David Reed has written an insightful insightful short piece on the limitations (and inevitability) of scientific metaphors. For example:

Most radio engineers work in a particularly inapt metaphor - but they don't know it. That metaphor still includes the essence of the idea called the luminiferous aether (except they call it the "spectrum"). The metaphor includes the idea that a "bit" is a unit of energy (rather than what Shannon defined it to be - which is something that represents correlated probabilities among parts of a system). This confuses the thing (bit) with one possible instance of the thing (a coded pattern of energy or matter).

Communications regulators work in an even more inapt metaphor...

My son Nathan, 11, yesterday asked:

Why can you sell your soul to the Devil but not to G-d?

Adina Levin makes the case that a sufficiently usable search engine that has indexed a sufficiently large text base — i.e., Google — in effect is a KM system.

Yup. In short: If you know where things are, you don't ever have to clean up.

A Mini-Bogus Contest: Now that Tivo has taken over our home — last night it deposed our atomic clock, frog-marching it out the door — how might this jealous god recast old programs in its image? For example:

Leave It to Tivo

Tivo Knows Best

I Love Tivo

Get Tivo!

Tivo in Charge

Tivo's Heroes

Your Tivo of Tivos, starring Sid Caeser

I'll be talking on the radio today about why I want to have hot monkey sex with Google. This is the first of what I hope will be a continuing series of tech commentary for WBUR's "Here and Now," carried by 44 stations. My piece will be on at 12:20pm. Come listen to me make a fool of myself. Again.

Yes, I'm on the road, but in about as anti-Kerouackian way as possible — fly in at night, stay at a Sheraton, watch TV, give a talk, leave the next morning. I don't think this qualifies me as a Dharma Bum.

I just gave a keynote at a conference on integration (think KM, portals, XML) put on by the Delphi Group. The conference is in Reston, VA, and has drawn people from around the country and even overseas, yet another anecdotal indicator that the conference business may be springing back. (Since I make much of my living as a speaker, this is of more than academic interest to me.)

Delphi surprise-inducted me this morning into their League of Honorees (I didn't quite catch the name of the group). I was in a post-speaking Zone of Confusion and missed the details, but I think we get capes and fight crime. (No, but seriously, I appreciate the honor.)

Jon Udell uses my blog about the phone number trick at Google to talk about what happens when what used to be public but obscure becomes public and easily accessible.

He also mentions the IETF ENUM initiative "which seeks a mapping between telephone numbers and the DNS." The official IETF white paper on usage scenarios of this mapping says:

In a pure IP environment, ENUM will allow end users to be identified by a commonly used name (i.e., their telephone number) for a variety of applications. ... [E]nd users can change IP service providers without having to change their destination identification. For example, an end user can change their underlying e-mail address from [email protected] to [email protected] but, with ENUM set up to handle e-mail ... still be reached by having ENUM-enabled mail clients send mail addressed to their �telephone number� (e.g., mailto:+1-973-236-6787).

For voice services, ENUM will allow the easy end user identification described above as well as interworking between terminals [phones] on the PSTN and on the IP-based network. It may also allow for the implementation of more advanced services, such as find-me. For voice communication starting on an IP-based network, ENUM can be used on each call to determine the preferred type of destination based on the priority of the network termination available. For voice communication starting on the PSTN, ENUM is more likely to be used where at least one of the destinations of the call exists on an IP-based network.

Frank Paynter responds to my blog about designing a new peace symbol:

In the age of hip-hop, local students didn't want to associate with the super-annuated ND semaphore signal. The symbol we adopted isn't as easily sketched, but it makes a nice logo. Here's what we came up with, after heartfelt discussions last winter (logo is in upper left corner)

Fox proposes televised coverage of arms inspectors in Iraq so "Viewers could decide for themselves if the inspectors are being allowed to do their jobs"?

A pyramid scheme for "Women Helping Women" sweeps the nation?

A movie featuring stupid, gross stunts is #1 at the box office?

Eric Norlin responds to the article in my newsletter on the DigitalID World conference. [Note: Eric reports that he's having trouble with his permalinks.]

Stanton Finley has sent a message to a bunch o' bloggers and others asking us to re-design the peace symbol.

The old peace symbol represents the letters "ND" in semaphor language. Since Nuclear Disarmament no longer tops the peacenik agenda, we could indeed use a new symbol.

I am refraining from suggesting my proposed "Forgive me" hand gesture, but I have no other ideas. Suggestions anyone?

Stan also asks us for a "manifesto" of peace. Here's mine. (Prepare for Hippie Resurgence Syndrome.)

All people are created equal. We all have an equal right to life, liberty, and the pursuit of happiness.

All people are created different. There is value in that difference and we need to preserve it.

All people are connected. We are connected by geography and responsibility, and, if we would let it, by love.

Peter Kaminski responds to my neologizing "Google URLs" ("A search query that puts a page at the top of Google's returns list"):

"Google URLs" are the same as the mechanism behind Robust Hyperlinks (For a joke intro, start here.)

Robust Locations there is a pretty cool trick, too.

You make a page robust, according to this paper, by running their free, open source software that adds a "lexical signature" about five words long, a hash of your document content. People can find your page by searching for its signature, so even if you move the page, Google (or whatever) will find it for you.

The problem is that the signature isn't necessarily memorable. For example, the signature of www.cluetrain.com is "html intranetworked uznajut happytalk stemmens" whereas its Google URL is "cluetrain."

Seth Johnson points out that two senators, on this sad day for that institution, "are starting to show some truly helpful cluefulness" :

"Digital media simply shouldn't be more restricted than other copyrighted items," [Ron] Wyden [D-Oregon] said. "Digital technology is a great step forward, and it would be a shame to take a big step backward on consumers' rights when it comes to using this material."

He and Chris Cox (R-Calif) are sponsoring a bill to make this idea all legal and everything.

Senator Paul Wellstone's death has narrowed our political vision yet further. We're down to about what can be seen through the sights of a gun.

I was at the Newseum, a site that thumbnails newspaper front pages from around the world. (Thanks, Dan Pink.) I clicked on the Australia's Courier Mail and saw their tag line:

Unfortunately, I freudianly misread it as:

Two letters can make the difference between marketing and truth, eh?

Adina Levin, having read my ramble about Stephen Wolfram's presentation at PopTech, recommends Kurzweil's appreciation of him, which she has summarized here. The Kurzweil piece is well-written and leave us humanities majors behind about a third of the way in.

There's also a good article — again only two-thirds beyond my comprehension — by Steven Weinberg in the NY Review of Books.

Steve Yost writes pithily about reading Wolfram. He says:

The repetitiveness of Wolfram's style led me to think that near the end he'd reveal that the book was generated using his main thesis as the initial condition of a CA algorithm. Now that would be a substantial example.

Proposed neologism:

Google URL (n) A phrase sufficient to bring a desired Web site to the top of the returns list at Google. E.g., "My real address is weird, so I gave him my Google URL: 'Locke die cast'"; "I couldn't remember the dictionary's web address so I used the Google URL 'American Heritage'"

I learned a lesson from Ernie the Attorney at PopTech.

Which should you bring to a conference if you want to be incredibly popular?

More than a couple of people noticed me and Ernie the Attorney blogging next to each other from PopTech. "Why?" they asked. "Why are you so focused on blogging the conference?"

Durn fine question to which I have less than durn fine answers.

1. I blog conferences for the same reasons that I blog in general: I don't know.

2. blogging forces me to pay attention, just as note-taking in general does.

3. Insofar as I'm engaged by what the speakers are saying, I want to be talking with them. Since conferences insist on maintaining a distinction between "panelist" and "audience member," blogging lets me participate. Best of all, I always get the last word.

But why real-time blog since post-session blogging enables me to reflect on what was said and write more thoughtfully? But post-session blogging means that after a full day at an intellectually intense conference like PopTech, followed by an intellectually intense dinner, followed by an intellectually intense dessert, I have to go back to my hotel room and write a @#$%!-ing blog entry. So, real-time blogging is better for me but worse for my readers.

[From PopTech] Alexander Shulgin, the chemist who invented ecstasy, carefully creates new molecules and then more carefully ingests them on the grounds that there's no other way to see the effect on the mind. He said wants to raise the question of the mind rather than the brain. But he never quite got to it, at least not explicitly. Shulgin does not believe that computers will ever imitate the mind, although they'll be good at imitating the brain. "The mind process is what I'm primarily interested in."

He's a charming presenter. Just when you think drugs have smoothed off the edges of his mind, he comes through with enough detailed chemistry to remind you that he is in fact a rigorous scientifist.

He gave a personal accounting of drugs he has known and loved, although he talked about them in their chemical names ("dimethoxyladida" is about as close as I can get) so it came out as a list without punctuation.

He railed against the "analog drug bill" that says that if a drug is "substantially similar" (vague enough for you?) to a schedule one drug (one with no medical use and a high potential for abuse), then it shall be treated as a schedule one drug.

He estimates that there are 200 known psychedelics. Given the growth rate, there will be maybe 2,000 in the next 10-15 years.

Now, he says, he's ready to begin his talk. (The clock has run out.) When he first took mescaline he realized there were parts of himself he had not been in contact with. Then he took another drug with just a small chemical difference and found important differences; for example, the flower he'd meditated on when on peyote he now tore apart to see what's inside. (He talks about drugs the way others talk about wines, I think on purpose.) A small brain effect can have a large mind effect.

He ended by describing a new molecule. "Here's a compound that's never been made. A whole new area of chemistry...I've not tasted it yet. I don't know what the interesting properties are. ... It doesn't know me either." So you have to do a careful introduction. "You start with a few micrograms. In case you make a mistake." There's an unlimited number of these. He finds this fascinating "and wants to do this for the rest of my life."

Fascinating person. Needs more time, and maybe an interview format. (I heard him and his wife on Terry Gross'ss Fresh Air a few months ago, and she drove the interview masterfully.)

[From PopTech] Vinge, the science fiction writer, talked about nutty stuff, presenting seriously insane ideas with the right mix of conviction and humor to suspend our disbelief.

Vinge began by giving us a Google URL, i.e., telling us to find a site by looking up "vinge technological singularity" in Google. That leads to a page about "the singularity," the extropian notion that "...we are on the edge of change comparable to the rise of human life on Earth." What will life be like for those before and after the change? This lead him to talk, with only a glimmer of a smile, about life for the "early post-humans."

"The surface of the earth may not turn out to be the best place to think." Maybe we'll have to seek out better places. (Better in what way? Better reading light?) Vinge points us to Hans Moravec's "Pigs in Cyberspace" that suggests that we convert the universe into a place that computes. [What a lead in to Wolfram, who speaks in the next session.]

Now for the next step. Vinge suggests that the "principle of mediocrity," along with Occam's Razor and entropic rules, are ways of getting answers when you don't have facts. The Principle of Mediocrity says that if you don't know what's going on, assume you're an average case. E.g., the earth isn't special, so the planets probably don't revolve around us. Moravec's conclusion is that the principle of mediocrity says that it is almost certain that we are ourselves living in a simulation. Says Vinge: "This is a moderately logical argment, especially if you are into this sort of thing" (i.e., if you're a nutcase).

But is this an "operationally significant issue"? Vinge says we might actually be able to tell if we're living in a simulation by "looking for the jaggies." (The "jaggies" is the stairstep effect you get with straight lines painted in inadequate resolution.) Perhaps, he suggests, the jaggies are the quantum mechanical anomalies. Apparently there are physicists who take this seriously.

[From PopTech] Sherry Turkle, who did the missionary work in the effect of social computing on the sense of self, talked on the psychology of artificial worlds. Brilliantly. In particular, she talked on the natuure of authenticity. The old concept, she says, doesn't hold. "We need to take our new relationships on their own terms." And these terms include accepting that the digital world's perfection is teaching us a new sense of imperfect human perfection.

The number one question journalists ask Turkle is if children will come to love their objects more than their parents. Turkle instead is interested in how love might change. What kinds of relationships with technology are appropriate? What is a relationship? What models of the self, intention and emotion are suggested by our current technologies? What habits of mind?

Turkle said that we think about our minds as machines more than ever (as in robots and psychopharmacology). She's concerned that with the shift to a computational model of the mind, there's been a diminishment of our appreciation of ambivalence (i.e., holding more than one idea in your head). In artificial worlds, the rules are too clear. But resistance is coming from a changing notion of human pefection. We need richer language for talking about our increasingly rich relationships with artifacts. [This is a topic near and dear to me.]

During the Q&A she said that rather than asking about the effect of video games on kids, we should be talking about what habits of mind games inculcate.

Killer soundbyte: "Windows is a powerful metaphor for the distributed self"

[From PopTech] Alvy Ray Smith has won two Oscars for technical achievement and was the founder of Pixar. He gave a terrific presentation on a single idea: "The simulation of human actors will not happen at any known time in any known way." Smith addressed the question in a far more interesting way than I'd expected. Yes, says Smith, there are technical issues that will be overcome via Moore's Law. But, the more important issue is that acting is an art.

Using as his reference Antonio Demassio's "The Feeling of What Happens: Body and Emotion in the Making of Consciousness", Smith made the case that machines can't have consciousness and thus can't do what actors do. They therefore can't generate graphical representations of convincing actors: There can't be a "React to the De Niro character's confession of adultery" subroutine that results in the good acting that Streep would do because that would require consciousness. Ray thinks that in his lifetime we will see a convincing feature film that's entirely digital, but it will be done by digitally representing a live actor. (In fact, Pixar hires animators based on their acting ability.)

Killer soundbyte #1, on the assumption that we'll build conscious machines: "It's a leap of faith that many people here are willing to take, but I call it faith-based science."

Killer soundbyte #2 on why Pixar has so far backed off of representing humans: "We have a word for almost human but not quite: It's 'monster.'"

Killer soundbyte #3: "Reality begins at 80 million polygons." Per frame. Toy Story had 5-6M polygons per frame, and Toy Story 2 had double that. "Then you have to model reality and map it onto those polygons." Woody had 100 controls in face. Al, the most complex guy in Toy Story 2, had a thousand. For an accurate human representations it might be hundreds of thousands.

[From PopTech] Paul Stookey, the Paul of Peter, Paul and Mary, opened Saturday morning with a great set. "Virtual Party" is intensely clever and "Love Rules" is genuinely moving.

[From PopTech] The afternoon led off with two interesting presentations, moderated by MIT Good Guy, Henry Jenkins.

Lauren Rabinovitz talked about the social importance of amusement parks at the turn of the previous century when there were 1,500 of them. By 1920, 75% had closed. She presented lots of interesting ideas about a phenomenon I know nothing about. For example, rides weren't at the heart of them at first, in part because the mechanics weren't sophisticated enough. But when rides did come to prominence, they enabled people to reverse the usual human-machine relationship, giving themselves over to the machine.

[From PopTech] Dressed in a hippie's idea of respectable - purple pants, multihued shirt, sun belt buckle, painted klogs - Rheingold eloquently makes the case that hopeful and unpredictable phenomena emerge from simple technology. His example is, of course, "smart mobs": the social organization that springs up around cell-phone text messaging. Much hinges, he says, on the emergence of trust and mechanisms for managing reputations.

Rheingold ended by asking: Will those mobile technologies be shaped by users or will we be tuend back into consumers?

During the Q&A, he said that online is a great place for people to express themselves but a bad place to make decisions. He also said that rule-less places are fine, but some sites need rules, e.g., no personal attacks allowed. (Both points conform with my experience.)

[From PopTech]: Kurzweil is an excellent speaker. Wicked smart. Primarily he made the case that the rate of innovation, and the rate of "paradigm shifts" (in quotes because he doesn't mean the term in Kuhn's sense) is increasing. Then he paints a picture of life in ten and twenty years: By 2010, computers will disappear will write directly to the retina and we'll be able to dip into virtual reality whenever we want. By 2029, we will have reversed engineered the human brain, and nanobots will do the red pill thing of putting our brain fully into the virtual world. And, of course, we will have computers large enough to hold an entire set of data about a brain. The brain is hardware, and by 2029, says Kurzweil, we'll have deciphered the brain's software.

I got to ask him my question from the audience. I said last summer I stood in a wheatfield that 100M stalks of wheat. If we take left-leaning is on and right-leaning as off, for 5 minutes, that wheatfield completely represented Casear's brain state when he was stabbed. So, I asked, it seems to me that hw-sw is entirely the wrong paradigm for the brain, intelligence, consciousness. (Unfortunately, I chose not to draw the explicit connection, in order to save time, and thus sounded like a lunatic.) So, I asked Kurzweil, how confident he is that by 2029 - given the rate of p aradigm change he pointed to - the sw-hw paradigm of the brain will still be in effect.

Kurzweil replied by distinguishing intelligence and consciousness. Whether machines will be conscious is a philosophical question that he stays away from. [Well, not in The Age of Spiritual Machines.] But, he said, he'll make a political prediction: we will take computers as conscious, "because if we don't, they'll get mad at us." Plus, he said, our consciousness will be augmented by computers so the line will be fuzzier.v

I remain convinced that the brain is no more hardware and software than the liver is. The issue is that software is symbolic. The eight light switches in my house have on and off states, but they only become a byte of information if I choose to take them that way. And the specific number the byte represents depends on my deciding to read from the top floor down, bottom up, east to west or west to east. A computer that mirrors the brain state only does so because we have supplied symbolic meaning to it. It thus is a picture of a brain but is not a thinking machine. That's what I think, anyway

I'm at PopTech in purposefully picturesque Camden, Maine. We're five minutes from the official opening. I'm sitting next to Ernie the Attorney, the webloggin' lawyer, part of a delightful dinner party last night. I drove up with Pito Salas, which was a total treat. I also got to talk with people like Howard Rheingold, Dan Gillmor, Buzz Bruggeman, etc. etc. (I hate doing lists like this because I have a terrible memory and thus end up slighting people I care about.) So, so far I'm a happy guy. (Not to mention that they've got wifi. Woohoo!)

PopTech is a techno-humanist conference, this year focused on "artificial worlds," a topic I will undoubtedly find annoying in the best sense.

Bryan Field-Elliot of PingID suggests at NetMeme some proverbs for our times.

This has similarities to Gary Turner's Blogstickers, which I mention primarily so I can say:

Chip just sent me this poem by Ani DiFranco that apparently she's reading at all her concerts now.

According to Masha Geller's MediaPost column today, a report from WebMergers says that the dot-com failure rate has "declined dramatically":

So far this year...93 Internet companies have shut down, which is about a quarter of the number (345) that shut down in the first half of last year.

And here's what I expect tomorrow's headline to be:

Jessica Pink did just one thing yesterday: gave birth to baby Saul. The world's already a better place for it.

Microsoft posted a dumbass "testimonial" from someone (by coincidence, an attractive woman...what are the odds of that?!) who recounts why she switched from the Mac to XP. Everything about this "real life" testimonial is phony, starting with the faux voice:

To my surprise, the process of switching was as easy as the marketing hype had promised. I was up and running in less than one day, Girl Scout's honor. First, let me tell you more about why I converted.

After the page had begun circulating through mailing lists ("Can't they find even one real person to give a testimonial?"), Microsoft deleted it. We can't know the cause, but we can only hope that Microsoft recognized that such an obviously bogus ploy works against them.

Thank goodness for Google and its amazing feats of memory (= its cache).

Although I'm frankly afraid of Marilyn Vos Savant, the World's Smartest Person, she is wrong in her column in Parade today.

Which of the following sentences is grammatically correct: "I ran over him with my car," or "I ran him over with my car"?
-Lee McKimmons, Glendale, Ariz.

I hope you're just writing a novel, Lee. Anyway: "I ran over him with my car" is correct. This is called an "active" construction. Also correct is: "He was run over by my car." This is a "passive" construction.

Also correct: "My car ran over him but I have an alibi" which is known as a "defensive" construction, "Really? I thought he was speed bump because I'd done like nine tequilla shooters in a row, dude" which is known as a "penal" construction and "Sentences are like skid marks over the life the have just ended" which is called an "active" deconstruction.

But what I meant to say is this: Imagine two scenarios.

A. You are driving along an Arizona highway (which, by the way, is one of the great niche magazines). It's the middle of the night. You're tired. It's raining because you realize you're not on a straight highway in Arizona but a twisty swamp road outside of New Orleans. You feel a bump, stop the car, get out, and discover that you've run over a refugee from the Mardi Gras who fell asleep in the middle of the road. Did you "run someone over" or "run over someone"?

B. It turns out that Swamp Thing was sitting by the side of the road and saw you tenderize the meal he'd been eying. With a mighty, swamplike roar, Thing comes after you. You hop into the car and start to drive. Swamp Thing stands in the middle of the road. There's no way to turn around and reverse is broken. So, with a steely glint in your eye, you aim squarely at Swamp Thing. A sickening thud confirms that you..."ran him over" or "ran over him"?

Answers:

A: You done run someone over.

B. You think you ran him over, but you can't be sure because of the pain from the steely glint in your eye.

Larry Lessig, The World's Most Pessimistic Person^TM, waxes almost optimistic in his blog discussing the Eldred case he argued in front of the Supremes a few days ago.

Lessig is more optimistic about the outcome of the Eldred case than other reports I'd read because he focuses more on what the Justices did not ask during the hour-long argument, an indication of what they had accepted.

I was talking this morning with Mark Hallas, policy advisor to the government of Ontario, and he helped me understand why "federated" ID systems are an important topic here. I had thought that the draw of federation is that it enables identities spread across multiple applications and domains to be unified. But Mark explained that federation is important to him also because of the looseness of its binding of IDs. For example, one government system may need to check an ID in another system, perhaps in another province, to verify someone's age but should not have access to the rest of the information being stored. Federated systems can allow this type of filtered access.

Drummmon Reed, CTO of OneName Corp, talked about the XNS protocol his company has pioneered and is trying to get adopted broadly. XNS allows identity information to be expressed in XML documents and be linked, creating (as one of their papers says) "an Identity Web that can do for digital identities what the World Wide Web has done for content." Drummon said that this identity web will be capable of modeling the rich and complex interactions among identities via "contracts" based on an extensible set of attributes that includes permissions, purpose, policies on privacy and security, and signature.

Drummon listed the advantages of the system, including: global addressing and logical naming; access control and auditing; permission management; data sharing and versioning; persistent links; and workflow.

Sounds great, but I'm confused: Do we get these benefits because XNS is so extensible that it can handle the data usually handled by, say, a workflow system, or is it that a workflow system could use the identity services provided by XNS? Or both? If I came in knowing more, this would not be a question...

In speaking with other attendees, the main challenge facing XNS seems to be getting enough "traction" in the market since it requires large scale adoption to fulfill its promise. But I'm merely relying on the kindness of strangers when I say this. (Frank Paynter knows about this stuff and he's excited about XNS.)

Carol Coye Benson from Glenbrook Partners is giving an old fashioned, stand-up, Powerpointed talk. And she's very good.

She says:

People say that they want to be the "Visa of the identity business." But there's no equivalent in the identity business of the way the credit card system motivates merchants and issuing banks Also, transactions that are clear and simple with credit cards become unclear and complicated in the identity management world.

And, Carol says, there are no obvious non-transaction-based models.

Carol presents a second "bad idea": "The compulsive consumer fantasy" in which customers have complete control over the information disclosed in different contexts and will pay for the privilege. But this appeals only to the obsessive minority of users, so it's not going to happen on a scale sufficient to build a business.

So, where is money to be made? In the enterprise market. And there may be some opportunities for intermediaries, along the credit bureau model.

By the way, Carol says we'll have a variety of identities, not one clean, aggregated single identity. We'll have government-issued identities and employer-issued identities, ISP/bank-issued identities, etc. It's going to be messy. And that's a Good Thing.

Craig is CTO, Advanced Strategies and Policy for Microsoft. He's going to build the case for digital identity as central to the progress of the computer industry and the progress of computing itself.

He says MSFT got interested in "trusted computing" because they realized that if people no longer trusted their computers, they'd stop using computers. (Oddly, the notion that content owners could have access to what's on my computer makes me much less trustful.)

The "core tenets" of trustworthy computing, he says, are security, privacy, reliability and business integrity (i.e., the relationship of the vendor to the consumer).

Digital identity is a "building block" of trustworthy computing. It involves all four tenets. To get people to accept it -- and "MS identity technologies are opt-in by philosophy" -- they will have to be educated about the benefits.

["Opt-in" is a relative term. If in the future I need to use Microsoft Palladium to download Hollywood content, it is opt-in only if I'm willing to opt out of the entertainment mainstream.]

A MSFT product manager is about to give us a demo of a future feature of Passport that will tell you how secure the password you just created is. (Presumably, if you try to use your user name as a password or the words "password" or "god," the thing reaches out and physically slaps you.)

Now we're seeing a demo of how MSN8 enables parents to set up controls for what their children can do online. E.g., the parent can specify the subject areas of pages the child can see; it looked like a list of about 50 topics. Of course, filtering by topic is notoriously and conceptually unreliable. So, MSN lets the kid send an email to her parents asking for permission. (The MSFT guy calls it a "feedback loop.")

Craig is back. The demo was, in short, a waste of time.

He's now working on showing how deep and complex the issue of digital IDs is. IDs are needed not only for humans but also the identity of the machine and the software.

Digital Identity has many layers, he says:

a. Identity: collections of attributes, some provided by user, some inferred
b. Authorization: This user can perform certain activities
c. Personally identifical information: True name
d. Pseudonymity: Aliases, screen names, etc.

As computers more deeply networked, DigitalID becomes even more necessary for the computers are more intimate with one another.

Good job of complexifying the issue. (I'm not being snarky. It's good to see how deep the waters are.)

Craig is announcing the "Passport Shared Source Release," releasing the code that enables other companies to integrate with Passport. Big deal. As fellow blogger Frank Paynter, sitting next to me, just whispered, Craig referred to this as "enhancing the ecosystem," which makes sense if you assume that Microsoft is the ecosystem.

Craig is wrapping up by talking about "future scenarios," that is, places where the digital ID tentacles need to reach. It's a pretty grainy list: Secure extranets and e-government, sure, but "DRM of corporate and personal documents" is ominous in its implications. Of course, this is no surprise. There is an inexorable logic to DRM: MP3s today, your email tomorrow. Of course, the obstacle to DRM of individual documents isn't the lack of digital IDs but the difficulty (impossibility?) of designing a user interface that doesn't feel like your office has been taken over by kafka-esque, form-wielding bureaucrats. Indeed, Craig is concluding that we will indeed see an increase in regulation over more and more of our lives, and we will have to work hard to come to the right balance.

And, he points out, this is a trans-national issue because of the nature of government.

Craig is obviously thoughtful, smart and reasonable. And it is reasonable for operating system companies to be involved in digital ID issues. That they should be locus of these efforts still strikes me as inappropriate as allowing providers of media pipes to control the content that flows through them.

(Still blogging from the DigitalID World conference. The aggregation of blogs is here.)

Esther Dyson, the moderator, begins the morning general session by asking the conference organizer why there's no list of attendees. (Actually, she began by graciously saying to this audience of 250 that in a few years, 800 people will claim to have been at the first Digital ID World conference.)

The panelists are Michael Calhoun, Principal of CSC Global Health Solutions and Nikolaj Nyholm, CTO of ASCIO Technologies.

The topic is how policy and practicality collide. Everyone, Esther says, is in favor of privacy and people controlling their own information. But what about transparency, i.e., people knowing the business model, the use of the data, who requested it, etc.?

Michael talks about HIPAA, federal legislation passed in '96 that required the Dept. of Social Services to come up with regulations for handling patient information across the health care industry. It says that the individual owns his/her health information, possibly to a degree that can be burdensome to good sense. And the data is "tagged" so that if it's given to a third party, there are still agreements about how it can be used.

Nikolaj says that this is the most European privacy policy he's heard of in the US. Generally in Europe, he says, the data just isn't collected. Stores don't even collect phone numbers from customers.

Esther summarizes nicely: Europe is a bureaucratic culture while the US is a legalistic one. Nikolaj says that US law is very binary: you'll get your ass sued if you get it wrong. Just as security isn't binary, he says, - it's all about risk management - so, too, HIPAA will not provide binary, perfect privacy but can help enforce it.

Esther, the acknowledged master of moderating panels, has now asked all the people lined up to state their questions before the panel addresses them. As a result, themes emerge and some questions that are not as helpful will undoubtedly fall by the wayside. Nice technique.

Jon Udell of InfoWorld asks if HIPAA can actually be implemented and if it will be extraordinarily expensive. Michael replies that the answer is Yes to both questions. The "drop dead" date for implementation is April '03 and it's going forward.

Esther's closing comments: Most of the information about her hasn't been gathered from transactions. It's what has been written about her, what she's written, emails, etc. "In the future, we'll all have about as much privacy as a rock star." We need transparency to enable us to fight against the invasions we don't want.

This session (at the DigitalID World conference) was entirely conversational which is great to listen to but hard to summarize. Phil Becker, the moderator, is doing a terrific job of keeping the flow going.

Now he's pushing on Microsoft as having the worst reputation on which to build trust relationships. The Microsoft guy replies that they're trying to do all the right things, that Palladium lets the user control her/his own privacy, and Media Player 9 asks you out of the box how you want to set your privacy settings. (I've already written too much today about Palladium, but touting the settings screen of Media Player is really pretty lame, even as part of a list of steps Microsoft has taken. If Microsoft really wanted to do the right thing, it would support standards and independent authentication systems.)

An audience member asks: Can I delete my Passport identity from the Microsoft database if I want to? The Microsoft guy replies Yes (I think that's what he said), but points out that deleting an ID raises a whole new set of issues such as: How do you recreate an identity you deleted by mistake? (Why not just put up a bunch of warnings to make sure you're not deleting it by mistake?)

An audience member says that the biggest thing eroding consumer trust is spam because we have the sense that our personal information is being picked up and used. This doesn't seem right to me. Sure, spam means that your email address is being shopped, but the spam is so random that it's clear the spammers know so little about who I am that I feel my time has been invaded but not my privacy.

Audience member: Does Microsoft get value from Passport? Microsoft's answer: It enables Web services; MSN needs an authentication service; Microsoft sells more servers because of it.

Good question, and it was a crisp, blunt answer; Microsofties are usually really likeable. Yet the answer didn't convince me that there are no other reasons why Microsoft wants to be in the identity management business. The reply was short, of course, so maybe there's more to be said. Nevertheless, Microsoft getting into the trust business is a new definition of chutzpah. That they don't see this is continually surprising for a company so savvy about customers.

This panel was over my head technically, but I got my money's worth from a single comment from Doc Searls, the moderator. He said that none of this identity stuff will happen until someone comes up with an application that isn't pushed on users but that users actually want. I thought I heard some glubs as various software strategies sank under the weight of their own presumption.

If you're interested in the Open Source Identity topic, Denise Howell has blogged the session well.

Already too much to blog. I'm sitting in a session I came late to because I was engrossed in a conversation with Peter Biddle, a leader of the Microsoft Palladium team. He's an engineer and a partisan and able to remain good-natured even while being, um, pounded. Gives as good as he gets. I felt like we actually got down to some basic issues. Thank you, Peter: you're a good guy.

Palladium is a Microsoft initiative that will bring high security to your PC. It will act as a vault for contents and enable users and "content providers" to negotiate terms of usage. Palladium is neutral about those usage terms. It'll enforce any that are agreed upon.

I'm reluctant to present Peter's point of view since he has been thinking about this for a long and is quite eloquent. Nevertheless, I'm going to. As I understood the conversation, we got down to this: Peter is designing Palladium to be neutral to usage policies but also capable of enforcing them. So, if Eminem says that you can download his new song but you can play it once for $5 and ten times for $10, then, fine, Palladium will Make It So. And if Joe Mahoney says you can download his new song and play it as much as you like, but you can't resell it digitally, then that's just as fine. Palladium is neutral to policy. Yet, once terms are agreed upon, it builds that policy into the computing architecture (as Denise Howell put it during our conversation).

And that's one of two problems I have with Palladium. The real world is enriched by the leeway that's inevitable in it. Even as we assert our "intellectual property" rights over our ideas and expressions, we know that in the real world those rights are often unenforceable. For example, a report from Forrester Research may have "Do not photocopy" at the bottom of every page, but you won't get sued if you run off a couple of copies of the graph on page 110 to use at an internal meeting. We make these decisions all the time, and the world is richer for it. Further, as Denise pointed out, when there are infractions worth prosecuting, we have human judges and a legal system that comes to reasonable (usually) decisions. Implementing policy in silicon drives the leeway out of the system.

My second problem is that Palladium may be neutral in its architecture but it is being born into a world that isn't neutral. Content has been locked up by gigantic, greedy, stupid companies (mainly headquartered near Hollywood) and the company producing Palladium has been declared a monopoly. If Palladium becomes the only way that the entertainment industry can sell digital content according to strictly enforced rules of usage, then Microsoft will become the de facto entertainment player, forging the "unholy alliance" that so many of us fear. The fact that Microsoft is not committed to producing Palladium across multiple platforms in a timely way is certainly unsettling.

There was nothing that I said, with the help of Denise, that Peter hasn't heard before. His response was surprising to me. Palladium won't really lock down content that well, he said. Pirates will still be able to get unlocked copies of whatever they want. You'll still be able to find a copy of the latest Eminem song to download because some hacker somewhere will crack the encryption. My response was that Palladium will eliminate the gray area so that those who download a song will have to become pirates. Peter, of course, thinks that we're already pirates, so that got nowhere.

Ultimately, I think you have to ask what world will be better, one with enforceable usage rights that drive out the leeway and hard-codify fair use, or one in which there's reasonable (and even unreasonable) leeway where some genuine piracy happens, a lot of genuine cash-for-use happens, and a whole bunch in between goes on at every level of society.

Phil Windley, CIO of Utah is talking about the complex ways in which governments deal with ioentities, from birth certificates to death certificates.

He points out that vehicle titles are in a one-to-one relationship with the vehicles: each vehicle has one and only one title. Why? Because people want government to track this. But government has abdicated its responsibility as an issuer of digital signatures, which is why they're not as useful as they should be.

But (Phil says) people want the services that a government-based identity program could bring. For example, supose you move to Utah. There's a long list of things you have to do, from registering your car to enrolling to vote to getting tax information. Utah wants to give you a single site where you say you're moving to the state, pay one tax, and everything gets done. But this is hard to do because the various information apps are not connected. (In Utah, your name can be in over 200 different databases.)

Phil just netted it out:

Governments are in the identity business but don't recognize it.

Governments have abdicated their responsibility.

Digital certificates are not the answer.

Citizens are fearful of government collection of data but still demand the connected services that require that aggregation.

The real problem is that these are public policy questions and technology can't solve them, Phil says.

The general session presentation by a guy from GM is in process. After spending too long telling us that GM is a big, honking company, he's talking about "identity as a business issue." But so far he's pointed to two basic situations where identity management would help: First, consolidating information about customers so that the dealer has access to the same information as the marketing department, etc.; Second, consolidating the six different identities (= email addresses?) the speaker has at GM.

This type of consolidation of customer information is the standard benefit of application integration that's been touted for years. Certainly it's valuable and important (and just a little bit scary), but it's not what I think of as central the issues of digital IDs: the creation of a permanent, traceable link from Web presence to the real world body sitting in front of the monitor, and vice versa. Yeah, of course we want to consolidate our customer information. But do we want a digital ID across companies? Who manages it? Who controls it and the record that it accretes? That's what I'm here to learn about. Maybe he'll get to it...

Nah, now he's telling us why automating processes over the Web saves GM yada yada bucks. Presumably his standard presentation.

Now he's getting to it! Three major digital ID issues.

1. Single sign-on is no longer GM's focus. Other factors have taken precedence at GM, including security and privacy. GM feels it needs a more comprehensive approach, rather than focusing on single sign-on by itself.

2. They've joined the Liberty Alliance, which is the everyone-but-Microsoft response to Redmond's attempt to own digital ID management. But he just said that the Liberty Alliance has nothing to do with countering Microsoft. (Bushwa! As if there's anything wrong with countering Microsoft on this!)

Hmm, I don't know what the third thing was.

Take-away: GM is doing lots of interesting things webbing people together. The question of digital ID for them comes down to how they can consolidate information about the variegated interactions each customer and employee has.

I just got in, an hour late. (And I had to get up at 4:30AM to be this late. Sigh.) So I missed the opening presentation. But I'll be able to get the gist from the many blogs being written here. For example, Denise covers it really well. Nice on-the-fly writing!

But we're off to a great start: I ran into AKMA at the airport so we took a cab here. In the hall at registration was Eric Norlin, high on ideas and DayQuil. AKMA opened the door to the hall where the keynotes are being given and who's sitting on the aisle but Doc, I believe next to Denise Howell (whom I've never met). This will be fun. [I just read Doc's blog, and, yes, he says he's sitting next to Denise. Howdy, Denise!

I'll be blogging the conference for the Boston Globe over at DigitalMass. Sorry to split between two sites, but I'm a writer so I like to be published. (No, I don't know the relationship between globe.com and digitalmass.com. There is some relationship. If I had to guess, I'd say that digitalmass is the tech site for globe.com)

One good incidental effect of the Globe.com system is that they aren't yet set up to let me post directly, so I have to do it by sending email to their very helpful Web guy. So, I'll have to do batch posting as opposed to real time blogging. This is good because it will discipline me into writing retrospective chunks ("Watch out, dude, that guy just spewed retrospective chunks!") rather than a set of running notes; I tend to get unreflective when taking notes.

The guy from GM is now talking, describing the complexities of GM's IT/Web infrastructure: over 2M subscribers to thheir Business Web, etc.

Gary Turner is starting an anthology of quotations from unfamous people (where fame is defined by mass culture or acceptance into the Canon of Important People). For example, from Marek we have, among other bon mots, "The truth will set you free, but first it will piss you off." Gary is soliciting contributions.

Euan TheObvious Semple responds to my blog about the way in which the Internet is like my brain by wondering if the Internet could convey moods.

I listened to the first half of W's speech on the radio. Then I switched to TV because his performance was so lackluster, his tone so monotonous. It was worse watching him. My working hypothesis: Before he went on, W said he really wished he weren't a recovering alcoholic so he could have a drink "jes' to steady mah nerves," so the White House doctor gave him a mild sedative.

The new issue of David Isenberg's Smart Letter is not only must reading for anyone who uses telephones but is also a delight to read, beginning with Buckminster Fuller's explanation of why pirates had a better business model than kings. At issue is what's right and wrong in the fundamental stance of George Gilder, the telecommunications guru. But David uses that to dig into the roots of the telecommunications industry and its coming meltdown.

Now compare that with a recent conversation the FCC sponsored in which the operating assumption is that the only way to ensure continuity of service and innovation is to prop up the current telco cadavers.

Richard Seltzer at B&R Samizdat Express has reviewed Small Pieces Loosely Joined, quite enthusiastically.

Richard reviews a wide variety of books. I'm in the middle of Fagles' new translation of The Iliad and found Richard's review both learned and helpful, a rare combination. (Think I'm just reciprocating? Read some of his reviews for yourself, you cynical, jaded, suspicious, person! Take "Trying to Enjoy Bellow" as a fr'instance.)

Someone heard me on CNET radio over the weekend and was stimulated to send me an email suggesting that the Internet is becoming conscious. I think the Internet lacks two essential ingredients: First, consciousness has everything to do with the ability to pay attention, and attention requires that we care about ourselves and our world. Second, consciousness needs a body because thought isn't a formal process. Nevertheless, there are some important ways in which the Internet and consciousness are similar. For example:

We've had four loose 2" tiles in our bathroom for a year. Every couple of months I try a different way of cementing them in. Every couple of months they loosen and come out again. Two days ago, I scraped their backs down to the original tile and chiseled the wood to which they attach in order to get a nice, clean fix.

They are already loose.

This morning I woke up, trembling with excitement, for I have realized how to fix them for good.

[SPOILER BELOW]

From Peter Kaminski, one of the leading sources of leads to sources, comes a page with links to the world newspapers he likes to read. And there's more here.

Sites include:

Le Monde
Die Welt
El Mundo

And, in English to see how we look to others:

Yomiuri
Ha'aretz
Mail & Guardian

I was interviewed bythe CNET WebTalkGuys a couple of days ago. It will air on Saturday. You can hear it here with the RealAudio player or here with the Windows Media Player.

They asked good question, to which I responded, as best as I can recall, like a rambling, self-important asshole.

Here's info from the WebTalkGuys about play times for the actual broadcast:

The show will broadcast on CNET Radio Saturday, 10/5 from 10-11am PST in San Francisco/San Jose (910 AM), and the XM Satellite network on channel 130. It will repeat on Sunday, 10/6 at 5pm PST. It will stream live on CNET at http://www.radio.com at those times. It will be available on demand for listening on any phone via the Mobile Broadcast Network, SonicBox and at our website. It is also broadcast in Seattle on KLAY 1180 AM on Saturday at 11 am. CNET has put up the brief description of your topic at the Weekend Shows page.

JD Lasica blogs the O'Reilly OS X conference where Digital Rights Management is the topic du jour. He quotes Tim O'Reilly, who is high up on my God Bless 'Im list:

"If I'm a publisher and given the choice of piracy or obscurity, I choose piracy."

Dan Gillmor is blogging the conference, as is GlennF (taking notes as people speak). As JD points out, Glenn ends his entry with the hopeful phrase:

Fair use has a posse.

Let's hope. And it's about time.

Boston, Oct. 2. In response to NBC's decision to extend Friends two minutes in order to hold viewers past the start of other networks' 8:30 shows, JOHO the Blog today announced that it is moving to a 23.6 hour publishing schedule.

My Top Ten picks to replace Sen. Robert Torricelli in New Jersey:

10. Bill Bradley
9. Bruce Springsteen
8. Corrado "Junior" Soprano
7. Thomas Edison
6. David Copperfield (magician, not waif)
5. Allen Ginsberg
4. Bud Abbot
3. Philip Roth
2. Dorothy Parker
1. Pocahontas