A billion? You're telling me a billion twenty five to forty years from now? A billion! you're pulling my leg, right?
Rich smiled. I did the same and added
25, maybe 50 years. If I had to bet closer to 25
Two hours with the patent attorney. He didn't buy our story. Even if we were right, he said, the time frame was wrong for the company. The focus had to be much shorter. It was 1989, not the pre-divestiture days when we were actively inventing the future. Such a change from the good old days..
We went for a long walk. It was a hot day with loads of blueberries along the path by the power lines. Our early morning realization that one watt short range cellphones could - would - explode in use had become a driver for more interesting conjectures. It was one of the more electrifying points I can remember. Blueberries still stir the memories.
We were questioning the wisdom of AT&T leaving the cellphone market. The decision was part of divestiture, but we had learned internal projections showed mobile telephony would always be a tiny market. Cellphones were an easy chip for the company to abandon in the negotiations. Perhaps that was the right decision for the decade, but Moore's law was relentless and new doors were showing a bit of light around the edges. We started doing what we do - back of the envelope calculations. At first you come up with a dozen reasons why it has to be expensive, but then you decide to go to basics and rethink your models. What other technologies are maturing and how can they change the game?
In an hour we were into the basics of one watt radios with digitally encoded voice. It was looking really cheap. Our head of research was talking about MIPS/milliwatt. Digital transmission and speech encoding was getting inexpensive. Moore's Law was giving us a different world. Throw away all of the artificial rules we had been using and you could probably make money with a service that cost a buck a day. Radios that cost $50 to $100 in parts. Maybe a billion users. Maybe more.
That isn't the exciting part. What you can do with a billion phones? What can you do with a billion devices where you know where they are? What if they have sensors? What if they have the compute power of a Newton or a Palm Pilot - or more. We mapped out several services that waited a decade or two and later some of us got involved in fleshing out the ideas more deeply.
Other people had thought about very high mobile phone densities - it was the subject of stories, science fiction and even cartoons in the teens and twenties.1 Human communication is a basic function, but sensors and computation were now part of the game. That realization had ignited our imagination.
° You could add a simple infrared temperature sensor and take the temperature of people with knowledge of time and position potentially allowing early flu detection.
° Air pollution could be mapped at great detail where it matters
° Cameras would allow you to look at bar codes on products to price them. You could even use these phones as credit cards with cryptographically secure communications.
° A portable video phone was thinkable as was a music player (we were getting involved in music compression)
° If GPS could be miniaturized, and that seemed possible, the universe of possibilities would explode. DARPA was very interested in personal maps and that seemed potentially revolutionary.
° This could be done over the Internet rather than relying on plain old telephone service. Would mesh networking or some other form of intranet emerge? There were several interesting possibilities that seemed serious threats to phone companies.
° These were any number of problems, but all we looked at seemed to be business, social or engineering problems rather than anything fundamental. Batteries for example.
The list had over twenty items when we walked into the attorney's office and another ten came during the meeting. Some of it was used for the You Will ad series that came years later. Many are common today. Of course we missed much of what has happened - particularly social changes - but it was clear inflection points in how people thought about communication and computation was approaching and perhaps it was wrong to think about them separately. A serious blueberry day.
Of the hundreds of new uses being built heath and wellness are among the most interesting. These days people are cobbling together basic telemedicine, using the computer as part of a $100 EGC machine that can send results anywhere in the world, camera adapters that allow remote dermatology, a $34 attachment that allows accurate AIDs tests. Diabetes and other chronic conditions are getting some attention. Smart clothing that monitors temperature, heart rate, blood pressure and other parameters that are important but only known during brief moments of your life. Pollution monitoring is emerging and perhaps social pressure and real change in environmental policy will come. And this is just one narrow area.
My family used to vacation in Banff, Alberta. One of the touristy activities is the cable car ride up Sulphur Mountain. You get a beautiful view, but there was an interesting building to the North on the next peak of the mountain. It was an walk and more interesting than hanging around with the family and the horde of German tourists.
A guy in his twenties opened the door. He looked like he had been up for at least a day and electronics I couldn't identify were scattered about on a work bench. I knew a little about radio, but this was exotic. I did what any geeky kid would do. I asked.
Physicists an almost primal need to explain as deeply as they can. It seems strange as physics is not about answer, but if something is already known, there is an urge to get anyone who is showing remote interest up to speed. Perhaps then they can help you with the interesting part - the questions. I spent the better part of an afternoon learning about cosmic rays, muon showers, scintillator, photomultiplier tubes and so on. Hours went by before my father was knocking. They had been looking for me and he was really pissed. I left knowing I would be either an astronomer or a physicist.
Cosmic rays are tiny bits of matter - protons or atomic nuclei - that have enormous energy. A single proton can have as much energy as a major league fastball. When they strike Earth's atmosphere a shower of electromagnetic radiation and other particles spreads. The greater the energy of the incident particle, the greater the area of the cascade - they can cover tens of square kilometers, but the big ones are really rare. Less than one per square kilometer of the Earth's surface per century.
One way to detect them is to spread an array of detectors across a large area. They don't have to be contiguous - you just want to cover enough area to sample the final products of the shower so you can estimate the energy of the incident cosmic ray. You can measure its energy, time and position and correlate that with other detectors. The greater the area, the greater the chance of seeing a shower and accurately measuring the energy of the incident cosmic.
A big detector would be very expensive, but what if distributed detectors could be super inexpensive and each linked to a smartphone? The smartphone knows position and time. It also has storage, a computer and a way to communicate results to computation that can sort out the noise from real events. Get this far and you can begin to do some physics. But there is that issue of the detector... it needs to be cheap and somehow attractive enough that people will use it. A lot of people would be necessary.
The silicon imaging array - the active element of the camera in your cellphone - will generate a signal when a charged particle strikes it. This property of silicon is exploited in some of the detectors found at CERN and other particle physics facilities. The arrays are tiny, but there are more than a billion phones spread all over the world. What if you could get a million of them waiting for events?
Cosmic Rays Found in Smartphones - CRAYFIS - is a project designed to do just that.2 Apps have been built and beta-testing is underway (I'm part of that). The easy problem is building the data logging facility. The more difficult problem may well be enough people to do meaningful science. It needs power and the phone needs to be powered on with the camera active. It probably makes sense to do this with the phone plugged into a charger when you're sleeping. There are some issues with this and it will probably be a tough sell, but the project is sweet - a wonderful idea that taps into what the smartphone has become.
There are so much potential for wireless, portable, sensor laden computation at this point. Smartphone has become obsolete and misleading term.
I'm at the end of my hour, so I'll mention something basic. If you go back to the basics of what a radio wave is you realize the spectrum crowding issues we have now are artificial - a property of policy that came from how receivers of the 20s operated - a policy that later turned spectrum into property. If you ask a physicist what the real limits are if you don't constrain yourself to current radio design you get a much larger number. Without being terribly clever a couple of orders of magnitude should be possible. With cleverness probably much more. There are people playing with these extremely disruptive ideas. How long you ask? For that I need to have another handful of blueberries..
twenty five, maybe fifty years ... I'd bet closer to twenty five than fifty
1 The cartoon is by Karl Arnold published in the German magazine Simplicissimus in 1926. Wireless personal communication ideas had been around for at least two decades at that point.
2 If you have a science background the paper is on arXiv.
Fresh tomatoes are terrible this time of the year, but you can do what a good restaurant does off season and use high quality canned tomatoes. Here's one of my simple go-to sauces. It is just a starting point, so modify as you see fit. I'm guessing amounts from the last batch.
° 1/4 up extra virgin olive oil
° 1 large Spanish onion
° 1/2 tbl kosher salt
° 1/2 tsp black pepper
° 5 or 6 garlic cloves chopped
° 1/2 a carrot pealed and finely chopped or shredded
° about 3 tbl thyme leaves (try for fresh)
° 2 large (28 oz) cans of San Marzano whole peeled plum tomatoes. Don't scrimp on this part!
° pour the oil into a large pot and bring up to a medium heat. Add the onion, salt and pepper and cook until the onions are soft and translucent.
° add the garlic and stir for about a minute - don't let it brown.
° add the carrot and thyme and cook until the carrot is tender - maybe 5 or 6 minutes
° pour in the tomatoes and juices and bring to a boil. Reduce the heat to a simmer and cook for about a half hour stirring every few minutes so the tomatoes don't burn,
° check and adjust the seasoning. (you'll probably want more salt - I started with just a bit)
° use a stick blender or run through a food processor to smooth. If you like a chunky sauce get busy with a big fork.