Cost of personal genome sequencing plummeting: currently $48k (down from $2mil) and expected to be $5k soon http://bit.ly/CGP3r
Next Generation Mobile Device / “Killer App” (May 08)
by Eric Sheline - February 5th, 2009. Filed under: feature.
Editor’s Note: I wrote this paper for Professor Mark Bryant’s Communications Technology class while I was a student at the University of Texas at Austin back in May 2008, so read with that in mind. I am posting it now without updating it for the time difference, saving that for a future post.
A new generation of mobile devices is just around the corner. Though Moore’s law is continuing [1] for the time being, right now the most exciting developments are arguably in the miniaturization of processing power. This miniaturization means higher power, higher efficiency devices able to run applications unheard of on today’s mobile phones.
For example, the Apple iPhone’s processor already runs at a speedy 620 MHz. By using Moore’s law, since the iPhone was released in the summer of 2007, we can roughly predict that before the summer of 2009 there will be a mobile device released with a clock speed around 1.2 GHz (1200 MHz). Such a speed is as fast, at least in terms of clock speed, as some subnotebook computers today, and easily quick enough to run most traditional “desktop-level” applications.
Beyond the device itself, any modern computing platform must also have a solid network connection. While Moore’s law will inevitably solve any shortfalls in processing capacity, a high-speed mobile network is needed to make any future device viable. Again using the Apple iPhone as an example, the rollout of the “2.5G” EDGE network was largely viewed as inadequate for full web functionality, even by CEO Steve Jobs himself [5].
Unfortunately, the network problem is harder to solve than by simply waving the Moore’s Law Magic Wand. While Europe and Japan have largely complete 3G networks capable of (mostly) full mobile Internet usage, North America still lags behind, with major carriers like T-Mobile only starting 3G in 2006 [6].
However, even considering network difficulties, the primary stumbling block with such small devices is the interface. As shown by the success of the iPhone, traditional (pushbutton-driven) mobile interfaces can’t really compete anymore. However, even the mighty iPhone’s touch-based interface may not be enough to fully utilize a device’s power when they start to push 2 GHz within three years.
New interfaces are already being tested and released by major consumer technology companies. The most well known of course are Nintendo’s Wii gaming console and Apple’s iPhone mobile communicator, but Microsoft is also definitely testing out next-gen interfaces in a big way. Their Surface [2] is a prototype computer used for everything a standard computer can do – browse the internet, organize digital media, collaborate electronically, etc. – except it has no mouse or keyboard. The device is based around a large 30-inch tabletop touch screen interface that can even intelligently detect when a device such as a camera is placed on it, and then wirelessly download photos.
Voice
However innovative and headline-friendly the Nintendo Wii, Apple iPhone, and Microsoft Surface are, I think that one interface not getting enough publicity is voice. This may be because voice recognition and interface has been a background fixture of many applications even since the 1990’s, but never gained popularity because of the lack of polish and a true “killer app” (application) for the interface.
A voice interface is actually one of Microsoft’s other big, public projects, starting now in Ford cars as the Sync platform. An automotive platform makes sense for the debut of a new voice interface, as a car’s environment is uniquely suited to it. First of all, a voice interface is needed for safety, as the user’s hands and eyes are occupied by driving. A car can also provide the two things currently lacking in mobile devices, at least for voice recognition, processing power and battery life.
The future voice interface will be implemented on faster, next generation mobile devices as well as cars. Of course this is already in use today in many mobile phones [3] for simple commands like “call John Smith” and even more advanced ones like searching Google, but I hypothesize that the voice interface can and will be used for much more aggressive applications.
“LifeLog” (term credit DARPA [7])
All that’s needed for this next “killer app” is a good speech recognition program, which was already available back in 2006 [3], and a faster mobile CPU to run it, which is coming within a year or two. When those come together, a digital, searchable “life recorder” becomes a distinct possibility.
Basic life “recorders” have been done before and people even use them now, but these are niche devices that record all audio in analog (non-searchable) for the day, or take photos constantly [8] at certain intervals. I propose a device that does not record waveform or any type of analog format, but instantly digitally translates all voice input it hears into text. When that is done, it becomes searchable through a Google-like interface.
Through this searchable interface, the device could give one a perfect memory of all spoken word said and heard, forever. This obviously would aid not only medical patients such as those with Alzheimer’s, but also, people of all backgrounds and jobs. Such universal appeal and utility is what make this the true next-generation “killer app”.
Of course with such a recorder comes massive privacy concerns, but those are most easily addressed by looking at the history of recent technologies. What has happened in almost every case, such as online shopping, computerized databases, and online banking, is that the convenience and utility has far outweighed the potential privacy and security issues. In the same way, the massive utility of the life recorder would outweigh privacy concerns. Adopting the technology would be slow at first, but eventually follow the diffusion of technology curve, used by society at large as the advantages of the device become known.
Such a “Life Log”-like application was actually thought of back in 1945 by Vannevar Bush when he proposed a “Memex” and is now being explored by Microsoft Research with their “MyLifeBits” project [4]. The problem with MyLifeBits is that the concept is overaggressive and reaches too far. All that is needed for the first step towards this life recorder is voice recognition installed on a mobile device. Microsoft researchers are correct about the future, there will be high-level integration with myriad other technologies including RFID, GPS, and more. However, for now, the first step is within reach using today’s software and technology.
Beyond “Lifelog”
Though this is not the main subject of this paper, further into the future this same voice recognition technology coupled with Moore’s Law-driven processor advancements could lead to mobile Artificial Intelligence (A.I.). In a way, a mobile A.I. “helper” would be the ultimate interface advancement until the development of a mind- or thought-based interface.
A handheld A.I. could probably start as a highly evolved voice interface. As voice commands evolve from the simple mentioned earlier to the more advanced, such as “Book me tickets this Saturday for Iron Man at the Alamo Drafthouse”, the device will have to be able to intelligently recognize human speech and not just computer-approved command prompts.
Conclusion
The next big upset in mobile communication isn’t the Apple iPhone referenced so frequently in this paper. While the iPhone is an extremely well designed device combining two of the most popular consumer electronics products in the last decade (the mobile phone and the iPod), it is not a radical revolution.
The revolution will come when the next true “killer app” is released, not another “killer device” like the iPhone. In my opinion, a voice-recognition driven, searchable “Life Recorder” is one of the most likely possibilities in that regard. It obviously has the ability to change the way people live, and most importantly, can be achieved with current or very-near-future technologies.
Sources
1. BBC – “Law That Has Driven Digital Life”
http://news.bbc.co.uk/2/hi/science/nature/4449711.stm
2. Microsoft – Surface Official Site
http://www.microsoft.com/surface/index.html
3. CNET – Voice Signal Demo
http://reviews.cnet.com/4531-10921_7-6637293.html
4. Microsoft Research – MyLifeBits
http://research.microsoft.com/barc/mediapresence/MyLifeBits.aspx
5. Gizmodo – Steve Jobs on EDGE
http://gizmodo.com/gadgets/apple/jobs-tells-wsj-that-att-edge-network-isnt-fast-enough-273558.php
6. GigaOM – T-Mobile 3G Rollout
http://gigaom.com/2006/10/06/t-mobile/
7. Wired – Darpa Kills “LifeLog”
http://www.wired.com/politics/security/news/2004/02/62158
8. Microsoft Research – SenseCam
http://research.microsoft.com/sendev/projects/sensecam/memory.htm