At this year’s E3 Microsoft announced Natal (check out the link but prepare for bad acting). Natal offers a new controller-less gaming interface - presumably an attempt to 1-up the wii, and take over your living room. The technology appears to be a depth sensing camera, something which my colleague Derek wrote about in “You may never need a controller again“. It’s a move that suggests that Microsoft got a pretty good deal when it ended up purchasing 3DV systems (a maker of depth sensing cameras) for an alleged $35 million. Here’s their video from E3:

Other aspects of Natal include identity recognition, interaction with virtual characters, and some nice augmented reality tricks like fishing (reminiscent of Reactrix or GestureTek’s virtual fishponds) and being able to pass a piece of paper into a virtual world (see what appears to be a carefully scripted video below):

Continue reading…

…my first confession: I really like those URL shortening services. They’re helpful for taking an otherwise gigantic URL and compressing  it into a tiny one  to tidy up an email link, but even more valuable for squeezing a hyperlink into the 140 characters permitted by twitter.

Yet, what bothers me about is that these sites completely undo one of the best features of the web, a publicly available address owned by the content provider. The problem is that what used to be an open public resource now falls into private hands. If  any one of — tinyurl.com, is.gd, short.ie, ow.ly, tri.im, or kl.am (or numerous others) fail — then they will bring down tens of thousands of URLs down with them.

It’s a no win situation because either these services have two choices:

 1) Operate these services as a public good, funding them out of their own kindness

 2) Figure out a business model that makes sense

In the first case, we have to ask whether that generosity might someday end, and even if it doesn’t - whether the lack of funding might one day cause the link service to fail anyway. In the second case, once we have a for-profit business model, suddenly we have for-profit proprietary extensions to the web itself. For example, some sites like ow.ly have an option to put the underlying page in a frame where you never get to see the original URL in the address bar at all. Now, that’s handy for retweeting, rating the site (and yes, the full address is still available via clicking if you want it), or sharing with others - but turning a public address into a private index that takes over the web’s basic interface seems to run counter to the Internet’s open beginnings.  It reminds me a bit of stumbleupon.com for example, useful, but ultimately shielding us from the open structure of the web by hiding its addresses.

Maybe I’m overreacting, and this is a little like people clinging to the old command line interfaces over a GUI, or perhaps insisting that the headers of email remain visible. In both those cases, most of us manage to get along just fine without a connection to the stuff that’s “under the hood”.  But does moving the structure of the web itself “under the hood” start to cross a new line? Are we in danger of losing the Internet’s open beginnings?

Part of what makes it such a wonderful resource is the fact that content owners (not distribution channels) allow a direct line of communication with end users. Search engines certainly preserve that connection. But URL shortening seems to be more of a slippery slope - one that could ultimately disintermediate web pages themselves.  Thoughts?

Thanks to a recommendation by my colleague Naumi Haque, I recently finished reading, “World War Z: An Oral history of the Zombie War.” Great book.

What I liked most about the book was the clever variety of well thought out characters and scenarios. How would people, companies, and institutions fare if plunged into a full scale zombie war - an unusual question, but one that the book tackles with great relish.  Interestingly, one of the scenarios described in the book gave me a new insight into wikinomics (see other wikinomics zombie mentions here, here, and here).  [if you plan to read the book, stop here, minor spoiler ahead] 

There’s one event described early in the zombie wars which pits America’s military might and technology against an invading army of mindless zombies. The book does a great job of describing a modern battlefield technology which allows combatants in the field to communicate and exchange information with one another (including live video transmitted by others). The description got me thinking… wow, great stuff, wikinomics, distributed collaboration, and collective intelligence in action. Cool.  Sure enough, I was duped. It turned out (in this fictional case at least) that this advanced battlefield technology contributed to the mighty army’s defeat. Instead of spreading intelligence and coordinating responses across the troops, it helped spread panic. Frightening visuals of defeat and voiced fears were instantly shared with everyone… and the result was a frightened mob, not orderly action from soldiers. In the end, the higher ups had to turn off the video feeds and try to reassure people that all was well (and by that time it was not).   

The lesson for me is that succeeding with wikinomics takes much more than just simply connecting people. A crowd can be “smart”, or it can be a mindless mob too. While James Surowiecki wrote about “The Wisdom of Crowds,” in 2004, I highly recommend Charles Mackay’s counterpoint written in 1841 called “Extraordinary Popular Delusions and the Madness of Crowds.” Crowds, even when highly connected via digital technology, do not automatically become smarter. In fact, as Mackay observed, sometimes they become mad and delusional too.  The difference is less about the process of magically connecting people, but more about the architecture for connecting them. HOW people are connected, seems to make the difference between  wisdom and madness… call it appropriate collaborative architecture. At the moment, finding the right collaborative architecture appears as much art as it is science, but we have lots of examples (and smart community builders) to point the way. Perhaps readers of this blog can share their own insights into these principles.

Onlive.com just announced an interesting twist on cloud computing applied to gaming. See NYTimes article for a snapshot or check their press conference for the full details.

In a nutshell: power-hungry video games are run on a server far away, and then high-definition video of gameplay is piped over the internet back to the person playing the game. What that means is that games like Crysis (which are virtually unplayable on anything but a high-power PC) can be enjoyed on a low power laptop… or even piped directly to a TV with the help of a small box that can do the video decompression.

…think of it as inserting the Internet between a computer’s video output plug and the computer screen. Here’s their trailer/demo video:

The enemy of such a system is “lag” of course - any small delay induced in the video compression/decompression or via Internet transmission will kill first person shooter experiences. These are people that like 60 frames per second after all. But OnLive says it can combat that via the fastest video compression I’ve ever heard of (single digit milliseconds). To be honest, that’ sounds too low to believable - I don’t think Cisco’s telepresence system approaches that.  But if it really is anywhere close to that good, they’ve accomplished something spectacular.

It’s really interesting how most Internet-delivered multiplayer games start from the premise that you have lots of smart hardware and a narrow pipe. A lot of computing power goes into contructing and rendering physics and virtuals worlds that come out of those pipes. Onlive’s disruptive approach assumes the opposite. It assumes “dumb” hardware at the other end, but a fat pipe. That way it transmits a ton of “dumb” video pixels that can be displayed anywhere.  That’s not entirely the case of course, since there’s gotta be a ton of “smart” work going in in that video compression/decompression hardware, but it does give pause on how much smarts we really need at the edge of the cloud. It’s great to see OnLive take a stab at delivering games on demand - it creates interesting possibilities for controling piracy, try-before-buy, or even delivering a new breed of high-res games that even the highest end machines are incapable of.

But this is the tip of the iceberg when it comes to the larger possibilities for delivering cloud computing.  Games are just the beginning, the same model of delivery can be applied to anything that hits a computer screen. Regardless of whether OnLive itself succeeds, this is definitely an important turning point in our computing model.

As part of a stimulus package, every city in Los Angeles county was slated to receive $500,000 from the Metropolitan Transportation Authority (MTA). The problem is, many of these (especially smaller cities) didn’t have any “shovel ready” transport projects for this unexpected injection of funds. 

So what did they do? They took matters into their own hands and auctioned the funds to the highest bidders -  other municipalities who had projects ready to go paid them cash for the rights to the MTA funds (in one case just $.61 on the dollar). The city selling the MTA funds would then be able to take the cash and use it for anything it liked.

Three interesting things about the situation stand out:

• First, the swaps/marketplace itself is quite innovative/creative. An interesting case in optimizing results in a grass roots way. Grade: A-
• Second, the low prices paid for the funds suggest that, in some cases, stimulus dollars are being weilded very inefficiently  (A $.61 price suggests a 39% inefficiency) Grade: C
• Third, the MTA appears to have put the brakes on these swaps. Maybe because of the publicity they generated? Who knows. But cities must now build (invent?) a case for their use, or risk losing them altogether. This seems to create the wrong message and set of incentives. Grade: D 

Dear Music Industry:

… please please let me buy something from you!   I hit the “buy” button, logged in, agreed to the terms, and: 

Everyone hates juggling usernames and passwords. So all the great activity around OpenID, Facebook Connect, and more recently OpenID and facebook - all which suggest that mainstream use of open web authentication schemes are reaching critical mass.

I like the idea, a lot. However, I think it’s a bit early to bet on one horse - so why not add more to the mix. I like twitter’s generally open approach, so why can’t they play in this space.

So, here’s a proposal on how anyone can use twitter as an open authentication scheme to log into their site:

The first step is a login page (screenshot below) which gives you a unique one-time authentication key that is used to identify your session. In this example the one-time code is “82kjx_OneTimeAccessCode_IeZh9els” and it is designed to be tweeted (probably best to DM) to the web site owner’s account (”SiteTwitterName”in this case). By DM’ing the one-time code to the site owner you link your session to a specific twitter account, and by DM’ing it, you provide proof that you own that twitter account. To make this easier to tweet, you could add a “copy to clipboard link”, or “tweet to login” button/link which would automatically prepopulate the tweet in a browser window (see next screenshot).

Below is a sample of what the page might look like after you click the “tweet this to login” button.  You can imagine the button creating a popup window like this (if the browser allows popup windows). On twitter, it’s easy to prepopulate a page with a ready-to-tweet message like this. Just open a page with the URL:

http://twitter.com/home/?status=d%20SiteTwitterName%2082kjx_OneTimeAccessCode_leZh9els

And that link should give you a page similar to the one below:

Then, once you send the DM through twitter. The website can use the twitter API to read the DM and then make a connection between your twitter ID and the unique session key in order to authenticate you. At that point, your original login page can be refreshed, logging you in automatically. Voila, you are logged into a website using your twitterID as the account name:

A login scheme like this would work with twitter, but equally well with any messaging or IM service that’s sufficiently quick and also has an API. One of the best things about it is that it doesn’t require any endorsement of the service provider in order to use it for authentication either. You can even imagine doing this via a mobile phone too (either through cameraphone image, QR code (discussed here and here), IVR, OCR, or even a “sound” produced by the website that you could hold your phone up to).

Any suggestions about holes or problems with this scheme that I may be missing? Or ideas for improvements?

If anyone would like to implement the first working demo of this scheme it would be a great contribution to the public good.  I’d love to credit you with it here. Happy to share any demo code for it too if you wish.

…please contact me via twitter @crasheral if you would like to help kickstart this.

If there was any doubt before, yesterday’s plane crash in the Hudson river provided ample proof of how useful Twitter can be for emerging news. Here’s a truly amazing photo that @jkrums snapped and linked to via twitter while his ferry was en route to help:

Three minutes earlier than that, @manolantern sent out what appears to be the first tweet on the plane crash:

I just watched a plane crash into the hudson rive in manhattan

I love how our earliest news is most certainly going to contain typos from now on. Same case with Mike Wilson (@2drinksbehind) who in December twittered about being in a plane crash. Understandably he put it quite strongly:

Holy fucking shit I wasbjust in a plane crash!

I enjoyed his followup tweet just as much though:

You have your wits scared out of you, drag your butt out of a flaming ball of wreckage and you can’t even get a vodka-tonic. Boo

Twitter truly connects us to the experiences of others - and terrific first hand news is just part of the fun. While I hope there are no plane crashes in my future, you can follow my own twitter updates at @crasheral (gee, hope I’m not tempting fate with that twittername).

Speaking as someome who has pointed cameras at bird houses before, I suppose it’s about time they got even with us. Here’s a cool shot of security cameras dressed up as bird houses:

(From Pasta & Vinegar via flickr )

Here’s a photo of the Tuttuki bako:

The Tuttuki bako (or “poking box”) is so strange that when I first heard of it I just had to order one and see it for myself. Technically, you might describe this as a finger-based augmented reality system. But to put it bluntly you stick your finger in the hole in the side of the box and prod a variety of virtual characters - sort of how you might play with ants or something. Here’s me dueling with a stick person (watch out it knows karate):

Continue reading…