Thursday, October 28, 2004

Slashdot interview with Neal Stephenson

On Slashdot there's a great interview with Neal Stephenson that I recommend to anyone fond of geek fiction. For those that don't know, Stephenson's brilliant novel "Snow Crash" was the primary inspiration for "The Matrix" movies. (Though William Gibson fans might contest that). I became a Stephenson fan when I was in the avatar business about ten years ago, and even had a brief but distant brush with his work.

In the mid-late 1990's I was invited to Industrial Light and Magic to give a presentation and demo of the state-of-the-art in PC-based avatar technology. My company, a London-based simulation developer, was at the time way ahead of the industry in producing photo-realistic 3D avatars that could run in real-time over the internet on an unaccelerated PC. ILM's virtual characters were all high-end SGI-based graphics that took hours of processing time just to render a few frames. ILM built for the hi-res silver screen; we built for a pixelated 14-inch monitor. They had been tasked by a company that had just acquired the movie rights to Stephenson's Snow Crash to come up with a production design for a movie to star Keanu Reeves, during much of which he would be an avatar, a virtual character existing inside cyberspace. ILM needed to know what the near-future would actually look like.

In those days ILM's San Raphael premises were well-camouflaged, looking for all the world like a furniture rental business. On walking through the door the full-size Empire stormtrooper parked in the reception area dissipated the illusion. The place had bits of movie models strewn around everywhere: the death star model was sitting on a bookshelf in a passageway, the chunnel train from Mission Impossible gathered dust on top of a filing cabinet. Out in the back lot were some less-shabby models for current productions that I was asked not to notice.

I walked into the boardroom in my NY-LON look: dark suit, tie and shades, and discovered the CEO and everyone else on the production design team in shorts, sandals, and polo shirts. Somebody politely said they were expecting someone more weird, and that I looked like an anti-geek from the bizarro world, dude. Perhaps I inspired the look of Agent Smith. Perhaps not.

The Stephenson interview is interesting, as is his writing, because he is not an author trying to imagine a sci-fi future: he is first and foremost a hacker, with a geek's understanding of technology.

Wednesday, October 27, 2004

Most Search Marketers are unsophisticated

According to ClickZ's analysis of a new Jupiter report most search marketers are unsophisticated. The study says there are two kinds of search marketers, Sophisticates and Unsophisticates, and they each have characteristic approaches to their work.
Only 25 percent of search marketers use sophisticated bid strategies and measurement techniques, and those who do are more likely to be old-timers, big spenders and direct marketers.

Sophisticates have bigger budgets, smarter bidding strategies, and buy in many more words than unsophisticates (39 percent will buy more than a thousand words in a campaign, while only 14 percent of Unsophisticates will go that high).

What disturbs me about Jupiter's segmentation is that it's all about bidding for keywords. Looks like search marketers have rapidly descended from the dizzying heights of e-marketing gurudom to the less glamorous role of media buyer.

Tuesday, October 26, 2004

European e-learning: “experts” get it so wrong

Prompted by a scathing review of ACTeN’s most recent “E-content Report” by Heike Philp of Kolabora, I downloaded the report to see what the fuss was all about. I don’t find anything to support Heike’s conclusion that in ACTeN’s vision real human-being teachers and trainers in classrooms will effectively become redundant. But I have to agree that the report is woefully inadequate in a number of key conceptual areas.

ACTeN (Anticipating Content Technology Needs) is an EU-funded project that
aims at stimulating the development of a European e-content industry by monitoring the digital media market and by transferring know-how in Europe.

Given ACTeN’s content-centric brief, it is not surprising that it ignores or dismisses much of the value that internetworking brings to learning and focuses only on “e-content.” But it is alarming, and to my mind outrageous, that the EU can set up a project that is doubtless spending a great deal of taxpayers’ money investigating the wrong thing and focusing the attention of learning professionals in the wrong direction.

Having lived in Europe for much of the past two decades, I am aware of the disdain that many Europeans have for the experience of their colleagues in the US. But Europeans have come to the e-learning party rather late, and they would do well to learn from the many mistakes made by the industry across the Atlantic.

One of the central mistakes made early on in the evolution of e-learning was to focus on learning content instead of on learning process, and to see the internet as networked technology instead of as networked people.

That core conceptual blunder resulted in the metastasis of vendors peddling pedagogically impoverished courses-by-the-Gigabyte along with Learning Management Systems exclusively designed to serve it up. It has taken (in internet time) an awfully long time for learning professionals to start backing away from that model and to start moving toward a more enlightened view of technologyahanced learning.

Heike Philp is clearly a fan of synchronicity, and from that perspective her condemnation of the ACTeN report is quite valid. She feels that teachers and trainers will migrate to a virtual classroom. The report, bizarrely, ignores this big chunk of the e-learning industry, possibly because it doesn’t fit ACTeN’s concept of industrially packageable content. But the report does say that most companies have adopted various implementations of blended learning rather than “pure” e-learning, so the classroom is still in their picture by implication.

Just some of the statements in ACTeN’s report are worth commenting on:

Content, technology and services are the three key segments of the e-learning industry, where content still seems to receive the main focus.

This is not only an over-simplification, it is just plain wrong. It seems that ACTeN is supposed to help define how Europe can develop its e-learning industry, meaning define how vendor companies can make money out of the e-learning opportunity. So, they see three segments in which companies might want to compete: they define “technology” as LMSs, “content” as the kind of courses provided by NETg and SmartForce, and “services” as ASP services.

I appreciate that the project is trying to help evolve vendors in the industry. But should it not also be trying to help European employers find the most effective way to benefit from e-learning? Encouraging the growth of a vendor base along the lines described above will simply replicate the same dire misuse of opportunity that already occurred in the US.

Companies that want to be successful as vendors in e-learning should be looking for ways to facilitate truly effective value-added performance-improving learning experiences, not join an already commoditized course production line. The “learning supply chain” model that is entrenched in American e-learning is a 1970’s music industry publishing model that has no place in 21st century e-business.

In order to build a successful e-learning business scenario the cost for an e-learning programme should be lower than an alternative classroom-based, instructor-led training.

This assumption, bought into by trainers and learning decision-makers early on in the e-learning industry, is one of the reasons that so much e-learning today is so bad. Technology should be used to enhance the learning output, not merely reduce its cost for an acceptable loss in quality.

As e-learning becoming more mainstream, learning practitioners are more and more positioning themselves for the next significant movement in the use of technology in learning, namely simulations.

While simulations will play a part in some aspects of acquiring some skills, they will not be anything like a mainstream next phase in e-learning. Anything but the most basic pre-programmed simulations are currently way too expensive to design, build, and maintain. The simulations that will become more popular invoke real people interacting virtually with other real people, rather than people interacting with a machine programmed to replicate such reality.

The internet has a lot more in store for e-learning, including collaboration, deep blending, peer-to-peer, mentoring, social networking, synchronous and asynchronous groupwork, dynamic communities of practice, wikis, blogging, workflow learning and learner-generated content. I’d be encouraging would-be vendors to develop products an services that facilitate learning experiences such as these, rather than pursuing the supply of simulations and canned courses.

Monday, October 25, 2004

Visualizing power and creativity

Here are two sites worth wasting a bit of time on, each adopting interesting "plastic visualization" approaches to presenting complex data-rich interrelationships.

Power: Who runs corporate America, and how are they interrelated? There is some fascinating content at They Rule. They Rule plots the boards of directors of American companies, and shows you which other companies those individuals have board seats on. You can drill down into each individual, link to deeper data on the web, and plot interrelationships. They Rule uses some very friendly use of drag-to-place graphics.

Creativity: An interesting way to pull together the different organizations involved in different fields of creativity around the world appears on NGf - Map of Creativity. The interface takes a bit of getting used to, but it is a very efficient way of packing a huge amount of data into a single small space.

Tuesday, October 19, 2004

Thought-mail: a paralysed man controls a computer by thought

The cyborg era is drawing closer as scientists discover more and more about the electro-mechanics of our minds and bodies. Five years ago, we were able to "see" through the eyes of a cat. Then wireless chips were used to bypass damaged nerves and stimulate muscle movement in previously paralysed limbs. Now a quadriplegic has had a tiny chip implanted in his brain that allows him to control his computer and send e-mail simply by thinking.
In June 2004, surgeons implanted a device containing 100 electrodes into the motor cortex of a 24-year-old quadriplegic.. Each electrode taps into a neuron in the patient's brain. The BrainGate allowed the patient to control a computer or television using his mind, even when doing other things at the same time.

The chip's 100 electrodes can tap directly into 100 neurons, and theoretically can transmit whatever each of those neurons is communicating. The learning curve is apparently rather short compared with other systems used by paralysed people, such as eye control or tongue-control.

Now, when the direct communication between mind and computer gets to be two-way, you won't need a keyboard OR a screen. In fact, your PDA might simply become a wireless supplement to your mind. And if it's a PDA connected to the web, your potential mental capacity is scary, to say nothing of your ability to communicate by thought -- imagine the future of IM. But I'm getting ahead of reality just a bit. Or am I?

Friday, October 15, 2004

Googling your PC just like the web

Yesterday Google released the beta version of its Google Desktop Search tool. It's available as a free download here. When you first install the tool, it sets about indexing every file on your system. In my case, that took nearly a day. Indexing takes place only while your computer is idle, and involves spidering the contents of each and every file it can get into -- including e-mails, documents, spreadsheets, databases, Instant Messages, images, and system files.

Once your drive has been indexed, Google Desktop Search searches all of the files on your computer in the same way that Google searches the web. Just enter the search term and click the button, and instantly a listing that looks remarkably like a web search results page lists all of the files on your system that contain that term. No more frustrations trying to get Explorer to find a document (or was it a spreadsheet, or maybe an e-mail?) that you last used months ago. It's brilliant.

There have been other desktop search tools, of course, but this is the big league and, if you are a regular Google user, it feels comfortable. And it does something that other desktop search tools don't do: it integrates your desktop search with your web search. If you search for a term on the web, a summary link to all of the files already on your computer containing that term appears at the top of the list.

What's more, Google Desktop will search all of your previously viewed web pages, and show you thumbnails of the relevant matches.'s A9 and AskJeeves recently launched personalized search features that let you store viewed web pages for future reference, but unlike Google Desktop, they don't do it automatically. I prefer the A9 system, because I try to keep my system as clean as possible, and purge history files, cookies and so on daily. That means Google Desktop is never going to find past-viewed pages; A9 keeps its own folder of only those pages you specifically ask it to save. But combine Google Desktop Search with LookSmart's Furl, and you get the best of both worlds.

Google's Desktop Search does not serve up ads, and if their privacy policy does not change, in-context ads will not be possible. Google does not feed any data about the contents of your desktop, or any personal data that you do not authorize, back to its systems.


Lycos becomes a social network

Lycos has released a beta of a consumer service that, at first glance, may be an excellent prototype for a corporate toolset that makes collaborative social networking among members of project teams (or within restricted communities of practice) easy and intuitive.

Last year, social networking was the Big Buzz. Friendster, a peer-to-peer networking environment became instantly popular among younger web users. Spoke made some inroads into professional users. Even Google launched a social networking system called Orkut. None of them set the world on fire, but maybe that's about to change.

Earlier this year, Lycos said it was going to change its business model and become a major player in social networking. Lycos changed ownership two months ago when the South Korean company Daum Communications bought it from Terra Networks, but that transition has not affected its change in strategy. In fact, a Korean perspective on networking may be just what is needed. Yesterday, Lycos released the beta of its new incarnation: Lycos Circles.

Lycos Circles is a "social sharing platform" designed to help users users stay in touch with their different circles of friends and acquaintances. Other social networking tools have tended to focus on broadening your network of contacts, or helping you connect with others via mutual acquaintances. Lycos Circles seems to be focused on adding value to existing social relationships.

Unlike the major blogging tools, which allow anyone on the web to see your content, Lycos Circles lets you restrict access to different circles of people. Lycos Circles gives users a set of instant personal communities within which users can choose what to share or discuss and what not to share or discuss, and who to include in which circles. You might have a family circle, a friends circle, a professional circle, a neighborhood watch circle, and so on, each with different priveleges and levels of access.

Lycos has a number of properties that can be invoked to make these social networks more feature-laden and content-rich: Angelfire and Tripod allow blogging, personal websites, and photo sharing, and People Search and Discussion Search allow you to find and connect with others. It's all about creating and sharing dynamic, relevant content that is not only personalized, but "circle-specific".

As with all new concepts on the web, the new Lycos business model is a little unclear, though apparently in-context ads a la Google AdSense will appear within your content. Lycos Circles is also partnering with outside services, such as the truly excellent online photo service Shutterfly.

Thursday, October 14, 2004

An RFID chip on your shoulder?

A little over a month ago I suggested that fast-food operators and coffee chains give their frequent customers a card with an imbedded RFID chip. Scanners at the door would recognize the customer and would transmit his/her most likely order to the kitchen well before the customer gets to order in person. Some interesting privacy issues there, but the fast-track VIP status would be rather appealing to many customers.

Now, amid all the controversy about WallMart’s obliging its suppliers to RFID-tag their products, and the resulting murmurings about potential invasions of consumer privacy, comes yesterday’s announcement by the FDA that it has approved the implantation of an identity chip under the skin of human beings.

The same technology that has been used for fifteen years or more to “chip” domestic pets is finally available to people in America. The FDA hopes that the chip, encased in a glob of glass smaller than a grain of rice, will be used to make vital medical information instantly available in emergencies. The chip is syringe-implanted in the upper arm in seconds, and when a scanner is waved over the body, the chip transmits its imbedded data.

The applications of this technology already go beyond the healthcare concept. A few weeks ago, an article on the BBC news site talked about a beach club in Barcelona that offered implanted chips to its patrons as a way of controlling access and instantly paying for drinks, like a tollbooth smartcard. For a beach club, this makes perfect sense: you get to leave your wallet and your ID card at home.

The time will probably come when we all have a dozen or more chips in our arms, one for every piece of plastic we now carry in our wallets. The state of Virginia is already looking at imbedding RFID tags in driver’s licenses. Can the move to arm implants be far behind? And speaking of driving, do you really need car keys if your car can recognize your chip?

Where is George Orwell when we need him?

Parkin Space - a new weekly column on the UK's TrainingZONE

Claire Savage, the editor of the UK's TrainingZONE has asked me to write a weekly column for the forty thousand or so members of that online community, and the first Parkin Space piece went live last Friday.
Sometimes controversial, sometimes combative but with a liberal dose of common sense, Godfrey Parkin's columns take a sideways glance at topical training issues.
It's a challenge to remember to spell in English English, but Claire's editing is excellent, so I'm not sure why I bother.

If you are not a TrainingZONE member, registration is free, and the community provides a uniquely British insight into training that is sometimes inspiring, sometimes a little scary.

Some changes in the blog

I've had a number of e-mails asking if I'm still blogging or if I've abandoned the effort. Good to know there are actual readers out there, not just stats on the tracker! I just took a week off to get reorganized, and am back with some minor changes..

I have drafts of eight to twelve topics for every one that gets posted, partly because I have been reluctant to do the typical blogger thing of saying "here's a link that is interesting, go read it" and partly because the Blogger default keeps putting that irritating "read more" message in even when there IS no more.

I am now persuaded that providing a link and brief introduction is better than the items not making it at all. In the meantime I'll try to hack a solution that prevents the read more message from showing up every time. Anyone who has a ready-made solution, please e-mail me!

And maybe if I am less long-winded there will be more comments within the blog itself (though please feel free to continue e-mailing me if that's what you prefer).

Tuesday, October 05, 2004

Search sucks. Foraging rules.

Juan Dürsteler at InfoVis suggests that
The behavior of human beings when searching for information intensively resembles that of the hunter-gatherers of our past and that of the foraging of animals.

The way we present information on the web should acknowledge and exploit what we know about this foraging methodology.

In his article, he points out that biologists believe that animals go after a food target that maximizes ROI, delivering the most nourishment for energy expended; so they will target something young and frail rather than something large and robust, even though the total reward is less. They will move on to fresh hunting grounds when the energy involved in foraging outweighs the probable payback of finding anything to eat. We humans would do that with complex decision-making algorithms; other animals just do it.

The theory goes that data such as hyperlinks have an "information scent" that implies more than is overtly revealed about what lies beneath, and we have to learn to sniff out which paths may be the best for us to search down. Some of us get rather good at it; others flounder.

The lessons for information architects (as they like to be called) are clear: populate your sites with strong but un-confusing information scents, make your search paths clear and inviting, and make sure your information ecosystem provides more easy prey than hunting effort. Clear out the unnecessary information and gratuitous design elements, and build an environment that makes foraging a rewarding experience.

The last line of the piece is interesting.

Maybe we’ll have to begin thinking about the creation of an Information Agriculture.

I used the term "information agriculture" in a debate at a conference in Paris back in 2001, where I criticized then-mainstream approaches to Knowledge Management as unproductive foraging-and-storage rather than productive planned agriculture of information. Of course I was roundly rebuffed by a Knowledge Management professional from the European Commission for failing to understand that the purpose of KM was to gather, catalog, and make accessible the knowledge inherent to an organization -- not to proactively do anything with it. But I think things have changed since then.

Thanks to Maish Nichani of elearningpost for pointing out this article.

Asymptotic freedom in the color force

This has nothing to do with marketing, training, or the internet, but it's very cool. The 2004 Nobel Prize in Physics is going to three Americans, David Gross, H. David Politzer, and Frank Wilczek for discovering that, in the tiniest of particles, gravity appears to go into reverse and act like a rubber band.

The closer to each other that the quarks inside a proton and a neutron get, the weaker is the force that pulls them to each other; the further apart they move, the stronger the force. When the quarks get very close to each other, they act almost like free particles that are not bound to each other by any force at all.

The force attracting quarks to each other is known as the "strong force" or the "color force". The reverse-gravity phenomenon is called ”asymptotic freedom”.

Why does it matter? Because matter matters, and quarks are at the heart of matter. The discovery is a foundation of the theory of Quantum ChromoDynamics, which in turn is important to the dream of building a theory of everything, a model that explains the interactions among everything from the largest objects in the universe to the smallest. And all of us in between.