Starting up – that would be the fourth

Well, well, well.

I guess it’s some kind of a medical condition, but I’m leaving a great job at Síminn (Iceland Telecom) to start up a new company once again. This will be my fourth start-up, and I’m as excited as ever.

It will be a relatively slow migration as I’m finishing off a few projects at Síminn for the next couple of months, while at the same time setting up the new company, assembling a core team and refining the strategy for an idea that has been with me for some 18 months now, gradually getting more and more focused, until I became so obsessed that I simply had to go for it.

Some would claim this is a horrible time to start a company, with a gloomy economical outlook and a lot of turmoil in the world of business and IT.

I – however – see this as an opportunity. Due to these very conditions, highly qualified people are looking for exciting new opportunities. This is especially true here in Iceland, where the financial sector has drained the market of IT talent for the last 3-4 years, and those adventurous people that really would rather be working on something new and innovative have been tempted by the lucrative salaries and “never-ending party” of our booming banks. Now the banks (and others) are scaling down and being a lot more careful, so these people – many of them not necessarily in danger of loosing their job – might very well want to flex their start-up muscles again. Actually I know for a fact that this is the case.

Secondly, booms and busts in economy seem to come at an interval of 6-8 years. It takes at least 3-5 years to build a great company, so those starting now are likely to catch the next upswing, without having to run too fast for their own good – as long as they can build sufficient income or find the venture capital to fund their operations in the meantime.

The concept I’m working on – and my situation in general – is such that I believe I can pull this off.

I’m not willing to share publicly – just yet – what the concept is, but I’ll surely blog regular updates as things progress.

Long nights and fun times – here I come 🙂

The Government API

A couple of weeks ago I attended a conference where there was a lot of talk about Business Process Automation (BPA) and Service Oriented Architecture (SOA).

Disclaimer: Yes, I do lead a very exciting life, even if attending such a conference may indicate otherwise.

You’re probably familiar with the story: Companies are increasingly wrapping their legacy systems, back office processing and other components of their day-to-day operations with web service interfaces in order to integrate with a newer breed of software that usually comes with some flavor of WS interfaces out of the box. Then, using Business Process Management they link all of these together, practically describing their businesses as elegantly written computer programs – the cafeteria included. The holy grail is then to automate as much of this as utterly possible – all leading to an increase in productivity, lower costs, more revenues and happy investors.

The concept is great and several companies have excelled at this – while most are probably struggling. This is after all a huge task and requires a change of mind no less than a change of technologies. But that’s besides the point here.

At the conference the following truth dawned upon me: As companies automate more of their operations, not only within, but also in interacting with other businesses (electronic bills, automated orders, etc), a critical part of their ecosystem is being left out: The Government.

So here’s an idea: Let’s wrap a government in an easy to use API – a Web Service interface that will fit nicely into the world of Business Process Automation.

All the mundane, numbing and often error prone tasks of running a company could be automated and integrated with the company’s own infrastructure. Filing of tax reports, filing salaries statements, filing summer vacation payments, filing changes in corporate ownership, issuing new stock, etc., etc. and doing it all in the right way at the right time without anything falling through the cracks – all taken care of automatically by the rules set up in the Business Process Automation system. Sweet!

In this lies a huge opportunity for some small, technically advanced nation that is well integrated in the international legal and business environment. It could really set itself apart in the world by catering to companies that are far along the path of BPA – and new companies that are set up in such a way from day one. These are probably most often sophisticated companies in the IT, financial or other high value industries. If done correctly, this could be no less valuable than tax incentives, already offered in various havens around the globe. Obviously it would need to be competitive there, but not ridiculously so – which in turn would keep the country on friendly terms in international politics, where tax havens are often frowned upon.

This could even lead the way to the world’s first fully automated company – a concept that is probably material for another entry.

Once in place – there is a plenty of derived services that service companies could offer using the same methodology. Companies need banks, financial and auditing services, legal services, IT-infrastructure (email, web hosting, etc) and a whole lot of other stuff a lot of which could be offered as web services if approached correctly. This would make the deal even more attractive to the companies and bring even more value to the nation in question as it would fuel local busines.

It so happens that Iceland – my very home country – seems in many ways well suited to take this step. Good IT infrastructure, many government tasks already electronic at least to some extent, well integrated in the international legal environment, good financial infrastructure and most importantly – a small and highly interconnected society were everybody knows everyone else, essential in order to pull the API-fication off relatively fast.

In any case, I’m sure this will happen somewhere sooner than later and I for one would certainly look into founding my next company where such infrastructure is in place.

Can’t wait to invoke the FormCompany() method, let’s just hope that the FileForBancruptcy() call will not have to be used 🙂

The Polar Express … for data!

I was at a nerd party last Friday and as it goes, ideas became wilder as the beer supply diminished. One of the wilder ones stuck with me: Jarl brought up the possibility of a submarine cable across the Arctic region, properly connecting East-Asia and Europe.

This is certainly a wild idea, but as a matter of fact it may have a great potential. The current routes between Europe and – say – China or Japan are flat out lousy. Ping times to Japan range between 300 and 400ms and to China close to half a second (brief and unscientific tests gave me average ping times of 320 and 420ms respectively). And for a good reason – the traffic has two equally lousy routes to choose from:

  1. Across the Mediterranean, through the Suez Canal and via the Red Sea into the Indian Ocean where it zig-zags its way to the region, probably through a major peering point, located in Singapore.
  2. Across the Atlantic to the Atlantic Coast of the US, across entire North-America and then across the Pacific! Incredibly this is the route I usually saw when tracerouting Japanese and Chinese servers.

Either of these two routes is at least 16 thousand kilometers and probably closer to 20 thousand, with a lot of peering points along the way. (See Wikipedia map of submarine cables for reference)

Iceland is finally becoming pretty well connected as we will soon have at least 3 major submarine cables, each with bandwidth in the 1 Tbit/s range, and directly linked with the main peering points in Europe, namely London and Amsterdam (the latter one is unconfirmed).

From here, the distance through the Arctic region to Japan is less than 10 thousand kilometers as drawn below: From Iceland, north of Greenland and then straight across the North Pole to enter the Pacific Ocean via the Bering Strait and on to Japan. It is probably possible to lay a cable this route using submarines. It has been done, but I bet maintenance is a bit tricky. Somehow I just don’t see a submarine taking in a cable to splice damaged fiber.

A more likely route would therefore be the fabled North-West passage that presumably is now open for cable ships just as well as other vessels, allowing for relatively normal maintenance on a cable lying there. This variation is a little further in total (counting from mainland Europe). Both routes should nevertheless be able to bring ping times from Europe to East-Asia down to the 100ms range (laws of physics, like the speed of light start to kick in at these distances).

This is not perfect, but importantly this is below the threshold acceptable for VoIP traffic, meaning that bandwidth on this route should be hot property. So maybe, Iceland will one day become a major peering point for IP traffic to Asia…

Certainly a big, crazy, wild idea – but worth further investigation.


How far is it?

I’ve always been fascinated with those center-of-the-world type sign posts that tell you how far it is from where you are to various distant places.

I just made a fun little tool so that anybody can do the calculations needed to create one.

It’s fairly straight forward to use. The input fields accept most formats used for geo-coordinates (including those given by your GPS). Alternatively you can pick from a list of major cities around the world.

If your city is not listed and you don’t know your location, you might be able to find your coordinates by looking up your town in Wikipedia.

Any feedback is welcomed.


How far is it?

Visited Countries – Revisited

When I read Bill Bryson’s fantastic book “A Short History of Nearly Everything“, one of the things that stood out, was a reminder that the world is still a really big place.

Even though we feel that we can – with a credit card, and a solid visa – get pretty much anywhere in the world within 72 hours, the world is still HUGE on the ground level. There are large areas where – to our best knowledge – no man has ever set foot; it still takes days or even weeks of trekking to get to much of Earth’s surface; and humans are still to explore even 1% of world’s oceans.

This made me think of Douwe‘s Visited Countries project. You surely know this one – everybody and their grandma’s have been posting these maps on their blogs in the last 3 years, coloring the countries they’ve been to. But that’s not really fair – is it? I’ve been to Hong Kong, so I color China. I’ve been to a dozen cities or so in the US, that gives me half a continent. Looking at the colored map, it really feels like I’ve pretty much covered it – only some minor continents to go :).

Well, think again!

Above is a typical Visited Countries map of the countries I’ve visited. But look closer to see where I’ve REALLY been. The yellow dots are (roughly) the dots I’ve set foot in. I’m even pretty generous, I colored dots that I’ve only zoomed through in a train or on the motorway, (but skipped those I’ve seen out the plane window). And each dot is actually huge. Roughly estimated, a single dot on this map is about 3,000 square kilometers on average!

It now looks like I’ve still got some planes to catch!

P.S. I wonder if anyone will come up with a tool to make this type of map, as easily as Douwe’s original?

Massively Multiplayer game as a work place

More than two years ago, I wrote a post titled When People are Cheaper than Technology. The basic premise there was that the cheapest and best solution to many problems we are trying to solve by building software systems could be to make people part of those systems – basicly what Amazon is now trying to do with the Mechanical Turk.

I also wrote another one on what I called Games With a Cause – trying to make simple, fun games that whose use would result in data with some research or even monetary value. I even made one, with limited success 🙂

But earlier this week, in yet another nutty conversation with my friend Haukur, we started discussing the incredible amount of time that people all over the world are spending, playing Massively Multiplayer games (MMORPGs). What if just a tiny portion of this playing time could be turned into some “real work”?

The numbers are staggering:

  • A MMORPG game player will spend on average between 12 and 21 hours per week playing their favorite game. (source)
  • Ultima Online players spent more than 160 million man hours playing the game per year around the year 2000
  • World of Warcraft recently reached 5 million subscribers. If the 12 hour number from above holds true that’s more than 3.1 billion man-hours in a year.

A typical man-year (must they be called person-hours to be politicaly correct?) has about 1600-2000 man-hours in western coutries, so Blizzard (the maker of World of Warcraft) is controlling what equals a work force of 1.5 million people! Even if you could only turn a miniscule amount of this time into monetizable work…

And what could that work be? Most of these games have advanced trading systems. Maybe some financial simulations could be run in the game world, if I remember correctly the Nobel prize was awarded a couple of years ago for work in experimental economics. The MMORPGs could be ideal testbeds. Or maybe “stealing” a few computing cycles in the vast grid of connected computers. Or maybe somehow building tasks like image recognition (a la Amazon’s Mechanical Turk) into the gameplay. Or…

My hunch is that simulations of how markets and societies respond to different settings, rules and events would be a good bet.

Imagine – if you could make Blizzard a 0.30$ per hour, that would amount to about 15$ a month, which is incidentally the same as the month subscription – and you could play for free!

Using AJAX to track user behavior

Here’s a thought. With the rise of AJAX applications this is bound to happen, and may very well have been implemented somewhere, even though a quick search didn’t reveal a lot. So here goes:

By adding a few clever Javascript events to a web page, it is possible to track user behavior on web pages far more than with the typical methods.

Similar methods have for a long time been used on some sites (including my own to track clicks on links. A very simple script could also log how far down a page a user scrolls, over which elements he hovers the mouse and even infer how long he spends looking at different parts of the page and probably several other things usability experts and web publishers would kill (or at least seriously injure) to learn.

Even though it’s nowhere near as precise as the famous eyetracking heatmaps (here’s the one for Google’s result pages), it provides a cheap method for web publishers to help better position ads and user interface elements, experiment with different layouts, etc.

Does anybody know of implementations like this? Does it raise some new privacy concerns?

P.S.: I just saw that Jep Castelein of Backbase posted some related thoughts this summer.

Massively Multiplayer Robot Game (virtual reality without the “virtual”)

Here’s an idea that has been cooking in my head for years: Making the most real computer game still to be seen anywhere. How? By actually making it happen in reality.

I’m not thinking about games that blur the boundaries to real life and make yourself a character in the plot (Alternate Reality Games, they’re called) like EA’s Majestic. What I’m talking about is making a fairly typical computer game that actually uses reality as a rendering engine. What better way to get realistic effects? And what better way to avoid predefined rules that limit what players can do to the imagination of the game designers?

I will use an example to explain the concept. Let’s call my imaginary game “The Robos 2004” or something. In “The Robos”, you control a robot in a similar way as you control a character in a first person shooter or a massively multiplayer game. Your “Robo” lives on an island together with a ton of other robots, all of whom are owned and controlled by other players in the game.
Continue reading

Introducing: I like! – Collaborative Approach to the Web

I’ve been doing some coding, hence not much written on Wetware in the meantime.

The hack is called “I like!” and is a very simple, yet powerful service that allows you to mark web pages you like, by a single click of a button. In return you get several things:

  • First of all it recommends other pages you will probably like as well (matching your “liked pages” with other profiles).
  • Secondly it allows you to syndicate any list from the service and put it on your blog or other web sites. It is based on a JavaScript, so it works and is up-to-date even on static HTML pages. This can be used in several ways, e.g.:
    • To put a “Most liked pages” list on your web page
    • To put an “I like! recommended:” list on your web page (see example)
    • To put a “Pages that grabbed my attention” list on your web page (see examplein Icelandic)
  • Lists are also available as RSS feeds for those of you that want to do more advanced stuff or just want to put some of the “I like!” lists in your RSS reader.

I actually wrote about this idea under the name of “Interesting Content Indicator” just over a month ago, but my lazy programming didn’t seem to be working, so I had to do it myself 😉

“I like!” is anonymous and doesn’t require any installation. The aforementioned “I like!” button is a link you put in the Links bar or the Favorites in your browser for easy access (or corresponding places in non-IE browsers). It seems to work with most types of browsers and on the most common OSs.

Keep in mind though that you might have to convince your pop-up blocker that the link is OK.

This is still just a BETA, so bear with me for the first few days if there are some minor issues (and please let me know about any bugs or quirks you might find). I have a long list of neat features that I’m implementing at the moment, so it will only get better. All suggestions welcomed.

Like it? I like! can be found here

– – –

Note: When using Internet Explorer to mark a page with frames, you’ll have to click somewhere on the page in order for the “I like!” button to work.

Games With A Cause

Some time ago I wrote about various attempts to gather common sense, the lack of which is believed to be one of the main hurdles to creating successful AI systems capable of human-like interaction.

A few weeks later I wrote about people as parts of computer systems to make them cheaper or more intelligent. Some of the best examples are when you can tap into the normal usage of Internet users to create something valuable as in the evolving banners example.

I began wondering if there was a way to make an incentive for Internet users to build a common sense database and actually giving them something in return. My suggestion: “Games With A Cause” – you play, we get a bit of common sense.
Continue reading