Internet State Change

My working hypothesis for the past thirteen years is that, in the long run, the Information Revolution will form an S-curve like previous industrial revolutions. The three major areas of an S-curve are a) an increasing rate of growth, b) a declining rate of growth and c) stagnation or shrinkage of growth.

Google's traffic has entered stage c).
Craiglist's traffic has entered stage c).
Google searches for "porn" have entered stage b).
Technorati's traffic has entered stage b).
Social networking site traffic has entered stage b).
Wikipedia entered stage b).

Part II, additional graphs

Tukwilla Trace

YOU!
71.217.70.26 - - [17/May/2007:16:11:08 -0400]...
(compatible; MSIE 7.0; Windows NT 5.1; stumbleupon.com 1.927; .NET CLR 1.1.4322)"

Yes, YOU!
Hiding there behind the anonymizer.
216.127.72.7 - - [17/May/2007:15:02:25 -0400]...
(compatible; MSIE 7.0; Windows NT 5.1; stumbleupon.com 1.927; .NET CLR 1.1.4322)"

You in Tukwilla!
Quispy Queepy Wiccans!

How many times do I have to quote Ferris Bueller?!!

"You're still here?
It's over.
Go home."

Divorce

My divorce took twelve minutes.

My first offer was $700 per month for four years.
My second offer was $600 per month for four years.
My last offer was $400 per month for four years.

Now my ex-wife gets zero money.

The Squeeze

I've been all over this country during the past few years; Oregon, Seattle, Las Vegas, Phoenix, Boise, NYC, New Jersey, Virginia, Atlanta, Miami. In Boise, I ate dinner with an Australian and I asked him how the U.S. differed from other countries. He said, "Everyone is in a hurry and has no time or interest in other people".

In Seattle I asked an ex-Marine the same question, he'd been living in Australia for the past decade.
"Electronic gadgets. Cell phones, mp3 players, cameras. Everyone is always fiddling with them".

And that's what I've seen.
Self-involved people without much introspection or integrity.
I can't think of many people that I respect.
Maybe that's just the IT industry in general now.

Last week I wrote "prices for everyday goods, commodities, utilities, etc, will rise faster than incomes and people with little or no disposable income to absorb the loss will... fail".

And today I read -

"Wages are not keeping up with costs. There is no other way to look at it."

I'll be glad to leave Miami.
It's not worth the trouble.

Forever Autumn

JOURNALIST: For three days I fought my way along roads packed with refugees, the homeless, burdened with boxes and bundles containing their valuables. All that was of value to me was in London, but by the time I reached their little red-brick house, Carrie and her father were gone...

Like the sun through the trees you came to love me
Like a leaf on a breeze you blew away

As if to hide a lonely tear
My life will be Forever Autumn
'cause you're not there
'cause you're not here
'cause you're not here

JOURNALIST: Fire suddenly leapt from house to house, the population panicked and ran - and I was swept along with them, aimless and lost without Carrie. Finally, I headed eastward for the ocean, and my only hope of survival... a boat out of England.

28 Days Later

I headed westward out of Miami's infected zone towards my only hope of survival, a fortified compound in Idaho. Or aboard a 767 bound for Saudi Arabia.

Few people understand the nature of the S&L Crisis of late 1980s. During a period of abnormally low interest rates, S&Ls borrowed short-term money and loaned it out in the form of long-term mortgage loans. The S&Ls profited from the spread between short and long money. Borrow for 5%, loan for 8% equals 3% profit.

But when short-term interest rates rose higher than rates on those long-term loans, continuous profits turned to continuous losses and after a few years the S&Ls began to fail en mass.

This time, consumers (and to some extent, the Federal gov't) play the role of the S&L, having borrowed short (ARMs, credit cards, t-bills) but loaned long (30-year mortgage commitments, 7-year car loans, Medicare and Social Security obligations).

As short-term rates rise, the short-to-long spread will shrink and then turn negative. In other words, prices for everyday goods, commodities, utilities, etc, will rise faster than incomes and people with little or no disposable income to absorb the loss will... fail. It won't become a full-blown crisis for another year but it will be the most important issue of the 2008 presidential election.

It's possible that the Federal Reserve can reduce short-term rates to new lows and temporarily prevent a credit collapse but there's already widespread abandonment of the U.S. dollar in process. Adam Hamilton has made several good macro trend calls since the
start of the Crash in 2001...

" as far as I am concerned the USDX breaking below 80 to new all-time lows is all but a fait accompli. It is inevitable. Investors not ready for it could really be hurt, especially if dollar and US-asset selling intensifies under 80"

The good news is that as gasoline becomes more expensive, so does outsourcing.

What People Email

It's interesting to see which links are forwarded through email...

76.110.165.110 - - [09/May/2007:21:13:13 -0400] "GET /roller/images/florida/AirSeaLauderdale2.jpg HTTP/1.1" 200 64578 "-" "Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 5.1; FunWebProducts; .NET CLR 1.0.3705; .NET CLR 1.1.4322; .NET CLR 2.0.50727)", i.e.

The Link

Here's some better pictures from today -

Ft Lauderdale Air & Sea Show

This weekend I got sunburned...

Dominique Revisited

Yes, I can see you logging in and reading, Dominique...

64.27.13.94 - - [29/Apr/2007:05:48:21 -0400] "GET /roller/page/realmeme?entry=mono_project_meme+ HTTP/1.1" 200 15305 "-" "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; .NET CLR 1.1.4322)"

Yes, I can see you double-checking "the cloak" IP address shortly after I mistakenly called it "anonymouse"

64.27.13.94 - - [07/May/2007:01:36:09 -0400] "GET /roller/page/realmeme?entry=www_the_cloak_com HTTP/1.1" 200 15319 "-" "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; .NET CLR 1.1.4322)"

Yes, I can see that you read at least 50 pages this weekend over several hours...

64.27.0.108 - - [05/May/2007:17:34:39 -0400] "GET /roller/page/realmeme?entry=status_update HTTP/1.1" 200 14915 "-" "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; .NET CLR 1.1.4322)"
64.27.13.94 - - [06/May/2007:00:31:11 -0400] "GET /roller/page/realmeme?entry=recession_meme HTTP/1.1" 200 16780 "-" "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; .NET CLR 1.1.4322)"
64.27.13.94 - - [06/May/2007:00:31:13 -0400] "GET /roller/page/realmeme?entry=the_strange_change HTTP/1.1" 200 13993 "-" "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; .NET CLR 1.1.4322)"
64.27.13.94 - - [06/May/2007:00:31:14 -0400] "GET /roller/page/realmeme?entry=myspace_meme HTTP/1.1" 200 15886 "-" "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; .NET CLR 1.1.4322)"
64.27.13.94 - - [06/May/2007:00:31:25 -0400] "GET /roller/page/realmeme?entry=my_first_topping_update HTTP/1.1" 200 14171 "-" "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; .NET CLR 1.1.4322)"

Yes, I can see when you read my funnier posts and messages to you...

64.27.13.94 - - [30/Apr/2007:16:49:16 -0400] "GET /roller/page/realmeme/?entry=do_the_maca_lena HTTP/1.1" 200 15103 "-" "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; .NET CLR 1.1.4322)"

And yes, I can even see when you click on the main page of the site but not the blog...

64.27.19.59 - - [07/May/2007:20:13:02 -0400] "GET /Main HTTP/1.1" 302 - "-" "Mozilla/4.0 (compatible ; MSIE 6.0; Windows NT 5.1)"

Come on.
Decide what you want to do but don't keep baiting me.

IBM Layoffs - My Extrapolation Confirmed

Wow. I wrote Crosscutting Information Domains two months ago. I may have struck closer to the mark than I thought...

"IBM and IBM's Businss Partners's... strategy is moving in the opposite direction if we want higher success rates for increasingly complex projects... it appears that IBM has shifted their own internal strategies towards increasingly specialized workers, so that more workers are involved on a given project, but for less time and work.

That's just can't work right over an increasing scale of complexity and project size, since it multiplies communcation costs at an exponential rate."

Today's IBM rumor from I, Cringely - "LEAN is about offshoring and outsourcing at a rate never seen before at IBM... The problem with LEAN is that offshoring on this scale creates huge communications and logistical problems, doesn't generally improve customer relations, and won't save money for years... Lopping off half the technical staff, as Global Services is apparently about to do, will eliminate much of the company's traditional wisdom and corporate memory"

I have personal experience with IBM Business Partners engaged in office politics and deceptive sales techniques. And I'm currently working for a customer that wanted an "independent evaluation" of their mistrust of IBM.

I like IBM.
I like their products.
But I'm glad that I'm cross-developing on cheaper, competitive alternatives.

I Often Wonder

I often wonder why I bothered when so many Americans around me don't. There is so much greed, fear, selfishness or hate. I feel like I'm surrounded by ignorant monkey people whose main purpose in life is to eat, drink, fuck and grunt.

BotNet Talking Points

[A building block of my Defcon presentation proposal]...

Recent Google article which dissects "Clickbot.A.", a botnet engaged in click fraud during the past year.

My notes & thoughts -

1) The authors make no mention of memetic analysis to detect click fraud, apparently relying on "click pattern analysis" to uncover robotic patterns. It's possible that I've assembled an alternative method of fraud detection before Google! Awesome if I really did it.

2) The article claims that "IP diversity usually is not enough to hide click fraud... traffic generated by them is identifiable". This matches my own experience but I strongly suspect that it will not remain true for much longer. I now know that I could substantially camoflage the click patterns of the botnets hitting my own website.

3) Google's description of Clickbot.A's methodology confirms my own beliefs... even today they are surprisingly sophisticated, relying upon low-level traffic, feedback loops, redirections and secondary traffic mimicry.

4) It's probable that botnets were originally designed to defraud porn sites and they retain vestiges of their ancestry. It's possible that porn sites are suffering far greater fraud than Google. Porn sites may be more lucrative customers for memetic fraud detection techniques & software than major search engines.

5) The authors estimate Clickbot.A's fraud potential in the $50K range without specifying a time period. I suspect that this could easily be a monthly loss, adding up to yearly losses in the $500K+ range.

Gaming The New Miner Model

[A building block of my Defcon presentation proposal]...

Meme Miner originally used Dejanews.com keyword traffic as a proxy for real meme propagation. I recently revised my model for increased accuracy by cross-checking memes within other venues. I chose Alexa.com for network traffic and Google/Trends for search engine traffic. If a meme is propagating through a human population, network traffic and search traffic should show growth which is proportional to the Dejanews.com proxy traffic -

The new model did detect traffic anomalies, specifically at IAFF.com. My original purpose wasn't click fraud detection but those anomalies led to circumstantial evidence of botnet manipulation. The more I work with this new model, the more I realize how susceptible the Internet is to click fraud.

My current model is easily deceived by a simple addition -

The S-Curve, Meme Propagation and Rate Of Change

[A building block of my Defcon presentation proposal]...

The progression of a meme through the Ideosphere is typically described as an "S-Curve", an exponetial growth of exposure as the meme is transmitted from one person to two, two to four, four to eight, etc, until saturation is reached...

In abstract terms (or mathematical terms, ho ho), the propagation of a meme through the general population looks like this...

Why do we care about the S-curve? Because it represents rate-of-change which is a predictor of future behavior and return on investment. The optimum entry point is slightly prior to the "inflection point" (i.e. maximum rate-of-change, which is the maximum rate of meme propagation through the Ideosphere)...

If we can measure an S-curve for a technology or trend, we can predict entry and exit points for that trend which maximizes profit and minimizes risk...

Gaming Alexa.com With Botnets

[A building block of my Defcon presentation proposal]...

Statistically, certain percentage of hijacked computers in a Botnet have the Alexa.com toolbar. If the Botnet creates illusionary traffic for Google Adsense click fraud, a side-effect is an increase of illusionary traffic to Alexa.com as well.