December 23, 2001

The software industry is stuck in a rut. The PC platform has reached its limits. As I expected, the Palm-thingy fad seems to have crested. Web-enabled cell phones were dubious in concept and worse in execution.

Here is a better idea for the Next Big Thing: Headsets.

I can imagine a world in which everyone spends several hours a day wearing a headset. There will be a software industry devoted to building applications for the headset platform, which consists of earphones, a microphone, and something that I call a "tuner."

Headset-based applications have an obvious role in telephony and listening to music. In addition, intelligent headsets could help make meetings and classrooms more productive. I can even see potential for use in household and social settings.

I envision the "tuner" as something like a cross between a remote control, the buttons on a car radio, and a hand-held game machine. The tuner will be the device with enough software to give the headset flexibility. As with a hand-held game machine, one might envision different cartridges with different sets of functions.

Hands-free Telephony

One obvious application for headsets is to free people's hands when they use the telephone. This is desirable for people who talk on the phone while doing other things, such as computer work, driving, cooking, and so forth.

I can think of no reason to prefer a hand-held phone to a headset. All of the dialing functions that currently are performed using the phone's keypad could instead be performed by the tuner.

Tuners might be programmed to negotiate priorities with one another. For example, I might provide my tuner with a set of weights that gives high priority to a call from my wife or some key business associates. If I am waiting for a call from a doctor's office, I can temporarily assign that a top priority. I might give prior approval to some organizations, such as the staff where my daughters go to school or the police, to assign a high priority to calls to me when that priority is appropriate.

The people who call me also might assign priorities to their calls. The school might use high priority when my daughter is sick and needs to come home, but for other calls they might assign a lower priority.

Tuners would have embedded intelligence to sift through these priorities. One would hope that they could be programmed to filter out solicitors. If so, that in itself could be a killer application.

Another natural application for headsets is for listening to music. If high-speed wireless connections to the Internet become widespread, then listening to music could be as simple as putting on the headset and using the tuner to select the music that you want. You could select either specific tunes or the recommendations of a DJ. Music services might be funded by subscriptions, advertising, or a combination of both.

Without high-speed wireless Internet, you still might be able to use a headset with a tuner as a radio. In addition, if you were near a computer, you might be able to use the tuner to listen to a CD that is played on the computer.

My current occupation is teaching in high school. Based on this limited experience, I am becoming convinced that feedback is at the core of education. This may be the subject of a separate essay.

If I am correct that feedback is the essence of education, then technology in the classroom should be designed to facilitate feedback. Headsets could do this.

Suppose that every student in the class were wearing a headset, and so was I as the teacher. I could give students an in-class assignment, and each student could ask me questions as they come up. If I think a question might interest the rest of the class, I could press a button on my tuner and "broadcast" to the class. Otherwise, I could answer the student individually.

With the headsets, I could give individual attention to students without having to wander around the whole room. I also could manage priorities intelligently. I could have my tuner give a higher priority to students who I know are having trouble or who I know are shy about asking questions.

I might have students that have different abilities in a classroom. With headsets, I could switch back and forth between groups. I could direct one set of remarks to one group, and then make a different set of remarks to a second group.

Another possibility would be to set up several small group discussions. I might divide the class into groups of four students each. By setting the tuners appropriately, each group of four could hear one another but not be disturbed by other groups.

Because the ability to hear would not be rigidly tied to physical proximity, I could move students in and out of groups with the flick of a switch. Also, I myself could monitor and contribute to different groups while sitting at my desk.

What about the risk that students would set their tuners to talk to one another or to listen to music? Well, the fact is that this happens now in school. One approach for dealing with it might be to create a "transparent society" in which every student's tuner settings are visible to the teacher, so that the teacher can set and enforce consequences for misbehavior.

One software challenge in the educational setting is to build in the functions that would be useful to a teacher and make them easy to use. I do not want to have to interrupt my train of thought to go through a long and complex procedure to set up group discussions.

Large conferences, such as the annual meeting of the American Economic Association, could be made much more efficient if everyone were wearing headsets. At such conferences, people have objectives of meeting with specific associates, but the crowded conditions can make such meetings difficult. With a headset, you could talk to someone that you bump into, then switch to listen to a conversation taking place across the room among some important colleagues.

With a headset at a conference, you could attend a session but flip your tuner to other competing sessions. Using this approach, you could try to pick the best session or even try to keep up with more than one session at a time.

Business meetings also could be more efficient with headsets. Often, an issue comes up that affects only two or three people at the meeting, but it would be efficient to settle the issue right away. With headsets, you could allow them to hold a conversation about the issue while the meeting proceeds on to another topic.

Sometimes, you will go to a meeting and see someone with whom you need to speak about a topic that is not related to the meeting. With headsets, you could do this without disturbing anyone.

Sometimes at a business meeting, you want to take a phone call but you do not want to leave the meeting. If everyone were wearing headsets, you could do this.

One software challenge for business meetings is to make it easy for people to negotiate appropriate volume and target adjustments. If you and I want to have a side conversation, then we should turn up the volume for one another on our tuners. It gets more complicated if one of us wants to have a side conversation but the other one does not.

If people wear their headsets and are connected to the Internet all of the time, then something like Instant Messaging capability will always be available. This would have many applications.

For example, I might be able to help my daughter with her schoolwork during school. She can tune me in and ask a question about math.

Suppose a friend of mine goes to a party that I cannot attend, and people at the party use headsets to converse with one another. If my friend gives me eavesdropping privileges, our tuners can be set so that I am able to listen to what he hears and speak with the people to whom he speaks.

The time may be ripe for a new platform for the software industry. A platform needs to be flexible, with many possible applications that can be developed for simple hardware.

If the flexible tuner could be practical, then the headset could become a universal platform. Almost everyone can listen through earphones and talk into a microphone. If the tuner also can be made easy to use, then I believe you have a compelling device.

I have been quite un-specific about how the tuner would work. Some features would be hard-wired. Others would be programmed by the user. The programming issue is worrisome: you do not want to create something as painful to use as a programmable VCR or a Honeywell thermostat. I would hate to think that one would have to create an operating system as complicated as those used for personal computers.

The tuner is one of the biggest question marks in my vision for headsets. (Other question marks would include spectrum availability.) I am counting on a lot of different physical capabilities (adjusting volume, rapid connections, conference calling, etc.). In addition, I am counting on software developers being able to endow tuners with many sophisticated functions, such as negotiating priorities for connections.

Many futurists have said that we need to put computers into the background, where we can take them for granted. The headset-with-intelligence that I have sketched out here represents a model of this type of device.

Overall, I would argue that we have reached a plateau with what we can do with a device that is designed primarily for keyboard input and screen output. It is time to discard the legacy of the screen-with-keyboard and start over. Strip the hardware down to a headset and a tuner, and begin from there.