You may try to model me, but you can't define me. I'm larger than a tidy form.

That's why God and evolution gave us the ability to lie. And posture. And pretend. And choose our words and body language.  

I am not my business card. Or my resumé. Or my orkut profile.

They are merely shorthands, placeholders, for the real thing.

And you never get the real thing.

The closest you come is by interacting with me: self as black box. Not by description.

Second best to getting the real thing: ethnographic observation of my real life behavior.

Third best: following my narrative.  

Fourth best: analysis of my online behavior.

That hamstrings YASNs.

The prob, of course is that I am many people in one skin and we all change who's in charge with the ebb and flow of blood sugar, brain chemistry, and the damned cat that peed on the carpet and I'm one way in A's company, another in B's company, and some awkward way when in the company of both A & B since I'm working with B, dating A, but had horribly bad sex with B but I can't remember whose fault it was.

I am complex, not one persona but many, changing over time. 

We become self aware of this in puberty.

And spend adolescence learning to navigate ourselves, to choose, to actively dream versions of ourselves into being. For the objects of our infatuation. For our authority figures. For our parents. For strangers. Hopefully for ourselves.

Then adulthood calls for settling upon an outer persona. We simplify, most of us, at least our outer affect.

But the other me's are still inside.

And each of the million other Ryzers/YASNSers are the same way.  

If I'm that messy and convoluted, can you imagine the relationship complexity? personas*(personas-1).

I have a few broad suggestions for social network improvements.

Model sociology. Not just nice stuff but all the icky horrible interactions we see in the office, in school, in gangs. Rites of passage. Flirting. Insults. Combat. Cliques. Authority. Power.

Model psychology. Squeeze in Maslow's Hierarchy. Piaget or someone else who models childhood development, especially arrested development.

Let me do more. Solve real life problems virtually. Design tools for tasks I really need. Group formation. Group destruction. Group work. Better meetings. Prioritization of communication. Help people be useful to each other.

Do less. Email, texting, http, phone calls are all pretty dumb systems. They just move content, so the human content becomes paramount, richer, engaging. Find your core and strip away the rest.  

Admit your limitations and open up the plumbing. We need APIs so programmers can extend the models and tools. 

I should be able to write an extension that lets you see a combined orkut profile and Google references.

Or find what the people in my virtual sub-community are buying on Amazon.

Or authenticate access to my calendar using friendship degrees.  

Or sync my mobile phone behavior (who I text and who texts me) with my buddy list.

Or turn orkut into my smart phone's Caller ID system.  

Or bounce new friends-of-friends against my interest profiles and nominate a few for acquaintanceship.

Or whatever. The world knows more than you do, so let them in.

If you want social software to endure, ...  

A guy asked what it takes to scale your rolodex to 100,000 people. Then built Ryze to find out.

What's your question?

[a klog apart]

My second thesis for blogtalk was that machines will blog. DARPA is now funding an engineering research project.

Halvais reports "DARPA is funding the uberblog." Capture everything that goes in someone's life and make that information useful to the user. It is fodder for personal diary. Or for understanding what's going on in a subject's life. The project's real world challenge: wear the unit on a trip to Washington, D.C., and demo it for the brass.    

As your physical world is increasingly overlaid by cyberspaces, you need tools to sense it wherever you are. Tools to present its many digital elements for humans. Tools for navigation through and manipulation of the new sensations and history. Tools to distill the information overload through pattern recognition, clustering, semantic analysis and data visualization.

Oh, and make it unobtrusive.

The Information Processing Technology Office (IPTO) of the Defense Advanced Research Projects Agency (DARPA) announcement. The LifeLog Proposer Information Pamphlet (BAA # 03-30).

Get your proposals in by 23 June 2003.

Newsreaders like NewzCrawler and Radio UserLand do TiVo things. Time shifting. Easier, more complete channel and program selection.Season pass for your favorite shows. Record in the background while playing in the foreground. Save a post to your blog instead of to your VCR.

TiVo needs blogspace community tools: add social filtering (recommendations), feedback, and threads of commentary.

Is your project plan suffering from outline-centric thinking? Break out of the box with Julian. Julian is using a mind-mapping tool for project start-up. If you're into project visualization, read Synesthesia about mind-mapping integration with MS Project, MS PowerPoint, and wikis. Julian is on to something.

Windley, formerly Utah CIO, blogs his @Task product briefing.

@Task combines workflow, collaboration, and project management to actually drive the project or process from the tool rather than using it as simply a static document that records progress. Because it assigns tasks to group members based on the workflow and tracks completion, a manager can see at a glance where the project is, who's got unfinished deliverables, and so can everyone else. Consequently, the tool functions as a dashboard and uses tranparency to enforce good behavior. They're using SOAP to interface with traditional CRM and ERP systems like PeopleSoft so that hours, for example, can be entered into the accounting system from the project tracking tool automatically.

All in all, I think it sounds like a great improvement over other project tools I've seen. One of the things I was thinking is that it would be a great tool to build a lot of software development processes. Most of these are enforced by culture and training with little tool support to help foster the correct process.

Ross Mayfield marks up the social network analysis (SNA) Strategy + Business article. Ross compares the short evolution of SNA's models with that of longer and more widely attended financial models. It is relatively early for SNA but not new.

Social network analysis isn't exactly new. Just look at epidemiology and heraldry, discipline for centuries. What is dramatically new is the digital network.

The net makes SNA cheap. Health and social scientists analyzed social networks for decades. Their process, using phone and face-to-face interviews, paper forms, call-backs, and methods to avoid system gaming, was slow and cumbersome. The costs were so high and the human intrusion so deep that projects were done selectively and with limited scope. Social maps were out of date before they were finished. Still are for the most part. You might do a few dozen projects in a full career.

Online networks change all this. With tools like Valdis Kreb's InFlow, digital networks:

• Speed things up. Projects can be executed in a day, updated in near real time.
• Scale large. Big networks, with hundreds of thousands of people, can be mapped.
• Itirate. Time lapse studies show a network's dynamics. Now longitudinal trends and changes can be revealed.
• Ingratiate SNA. SNA tools become polite, less intrusive by using existing digital artifacts and debris.

More, better, faster.

The other change is qualitative. Let me invoke the Black Scholes options pricing model to explain.

One of the great simplifications of securities analysis is that of the efficient market, a market where prices reflect all information. This is useful because money is the ultimate fungible item: everything can be expressed as money and transactions don't take place unless both parties believe they are getting a fair exchange. Money and markets reduce commerce to math.

And then you complicate things; there is a real world to accomodate. Not everyone gets the same information, information is dispersed unevenly, markets interact with each other, and the list goes on. Modelers refine their explanations of how markets work, the better to plan markets and exploit their weaknesses.

SNA, on the other hand, can look at the full range of human behavior, thought, expression, and interaction. Happily, humanity is not entirely expressible in one dimension. SNA on the net offers the hope of diving deeply into the rich colors and textures of human relationships.

So the net amplfies SNA both quantitatively and qualtitatively.

The world of microcontent has a role in this.

The web's technical structures (links, web pages, addresses, etc.) are the skeleton of the new SNA. Microcontent's new metadata, substructures (the elements that make up a CV or a blogroll, for example), and contexts are SNA's flesh and blood. Microcontent will help us find clusters of people with similar skills, clusters who attended the same Sheryl Crow concert, who are alumni of the same firm. Annotation of time and space, topic labels, and taxonomy markings add relevance and enable new applications.

Blogging is one of the first sustained microcontent movements. Weblogs concentrate the content and behavior of a person. My blog is of, by, and about me. I link and write, therefore I blog. Add a thousand thousand fellow bloggers and the blogosphere becomes a fertile space for mapping social interactions and relationships.

That's why I want blogging and metablogging tools to embrace microcontent metadata and substructures; the better to support new contexts.

[a klog apart klogs]

Techdirt

A few months ago we wrote about the problems companies are facing due to a 1979 regulation saying that they need to record the race and gender of every job applicant who comes there way. The problem, of course, is that in the day of the internet, thousands of people apply for jobs, many of whom have no real interest in getting those jobs. The internet makes it more difficult to determine who is really an applicant and who is not - though, the law states that just about anyone is to be considered an applicant. This has placed a huge burden on companies that often need to hire people just to track down the race and gender - which often upsets many of the candidates who feel that it's being used for discrimination (and not the other way around). Of course, some are now saying that companies are fighting back against this rule in order to make it easier for them to discriminate again.

Black and white old school law meets infinite scale and shades of gray. It's a real problem. The law assumes an old model: workers responding to classified ads. Today, both parties see more of the labor marketplace (more jobs, more candidates). And both parties have many styles of interaction at their disposal beyond the application form: mailing list sign-ups, reading or subscribing to a candidate's blog, meeting in a chat room or trade show, schmoozing at Ryze. And the data employers collect is often temporary: If you download a thousand resumes and delete 950 of them five minutes later, are they all candidates or just the final fifty?  

How to respond?

First, this is a great opportunity. If you run private job boards for companies, act as an infomediary. Collect and protect candidate race/gender data and sanitize their profiles before employers see them. Share aggregate statistics with employers, so they are covered by the law. Protect employers from selecting candidates based on race and gender. With this regime, employers don't pair up race/gender with a specific candidate until selection for an interview.

Second, define "candidate" in your policy handbook as someone who submits a résumé or profile for a specific position. This ties data collection to a well-defined process.

Third, participate in the regulatory process. Share your real world problems with state and federal regulatory panels.

Fourth, keep working the labor market using your full toolkit. The laws will catch up.

Sometimes push works better.

For your nightly email of updates to Phil Wolff's "a klog apart", use the form below and Bloglet will get it to you. Click here if you're a bloglet regular.

I use  to get about 20 feeds sent in one big html mail every night. Clean formatting. Easy to read. Easy to admin.

I met Bret Fausett at Digital ID World. Brett asks about the idea of a .blog top level domain (like a .com or .net). Lots of good comments; here are mine.

I concur with the general objections stated before. The central identity of blogging is that a human voice (sometimes a small tribe) is found in one place, instead of being scattered in many large community sites.

That said, what creative ways could we exploit a .blog tld?

Might this help with search?

In a world of microcontent, could DNS help each post be unqiuely identified or found? If so, would DNS take part in update notification, perhaps helping with threading of conversation?

I know a number of people who blog anonymously or pseudonymously. Could the registrar help assure privacy of domain ownership?

Assuming everyone will get at least one personal domain for blogging, this could be a very active registry. Compared to businesses, they have short lives too.

Blogging is a form in transition.

Personally, I think blogging as a form will merge with all the other forms of digital expression. With email and IM first. With voice/video conferencing, streaming videos, browsing, and PowerPointing later.

Watch it change:

• as more people blog from their foto-mobiles
• as devices start to blog ("My car's day")
• as audiobloggers create radio shows and videobloggers create televsion programming
• as Sims characters start to blog.

Moving forward, see a convergent software client emerge.

Source: evanwolf group, 2002.

A lot to shovel into one bucket.

Why bother?

Synergy and Usability.  

Synergy because each of their abilities are horizontal and complementary to to the other functions. This is why spell checking rolled into word processors.

Usability, because with great design, learning one feature makes it easier to learn and do the next. Once you learn spell check, any sort of text editing or proofreading tool is a snap. This slashes the incremental user burden of new features. So spell check, for instance, can cross into spreadsheets, presentations, email, even project management tools without taxing the new user.

What else do you need from a converged user experience? What are our collective design goals?

Simplicity. Unity. and Adaptability.

The surfaces presented to a user will adapt to each medium and form. Perhaps I need a storyboard for planning a video; maybe it can also be used for planning a presentation, an extended blog post, an interaction with a customer. Are you presenting on a computer projector, a video stream, or paper? The software should understand how to adjust.

The converged client should also adapt to people. A person's culture, experience, goals, interests, and skills. This is hard as adamantium, but it is what allows robust tools to work for most people in many situations. Some people need help and wizards and automatic spelling correction (think Microsoft Office), others need directly manipulable affordances (think Kai's Power Tools). Small children need different environments (Power Puff Girls) than teens than adults. Grokking world cultures and subcultures, and reflecting those in software, is a fine art.

Adjust to hardware platforms. How do you incorporate the strengths and limitations od the PC fat client, game box, TV set top, and thin clients on the mobile phone and web browser?

Embrace specialized content. Some database tools can automatically generate editing screens and menus and even workflows from data structures and definitions. We need to do the same thing, but across many kinds of content and activities. From blogging movie reviews (with extra metadata and internal structure) to IM conversations guided by scripts. 

A constraint: Adapt while preserving and leveraging the user's prior knowledge, skills, and abilities.

So.

Contrast this with Anil Dash's microcontent client. I'm seeing the converged client as a conceptual superset or framework for building microcontent clients.

Can you imagine the plumbing?

You'd want to design for:

• Flexibility
• Interoperability
• Extensibility
• Scalability
• Polylingual

Your architecture would need:

• Shared services. A common chassis. 
• Open APIs. So third party's can connect, communicate, and interact with the client.
• Plug-in sockets. So tool makers can add their own features and extend the client's abilities.
• Standards support. To increase interoperability.
• Heavy transcoding. Transcoding is a fancy term for converting content from one platform to others. The converged client will have to handle a wider range of content than most. From story outlines to storyboards. From audio tracks to text subtitles. From IM threads to blog posts.

So what?

We're on our way. Blogging tools are starting to interact with email and sounds. PIMs are managing contact information across multiple applications. Community and collaboration features are as critical to games as traditional gameplay.

I'm calling it: 2003-2005 will see many clients converge, weblogs among them. The challenges? Immense. The rewards? Many and rich. The fun? Deep and lasting.

[a klog apart klogging]

A friend of mine asked: how many webloggers are there? This is like "How big is the Internet?"

I searched through Nua and a dozen other internet sites and haven't seen any research on the size of the blogosphere.

I ask you:

1. Do you have an educated guess?
2. Do you know of any prior work in this area?
3. Can you think of a methodology or two to create useful measures of the number of bloggers and the number of weblogs?
4. What related questions would you want answered?
5. How might you use this information?
6. Pitfalls to avoid?
7. Would you join a BlogCensus.org to provide and share stats?

My wild stabs:

1. Do you have an educated guess?
• Not yet.
2. Do you know of any prior work in this area?
• No. I've looked.
3. Can you think of a methodology or two to create useful measures of the number of bloggers and the number of weblogs?
• Some vendors host weblogs and have relevant stats. We could add those up.
• We could look at download and registrations from the top 5 vendors, and add fudge factors to cover other tools and disadoption rates

4. What related questions would you want answered?

• LiveJournal.com, has a statistics page: (numbers as of 10 November 2002)
• Total users: 770910
• Users that have ever updated: 635168
• Users updating in last 30 days: 280213
• Users updating in last 7 days: 200543
• Users updating in past 24 hours: 72587
• Gender:
• Male: 201452 (36.3%)
• Female: 354085 (63.7%)
• Unspecified: 131153
• Account Type
• Free Account: 718109 (93.2%)
• Early Adopter: 14282 (1.9%)
• Paid Account: 36718 (4.8%)
• Permanent Account: 1218 (0.2%)
• Country of origin (Mostly English-speaking)
• US state of origin (California, New York, Florida, Michigan lead)
• Age distribution (mode=17)
• Client usage (90% web)
• Activity: posts by day overall (147k posts last Wednesday) Per-person would be interesting too.
• New accounts per day (eyeballing a chart it looks like 900-1400 new LJ users per day, averaging about 1100)
• I'd love to know:
• How many entries have ever been blogged? (the cumulative number of posts).
• How many links in posts? (excluding blogrolls and navigation)
• What blogging tool or service they're using?
• Blog lifecycles:
• How long to bloggers of various stripes blog?
• How many change hosts? Change tools?
• Why do people abandon blogging?
• Is there a critical mass, a minimum number of posts per day/week/month that separates those that blog from those that fail?
• Of people who take a break, how many start again?
• Number originating within a company or operating behind a firewall
• Connection speed (does broadband make it easier to blog?)
• Payload distribution. How many people include pictures, sounds, flash games, or movies? How many bytes are home pages?
• Syndication. What percentage syndicate their sites?
• Duplication/Overlap:
• How many blogs per person?
• Do you post to them equally? How many are updated daily/weekly/monthly?
• How many tools do you use?
• What ancillary tools do you use?
• Graphics and other media
• News readers
• HTML editors
• email clients
• blog-specific search (daypop, google)
• blogosphere navigation (blogdex, blogtree)
5. How might you use this information?
1. As a blogger.
• Always good to know where I stand in relation to the pack.
• Trends might tip me to new capabilities
2. As a consultant or IT leader.
• Make better choices about deploying blogging and community tools
• Use the "bandwagon" sell when appropriate
3. As a blog tool maker.
• Understand the markets I serve vs. the ones I don't 
6. Pitfalls to avoid?
• Hype
• Irreproducible results
• Bias - vendor, country
7. Would you join an BlogCensus.org to provide and share stats?
• As a user, with anonymity.
• As a vendor, sure.

What say you?