If robots do all our work, what will happen to us? And a tiny oscilloscope sparks the imagination.
And so, in the arrangement of the family, a slave is a living possession, and property a number of such instruments; and the servant is himself an instrument which takes precedence of all other instruments. For if every instrument could accomplish its own work, obeying or anticipating the will of others, like the statues of Daedalus, or the tripods of Hephaestus, which, says the poet,
of their own accord entered the assembly of the Gods;
if, in like manner, the shuttle would weave and the plectrum touch the lyre without a hand to guide them, chief workmen would not want servants, nor masters slaves.
(From Book 1, section IV of Aristotle's Politics , as translated by Benjamin Jowett.)
|Click image to go to digital edition.|
Bertrand Russell thought Aristotle's ethics were “repulsive” and indeed sections V and VI of Aristotle's book uses dismal arguments to justify slavery. It's sad that two and a half millennia later that institution still ravages parts of the planet.
The quote above does explain a prime reason slavery arose in the United States. Cotton didn't pick itself, and the plantation's entropy increased daily unless constantly tamed by labor, labor which the rulers were unwilling to supply themselves. Slaves were used as machines. To paraphrase Aristotle, if the work did itself, the master would have no need for human machinery.
I think the inevitable result of the evolution of embedded systems is a new kind of slavery, one without moral objections: the subjugation of machines. Since the Industrial Revolution machines have increasingly replaced human labor. But those machines were dumb; devices that repetitively performed only simple operations. The new paradigm is one of increasing smartness.
Isaac Asimov in his I, Robot books envisioned devices that were largely indistinguishable, at least in a Turing sense, from people. Surely the path to that end is murky at best. But that sort of automation is hardly necessary to completely reinvent the nature of work.
Frank Tobe, the always-interesting editor of The Robot Report (www.TheRobotReport.com), was interviewed at http://find.botmag.com/031121. He says: “A thriving robotics industry provides jobs, helps the nation increase efficiency, profitability and productivity and upgrades the mix of workers involved.”
Well, yes. And no.
When a country exports manufacturing, the immediate benefit is lower prices. But ultimately it also exports knowledge, eventually undermining business and creating a new highly-effective competitor. There's a short-term gain and a long-term problem. “A thriving robotics industry” is similar; there will be a lot of great engineering jobs inventing the technology. In the long term, we can expect a total reinvention of the nature of labor and work in general.
In a few decades I believe we'll see autonomous robots that, while intellectually dim, will perhaps look vaguely humanoid and be able to do all of the routine tasks in life, like collecting garbage, fixing plumbing, serving hamburgers, and much more.
A lot needs to be done; for instance, battery technology has to evolve quite a bit. But the automotive industry is pouring money into that, so surely advances aren't far away. The computational capabilities are nearly here and machine vision is quite advanced today. Motors and servos need some evolution, but Honda's Asimo and U3-X signal how close we're getting.
Once machines can perform routine tasks, they will be able to build robots. Mining and production of raw materials will be automated. The distribution and movement of goods will all be taken over by the machines. (How much worse can a robot driver be than the crazies that now populate the freeways?).
Business will follow the usual pattern. A startup or ten will invent the needed technology. A few companies will buy early prototypes in their zeal to cut costs (in other words, reduce the cost of labor). Competition means others soon follow suite. The Robot Bubble, another tulip mania, inflates, more people lose their jobs, and just as cell phones went from novelties to teenagers' essential fashion accessory nearly overnight, robots will quickly be everywhere. Eventually they'll become a commodity so the bubble will burst.
Today, it feels like the world is collapsing as we face monumental financial, environmental, and cultural issues, the resolution of which will inevitably cause us to confront some very difficult decisions that I doubt we as a society have the skill to manage. These will seem but trifles as the age of the robot dawns. For if robots can make themselves, starting from quarrying the mines to final assembly of the silicon bits, they will have no monetary value. They'll be free. (The only element of value will be the ownership of the mines and other sites of raw materials, and it's hard to see how that will play out). What will happen to those people who now earn their living mining, plumbing, collecting garbage, and doing pretty much any blue-collar job?
Politicians of all stripes spout a mantra of “worker reeducation” whenever a city loses a manufacturing industry. Not many can be retrained to be an engineer or a scientist. And there is not enough demand for these retrained factory workers to have any significant impact on employment. It's hard to see that changing, especially as vast waves of people are made redundant by smart machines.
Will our machines get some sort of creative spark, like Sonny in the film version of Asimov's book? It's awfully hard to see a technological path to that, though the modern miracles we take for granted were impossibilities not long ago. But if robots got “the knack” then most other jobs are toast. I'm pretty sure my doctor is already a robot as he hasn't looked at me in years, so busy is he entering data into his computer. The New York Times recently had an article (www.nytimes.com/2011/03/05/science/05legal.html?_r=4) that showed how computers are replacing paralegals and associate attorneys. This will surely continue, and automation will encroach on other jobs once thought immune from information technology.
What does money mean when human labor is no longer important? Goods have value is because labor is required for their production. When we want something, we exchange the money earned by our labor for that goodie; money is merely a representation of work.
What will happen? I see three possibilities:
First, there could be a rising of the masses against technology and maybe even technologists. In Walter Miller's A Canticle For Leibowitz what's left of society after nuclear holocaust blames the techies for creating such destructive weapons. Engineers are hunted down and eliminated. History suggests such a rise is possible: the Luddites revolted because their work was replaced by machinery.
Or there could be some sort of collapse; With no need for workers, will those who control the raw materials become a new royalty with the rest mere vassals? With nothing to do will the displaced find solace in Huxley's soma?
The final possibility is one of a truly bright future, a Star Trek-like world where, free of the need to work to put food on the table, people pursue self-improvement and their passions. Possible? Perhaps. Most of the people I know have hobbies or other activities that they'd be happy to devote more time to.
Norbert Weiner expressed it well: “Let us remember that the automatic machine is the precise economic equivalent of slave labor. Any labor which competes with slave labor must accept the economic consequences of slave labor.”
The future faced by our kids will be interesting indeed. I wonder if they'll look back and cheer or curse today's engineers?Xprotolab
In recent months, I've written about some amazing oscilloscopes that you'd be lucky to have adorn your lab bench. (www.eetimes.com/4214706 and www.eetimes.com/4027601). These are professional units for demanding applications and cost thousands of dollars.
But I found a new scope that completely flabbergasted me. It's an unexpected but natural outgrowth of the same embedded upheaval that will shape robotics. The Xprotolab from Gabotronics (www.gabotronics.com/development-boards/xmega-xprotolab.htm) seemed at first the product of a soma-induced hallucination.
Three things blew my mind about this thing: the price, the size, and the surprisingly-deep feature set.
Forty-nine bucks a pop. $35 each if you buy 10. That's not a misprint. To my knowledge, it's the cheapest scope on the market. And it includes a display; this is not one of those USB dongles that digitizes data and passes the results to a PC application.
It's 1 inch x 1.6 inches, or about the size of two postage stamps. Did I mention there's a display? That's included in this form factor.
Then there's the feature set. The Xprotolab implements most of the features of a modern scope. This is even a mixed-signal scope: it has two analog channels and 8 digital. It'll compute FFTs and has an arbitrary waveform generator that can run in the background while the scope is being used. There's a DMM and frequency meter as well, cursors, all of the usual triggering options, and some math capability. There's a protocol sniffer, but since my unit's firmware was old I couldn't test that feature.
Did I mention it's $49?
Click on image to enlarge.
The unit has plenty of limitations, of course. It has a 200-kHz analog bandwidth with a max sampling rate of 2 MSPS. The buffer size is just 256 bytes per channel, and the input voltage range ±20V. The screen update rate runs around warp 0.00001; it redraws the traces about once per second. This is not a substitute for a lab scope.
Modern oscilloscopes tout their large screen sizes. Bigger is better. The Xprotolab's display is a bit smaller than a wide-screen TV. It's 0.9 inches. I thought about putting a magnifying glass in front of it, as people did for the tiny displays used on the first televisions. It's frankly hard to read, at least for the bifocaled, though the OLED display is very bright and crisp. The scope's astoundingly small size mandates a lot of squinting. But it is indeed readable. With 128 x 64 pixels, there's some distortion of signals; square waves may have a bump in the rising and trailing edges.
Just four (tiny!) momentary switches control all of the scope's modes. There's a brilliant menuing system to select and change parameters. On a lab scope a quick twist of the wrist selects a sweep rate or amplitude setting; on Xprotolab you'll have to click a few—or sometimes more than a few—times to set things up, but with practice it's pretty easy to control.
A user's manual was not supplied. I found what passes for documentation here: www.gabotronics.com/product-manuals/xprotolab-manual.htm. The manual is poor enough (it's basically just a flow chart of the menu operation) that I found myself referencing the schematics and the code, which are also on that site.
This Lilliputian scope will plug directly into a prototyping board, but it cries out to be integrated into a product. I can see including it in an instrument to show the data being collected. Or as a tuning indicator for a radio. Maybe a weather station barograph or an energy consumption monitor. The possibilities are endless. The scope is sort of marketed as a development platform. With the code and development tools all available it's an easy matter to tune it to your application.
Check it out. You may find yourself enslaving a scope in your next product.
Jack Ganssle () is a lecturer and consultant specializing in embedded systems development. He has been a columnist with Embedded Systems Design for over 20 years. For more information on Jack, click here.