Neuland

Why Work Sucks

Not my headline, but the title of a new book by work consultants Cali Ressler and Jody Thompson. I met them for a brand eins story about Best Buy's Results Only Work Environment (ROWE) a year ago. After spinning their in-house rebellion out in to an independent consulting shop, the two have now released a book called "Why Work Sucks and How to Fix It."

It expands on what it takes to free employees from the grueling meeting marathon, senseless deadlines, and mandatory office presence that most companies still cling to -- or have even brought back. The flip side of more freedm is, of course, being responsible for one's own workload, scheduling, and to make the judgment call between checking mail, signing onto IM and switching off to play.

It sort of flies into the face of another hot and hyped idea -- the 4-Hour Work Week propagated by Timothy Ferriss. His is a post-industrial, utopian recipe for personal outsourcing and info dieting. It sounds tempting, when Tim explained it to me at this year's SXSW in Austin, but its bare-bones approach does not work for most of us, I am convinced.

Why not meet in the middle, and leave it to each knowledge worker to say when? Some of us are long-distance workaholics, others thrive in meetings, yet others make do with quick bursts of activity. In fact, most people do not shun the trip to the office when given the choice. Just knowing you are in control makes a big difference, though. Just don't let management tell adults when to be where if it's not vital. Or, in the words of Ressler and Thompson: "People across the planet want better lives; employers want better results; neither side has to compromise." If you're curious, they offer a free PDF of their intro and first chapter here.

Start making sense

Companies large and small are betting that the semantic web is ready to party parse. Now that the building blocks and standards are falling into in place, it's time to develop business plans and use cases for a web of deeply linked data. That's the focus of my most recent piece in The Economist.

A key player in this promises to be Reuters, the traditional news service, which opened up its recently acquired text mining service Calais via an open API. Any website can send its raw text data through their natural language engine -- very soon even a lonely blogger. It's a win-win situation, as Calais evangelist Tom Tague told me. Reuters gets to see a lot of unstructured data and can train its algorithms via crowdsourcing. That allows it to build understanding and context -- and over time develop a clearinghouse of meaning. The users meanwhile can piggyback on a scalable technology that returns context in under a second when it's working properly.

That's the plan, but until we get to the point where a service like Calais can understand and make sense of various domains beyond business and finance, we got some ways to go. If it works, though, another highly complex task will have been commoditized. It begs the question what the value proposition of startups such as Radar Networks will be, which have spent years building a homegrown platform to perform similar feats of sorting, connecting and interpreting.

Right now, the whole semantic space is a bustling construction site, according to AdaptiveBlue founder Alex Iskold. He did a splendid job summarizing the key components in a recent post on Read/WriteWeb. Semantic technology can do search one better -- but it's not the killer app by default. Simply hunting for nouns in a standard search engine like Google works well, but it does not yield satisfying results for more complicated queries a human would ask another human, according to Alex. He thinks enterprises will board the semantic train before consumers do, for two reasons. They sit on silos of data that are waiting to be organized and monetized, and companies want to brag about the latest technology to look smarter than the competition. “Solid use cases are still rare, it is more about experimentation.”

Among the newer entrants is Qitera, a semantic engine (in beta) based in San Francisco, founded by three Germans. Joerg and Carlo walked me through their vision of how a web of meaning can be woven by tapping into services such as Freebase and Calais, and I've been testing it for about a month now alongside Twine. Instead of a pure top-down approach, says Qitera's cofounder Joerg Lamprecht, they want to leverage algorithms and human curation. Since it's a closed service right now, it has technophiles on RWW guessing whether it competes head-on with Radar's Twine.

Yes it does. In Qitera, users can build their own syntactic triples (A knows B, B lives in C, A has been on vacation in C and read book D etc.) -- basically drawing almost free-form connections between entities such as people, places, events and so forth, plus use a Firefox plug-in as they surf along. The human touch augments the entities and relationships the service can suck in from other sources -- like a mindmap on steroids. For practical consumer applications, says Carlo, semantic parsing and mark-up needs to go further and include concepts from popular culture: showbiz, sports, shopping. Users, after all, don't care how something works online. They will think "meaning ex machina" is cool when they see a service that understands what an album is and how it relates to a particular label, a promoter or related bands. If users slap those things on, so be it.

Ironically, those questions would all be no-brainers for a record store employee, but those guys have been killed off by the same technology trends that now claim to bring knowledge back in automated form. So let's start making sense!

Person Tom Tague

Right click for SmartMenu shortcuts

Blogging can kill you

Perhaps blogs should contain a health warning: What you are about to read was produced at the expense of the author's health. Seriously.

What I reported on for brand eins two months ago under the headline "Bloggen bis der Arzt kommt" is featured in today's NYT, In Web World of 24/7 Stress, Writers Blog Till They Drop:

"A growing work force of home-office laborers and entrepreneurs, armed with computers and smartphones and wired to the hilt, are toiling under great physical and emotional stress created by the around-the-clock Internet economy that demands a constant stream of news and comment."

Back then, I had spoken with high-profile writer-preneurs such as Om Malik and Matt Marshall about the incessant, self-induced stress to file and break stories 24/7. They saw the death of Marc Orchant and Om's heart attack as outlier events. I continue to think, though, they are symptomatic of the bigger problem with the big shift in how we produce, distribute and consume news and what passes as such.

A good case in point is the trigger for Sunday's NYT story -- the sudden death of Russell Shaw, yet another stressed-out tech blogger. Andrew Keen, the enfant terrible of the Silicon Valley blogging bubble, made an excellent point when I met with him: Today's bloggers are the Stakhanovites of the digital age, working themselves to death with mythical superhuman energy reminiscent of Stalinist propaganda. All blinded by the claim of building a new world. Not that this is exclusively true for bloggers, but the technology enables obsessive self-destruction more than journalism ever did.

In spite of the title of Keen's book "The Cult of the Amateur," these guys are mostly serious about generating traffic and making money. And very good at marketing even their plight. Just take Michael Arrington's comment in the Times:

“I haven’t died yet,” said Michael Arrington, the founder and co-editor of TechCrunch, a popular technology blog. The site has brought in millions in advertising revenue, but there has been a hefty cost. Mr. Arrington says he has gained 30 pounds in the last three years, developed a severe sleeping disorder and turned his home into an office for him and four employees. “At some point, I’ll have a nervous breakdown and be admitted to the hospital, or something else will happen.”

“This is not sustainable,” he said.

So, off we go to this weeks' sold out TechCrunch party in LA and a million other events until the TC 50 conference in September, which is timed to conflict with an established event, Demo Fall. What's a stressed  live-blogger to attend? Indeed, this is not sustainable.

Person Matt Marshall

Right click for SmartMenu shortcuts

Search This!

Social networks are a dud when it comes to ad revenue (WSJ story, sub. required), but they sure are a treasure trove for intelligence agencies. Ironically, we are all willing enablers in the big transparency experiment that threatens to digitally enslave us in the process. Datamining has never been easier, because each new "friend" sells you out to spy agencies and marketers without you ever knowing.

"People, especially in the U.S., say they are concerned about  privacy, but they seem to forget that as soon as they go online," tech-observer Nicholas Carr told me recently when we chatted about his new, excellent book "The Big Switch".

Steven Aftergood from the FAS Project on Government Secrecy relates on his blog how this symbiotic ecosystem works (via SECRECY NEWS):

The DNI Open Source Center, which gathers, translates, analyzes, and
distributes unclassified open source intelligence from around the
world, is steadily growing in capability and impact, according to Doug
Naquin, the Center's Director.

The Open Source Center, which replaced the CIA's Foreign Broadcast
Information Service, is doing more analysis and outreach than its
predecessor and is also exploring new media, said Mr. Naquin in a
recent speech [3 Oct 2007].

"We're looking now at YouTube, which carries some unique and
honest-to-goodness intelligence," he said.

"We have groups looking at what they call 'Citizens Media': people
taking pictures with their cell phones and posting them on the
Internet.  Then there's Social Media, phenomena like MySpace and
blogs.... A couple years back we identified Iranian blogs as a
phenomenon worthy of more attention, about six months ahead of anybody
else."

"But we still have an education problem ... both with the folks who are
proponents of open source but perhaps don't know exactly why, and folks
internally who are still wondering why I am sitting at the same table
they are."

"All of us have heard the statement by [intelligence community] leaders
at one time or another that 'Our business is stealing secrets.' Or 'Our
business is espionage.'  While I deeply respect that, and I understand
where that's coming from, from my Open Source perspective, I'm thinking
that's like a football coach saying, 'Our mission is to pass the ball.'
Or 'Our mission is to run the ball.'  Well, not exactly.  It's to win
football games."

The full text of Naquin's address to the Central Intelligence Retirees' Association is available as a PDF.

Boohoo Yahoo

So it has come to this: Battleship Ballmer makes a grab for Net pioneer Yahoo! in order to plug gaping holes that Microsoft cannot fill on its own. As plenty of analysts have pointed out, it is a parting shot in an old war -- the fight over ad-supported search where Google is far and away in the lead, seemingly unstoppable. For Microsoft to buy their way into something they never fully understood -- the Web -- is indeed reminiscent of the ill-fated logic behind the AOL-Time Warner deal.

It's also fair to point out that all the hoopla over search, market share and the advertising billions all three players are after leaves out the key piece. Innovation is happening somewhere else, most likely in places we do not even know about. This type of innovation, Silicon Valley-style, happens bottom-up, in small outfits, start-ups and most importantly in utterly unexpected places. It will most certainly come from the very people who get kicked out of Yahoo! or will leave the company once Redmond takes over.

What Yahoo! does have, but hasn't fully leveraged, is the deep pool of users and knowledge of their social activities, a good half billion of them. They could have built smart applications that combine the insights of email with social networking, life-streaming and rising semantic capabilities. That they didn't is a further hint that this new giant -- if it comes to pass -- will not drive innovation towards the semantic web and the attention economy. It's no wonder that flickr and del.icio.us were acquisitions.

The fatal flaw would be to turn these new ideas into a business proposition too soon -- Facebook fell on its face with its big sell-out called Beacon. Once the network has become a social organism, the backlash is swift, and the revolution will not immediately be monetized.

So where do we go exploring while the elephants are wrestling?

If you look closely enough, there are plenty of small pieces, loosely joined, of the puzzle that can make man and machine understand each other's actions and intentions better. That can be location-based enablers such as EarthMine, new types of search which combine social aspects with algorithms such as just-launched Delver, semantics-enabled social media apps such as Qitera or ShareThis, and analytics tools such as Rescue Time. And finally, the still unsolved challenge of fluent text-to-speech and speech-to-text which will unlock so many areas, from mobile to navigation to all kinds of new social web services.

I am rooting for the small experiments that the Ballmer-Yang wedding party will not have time or eyes for. Why not let a novel type of software agent analyze my productivity stream, while another agent sifts through my mail stream, and yet another one parses my travel and shopping streams and -- with my explicit permission -- exposes it to some select people and services. Blue Organizer and Twine are baby steps into areas where the giants fear to tread, but the masses are ready to work and play.

Meaning Trumps Social, Take 2

Tim Berners-Lee has a great post up on the development of the Semantic Web, the "social graph" and how it all comes together when the goal is meaning for humans.

In a nutshell, Berners-Lee writes:

So the Net and the Web may both be shaped as something mathematicians call a Graph, but they are at different levels. The Net links computers, the Web links documents.

Now, people are making another mental move. There is realization now, "It's not the documents, it is the things they are about which are important". Obvious, really...

We can use the word Graph, now, to distinguish from Web.

I called this graph the Semantic Web, but maybe it should have been Giant Global Graph! Any worse than WWWW? ;-)   Not the "Semantic Web" term has been established for a long time, I'm not proposing to change it. But let's think about the graph which it is...

It is about getting excited about connections, rather than nervous...

I'll be thinking in the graph.  My flights. My friends.  Things in my life.  My breakfast. 

Lijit's Stan James takes up the idea and has a very insightful post on why information from people I don't really know shouldn't be meaningful -- hello, Facebook Beacon?

Stan writes that, to him, relationships between people and things reign supreme:

Who said something is infinitely more interesting than what has been said.

What does it benefit a user if the computer can understand the meaning of "Buy Viagra now for cheap! Enlarge your penis!", but it doesn't know who said it?...

Here at Lijit (the company which grew out of Outfoxed), we surface this idea in Search. But it will eventually permeate every aspect of our online lives: message exchange, discovery, shopping, advertising, religion, and more.

In other words, the semantic web as a bunch of enabling technologies or standards without factoring in social engagement and trust (whether it's among humans or their software agents and APIs) is a pretty dumb idea.

Gimme Me Some Meaning!

The question keeps rearing its confusing head: What for Pete's sake is the Semantic Web, and what will it do for the average user? Yet another attempt was last week's panel discussion held by Stanford's VLAB. The names of the panelists sounded promising, but after mulling over my notes I've come to the conclusion that I could have saved myself and the planet several pounds of CO2 by not going.

In the ring were futurist Paul Saffo, Radar Networks founder Nova Spivack, Alex Iskold from Adaptive Blue, VC-blogger Paul Kedrosky and Freebase cofounder Robert Cook. When they weren't entangled in all those geeky acronyms like RDF and OWL, they mostly talked about how search should and could be better and how they will do just that.

Spivack, whose company has been working on its service called Twine for several years now, gave another tantalizing preview of what feats Twine will perform to automate collaboration. But the beta is still limited to a few hundred people and reminds me more and more of the ongoing vaporware performance for alleged Google challenger Powerset than of a real launchable product. Let's finally play with Twine's automated Facebook-ish features -- that's what a recent demo looked like to me -- and see what new meaning of online life we get.

Spivack, whose experience with semantic web technologies and concepts goes back to SRI's and DARPA's CALO Project, predicted that "in 20 to 30 years, almost all human knowledge will be accessible and stored in a machine-understandable fashion." And that appears to be the key to all attempts at harnessing AI for a nice and useful human computer interface: let machines and agents peruse the web so they can haul in, structure, sort and analyze all the stuff for us, bring back the nuggets, perhaps even pre-emptively suggest things we like, care about or should be alerted to.

Things a good butler, assistant or tavel agent would do -- if the geeks and lean and mean management types hadn't killed them off! I tried to make that point in an Economist article several months ago, only to start a debate among Tim O'Reilly and Co. what the proper definition of Web3.0 and the semantic web was.

It's not just standards -- that much I know after last week's disjointed discussion at Stanford -- it's all about meaning and better, i.e. meaning-ful usability for machines and humans alike. I agree with Alex Iskold's superbly insightful posts and comments that it makes little sense to go for the all-out AI assault when there is so much low-hanging fruit out there. We can harvest them by leveraging human ingenuity, our playfulness, self-service motivation, or even vanity in conjunction with existing data sets. Layer the human factor on top or underneath the new technologies and the results can already be pretty amazing.

Adaptive Blue's Blue Organizer understands dozens of concepts and preferences without many bells and whistles or techno-jargon. It's not so much about search and the next Google, but about finally having an able virtual clerk who hovers nearby, who can remind me, remember things, file to-dos, make connections and pull up items when I ask for them. In that sense: Yes, services like Sandy in combination with Jott's transcription offering are definitely part of the rising semantic web, just as TripIt and its trip organizer on steroids is. To deny that is sticking to techie jargon.

Paul Kedrosky (of Infectious Greed fame) made a good point when he said that financial markets are already swirling about in an overlooked semantic web. On any given day, 17 percent of all business and finance stories on the Reuters newswire are machine-produced, with no human involved  -- they are not even intended for human consumption. A whopping 55 percent of Reuters finance stories are only "read" by machines, according to the VC. "Wall Street is way out front," claimed Kedrosky, in the way autonomous software consumes, processes and extracts signals from information and turn it into actionable bits and pieces. That type of semantic web technology makes humans money today.

The AI hype has a sounder basis this time around because the data and the processing are decentralized and not just going on in a few big labs. Which raises an intricately connected problem.

How nice would it be if the data, my data, was actually open? If I could get my ratings on Netflix or any other site as a flow of personal insights, preferences and transactions that other services could reach into, as Dave Winer pointed out? Right now, chances are slim that such openness will really spread. It's valuable, proprietary stuff. And it lowers the barrier to entry and exit to let users or their software agents wander about with a cornucopia of transactions and a history of their engagement. Facebook's Beacon has it backwards, since it assumes opt-in to semi-open data and exploits the very users who entrusted their "social graph" to it. It's the latest twist on the Faustian bargain of "free Web services."

It's shame, because without such portability into which software and eventually their human masters can tap, neither the semantic web nor the so-called attention economy have a chance to go beyond baby steps and hyped-up announcements from start-ups.

In the meantime, the big Net players monetize our data by offering allegedly "free services" like beads to savages. We leave tons of functionality, meaning and money on the table. If we can harvest and control access to our own data and presence streams and give new apps access to it so they can wring meaning out of it (instead of Comcast & Co. selling our clickstreams behind our backs) -- then we can rightfully speak of the advent of the semantic web.

Or, to use Alex Iskold's closing words at the Stanford event: "My biggest fear is that AI will eat us." That evening at the VLAB event, it certainly didn't. It just yawned and showed some baby teeth.

The Blog Cascade, and How to Kindle it

Assume for a second that information spreads like an infection, that blogs are nothing but nodes in a scale-free network of bugs that want to proliferate. Which blogs should we pay attention to in order to read those posts that will spread like the plague and be the news fodder of tomorrow?

A half dozen computer scientists at Carnegie Mellon have developed an algorithm for "cost-effective outbreak detection"  called Cascade. It sifts through way too many feeds and finds those that matter by posting frequency, in- and outbound link density -- with quite amazing results when it comes to what constitutes the big, fat head of the long tail. The top 100 infectious news blogs run the gamut from politics to personal to outright strange.

The Cascade concept applies to more crucial questions, where power laws and tipping points lurk, such as pollution. From CMU's press release: "Carlos Guestrin and his students used the same algorithm to determine the optimal number and placement of sensors for detecting the introduction and spread of contaminants in a municipal water supply. Their report on the blog and water system case studies, 'Cost-Effective Outbreak Detection in Networks,' was presented at the Association for Computing Machinery’s International Conference on Knowledge Discovery and Data Mining earlier this year."

Here's the top ten list of news blogs ranked by their infectious quality, or biggest proportion of the population affected: 

1     http://instapundit.com   
2     http://donsurber.blogspot.com   
3     http://sciencepolitics.blogspot.com   
4     http://www.watcherofweasels.com   
5     http://michellemalkin.com   
6     http://blogometer.nationaljournal.com   
7     http://themodulator.org 
8     http://www.bloggersblog.com   
9     http://www.boingboing.net   
10   http://atrios.blogspot.com

Interestingly enough, when sorted by the question which blogs to read if we only had time to digest the top 5,000 posts, the list shifts:

1    http://sisu.typepad.com/sisu   
2    http://blogometer.nationaljournal.com   
3    http://www.watcherofweasels.com   
4    http://www.allthingsbeautiful.com/all_things_beautiful   
5    http://gevkaffeegal.typepad.com/the_alliance   
6    http://theliberalwrong.blogspot.com   
7    http://www.anglican.tk   
8    http://dendroica.blogspot.com   
9    http://gullyborg.typepad.com/weblog_archive   
10  http://trejrc0.blogspot.com   

Top news blogs will be less arcane source material once we carry better reading clients with us. Amazon's new Kindle reader, for instance, looks like it's out of a last-century sci-fi movie. It's certainly not the sleek silver bullet, but a nice step forward with its e-ink screen, EVDO connectivity, no wireless fees to struggle with and it's hyperlinking of potentially every morsel ever written. The device needs to be much, much nicer and sleeker. Kindle is to e-books what the first MP3-player is to the iPod: a proof of concept, and a courageous one.

Now imagine what somebody like Apple can do with the idea of an "always-on book" with multi-touch and cover flow control of documents, blogs, newspapers (what will be the difference between the two down the road anyhow?) and, yes, books purchased on iTunes... Or just wait until next year, I venture to bet.

Pathologische Vernetzung -- jetzt mit noch mehr Monetarisierung!

Nachdem die Details zu Googles OpenSocial Allianz endlich vorliegen, sieht der Aufstieg von Facebook etwas weniger beängstigend aus -- zumindest für all die anderen Social Networks.

Hier eine aktualisierte Grafik von Bill Tancer von Hitwise, die alle OS-Partner mit einrechnet:

Auf den ersten Blick beeindruckend. Aber es gibt reichlich Fragen:

• Für die (bisher wenigen) europäischen Teilnehmer wirft die Öffnung der "Social API" für so unterschiedliche Plattformen von hi5 über Orkut bis Xing neue Probleme auf: Wenn sich Daten zu Personen, den von ihnen gespeicherten und manipulierten Objekten und Activity Streams frei bewegen können, was heißt das für den Datenschutz, der europäischen Nutzer so viel wichtiger erscheint als US-Nutzern? Xing-Gründer Lars Hinrichs hat das in seinem jüngsten Blog-Eintrag nur kurz angerissen.

Das Thema ist umso brenzliger, als die Debatte um Vorratsdatenspeicherung in Deutschland tobt. Neben gesetzlich verordneter Überwachung bietet der neue Run auf Social Networking-APIs das perfekte Schleppnetz, gewoben aus dem gründlichen Überwachungsstaat und freiwilligem Daten-Exhibitionismus durch Lifestream-Seiten von Jaiku (ebenfalls Teil des Google-Imperiums) bis dem neuen Dienst Soup aus Österreich. Deswegen haben Verbraucherschützer in Washington bereits Alarm geschlagen (via San Jose Mercury News):

The potential is "Orwellian," said Michael Fertik, founder and chief executive of ReputationDefender, a Menlo Park-based company. "When you have a lot of traffic that comes from identifiable IP (Internet protocol) addresses that exhibit a lot of trackable behavior, you generate a staggering amount of rather specific information about individual users as well as classes of users. And in many social networks, the greatest part of their value is to identify users by name."

• Wird Facebook jetzt gezwungen, auf den Google-Zug zu springen und sich dem Standard anzupassen, oder riskiert Zuckerbergs Truppe, rechts überholt werden? Die Halbehe mit Microsoft, einem Meister des Walled Garden-Konzepts, macht die Sache nicht einfacher.

• Wie wird sich das boomende Ökosystem der ausschließlich für Facebook geschriebenen Anwendungen zu einer großen Nische entwickeln? Werden Programmierer nur nach der potentiellen Nutzermenge schielen, oder werden sie sich auf die etwas kohärentere Facebook-Zielgruppe von bisher rund 51 Mio. Nutzern konzentrieren?

• Wann steigen die anderen Besitzer von sozialen Daten ins Rennen ein: eBay samt seinem Skype- und Paypal-Netzwerk, Amazon -- und Yahoo!, eines der allerersten sozialen Netzwerke?

Richtig spanned verspricht die Debatte erst noch zu werden, wenn Facebook wie erwartet kommende Woche sein Social Ads-Konzept vorstellt -- Behavioral Targeting auf der Basis des so genannten "Sozialen Graphen." Das ist das Milliarden-Geschäft, hinter dem Google und Facebook/Microsoft her sind.

Die jetzt vorgetragenen Vorteile für Nutzer sind nur die Fassade, um die immer stärker fragmentierten Aufmerksamkeits- und Aktivitätsströme besser zu monetarisieren. Je mehr Daten wir digitale Krume um digitale Krume transparent machen, umso mehr Geld lassen wir alle auf dem Tisch. Unsere Aufmerksamkeit, gemessen in Freunden und Sekunden, ist die neue Währung der beinahe pathologisch vernetzten Welt.

"What are you doing" -- und wie viel ist Dein Status Update einer Marketingfirma wert?

Brussels, We Have a Brain Drain...

If you've read about a new "blue card," it's not a marketing ploy by Amex coming right after the Plum Card. It's the EU's attempt to attract more higher-qualified migrants, knowledge workers or members of the creative class. The goal is to entice 20 million workers from Asia, Africa and Latin America to move to the continent as "a magnet for the highly skilled," as EU commissioner Franco Frattini called it.

There' are plenty of reasons for concern driving the "blue card" idea (via the NYT, sub required):

"... high-skilled foreign workers accounted for 0.9 percent of all workers in the European Union, compared with 9.9 percent in Australia, 7.3 percent in Canada and 3.5 percent in the United States...

Figures recently released by the European Commission showed that 85 percent of unskilled laborers from developing countries went to the European Union and only 5 percent to the United States, whereas 55 percent of skilled workers went to the United States and only 5 percent to Europe."

Migrants with a good education and marketable skills prefer the US, Canada and Australia -- and the phenomenon feeds on itself -- it's the feedback loop of clusters in action.

A new research paper by Lynne Zucker and Michael Darby at UCLA explains why. It tracked the whereabouts of 5,401 prominent scientists from 1981 to 2004. Their final analysis in "Star Scientists, Innovation and regional and national immigration" is this:

• Top scientists are star innovators, and it's a regional phenomenon: "We argue ... that the geographic distribution of a new science-based industry can be mostly derived from the geographic distribution of the intellectual human capital embodying the breakthrough discovery upon which it is based."
• Top scientists are correlated to high-tech start-ups: "The statistical results are clear: ... regions or countries with more of these top scientists present in one of the six S&T areas also exhibit a significantly higher rate of firm entry..."
• Top scientists turned star innovators migrate to existing clusters of excellence, with the US in the lead: "About 93 percent of the world's star scientists and engineers in the world have residences in the top-25 S&T countries at the end of 2004 -- 62 percent in the US alone."
• The feared brain drain can reverse itself as soon as a scientist's home country develops enough critical mass in R&D and commercialization: "The four largest net immigrations of star scientists over the last quarter century were registered by the United Kingdom (-27 or 4.6 percent of all stars resident in the country at any time 1981-2004), the United States (-23 or 0.6 percent), Canada (-23 or 7.7 percent) and Germany (-11 or 3.0 percent)."

The authors conclude that "stars become more concentrated over time, moving disproportionately from areas with few peers in their discipline to many -- except for a countercurrent of some foreign-born American stars returning home... Our results on firm entry suggest that this is true across the broad range of high technology industries. The individuals at the very top of their scientific discipline are the ones most likely to fundamentally change how things are done in their science and in its commercial applications. America has profited greatly from providing a seed bed for producing and retaining most of these people for the last sixty years. There is evidence that recent policy changes such as on student visas have endangered that privileged status. It would be a very costly loss."