Sidebar to Jakob Nielsen 's column on microcontainers .

I have received some interesting user comments on my May Alertbox about the long-term strategic impact of the Web.

In This Article:

Cities Will Not Die

Joshua Lerner from New York City writes:

Re: your comment on large cities like New York experiencing a real estate crash within 5 years due to the Web making geography irrelevant -- I think you may be drastically underestimating the influence of other forces at work here. While I agree with you that the Web does in many ways make geography irrelevant, and that the Web's influence on society is greater than most people imagine, to predict the "death of cities" this way (without explaining how the Web will counter the forces that make cities desirable places to live) is not easy to accept.

Jakob's reply: I agree that cities will not die completely. I do think that many people will move to remote areas where they could not have lived in the industrial economy. If, say, 10% of New Yorkers move to Minnesota and similar places, then that will cause much more than a 10% drop in real estate prices.

Other readers have commented that the other side of the coin is an increase in real estate prices in beautiful but distant locations. This will probably happen, but not as much as the drop in prices in the overpriced areas. The different effects are due to the marginal impact of changes in the number of residents and potential buyers.

Mark Bernstein, Eastgate Systems writes:

Your prediction that ubiquitous networking will lead to a decline in urban real estate prices over the next decade is interesting, and may well come to pass. I think it is important to remember that work requirements are only one of the reasons that people live in cities. Other forces will continue to draw people to urban centers. Some of these include:

Sex

Urban places are crowded, filled with new and attractive people you haven't met. They're ideal places to find a new partner. Even people involved in settled relationships thrive on opportunities to meet new friends and see new people. Loneliness matters, and love is terribly important. Even if some members of a household feel free to move to Montana, others may long for the crowds, the bars, the parties, the bright lights.

Medicine and Education

While many services can easily be provided through the Web, medical and other therapeutic services need the presence of the body. Medicine matters to lots of people. Again, since people tend to make housing decisions as families, the expected need for convenient medical care for any member of a family may prove compelling for all members of the household.

Education tends to be conservative and slow to adopt technology, and so schools will lag the workplace in adopting collaborative technology. Some aspects of education, moreover, demand actual presence; this is true both for young children (who need to play with other kids to learn social skills), for advanced students (who need personal interaction with scarce experts), and for students in between (who need to learn about relationships; see above).

Finance

People will only move if they can afford to move, and consumer liquidity might well be a significant factor for some part of the next decade. Extrinsic factors play a role -- especially if recent dislocations in Asia do not prove to be short-term and local. Should demand for urban residential real estate decline dramatically in the next five years, moreover, many people who might wish to move would find themselves unable to find a buyer for their current residence. Unable to sell their house for enough money to pay off the balance of their mortgages, these households will remain in place until demand stabilizes (or until their mortgages are paid), reducing the outflux from urban centers.

Suburbanization required the invention of the train, the automobile, the active support of governments, and WWII. Even so, it required about 75 years to take hold. The network may change housing patterns yet again, but the changes may be more complex (and slower) than you predict.

Jakob's reply: I agree that it will take some time for the full effect of "the Death of Distance" to play out fully. Even so, I think the initial migration of maybe 10% of the people will start in about five years. I also agree that there will still be an uneven distribution of population, with some areas more densely populated than the average. I simply think that it will be less concentrated than it is now.

The factors mentioned by Bernstein are surely important, but may not turn out to be prohibitive for a more even population distribution. For example, some forms of medicine can be practiced remotely, particularly given the expected growth in telemedicine. Patients may get implanted sensors that transmit continuous readings to an expert system over the Internet, thus reducing the need for them to report for in-person tests at hospitals and improving treatment. Ultimately, the ambulance may be at your door two minutes before you get a heart attack.

Sex is a different matter, despite the many titillating speculations about the potential applications of virtual reality technology. However, statistics do show births in Montana, so I assume that associated activities do happen even in such places. Remember, I am not predicting that people will live as hermits in mountain cabins: simply that they won't crowd together as much as today.

Mike Garrison writes:

Sorry Jakob, but I think you are underestimating the very real forces described in the comments by Joshua and Mark (and by implication, many more people).

I live in the Seattle area. We currently have an anti-suburb real estate pattern which has been going on since the 80s. More and more jobs are being located in the satellite cities (Bellevue, Redmond, Kirkland, Renton), mainly because property is cheaper. But housing prices have skyrocketed in the city itself. Traffic on the "anti-rush hour" routes (out of the city in the morning, into the city in the evening) is now as bad or worse than the traditional bedroom community commute.

The fact is, people in this area are already ignoring their local geography and choosing to live where they want, regardless of where they work. And where they want to live is in the city proper.

It's true that patterns inside one metro area are not necessarily the same as patterns in the larger world (like Minnesota v. New York City). There are many people who would prefer to live entirely away from the city if they could. But there are many more people who prefer to live in the city but can't afford it.

The city is a lifestyle and a culture, not just a source of income. To a large extent, the car, the airplane, and the telephone are already allowing people more freedom to disassociate home and work than ever in history. And cities are as crowded and expensive as ever in history.

The web may change employment and shopping patterns, but it won't change living patterns nearly as much.

Jakob's reply: I agree that there are many non-job benefits to city living. Indeed, I am a great fan of London and Manhattan (Brooklyn is another matter, though). But in a highly networked future it will be possible to gain many of these benefits anywhere in the world. For example:

  • Going out to dinner with friends : future restaurants will offer remote dining that connects tables in different cities through a high-def video window. You can have dinner with somebody who is thousands of miles away and it will seem as if you are sitting at the same table. The two restaurants will even coordinate the meal service so that you get your plates at the same time.
  • Cultural activities like museums and theaters will be partly replaced by interactive art forms over the Internet. Once we get ultra-high resolution screens (1600 dpi), it will be possible to show better images on the screen than on a painting. And interactivity is more engaging than any passive activity , as known by anybody who has played Doom - it's not a deep plot, it's not beautiful graphics - but it's you shooting the bad guys. Think of what we can achieve once the best creative minds on the planet start making interactive works.
  • Shopping : well, that's already a strength of the Web. Amazon offers greater selection than any mega-bookstore. Currently, they don't have an espresso bar, but there are certainly many "community" vendors who are thinking about ways of enhancing online shopping with ways of connecting to kindred souls.

People Don't Buy Good Screens

Brian Rosen writes:

In this column, you make a prediction about high resolution screens, and you push out a spec (10,000 by 6,000 pixels). This is a subject I know something about; I started a company in 1985 who brought out a display monitor and controller which was 4096 x 3300 single bit. This the same size and resolution as a low end laser printer, 300 dpi. Later, this company, MegaScan Technology, provided a 2560 x 2048 x 12 bit gray scale system (200 dpi) for the medical community. Although the later product is still alive in close to it's original form, the ugly facts of resolution are that customers want more resolution, but they will not pay for it.

It has been feasible for 15 years to make 300 dpi displays. Probably, a color display at 300 dpi has been feasible for around 5 or 7 years. 300 dpi is very close to photographic quality with enough bits per pixel. We have taken photographs of the screen showing a scanned image and shown the original and the screen shot to experts. Only the distortion of the curved screen surface give away which is the screen shot. Yet, even a market you might think would demand quality, and not be particularly price sensitive, radiology, is still dominated today by the same 1280 x 1024 resolution that was developed in 1980. The 2560 x 2048 is relegated to micro-minority status.

There are 1600 x 1200 line (around 120 dpi) color display systems that have been available for 10 years, and today probably have a real 25-30% price premium over an equivalent 1280 x 1024. How many do you see? Very few.

This is a significant change in historical trends. Up until 1000 line displays, there were steady increases in real resolution from around 200 lines up to 1000 lines. All of the technology drivers for CRT displays (spot size, dot pitch, video bandwidth, scanning frequency) enjoyed almost lock step advances, and, the facts are that they still have the capacity to be increased as long as the economics to do so exist.

But they don't. It's a straightforward volume driven economics model; it starts out low volume and relatively expensive unit prices, and as the volume increases, the costs come down to below the prior levels (more image quality for less money). What happened in displays is that there were no markets that would accept more expensive displays at low volume, and therefore the volume never grew to get the cost down. This is as true today as it was. There is no doubt that if I offer you more resolution at no cost penalty, you will take it, but after around 100 dpi, you stop being willing to pay more for better image quality.

The same situation occurs in flat panels, with a big difference; the volume requirements for low cost are 2 orders of magnitude larger. So, yes, we will see 1000 line flat panels soon (we are at 800 line now, right?, around 72 dpi). However, don't be surprised when there just isn't that every-four-years-it-gets-better swing in resolution. All we are going to see is HDTV, around 1100 lines. It might be 20 years before we see substantially more.

Oh, and don't think that those really big wall sized flat panels do it; the ones available now have even less resolution that some laptops!

Finally, I'll admonish you to remember something you probably know; display quality is measured in dots per inch (or cm), not per screen. A 10000 x 6000 display 20 feet on a side would be impressive, but the image quality of a 500 x 500 at 2" on a side is much better, given equivalent contrast, brightness and gamma.

Jakob's reply: I agree that it is odd that users are so reluctant to spend money on the screen they stare at all day. All recent computers have enough computational power for most common applications and yet CPU speed is the one factor that gets most attention in trade press reviews of new computers. Many magazines publish elaborate laboratory tests of different models' ability to crunch floating point numbers, but they never measure the average reading speed from the monitors.

In the long term, I still believe that reason will win. We know that 300 dpi screens allow users to read 33% faster than current computer screens: the reading speed goes up from 75% of print to the same as print. The added productivity from faster reading speed means that companies that buy good screens for their staff will win in the long term. Also, as I briefly discussed in the column, high resolution is necessary for most collaborative use of the Internet.

I think the Internet may well be the event that pokes a hole in the dam and starts the flow of high-resolution screens.