Ross Mayfield's Weblog

Aloha

I'm in Hawaii, engaged in Wiki Wiki (cough) research.

Light if any posting this week. Give me a couple of days to get back into the conversation.

Are you a döt Net?

Socialtext developer Terry Brian Ingerson, also known as Ingy, has gone and changed his name. 

He is now known as Ingy döt Net.

Are you a döt Net?

The Ping War

Somewhere in the middle of the Google Books debate last night, it confirmed for me that we are in a ping server war.  It's a little hard to see, but the ping server will become the new center of the net.  Verisign's acquisition of Weblogs.com was the first salvo.  I'm not sure Robert Cringley is right about Google-Mart, but he isn't entirely wrong.  Google Base isn't just about volunteered structuring of data, but pushing pings (Mark Pincus isn't entirely wrong either, but that's another issue).  The important point is there is tremendous value being the first to have information pass through your central node.

As an aside, the reason it occured to me in the middle of last night's debate was how part of the copyright argument was shifting from opt-in to opt-out, regardless of the specifics of this case.  Google is suggesting that instead of getting permission from copyright holders first, others would contact them to opt-out.  The reason this is an issue at all is because there is no registry of rights holders that is kept up to date.  Maybe one day there will be a technical solution, and I'd imagine it would look something like an RSS page for each copyright holder that pings a server (or many) when ownership or contact information or license has been modified.  In a moment of madness, I considered Google Books to be a marketing ploy for Google Base.

Now, a supposed ping war may not matter.  Any time a network hubs around a node where the flow is valuable, an arbitrage exists.  Central power of a ping server leads to the recognition of alternative servers.  If that fails, the spiders start crawling faster.  The only drawback is the network becomes inefficiently loaded trying to find even distribution.  Such redundancy is a great reason to love the net.  They dynamics here hold far less ability for control than the lower levels, we should be aware of them, but more gravely concerned about the concentration below.

So I'll apologize for this post.  Can't make it groove like Steve Gillmor.  Just a musing.

Email is So Five Minutes Ago

Great article in BusinessWeek on how wikis and IM are solving the email problem.

Soar engineer Jacob Crossman says that's because the wikis eliminate the usual flurry of back-and-forth attachments and resulting document-version confusion that's rife in e-mail. At Dresdner, Rangaswami says that among the earliest and most aggressive adopters, e-mail volume on related projects is down 75%; meeting times have been whacked in half...

Recently, Lennard wanted an analysis of how to double profits on a particular trade. Instead of shooting copies of the same document to several people via an e-mail attachment, only to have to keep track of, merge, and archive all the fixes back into a central version, he threw the problem up on a wiki page where everyone could brainstorm, comment, and edit in real time. In the space of two days, entire e-mail conversations evaporated and Lennard had analytics that would have otherwise taken two weeks. Next year's budget practically wrote itself on the wiki page.

Only one problem with the article, no mention of Socialtext.  I guess I should be upset at this, but as long as we keep advancing wikis in general and delivering the above to customers, we keep on leading the market.

Community Lessons Learned

It's time to commend previous the whipping boys of the blogosphere.  Two companies that grew a bit too fast, made mistakes and seem to have learned from them.

Six Apart is working with their community.  You might remember the MT Pricing fiasco as a lesson for how not to work with a community that contributes so much to your value.  They suddenly changed the game on their community and they fought back.  This time, when managing performance problems, they shared the process.  Their lessons learned are good for everyone:

• Read what your customers have to say
• Ignore the tone of nasty complaints, but pay attention to the underlying messages
• Understand that the people giving feedback represent many who remain silent
• Don’t spend too much energy on distractions
• Don't be afraid to communicate
• Trust your customers

Technorati is fast again.  They shared the process throughout, kept at it and I think are winning people back.

The common theme is that good communication and sharing the process provides a way for your community to be included in the outcome.

Learning Through Trial and Error

The $100 Laptop has been unveiled in Tunisa, going into production next quarter.  I blogged about the need and means to connect these nodes earlier this week.  But there is a revolutionary aspect of the project that is barely recognized.  David Kirkpatrick:

Negroponte's team is seeking not only a technological breakthrough but also a teaching breakthrough. They believe that illiterate kids can, with a little instruction, learn to use computers on their own and then use the laptops to teach themselves to read. After that comes math, history—you name it. Alan Kay, a Xerox Parc veteran, is working with MIT mathematician and educational theorist Seymour Papert to build software that "watches" each student and makes suggestions. Papert's "constructionist learning" approach encourages children to reach conclusions through trial and error.

This breaks known conventions for education and technology, which could have a far greater impact than the commoditization at play.

The End of Process

If a knowledge worker has the organization's information in a social context at their finger tips, and the organization is sufficiently connected to tap experts and form groups instantly to resolve exceptions -- is there a role for business process as we know it?

My favorite Clay Shirky quote is "process is an embedded reaction to prior stupidity." That is, there was an exception to process and an expert designed a way for people to work together in one context that should fit all prior contexts. The problem is, the process becomes calcified and accepted as the rule. After all, it's a rule, and in corporations we follow them, even if it fails us or simply doesn't make sense. Because of constant change in our environment, processes are outdated the immediately after they are designed. The 90s business process re-engineering model intended to introduce change, but was driven by experts which simply delivered another set of frozen processes. Because participants in process are not considered experts in theory, they are empowered to make decisions on their own when something fails.

Organizations are trapped in a spiral of declining innovation led by the false promise of efficiency. Workers are given firm guidelines and are trained to only draw within them. Managers have the false belief engineered process and hoarding information is a substitute for good leadership. Processes fail and silos persist despite dysfunctional matrices. Executives are so far removed from exceptions and objections that all they get are carefully packaged reports of good news and numbers that reveal the bad when it's too late.

John Seely Brown and John Hagel point out that while 95% of IT investment goes to support business process (to drive down costs), most employee time isn't spent on process -- but exceptions to process. Further, competitive advantage comes from how we innovate in handling exceptions. When something fails, informed and empowered employees turn to their social network. The informal network, or heterarchy, where most business gets done.

Today, some staid corporations are abandoning process all together (I wish I could quote the source for this). Google is a more public example, albeit an exceptionally new large enterprise, where wikis and weblogs enable a culture of working openly in a flatter and decentralized organization. This is data point helps plot the trend of decentralized organizations that realize economies of scale, as described by Thomas Malone in his book, Decentralization.

Assume for a moment that the 25% of GDP that is search costs falls. Or the 50% of GDP that is transaction costs similarly declines. Coordination costs fall with rising connectivity. The cost of personal publishing and easy group forming are rapidly falling to zero. If a knowledge worker has relevant information at their finger tips, can form the right group to handle an exception, leverage the social context of information and contribute to memory as a natural by-product of getting work done -- what is the role of process?

A process is like a standard. It provides a common definition for others to build upon. This is generally a good thing. In technical systems it helps resolve complexity so higher order abstractions can keep things simple. But even in technical systems, efficiency comes at a cost of adaptation. In social systems, especially where not everyone helps design what they participate in, the constraints against adaptation are compounded.

At best, a process should serve as a reference model. Something that others can reference when completing a task. Something that can be leveraged for innovation, a boundary condition for experimentation at the margin.

As with many things, gaining greater participation and innovation requires sharing control. I do not believe we are near the End of Process, yet. I do believe the arguments for engineering organizations are being trumped by new practices and simple tools.  The first organizations bringing it to an end will have a decided competitive advantage.

UPDATE: Comments are starting to get interesting.  Euan Semple nails it: Process is the sort of word that grown ups in suits use to throw their weight around and to convince others that they know what is going on and that it makes sense.

The Print Debate

Some idiot put hard security on the Quicktime server for the debate.  But we hacked it.  Access the webcast here with the username and password: quicktime

Dorrian Porter:

The lawyer in me may be coming out, but I'm fascinated by the Google Print project and I'm going to watch the webcast of the New York Public Library/Wired Magazine debate. Titled "The Battle over Books: Authors & Publishers Take on the Google Print Library Project", the cast of characters is interesting, including Chris Anderson, David Drummond, Lawrence Lessig, Allan Adler (AAP), David Ferriero (NYPL), Paul LeClerc (NYPL) and Nick Taylor (NYPL)

If anyone else cares to join me, Ross has agreed to open up the Socialtext conference room so I can watch it on a big screen and pretend I was at the sold out event.  It starts at 4 pacific...

Read on for links about the debate at the center of copyright and tech, and maybe we'll see you tomorrow.

Tastebuds

Here's an idea for the lazyweb to solve how writers are lazy at categorizing their posts.  We've had lots of attempts to make categorization easier, but readers in aggregate do a better job.  I'm looking for something to leverage del.icio.us enclosed in a single post, a widget called Tastebuds:

1. Looks up the URL of your post
2. Grabs the tags people have used on your post
3. Displays the top 5 at the bottom of your post
4. Each tag is a link to a list of your posts with that tag

Double bonus points for enriching the list of posts with conversational threads like on Memeorandum, pivoting on the tags through Technorati and making the list sortable.

UPDATE: Adam Kalsey sayz that Tagyu could auto-suggest tags based on post content before someone links to it.  For example, here would be the tags for this post: del.icio.us api reference firefox programming blog

MORE: Here's the bookmarks of this post, care of the Tasty? bookmarklet I use frequently.  In comments, there is a Wordpress plugin that does Tastebuds with a convenient bookmark this post link.  However, the tag links go to all links with a given tag, rather than BlogURL+tag which gets you reader tagged archives for a blog.

Mesh the World

This week I participated in an off the record dinner with some of the smartest people I've met.  20 people and a single conversation without a whisper of Web 2.0 or GYM.  In the midst of our optimism that something was actually new, Gary Hamel called us on it, and pointed out the simple facts of abject poverty and unconnectedness for most of the world.  Besides the cell phone, hand-crank radio and $100 laptop there is little hope.  The former ambassador of Rwanda told us his story and imparted how poverty mattered more than security.  His plea was not to connect Africa to the Internet, but to help with local connectivity.

The greatest minds in tech do deeply care and try to work on solutions.  The first instinct is to connect everything to the Net, believing the rest will take care of itself.  But the cost of far-flung and rural connectivity, especially to impoverished regions, is simply too high.

There may be a different approach to this problem.  Instead of focusing on connecting the distant, help them connect with themselves.  In the absence of connections, nodes are state attractors.  That is, if you help a village connect with itself, the village will attract connectivity with others and eventually the Net itself.  In effect, when you create value in a local node, an arbitrage condition can drive enterprising individuals to make connection happen.

The cost for local connectivity is plummeting.  Wifi is in hypergrowth, leading to commodity production of radios, low power chips and innovation in software.  The $100 laptop comes with Wifi.  With advancements in Mesh Networking, you could gain an approach for networks that expand locally. Wikipedia may be available in their local language on a CD.  But also bundle in an open source wiki to let them build their own web.  As local nodes grow in strength, people will proxy for groups until they can be connected.

UPDATE: Windmill powered Wifi.  $20 per unit, designed for geological monitoring, but it makes you wonder.

Shai on Open Source

When I read this Vnnet article on Shai Agassi's comments on open source at the Churchill Club, it seemed too contrary my private conversation with him earlier in the week (disclosure: SAP Ventures is an investor in my company, Socialtext, a commercial open source company, that is).  Luckily, I was able to listen to a podcast of the conversation with John Markoff.  In a five minute span at 35:30, you will find the comments being taken out of context.

UPDATE: I see Jeff Nolan made a similar point to the above, biases aside, you decide.

FURTHER: Shai blogs on Open Source in his own words (thanks, Chris).