will go down. While there will always be churn and turnover, this should be a measureable artifact. It's kind of like the way that just-in-time inventory systems reduced the amount of stuff sitting in warehouses...
...associated wonder -- did all the Internet job sites reduce the number of unemployed, there during the bubble, through a similar effect? There aren't necessarily enough jobs for people (or people for jobs, for that matter) in the way that men and women are both approximately the same in number, and approximately the same in intent. So the two cases are clearly not directly comparable... but may be analagous.
It has just dawned on me that if forums like Match really are fundamentally more efficient at putting couples together than were traditional dating methods, we should see a measurable decline in the number of single people in America over the next 10 years as a percentage of the population, and possibly an absolute decline. Aside from some hard core of want-to-be-alone people, most single people are "waiting for the right one" and if you can find the right one faster, their numbers will go down. While there will always be churn and turnover, this should be a measureable artifact. It's kind of like the way that just-in-time inventory systems reduced the amount of stuff sitting in warehouses...
...associated wonder -- did all the Internet job sites reduce the number of unemployed, there during the bubble, through a similar effect? There aren't necessarily enough jobs for people (or people for jobs, for that matter) in the way that men and women are both approximately the same in number, and approximately the same in intent. So the two cases are clearly not directly comparable... but may be analagous.
Friday, July 25, 2003
The blogosphere increases... by one fine member
Robert Cringely (who, incidentally, is from Apple Creek, Ohio, a wide spot in the road just miles from where I grew up in Wooster, Ohio) recently posted a piece on "Snapster" -- his plan for reforming the music industry according to more rational lines. Since I'm no lawyer, I referred this piece to my favorite legal beagle -- Craig Calcaterra, my long-time friend, road-trip partner, and sometimes associate in vaguely dubious enterprises -- for his response. His response was so good that I urged him to start a blog -- which he now has done. Go check it out -- shyster.blogspot.com. On this issue, he's got Cringely nailed.
Robert Cringely (who, incidentally, is from Apple Creek, Ohio, a wide spot in the road just miles from where I grew up in Wooster, Ohio) recently posted a piece on "Snapster" -- his plan for reforming the music industry according to more rational lines. Since I'm no lawyer, I referred this piece to my favorite legal beagle -- Craig Calcaterra, my long-time friend, road-trip partner, and sometimes associate in vaguely dubious enterprises -- for his response. His response was so good that I urged him to start a blog -- which he now has done. Go check it out -- shyster.blogspot.com. On this issue, he's got Cringely nailed.
Wednesday, July 02, 2003
Subscribe to:
Posts (Atom)
