Shaping the evolution of the PC

Meet the guy responsible for developing IBM's famous "butterfly keyboard."

John Karidis, an IBM distinguished engineer, has what might be the coolest job at IBM's Personal Computing Division. Since 1994, he has been working with a number of different teams inside the division, from engineering to marketing, to turn sometimes wacky ideas into full-blown products.

At any time, the 43-year-old Ph.D. could be working on a handful of projects in various stages. Most recently, he worked on ThinkPad TransNote, a new notebook PC that can capture handwriting using a special notepad.

His most famous project, though, is the ThinkPad 701. With its foldout butterfly keyboard, the 701C is one of the most famous notebook computers and, like an automobile, has attained cult status.

These days, Karidis seems excited about a wireless IBM Micro Drive recently created by a fellow IBM engineer. The engineer fitted an IBM Micro Drive with a Bluetooth radio, allowing the drive to play host to a number of files, such as MP3s. Karidis believes the sky is the limit for this device, suggesting possible uses for the drive as a wireless MP3 player using Bluetooth headphones or as a form of wireless network storage.

In a recent interview, Karidis shared his thoughts on the coming generation of PCs and PDAs. His one overarching prediction: The PC is not going away anytime soon. But it is going to see some radical changes.

Q: Will the butterfly keyboard ever return?
A. The butterfly keyboard was no longer necessary, because people moved to larger displays, especially in this geography. Where the butterfly approach makes sense is where you want the largest keyboard possible in combination with an 8-inch or 10-inch display. We'll wait and see whether the market need develops (again) for that.

What PC trends do you see evolving over the next five years or so?
I think there's a couple of different ways to look at trends. You can look at the underlying technologies that go into products, or you can look at the applications and what people are going to do with them. One trend will be wireless, the shift to 2.5G wireless. The other will be ever-increasing processor power.

If you look at historical trends and you extrapolate them, history says that...you could get five times the performance at the same power in four years or so. So you could imagine a processor being five times faster, or it might not be five times faster--it might be 2.5 times faster, at half the power, in four years. So you could get the same computational power at five times less battery drain or maybe 2.5 times more computational power at half the battery drain.

Another trend is increasing storage capacity. That has been doubling every 12 to 18 months. That's allowing people to do more storing of rich media at home.

In the display area, I think we're going to see several parallel developments. I think we're going to see very high-resolution displays. We're also seeing the organic LED (light-emitting diode) technology being worked on in the labs. That could come to market in several areas. It could be PDAs, it could be notebooks. There are also virtual or projection displays. I think they are going to be important for providing large field-of-view displays for devices that are typically too small to give you direct, full view of a Web page.

You mentioned that storage capacities will increase dramatically. What will all that storage let people do that they can't now do?
The demand for storage is rather insatiable. People thought a 1GB or 2GB drive was too much a couple of years ago. But now, you fill it up with music, and I think people are going to start filling up their drives with video. They might not need all of that storage for Word documents. But when you have hundreds of gigabytes, you might not think twice about putting a copy of all your home movies on your PC.

To be sure, processing power is increasing at incredible rates. But do we really need it?
In the handheld space, we could use additional power for things like speech recognition. You could apply some processing power in a way that would minimize the disadvantage of a small form-factor device. Speech recognition isn't really feasible today on a handheld, but it is when you get into the 400MHz to 500MHz range. I'm told that you can do dictation kind of applications. A handheld of the future should be able to do a good job--as good a job as a PC can today or could last year.

On the desktop, it's a little more difficult to say exactly what people will do with the huge processor power increase. Many people might not need it. That may lead to a "good enough" scenario for a lot of people. However, people may start to do very intensive things at home, like the Hollywood "Toy Story" kind of animation. Home computing might be powerful enough for people to do very sophisticated things with 3D actors and personalities, and maybe some home motion-capture technology.

3D just to entertain the kids?
I've seen some very unusual stuff on the Web recently, where people are doing flash movies. Young people will get interested in 3D modeling and animation as the next level of those areas.

Another (use for processing power) would be software MPEG encoding, where you encode, for example, an analog video...directly to MPEG2. That takes quite a bit of processing power to do well. If you have more processing power you can do a better job...The video quality can be better.

What's IBM doing in terms of increasing security for people's data?
The IBM security chip, a chip on the motherboard of a PC, stores keys, certificates and passwords in a secure way on the motherboard. It's also used for encryption.

You will see (security chips) in IBM PCs.

Is that going to become more important?
Absolutely. People are accessing more and more sensitive data at home and at work. They need to make sure they know whom they are talking to, that the info is secure and that they can trust it. Any level of software-only security is not as good as a combination of hardware and software. Today, if you have online access to a financial account somewhere, it may be protected (only) by a password. There's a convenience to that, but that might not be acceptable for corporate applications or for personal applications.

What kind of effect is wireless going to have on personal computing in the next five years?
Wireless antennas seamlessly integrated--I think that's the way people really want. Because over time, we're going to see the wireless LAN (local area network) be used as kind of standard equipment, I think. (Right now, IBM ThinkPad A Series notebooks have integrated wireless LAN.) So you can expect notebook computers and probably handhelds, more further out, to have integrated wide-area wireless and notebook computers to have integrated local area, 802.11B. You're going to start seeing wireless networks pop up at corporations and in homes, executive lounges and gate areas. There are also hotels being wired.

We're going to see people moving between these pockets of wireless LAN connectivity to slower wide-area connections. People will want both, but people will try to utilize wireless LANs at work and home and airport, and if none of those is available, they'll certainly want the wide area.

Wireless really does change the way you work. When you can go to a meeting and have access to corporate information, people can go get relevant information during the meeting and not be fighting over one wired connection in a meeting room. For wide-area wireless, the key threshold is switching from circuit-switched connection (currently in North America) to packet-data connection (GPRS, or General Packet Radio Service), because then you get the equivalent of instant on and always on. Then you get things that are much more real time and synchronous, like instant messaging and instant e-mail.

What is there to do with instant e-mail?
I think a lot of what people want to do--at least with the small devices--is communications related. They will want to send instant messages. They want to deal with e-mail immediately. There are huge numbers of SMS (Short Messaging Service) messages sent around Japan and Europe, and they will want to do those things as well.

With technologies like speech becoming more prevalent in PDAs, will user interfaces also change in the next few years?
Within a few years, (there will be) things that you can't do in a phone or a PDA today. Once you get a few years out and you have the capability of a 1GHz processor in a PDA, you could use that to do speech recognition of a class that you can only do in a PC today.

In addition to speech, I think we'll see more use of natural input forms, things like ink, handwriting recognition, high-quality text-to-speech...We're starting to see more applications for speech interfaces in the telephony space, where you navigate...using speech as input, and speech as output is a natural approach for some info when you don't have a display. In the future, I think we'll have computers generate voice that is virtually indistinguishable from human voice.

How will speech recognition affect the daily lives of people?
Let's just take the case of a PDA or a small device that you're getting e-mail on. Today people don't use speech recognition to respond to e-mail on a desktop computer. But responding to e-mail on a PDA is much more different. An alternative that some people would probably appreciate is the ability to dictate a response and have that sent back via the device. Using speech when you don't have a keyboard is potentially more compelling.

So then speech could be the killer app for handhelds?
It could be. Speech could be used as a more natural way of querying and finding info on the Internet or a corporate database. You could ask a question in a more conversational way. You could end up getting at information primarily through speech as opposed to pointing and clicking on things on a screen.

Speech recognition will probably never be perfect, as in "Star Trek," but it gets a whole lot more valuable when the alternatives are lacking--when you don't have a good display or a keyboard.

Where is display technology going to take us? IBM recently introduced the T220, a very high-resolution display with a resolution of 3840 by 2400 and 9.2 megapixels.
This is initially a targeted, special application. But there's nothing that would prevent us from moving to higher and higher resolution in the future. I think there are two areas that we need to consider: One is direct view; the other is virtual display. People are conditioned to use fairly low-resolution displays. Palm and Pocket PC (screens) are at 100 dots per inch. You can make a display at 200dpi today, and when you do that, you can make information more readable and display more in the same amount of real estate. So the amount of information displayed on a PDA will increase in the future. Then you can get into these projector displays where a physical device very close to your eye can give you the equivalent of a large display at a normal viewing distance.

Take the IBM Cyber Phone, for example. With the display flipped out, you can see the equivalent of a full PC screen at a normal viewing distance.

With this kind of device, you're both hearing and seeing full screens of information. Through the combination of pointing, plus the speech for navigation, you can imagine interacting with very rich and complex data in a very tiny device.

So no more need for a PC?
We'll still have PCs, because that's the more natural way for people to do large amounts of content creation. We're not going to see the end of the PC. We're going to see more devices available. You'll see the traditional PC workload moving back to the server or out to the handheld devices, but we will still have full-function PCs to cover the high end of the computing market. We'll also see PCs coming out in a variety of different forms.

Any thoughts on the PC-as-home-server scenario, where the device is stored away in a closet somewhere?
I think some people might do that. But they may also have other devices. It may be a home gateway with network-attached storage, or you may have a TiVo-like box that has video. People are going to pick different combinations. It's not correct to ask what is the device that people are going to use instead of a PC. There are going to be a number of devices. The real question is what devices and services will meet the needs of different users.