Hi after a very long absence.
I thought I would post a message about my ideas I presented here in 1997 that someday soon there would be no work.
The idea was pooh-poohed then by most people who communicated on this thread.
Having had the recent opportunity to research and think about this subject again I thought it would be fun to pursue a discussion here again.
I think in 1997 I presented the idea wrong in saying there would be no work, I should have instead argued that there will soon be no jobs.
In 1983 my first PC had 29000 transistors in it's core processor. The transistor count on computers has increase from that to over half a billion transistors on the very latest Intel Itanium cores.
Today we manufacture over 2,000,000,000,000,000,000 transistors per year and at any given point in time there are more than twice as many transistors manufactured in the preceding year than existed world wide at the start of that year. The price per transistor is now measured in 10's of nano-dollars. About as cheap as the period on your screen, especially if it takes you several minutes to read this. To put 2 e 10^18 in perspective that is somewhere around 100-1000 transistors for each ant in the world. I doubt though that the ants care.
Enough about technology advances for the moment.
The increase or decrease in unemployment can be roughly predicted by this formula. (Rate of Unemployment = Rate of Unemployment + Productivity Increase - Demand Increase)
In each recession since probably 1900 the drop in the rate of unemployment has been delayed for successively longer and longer periods. It has been especially exaggerated it this recession and the 1992 recession.
At the present time we are beginning to experience the largest pent up demand for new technology and equipment in both industry and the office in history. As an example a company I work with is still using 100 MHZ Pentium computers as, just as, they were beginning to consider upgrading from these 1997-98 machines the recession struck, stopping almost all capital investment. They will soon upgrade to machines more than 20 times as fast with a hundred times more storage. (also at a quarter of the cost) This alone will allow for a tremendous increase in productivity for their office staff, and even their production floor people who must also continuously access their computers.
A huge exacerbating co-problem - we are now just beginning to face is the export of office jobs. (WARNING IF YOU TELECOMMUTE YOUR JOB IS READY TO BE EXPORTED) Recently there was a short report on CNBC of a London Company transferring 50 accounting jobs to India. The companies response when interviewed was "Why should we pay London Wages." With the present and soon to be upgraded office technologies and the present Internet capabilities (including voice and Conferencing) it is increasingly easier and easier to export office jobs.
Take a look at farm employment over the years. Even with the advances in technology from the advent of the steam engine etc. from the 1750's to the 1860's farm labor engaged approximately 80% of the population. 100 years later it engaged 16% of the population. Today farm employment is less than 2% of the labor force. Comparing manufacturing in 1860 it employed 8% of the working population that increased to 46% in 1960 and is presently below 16% and dropping both due to exportation of manufacturing to cheaper labor markets and continuingly increasing automation.
Office work has increased from approximately 10% in 1860 to 40% in 1960 to more than 80% today. That will be a be the peak in the United States.
Back to technology - The next "Killer App." from the software industry (probably Microsoft) will be natural language processing. Just talk to your computer like you would your secretary or your kids. For instance you may want to name your computer. So I could just say "'Jim' look up and prepare a chart on percentages of the work force from 1800 to the present by industry categories. Dress it up so I can make a presentation, I will review it and suggest changes. You pretty much know what my preferences are."
Co-advances in "Killers Apps" are self programming, genetic programming, etc. Self Programming in conjunction with natural language processing will allow computers to do the above example. Genetic Programming allows a computer to evolve a product. More than 15 patents to date have been issued for computer evolved products.
The resulting problems:
1. Jobs will soon start to recover indeed they may have already. But as productivity soars as it will as the huge amount of capital investment in technology surges. The formula I mentioned above will kick in productivity will soar at huge rates, especially on a dollar rate as jobs are also increasingly exported via the Internet. Then jobs will begin to permanently disappear.
2. Ultimately there will be no jobs at all. IBM is presently building the ASCI Purple with processing power equal to the human brain for delivery in 2004. They are also building Blue-Gene Lite for delivery in 2005 with 3.6 times the processing power of the human brain. Traditionally "Moores Law" ends up with the super computer of today being on your desk top in 10-15 years. With the power of the human mind on a chip what will there be for us to do?
Rod |
