Retinal implants are another matter. Thanks to new tech, several folks who would have been blind can see: a little.
- Retinal Implant: Low Resolution, but Better than Nothing
- Leaving the 1940s Behind: Heavily Interconnected, Biologically-Inspired Chip
"Metropolis," Tsukumogami, and the Roomba Revolution that Wasn't
The inventor Rotwang in Fritz Lang's "Metropolis" is more 'evil wizard' than 'mad scientist:' my opinion. It's still a good movie: but not, I think, a particularly realistic look at what we'll see in 2026.
Rotwang's Maschinenmensch looked more like the human she was built to impersonate after a high-tech makeover, but even without upholstery she was remarkably — human.
Rotwang wasn't the first fictional character to build himself a girlfriend. Ovid's Pygmalion predates Fritz Lang by about two millennia.
Robots have an even longer history: if you count Talos, Europa's bronze guard. That tale is at least two dozen centuries old.
With the year 2026 is still in our future, we have robots. Some of them, like Baxter there, can work on an assembly line: with a little help.
Others, like Boston Dynamic's PETMAN, test hazmat suits. (April 12, 2013)
Kawada Industries have been developing the dancing HRP series.
But so far, artificial intelligence has been quite obviously "artificial:" and emphatically not up to the task of leading a Roomba revolution.
Japanese folklore apparently includes tsukumogami, self-aware tools. I suspect that at least some speculative fiction about robots has roots in tales like the one about a zither that developed a mind of its own: and that's — not quite another topic.
IBM's TrueNorth Chip and Cockroach Brains
We may, eventually, have robots that can pass a Turning test. When or if that happens, I'm pretty sure that some folks will be upset: but I'm also sure that we won't be replaced by machines.
Some of our jobs: that's another matter.
In my 'good old days,' about a half-century back, folks who had earned middle management positions felt threatened by computers. They were very good at adding up columns of numbers, and writing the results in little boxes.
They were being replaced by information technology that did the same job: faster, more accurately, and without demanding vacation time.
The sensible choice for these folks was learning to use the new tech, or looking for a different sort of job.
A century earlier, folks with different abilities made a living by swinging a hammer. Their jobs were threatened by steam engines. John Henry's duel with a steam drill is probably more folklore than fact, but the conflict between old and new was very real.
I sympathize with folks who learned to do one job: and, after decades of routine, had to learn new skills. On the other hand, I don't feel threatened by today's Roomba, or 'Heimie:' my name for Hitachi's EMIEW2.
'Heimie' doesn't seem particularly useful, except maybe as an office mascot. But 'his' more utilitarian counterparts, like Kiva robots, are pretty good at stacking merchandise.
'Heimie' and company might threaten the careers of mail clerks who don't get along with new technology: but somehow I think humanity will soldier on.
(May 10, 2013)
What got me started thinking about real robots, and their more entertaining counterparts in speculative fiction, was news about a new computer chip with circuits that mimic the networks in living brains.
I don't think we're on the brink of a robot apocalypse, though.
IBM's TrueNorth chip only has 1,048,576 "neurons:" about as many as a cockroach's brain. We're still a very long way from robots like R2-D2 — assuming that human creativity, curiosity, and general quirkiness, is simply a matter of the size of our brain's neural network. That's a big assumption, and another topic.
Neurosynaptic Cores and a Creepy Implant
My guess is that tech like IBM's neurosynaptic cores will make next-generation prosthetic eyes and limbs more effective.
I'll admit that implants like Second Sight's Argus II look 'creepy.' Given a choice, I'd go through life without needing eye surgery.
But if I was going blind, I'd be willing to consider an implant. This isn't an entirely hypothetical question for me. I had defective hip joints replaced several years ago.
I'm pretty sure that some folks would rather go blind than submit to surgery. Some might feel that they were defying God's will, others wouldn't think it's worth the expense, and some might simply be queasy about a doctor poking electrodes into their eye.
Science, Faith, and Autopsies
As a youth, I read about the "dark ages:" when the fall of Rome had allegedly plunged the world into superstitious chaos. Later, I learned that it was "dark" only in the sense that most Europeans couldn't read or write: which meant that with few exceptions, documents were written by Catholic clerics.
The "superstitious" label seems to have stuck, but I think word's getting around that rational thought wasn't 'suppressed' by the Church. And that's yet another topic. (February 23, 2014; July 15, 2014)
We probably wouldn't have today's medical science and technology, if Christianity's attitude toward the study of nature hadn't made autopsies possible. Ancient Greek physicians Herophilus of Chalcedon and Erasistratus of Chios studied human cadavers in the third century B.C.: but for the most part the human body was off-limits in Classical antiquity.
Roman legal procedure loosened up a bit around 150 B.C., permitting autopsies in a few cases: like Julius Caesar's autopsy in 44 B.C..
England's another story. Dissection of humans was forbidden in that island nation until 1500s, and that's yet again another topic.
Since I'm Catholic, I have to believe that humans are supposed to be curious. Part of our job is taking care of this astounding creation: studying it and developing new tools. Science and technology are part of being human. It's what we do. (Genesis 1:27-31; Catechism of the Catholic Church, 31, 355-361, 374-379, 2292-2296, 2301)
Life and Health: Precious Gifts
As I've said before, the rules are simple: love God, love your neighbor; and see everyone as your neighbor. (Matthew 5:43-44, 22:36-40, Mark 12:28-31, Luke 10:25-30; Catechism, 1825)
As a Catholic, I have to believe that human life is sacred. I also must recognize that science and technology can help us help others, and that ethics are important. (Catechism, 2258, 2292-2296)
Since life and health are precious gifts from God, I'm expected to take care of my health: within reason. That means that I'm not allowed to kill myself, or anyone else, even if I'm sick and would probably die anyway: but I'm not expected to desperately cling to life by insisting on "over-zealous" treatment. (Catechism, 2277-2279, 2288-2291)
1. Retinal Implant: Low Resolution, but Better than Nothing
(From Second Sight, used w/o permission.)
(Argus® II Retinal Prosthesis System's external hardware.)
"Implant Helps Patients Blinded by Retinitis Pigmentosa"A phosphene is a ring or spot of light we perceive — even though light hasn't entered the eye. They're caused by pressure on the retina, mechanical, electrical, or magnetic stimulation in the visual cortex, or random firing of cells in the visual system.
Laird Harrison, Medscape (August 12, 2014)
"Retinal implants can provide some vision in patients blinded by retinitis pigmentosa, new studies show.
" 'The results are very encouraging in some patients,' said Stanislao Rizzo, MD, a retina specialist from Pisa, Italy, who presented some of the findings here at the 32nd Annual Meeting of the American Society of Retina Specialists.
"Retinal implants are the only treatment that can improve vision in patients with retinitis pigmentosa. The inherited condition leads to photoreceptor degeneration but leaves the inner retinal layers intact. This explains why electrical stimulation of the inner retina causes patients to perceive phosphenes.
"The Argus II retinal prosthesis system, developed by Second Sight in California, works by stimulating inner retinal cells. A glasses-mounted camera with a battery-powered video processing unit sends signals to the implant, which consists of a scleral band, internal coil, and retinal surface electrode array...."
Folks with retinitis pigmentosa generally have a fully-functional visual system: except for cells at the back of their eyes that convert light to nerve impulses.
There's no cure, so far, although research may give options involving transplants, gene therapy, stem cells, nutritional supplements, or drug treatment.
The good news is that we now have a working retinal implant that works. Today's replacement part gives the patient vision that's better than being blind: not much better, but better.
Second Sight's Argus II implant includes an array of tiny electrodes that stimulate the patient's retina. A photo of the implant shows what looks like a six by nine array of electrodes. That's enough for a 54-pixel image.
That's nowhere near the resolution we're used to, but it's better than nothing.
("Nine:" in an image 329 pixels across (left); and nine pixels across (right).)
"Preliminary Data"
"...The researchers tested the ability of the patients to locate squares of light on a computer screen and detect the direction of motion on a screen.From what Dr. Rizzo said, most of the 10 people who would have been totally blind now have limited sight: very limited, but real.
"Twelve months after implantation, square localization had improved in 9 of the 10 patients, and motion detection had improved in 7. 'Of course, we had the best results in patients with better visual acuity in the preoperative period,' Dr. Rizzo reported....
"...During the same session, data on the Alpha IMS implant, developed by Retina Implant AG in Germany, were presented by Caroline Chee, MBBS, from the National University of Singapore. Whereas Argus II is an epiretinal implant, Alpha IMS is a subretinal implant.
"Twelve months after implantation, enough vision had been restored to be useful in daily life in 46% of the 26 patients she reported on...."
(Laird Harrison, Medscape)
Besides a low-resolution visual image, patients got better at telling where sounds came from. Dr. Rizzo said that restoring vision might help patients locate sounds "by increasing their capacity for orientation in space."
Researchers had more to work with than what the patients told them. The patients were scanned with functional MRI to see what The researchers also used functional MRI to measure activity in the back of the brain, where we process images.
Sure enough, with the implants in place, showing the patient flashes of black or white rectangles triggered increased activity in their brains.
The bad news is that an error in one patient's surgery resulted in a shallow choroidal detachment. That's where the choroid, a layer with blood vessels and connective tissue that's just behind the retina, separates from the sclera, or white of the eye.
The MedScape article says that the Conformité Européenne ("European Conformity") approved the Argus II and Alpha IMS for use in Europe. I gather that CE has something to do with the CE mark we see. Apparently it's a European analog of America's FDA.
Background:
- "Argus®
II Retinal Prosthesis System Patient Manual" (redacted)
Second Sight, via FDA
(From www.fda.gov/downloads/AdvisoryCommittees/CommitteesMeetingMaterials/MedicalDevices/MedicalDevicesAdvisoryCommittee/OphthalmicDevicesPanel/UCM320423.pdf (February 24, 2013))
2. Leaving the 1940s Behind: Heavily Interconnected, Biologically-Inspired Chip
(From IBM, via BBC News, used w/o permission.)
("The chips can also be connected together to provide even more computational power"
(BBC News))
"Brain-inspired chip fits 1m 'neurons' on postage stamp"Computers aren't nearly as huge as they were in the 1940s and '50s, but the basic design hasn't changed. Data gets stored in one place, moved to another for processing, and moved back to storage for the next operation.
Jonathan Webb, BBC News (August 8, 2014)
"Scientists have produced a new computer chip that mimics the organisation of the brain, and squeezed in one million computational units called 'neurons'.
"They describe it as a supercomputer the size of a postage stamp.
"Each neuron on the chip connects to 256 others, and together they can pick out the key features in a visual scene in real time, using very little power.
"The design is the result of a long-running collaboration, led by IBM, and is published in the journal Science....
"...[Dr Dharmendra Modha, the publication's senior author] told BBC News the processor was 'a new machine for a new era'. But it will take some time for the chip, dubbed TrueNorth, to be commercially useful...."
That's a good way to handle sequential operations, but it's not how our brains work. Not quite, anyway.
Details in machine code, assembly language, and comparatively user-friendly languages like Autocode and Fortran have changed in the last half-century. But they all deal with hardware that works like the old Manchester Mark 1.
John von Neumann described a way to make computers work in 1945. Folks like Eckert and Mauchly used 'Von Neumann architecture' for early computers like Electronic Discrete Variable Automatic Computer (EDVAC) and Electronic Numerical Integrator And Computer (ENIAC). Nearly seven decades later, nearly all computers are still designed along the lines von Neumann described.
Computing with Spikes, Not Ones and Zeroes
(From IBM, via BBC News, used w/o permission.)
("Each of 4,096 'neurosynaptic cores' on the chip contains 256 neurons, which connect to 256 other neurons within and outside that core; this diagram illustrates links between just 64 cores"
(BBC News))
"...the heavily interconnected structure of biologically-inspired, 'neuromorphic' systems like TrueNorth is said to be a much more efficient way of handling a lot of data at the same time.A human brain has upwards of 86,000,000,000 neurons. About one fifth of these are in our cerebral cortex, where we do most of our thinking.
" 'Our chip integrates computation, communication and memory very closely,' Dr Modha said.
"Instead of binary ones and zeros, the units of computation here are spikes. When its inputs are active enough, one of TrueNorth's 'neurons' generates a spike and sends it across the chip to other neurons, taking them closer to their own threshold....
"The building blocks for the TrueNorth chip are 'neurosynaptic cores' of 256 neurons each, which IBM launched in 2011.
"Dr Modha and his team managed to engineer an interconnected 64-by-64 grid of these cores on to a single chip, delivering over one million neurons in total.
"Because each neuron is connected to 256 others, there are more than 256 million connections or 'synapses'.
"This complexity is impressive for a man-made device just 3cm across, but still pales in comparison with the organ it emulates. Biological neurons, packed inside the brain, send and receive something in the order of 10,000 connections each...."
(Jonathan Webb, BBC News)
As Jonathan Webb pointed out, each of our neurons has thousands of connections to other neurons: so although the TrueNorth chip is remarkable: it's not even close to being on a par with our brains' hardware.
Connections inside IBM's neurosynaptic cores are more nearly similar to neural networks in our heads than connections in my desktop's processors. The key phrase is "more nearly," though. We're not even close to production models of robots like C-3PO or the Stepford Wives: my opinion.
Even so, I'm pretty sure that it's just a matter of time before some writer decides to 'warn' the public that a cabal of TrueNorth chips may plot a robot apocalypse.
It wouldn't be the first time that media coverage was more imaginative than accurate. I've read that the June, 1946, Mechanix Illustrated magazine featured a picture of ENIAC superimposed on a human brain. ("ENIAC: The Press Conference That Shook the World")
But after nearly seven decades and the Y2K crisis, computers and their software are still occasionally-glitchy tools: not feared overlords.
New Tech: "Time Will Tell"
"...Dr Modha envisages myriad next-generation applications, from glasses that help visually impaired people navigate, to robots for scouring the scene of a disaster.I think Dr. Modha and Professor Furber are both right. New information technology should make prosthetic like Second Sight's Argus II smaller, lighter, and more effective. But it's early days: my guess is that we're decades, not years, away from truly autonomous disaster-response robots.
"But some of the gains might be overstated - or perhaps too eagerly anticipated.
"Prof Steve Furber is a computer engineer at the University of Manchester who works on a similarly ambitious brain simulation project called SpiNNaker. That initiative uses a more flexible strategy, where the connections between neurons are not hard-wired.
"He told BBC News that 'time will tell' which strategy succeeds in different applications...."
(Jonathan Webb, BBC News)
I'll be astounded if robots ever entirely replace search and rescue dogs. Dogs are very good at sniffing out people, and arguably more companionable than robots. So far, anyway.
It's likely, though, that we'll develop rescue robots for high-risk environments. Building and programming a fireproof robot should be easier than training a dog to use a hazmat suit.
Now, in case you still haven't had enough of my take on tech:
- "Fear, Foreboding, and Getting a Grip About Technology"
(April 27, 2014)
Particularly - "Getting a Grip About Dr. Moreau, Pigs, and Human Dignity"
(February 14, 2014)
Particularly - "Robots, a Martian Dune, and Mapping a Billion Stars"
(May 10, 2013)
Particularly - "Popes; Earth's Past; and a High-Tech Eye Implant"
(March 1, 2013)
Particularly - "Spaceships, Robots, and Being Catholic"
(April 12, 2013)
Particularly
- Neuromorphic engineering
Wikipedia - 2014 in science, August
Wikipedia - "ENIAC: The Press Conference That Shook the World"
Dr. C. Dianne Martin, IEEE Technology and Society Magazine (December, 1995)
(From www.seas.gwu.edu/~mfeldman/csci110/summer07/eniac2.pdf (August 13, 2014))
No comments:
Post a Comment