Social networks and the expected implications of exponential growth

Orkut is huge - in Brazil. Nowhere else I think. Myspace is humungous, growing from nothing to everywhere in no time at all.

That's exponential growth. One day there's one lily pad on the pond, a few weeks later the pond is half full, a day later it's full. The last doubling is the big one.

It's fascinating. Why does one network take off and another founder? What makes a movie a hit, or a toy "hot"? Why do some diseases simmer in the Heart of Darkness and others make it to Hollywood? Why does one primate inherit the earth (for now) and all the others die?

There's a fascinating mixture of determinism and contingency. Sure the primate had some handy features, but evolution could have taken a very different course. Orkut in Brazil and MySpace in the US had a combination of both value and serendipity that let them grow. (Survival, of course, is another matter.)

Is LinkedIn doing the same thing? LinkedIn is a snotty "exclusive" (hah!) network that seeks to amplify and extend social networks among "executive" (double-hah) types. Believe me, any club that has me as a member is neither exclusive nor executive. What catches my attention, however, is the rapid growth.

A friend (Jacob) sent me an invitation a few months ago, and since I respect his judgment I decided to play with it. I invited my wife. Then my alma mater invited all alumni and students to join up. Yesterday I received two separate invitations to join at an address unconnected to my existing account.

Clearly LinkedIn is now past the inflection point of the exponential curve. A few more doublings and my dog will be getting invitations (hmmm. No reason I couldn't give Kateva an email address and a resume ...).

It will be interesting to see where it goes next and how the stream is monetized. At the moment it is a handy way to keep in touch. They do limit invitations but if you're a friend/family and you want one just email me ...

Life on the net is making the exponential a familiar phenomena to many people ...

Update 8/4/06: I just got my first invitation to connect from a corporate recruiter. Ok, so now I see one way they'll monetize this -- the recruiters will sign up for a lot of expensive value added services. I made the connection -- the recruiters I know are smart and careful types who don't waste time. Of course if they abuse the connection I'll remove it. It will indeed be interesting to see how this evolves. The 21st century is all about identity, reputation and reputation management. (Credit to Charles Stross from writing about this quite well in Accelerando.)

Kansas returns to Reason

The infamous Kansas State Board of Education is once again in the hands of Rationalists. The election was fought in large part over whether natural selection was a well accepted scientific principle. I am pleasantly surprised to discover that the citizens of Kansas agreed that natural selection is fundamental to modern biology and deserves a place of prominence in science ducation. On the other hand Creationism/ID does not belong on the science curriculum.

I would, however, allow mention of 'design' as one of the solutions to the infamous Fermi Paradox. Of course the FP is not usually taught in high school.

The GOP votes to ban the Internet (from Libraries)

The GOP run House of Representatives has made it illegal to use most of Internet in a library ...

Good Morning Silicon Valley: All in favor of hysteria, panic and misinformation, say "aye"

Given the recent outcry over MySpace, I suppose I shouldn't be surprised that the U.S. House passed the Deleting Online Predators Act (DOPA) with an overwhelming majority. But I can certainly be disappointed. With a 410-15 vote (410-15!!!) Thursday, politicians approved the bill, which will block access to social networks and Internet chat rooms in most federally funded schools and libraries....

...Here's how DOPA defines social networking sites:

(i) is offered by a commercial entity;
(ii) permits registered users to create an on-line profile that includes detailed personal information;
(iii) permits registered users to create an on-line journal and share such a journal with other users;
(iv) elicits highly-personalized information from users; and
(v) enables communication among users.

Great work, Congress, you've just barred anyone who depends upon their local libraries for access to the Web from viewing eBay, Yahoo, MSN, AOL and Amazon.

I think they've also banned the upper-crust corporate-focused LinkedIn social network. Would a Democrat dominated House do something this stupid?

I admit I can't prove they wouldn't, and a lot of Democrats voted for this. On the other hand, I suspect they voted for it in self-defense (imagine the ads -- my commie opponent supports pedophilia!). I would, however, like to know the 15 representatives possessed of both safe seats and a positive IQ.

I'll bet a Democrat controlled House would not have done this. Think of that next time you get to vote ...

Why we don't remember the future, and other consequences of the 2nd law

Cosmic Variance is hurting my head again. Coincidentally, I've been listening to the IOT episode on the 2nd law of thermodynamics (excellent) and just the other day I tried to explain time's arrow to my 7 yo ...

Boltzmann’s Anthropic Brain | Cosmic Variance

... Suddenly, a thermodynamics problem became a puzzle for cosmology: why did the early universe have such a low entropy? Over and over, physicists have proposed one or another argument for why a low-entropy initial condition is somehow “natural” at early times. Of course, the definition of “early” is “low-entropy”! That is, given a change in entropy from one end of time to the other, we would always define the direction of lower entropy to be the past, and higher entropy to be the future. (Another fascinating but separate issue — the process of “remembering” involves establishing correlations that inevitably increase the entropy, so the direction of time that we remember [and therefore label “the past] is always the lower-entropy direction.) ...

After this the essay gets much harder. Bayes Theorem makes an appearance, though it is not labelled. Crossing Bayes with the antrhopic principle yields yet more disturbing implications. Now if only CV would toss the Fermi Paradox into the mix ...

Pandora: music radio and recommendations

Jacob Reider pointed to Pandora

Ad supported radio that suggests music based on one's likes. Oldest trick in the book (I think Firefly did this in the early '90s), but surprisingly rarely done. I'm suprised the Apple Store doesn't do this ... Amazon used to, but their recommendations service has all but vanished ...

Airport security 1995 - an insider's story

via Schneier. The beautiful thing about the web is all kinds of people can tell their stories, and people like Schneier tell us about the interesting ones. An insider's story about working airport security in 1995 is quite good. I liked the trick where he learned to carry any of the FAA test items through the metal detector without activating the alarm.

I wonder if things are different now; the security people I see at airports now seem a cut above this description. On the other hand, airport security is not terribly important. As has been pointed out many times, the most important post-9/11 security measures are:

1. armored cockpit door
2. pilots treat a hijacking as a suicide attack
3. passengers are inclined to resist (1 and 2 are more imporant).

Spolsky on pricing theory

Why does what you buy cost what it costs? Read Camels and Rubber Duckies - Joel on Software. I have a seriously dense textbook on pricing theory. Spolsky covers the key points in a single article. If you buy anything, it's worth reading.

It's non-trivial.

BTW, one technique he doesn't mention is sandwiching. Apple does this brilliantly. Sell one thing cheaply, one in between, one expensive. The cheap one is a bargain, but it's missing something everyone wants. The expensive one is wicked. The result is people will happily buy the middle one, and will pay more than they would have if the cheap one (which has a crummy margin) didn't exist.

The bottom price exists only to elevate the middle price.

An Interview with Charles (Bell Curve) Murray

About 10 years ago Charles Murray and Richard Hernstein wrote "The Bell Curve", a book that caused great outrage on the American left. I never read the book, but I dimly recall that the authors claimed that there was a general intellectual capacity that was roughly measured by IQ testing, that IQ was largely genetically determined, and that some ethnic groups had bigger average IQs than others. I think he was fond of South Koreans in particular, but I may be losing it there (I have a family relationship to Korea that probably affects my memory).

At the time I didn't say much, because, although I didn't care for the tone of the quotes I read, I suspected the substance would hold up. I thought the evidence even then was pretty strong that IQ was genetically determined [1], and that the main environmental effects were intrauterine and served only to lower the genetic limit. I also felt, with less data, that it was likely that there was indeed a general ability to synthesize and problem solve, and that it had some correlation with IQ test results. As to the ethnic relationships I was and am agnostic, but I didn't think the data was there to rule it out. So the book seemed plausible, albeit infected by an off-putting arrogance.

So I was interested in this Charles Murray interview featured in an obscure web site. Murry may even have mellowed slightly, though his religious devotion to Libertarianism is only mildly abated. It is interesting reading.

By the way, one way in which I believe I differ from Murray is that I don't think being smarter makes a person a more deserving human being. I value traits like integrity, compassion, mercy, wisdom, humility, curiousity, kindness, forgiveness... Murray seems to have adopted the Libertarian faith that intellect is the measure of perfection. (I wonder how they'll feel about our silicon heirs?)

[1] Yes, I've read some recent reviews claiming environment was more important than we'd thought. I thought the articles (NYT) were so dull and confused they weren't worth writing about.

Stress, disease and early aging

The NYT Magazine has a long article on disease and aging though I think the lead family photo is a big misleading (look yourself and contemplate). It doesn't have many surprises in it. It's one of a series of discoveries and reviews that show we age more slowly than we once did, largely because we're sick less often.

The main environmental influences are probably prenatal and certainly before age two (mothers should never smoke), but, there are later effects too. At least for the moment, we are spared much of the disease, parasites, and malnutrition that afflicted our ancestors. As a consequence, we age more slowly.

This is not a surprise to anyone who's studied old family photos. Our ancesters were old at 40 -- rather than 50-55.

FuturePundit, which pointed me to the NYT article, also features a related post about a gene that can be used to tell a person's biological age rather than their chronologic age.

I'm still waiting for the article that shows that aging is non-linear, and that we experience 'bursts' of aging after certain environmental triggers. The old folk tale about 'aging a year in a night' will be shown to be correct.

Geeks don't use Travelers Insurance

JOS calls out a comically stupid and offensive ad campain by Travelers Insurance. I wonder how much business they once had in Silicon Valley.

I assume that each year ad execs gather to nominate the absolutely worst ad effort of the year. This one is reaching for the decade award.

Dyer has about 10 new articles online

The journalist, military historian, arabist and eccentric web author has about 10 new June/July articles. This journlist doesn't do he said/she said articles -- and his record is good. Dig in.

The Ultra Portable Mobile Computer is doomed

The UPMC is crushed by a 10 year old Apple Newton in a head-to-head struggle. That's pathetic. The UPMC is dead. Move along folks ...

Black skin, SPF, and when a feature becomes a bug

[Update 7/31: read the comments. The BBC article really mangled the study, and my imagination was overly active. As our commentator notes, these melanomas aren't even clearly sun related. Detection was also not a problem, the involved skin is pale. I really don't know what to make of any of this, but you probably shouldn't waste your time on this post!]

Very black skin has SPF 13; I assume that very pale skin has SPF 1-2 (the BBC article didn't say) ...

BBC NEWS | Health | Dark skin 'does not block cancer'

...people with dark skin are more likely to die from skin cancer than those with fairer skin...

Dark skin has increased epidermal melanin which provides a natural skin protection factor (SPF) - a measure of how long skin covered with sunscreen takes to burn compared with uncovered skin.

Very dark, black skin has a natural SPF of about 13 and filters twice as much UV radiation as white skin, for example.

The BBC article is pretty weak. We're talking about melanoma. Very black skin is protective against melanoma, but it also makes it harder to spot early melanoma. The detection impact is only relevant when melanoma can be treated -- a modern phenomena.

In low tech societies black skin protects against death by melanoma, in high tech societies we can cure early melanoma. So the detection problmes mean that black skin has gone from being a 'feature' to a 'bug' -- at least in terms of melanoma protection. It's still a feature in terms of protection against sunburn, even the best sunscreens require frequent reapplication.

This is analogous to climate change and fur color. White fur is protective when there's snow, a problem when the snow melts. The difference is that our environment is technological.

Asians and other brown skinned persons may have the best compromise here. A reasonable SPF protection (maybe 8?) with reasonably easy detection as well. Sunscreen there just improves the odds. Good news for my younger kids. Getting black kids and adults to wear sunscreen will be damn hard. An effective SPF 13 sunscreen that doesn't wash is good enough to avoid a lot of sunburn, and without the scourge of sunburn kids and adults aren't going to put sunscreen on...

Hezbollah, collective punishment, and the falling cost of havoc

In the winter of 2001 I thought a lot about the falling cost of havoc. Technology changes the course and nature of conflict, and the relation between states and entities.

The theme returns:

Israel Finding a Difficult Foe in Hezbollah - New York Times

.... Never before in history has a terrorist organization had such state-of-the-art military equipment,” from medium-range rockets and laser-guided antitank missiles to well-designed explosive mines that can cripple an advanced tank, General Amidror said.

The cost of advanced weaponry is falling relative to income. Offense is cheaper than defense. The "IED" (improvised explosive device) is not really improvised, and it has changed the balance of power in Iraq.

In time missiles that travel hundreds of miles, with a great deal of "intelligence" and optional remote guidance, will become commonplace.

We still haven't thought enough about how technology is changing the nature of conflict. I fear collective punishment and police states will become the rule. I wish we thought about this more.

Landis: the odd thing about the test results

On NPR a specialist in doping described Landis results differently from the news stories. This is typical of what we're reading:

KRT Wire | 07/27/2006 | Landis denies cheating, admits Tour win will be questioned

...The test conducted at France's anti-doping laboratory in Chatenay-Malabry turned up high levels of testosterone, which medical experts say can increase stamina in the short run.

According to a statement posted on the Web site of Phonak, a cycling union official informed Landis' team that the rider's sample revealed 'an unusual level of testosterone/epitestosterone' when he was tested July 20.

The story on NPR was that his free testosterone was not in fact elevated. The epitestosterone was very low, so the ratio was elevated to about 11. Anything above 6 is out of the accepted range, so a value of 11 is extreme.

I was hoping someone on medlogs would do a more clinical analysis, but so far the press reports are just repeating the elevated testosterone line. I admit I'm harboring the faint hope it's something innocent, but I'm also curious about what would mess up the ratio. An effect of another banned substance? It's been too long since I knew my endocrinology ...