.comment-link {margin-left:.6em;}
The RoBlog
Thursday, December 23, 2004
 
MISWEB: Message from the future
I just finished reading an article on the MIS Magazine web site (based on a posting on Roland Piquepaille's Technology Trends) that gives some thoughts on the future from 5-25 years out. These always seem to get my juices flowing, and this time was no exception. Here are some of my thoughts on the article.


As I believe I've mentioned before, I think it's perhaps not a good idea to depend on people building a technology (especially the scientists and engineers that are solving the complex problems) to determine how the technology will be used in the future. This is especially true when we're talking about infrastructure technologies like nanotechnology. I think that these types of people are often too caught up in thinking about solving the challenges of the technology to spend much time thinking about how it will really be used. They are also, by the nature of working on the problem, more likely to be biased towards its importance (and often, a particular kind of importance) to be reliable. This often manifests itself in company PR statements claiming that a technology they are developing will revolutionize the way we do something but that something is either too narrow for a revolution to be interesting, or too broad for the technology to apply. So when I see statements like "Scientists portray a future in which...", assuming those scientists are the ones developing a technology, I tend to treat it with a grain of salt.

The author of the article, Helene Zampetakis, uses the term "disruptive technologies" a couple of times in a way that I think is just a bit off. Typically we think of disruptive technologies as ones that cause a paradigm shift making existing companies scramble or go out of business. While nanotech may well do this in some industries, saying that they will "disrupt the entire semiconductor fabrication industry" is an overstatement. As the existing major manufacturers of semiconductor electronics are likely to be the ones that will be creating nanotech electronics in the future, the disruption will really likely happen at the tool vendor level. The big guys will remain big and in charge, and most of us will probably not notice anything but the continued expression of Moore's Law. "Revolutionize" may be a better word, as few existing semiconductor companies are likely to have to scramble much more than they already are (and have been for quite a number of years).

In the section on quantum information processing (qip), Bob Hayward, vice-president and research fellow at Gartner is quoted as saying "It will give us an order-of-magnitude jump from the fastest computer of that time - it will not be an evolutionary increase." While he, no doubt, has access to much better information that I do, I do wonder that an order of magnitude increase from the fastest computers of the time (when quantum information processing comes online) will, in fact, be an evolutionary increase. I don't think anyone is saying that information processing ability follows a linear increase. To hear many tell it, we should see something significantly greater than that. I expect that qip will be used tentatively in the market at first while the bugs get flushed out, and by the time it has any real impact on the lives of regular people, we'll be able to point to some curve that more-or-less expected this kind of revolution; may, in fact, that a step-wise change was inevitable.

Zampetakis goes on to quote Hayward saying "As for security, whereas it takes a month for a supercomputer to crunch a DES encrypted code today, it would take just a couple seconds [using QIP]. But maybe we'll also get a corresponding advance in security algorithms." With the recent work done in quantum encryption (the subject of this month's Scientific American, coincidentally), it would seem that security, in the form of encryption, in any case, will still be strong enough to withstand the computing power of the foreseeable future. More interesting would have been a note about how qip will allow information on you from disparate sources to be correlated in real time, and how the vast sensor network will be able to track where you go and what you do (though this may not be as bad as it seems).

An interesting distinction that often gets overlooked when making predictions about the future is between when it will be possible to do something, and when doing that thing will be available enough to have any impact on our existence. For example, in the section on advanced materials, Zampetakis says: "Over the next five years, the development of nanoscaled sensors will allow intelligence to be built into many materials for multiple applications." This may be true, but (and this is highly dependent on the application) it is likely to be twice that amount of time before products incorporating this ability are directly interacting with or impacting regular people.

Dr. Peter Corke, autonomous systems team leader at CSIRO (I don't know what "CSIRO" is, and when I went back through the article looking for it, I couldn't find it, and I'm too lazy today to research it) is quoted as saying ""Robots will be fairly prevalent 10 years from now. We'll see them in offices and hospitals and shopping centres." I'm still doubtful on the regular integration of robots into daily life, but I do feel that they are probably coming SOMETIME soon, so 10 years is likely not that bad of an estimate. It's probably more accurate than other timeframes in this article.

Using robots to perform tasks like mail delivery and store re-stocking seems fairly unlikely to me in the near-ish term. Typically a technology like robotics encroaches into our daily life by taking over the excessively dangerous, costly, error prone, or tedious jobs from humans. Neither store shelf re-stocking, or mail delivery seems to fall into any of these categories with enough severity to warrant them taking human jobs. Further, both are significantly complicated in terms of object negotiation and barrier navigation that the technology will have to be pretty sophisticated (read: expensive) to do any good. More likely will be robots increased use in fire fighting, policing, mining (as mentioned in the article) and the like. No doubt there will be robots capable to doing most or all of the store re-stocking/mail delivery tasks in 10 or so years, but my belief is that they won't be deployed for those uses by that time. By the way, warehouse inventory re-stocking is a whole other matter than shelf re-stocking, and I expect we'll see widespread deployment of automation in that area (I'm guessing this already exists in non-trivial dollar amounts today).

I've mentioned NASA's efforts to help bring about a flying car in the past (although I can't for the life of me find the post...hmm...). Flying cars are one of those things we love to imagine, but I suspect we'd hate in practice. Even if we could conquer the noise, and the risk, we'd still be faced with an energy issue: It takes a heck of a lot of energy for an object to stay airborne. Probably about the time that we could have flying cars, the demand for them will be pretty low as virtual presence technologies increasingly mitigate the need for long-range travel (the most likely niche for flying cars), and automated ground travel (the self-driving car) and mobile connectivity mean that the time we spend in cars can be productive and entertaining for all.

The problem with allowing people to be further distanced from their places of work (the non-social problem, that is) is that if people migrate en-masse to wherever they like, congestion will become a critical issue where all of these commuters come together to land (given the urban airport scenario cited by the article), and then people will have to get from these urban airports to their places of work anyway. I have to admit that I'd LOVE a flying car, but I'm doubtful of it becoming practical enough to become anything but a toy for the rich. I suppose if it was fully automated, you might rent one for a family vacation, but the economics and convenience would have to work out pretty well even for that.

I've notices that this article uses "the next generation of" at least twice, implying that things like 3D holography are just around the corner. Most of these technologies are still several generations away from public consumption.

While we wait for 3D holograms to allow us a form of telepresence that allows us to walk around remote rooms (would we be walking around in specialized recording rooms in this scenario, or would a computer be adjusting where we are walking to make it look natural in the display environment so that when, for example, I walk around the desk in my office I don't appear to walk through a table in the remote location?), technology that allows us to meet in virtual space will likely become more commonplace, allowing us to interact more naturally an a wholly constructed environment than holography will allow for quite some time.

Five areas of business change are identified (rather awkwardly) at the end of the article, and I agree with every one of them. What we'll see, in my opinion, is the transformation of corporate IT from something that looks like the fleet department to something that looks more like the human resources department. In fact, I'd go so far as to say that we'll be seeing combined "Human and Information Resources" departments in the next decade or so as most of the "working" portion of IT becomes outsourced and the IT director becomes someone who manages contracts and regulations than software and computing infrastructures.

Something interesting occurred to me while reading this article. I get hung up thinking about the apparent paradox between things that seem unlikely today but happen anyway, and things that seem likely but don't. What occurred to me is that there is often a single breakthrough required for a technology to become mainstreamable, and that breakthrough is often not predictable. For example, holography is something that seems like it may yet take a while to produce even though it's been around for a while. At some point, some genius will re-envision some core component, and suddenly we'll all have 3D TVs. Perhaps a technology needs to mature enough (which, by the way, takes its own geniuses) so that a genius can study the system and make the requisite discovery. Food for thought.


Comments:
--------
You wrote an excellent commentary about the Helene Zampetakis's article. You also wrote you didn't know what was CSIRO. This is a major Australian research organization. CSIRO stands for Commonwealth Scientific and Industrial Reseach Organisation" and its root web site is at http://www.csiro.au/.

Happy New Year!

Roland Piquepaille (http://www.primidi.com)
 
--------
Thanks for the additional information Roland!
 
Post a Comment

Links to this post:

Create a Link



<< Home

Powered by Blogger