David Brin constructed a future in his novel Earth wherein nearly everyone wore continuously-recording TV camera/VCR "glasses" all the time, and privacy was considered vaguely obscene. While I can see the usefulness of this device from the standpoint of the ratings biz, it's a bit surprising to see it casually accepted by end-user snoopees.
I can see the events reconstructed by one of these watches...
7:30 PM radio WRNO (car)
8:15 PM disc jockey at Club Vinyl
8:47 PM conversation with female acquaintance
9:05 PM radio WRNO (car) and conversation
10:15 PM Bolero from CD
10:27 PM watch taken off
Hmmmm, I wonder if it keeps listening after you take it off too...
Whether the owner of a trademark can stop others from using it depends on such factors as:
whether the trademark is being used on competing goods or services (goods or services compete if the sale of one is likely to affect the sale of the other)
whether consumers would likely be confused by the dual use of the trademark, and
whether the trademark is being used in the same part of the country or is being used on related goods (goods that will likely be noticed by the same customers, even if they don't compete with each other).
I had a friend who bought two Philips Velos. Nifty little box for its day. He used the first one for about two weeks, then it crashed. Being embarrassed, having too much money to spend, and having liked what it did, he bought the second one, used it for about two weeks, and it crashed.
Enter Yrs. Truly. I can't get one working at all. Critical files are missing from the flash; they are not present on the CD, not available on the Web, and in general there seems to be no way to recover them. I carefully reconstitute the second unit, get it set up again to access his internet account, etc. It sits up for a week while I wait for him to come get it, during which time the primary battery dies. When I power it up, it crashes, and (get this) when I go to set up the internet account again, the user interfaces are different and no longer conform to the instructions I got from the Web for setting it up. I have yet to figure that one out.
In short, WinCE manages to bring to the palm everything we've come to love in Windows desktop apps: their bloat, their instability, and their occasional need for a fresh reinstall from scratch. Except that, unlike the desktop, you don't have a standard way of doing the reinstall.
Meanwhile, the fridge, stove, and toaster never crash on me...
can you really trademark 3 arbitrary letters which make up a common generic English word?
Yes, that is the essence of a trademark. You can enforce that trademark when it is used with regard to a particular service or business. And it develops that Aimster is a service very similar in some respects (even compatible with) AOL's AIM. It's not just about "confusion," but about implied endorsement and the value of the brand.
We need a new law. One which does not permit sophist parasites such as the legal profession to live off its back.
And how exactly would this work? I would agree that we have too many lawyers in America today, and the law is too byzantine and obscure so that they are more necessary, but do you really think it is possible for human beings to implement a legal system of any kind with nobody at all performing such a function?
No matter how clear and simple the law is, if there is a procedure for administering it (such as our network of hearings and trials) then someone will have to advise plaintiffs and defendants on how to adhere to that procedure, since most of us won't know all the ins and outs.
And even the clearest law may not be understood by everybody. Lots of people seem inclined to interpret the law one way when they are the aggrieved party but another way when they are being sued. It's the job of a lawyer to explain these folks' true position to them, so they can avoid going to trial and getting whacked because they didn't understand the situation (or, alternately, so they can sue for just compensation which they might not have realized could be theirs).
This doesn't mean we need a system like we have, but try to be realistic. When Shakespeare wrote "first, we kill all the lawyers" those characters were planning a revolution; this was recognition that lawyers are a linchpin in the rule of law, and the first step in subverting that rule of law would have to be to get rid of them.
Of course, I suppose we could try anarchy, but I don't think that would work too well with our population and our level of technology. One day all those wunnerful feedback mechanisms that drive the invisible hand of the market would overshoot and we'd end up living in caves again. So, have you some other idea? Just wondering...
There is an article in Wired about Aimster and much to my surprise the name ccomes from the author's daughter Aimee.
And of course we all know that (as they actually argued in court) the chain restaurant Hooters refers to the sound made by their owl mascot.
What's going on in The Bridge (spoiler alert)
on
The Business
·
· Score: 2
The entire novel takes place inside the guy's head. He is in a coma and deteriorating. Several details of the dream have bled in from reality (the circular mark on his chest is the bruise left by the steering wheel when he had the accident, etc.).
My favorite line is in the list of reflections he is making as he tries to decide whether to let himself sink further into the fantasy/coma or struggle out into the daylight world and the pain of his inevitable recuperation. He begins thinking of all the things he hasn't done, and finishes off with going to bed with three women at once!
It is distinctly implied that this thought pushes him over the edge into choosing life, heh.
Re:Iain Banks without the M
on
The Business
·
· Score: 2
The Bridge -- a tale told from the first person interior thought-space of a man in a coma
The Wasp Factory -- how can you resist something billed by one reviewer as "the most perverse book ever written?"
But district Superintendent John Fitzsimons said school officials followed disciplinary policies in this case, and although teachers and administrators are grieving the loss, they aren't responsible
...but not sorry enough to admit to our responsibility and try to change things.
Homeschooling. It's the only answer.
Thank Eris I prudently decided not to have kids myself. I thought my childhood was a living hell, but I had no idea. Sheesh.
You see, QBasic is what we call "native" to Windows which means it runs at a core speed which scales with the OS (operating system) kernel.
Actually Qbasic is native to Dos and breaks in significant ways under W9x, unless you boot in DOS mode. And some of the functions (serial ports, anyone) don't work at all in NT.
Incidentally, my experience with QuickBasic (the upscale non-bundled superset of Qbasic, with compiler) is that it generally makes faster code than C++ (no dereferencing to support OO).
Oh, and you forgot to work threading and XMP into your screed.
And most importantly: QBasic and QuickBasic are the development environments of choice for anyone who is hardware hacking bits out of the parallel port. (I am, in fact, embarking on such a project soon myself, to control a stepper motor which will drive my rock saw. An old 386SX16, dedumpsterized flat-panel VGA display with touchscreen, and and a little solder and WHEEE I won't have to run the damn thing by hand any more hehehe.)
Back before computers got fast enough to do 3D screensavers quite a few screensavers were written this way. I did one which is quite fun to watch while drinking beer that drew random lines on the text display, in random ASCII characters; it would draw a random number (20 to 50) in each character before cycling, so that characters would "splat" the screen. Originally wrote it to test my Bresenham's Algorithm code without switching into graphics mode. Was very engaging, and ran on a 8088. You can waste beaucoup time playing with the parameters of an algorithm like this to get it to look "right.":-)
The only reason you can assume that there is "no emotional hook" is that you know it is created by a computer.
No, what I mean is that there is no spark, no apparent purpose in the interaction of the women. It is quite possible to encode such meaning into line drawings like these, and such meaning is absent. Aaron gave the three human characters essentially random expressions, and it shows; they are not neutral enough to be passed off as an illustration, such as one might find in an advertisement; but neither do they make any sense.
Anyway, even if the argument can be made that the pantings elicit "no emotional involvement", that just makes it bad art, but art nonetheless.
That makes it illustration, as considered distinct from art within the art world. A good example of a human artist on the edge of the distinction is Andrew Wyeth. While there isn't general agreement on what specific works constitute art there is a pattern of agreement in the art community that a work must affect somebody in a meaningful way, even if you don't personally share their feelings, in order to qualify as art.
The problem with Aaron is that, while it may in fact have a great deal of "knowledge" about how to place objects (including human figures) in its "imaginative" world it obviously does not have any concept of how they interact on anything except the bare physical level.
This is absolute bullshit. Legal codes, like unsealed court documents, are in the public domain. It is incredible that not one but two courts have upheld this foolishness.
It doesn't matter how well it is done, how much work went into it, or whatever; once it is a matter of law it is the State's duty to make it as freely available as possible. Otherwise one has the Kafka-esque situation of being required to comply with laws which will only be revealed to you for a price.
Back in the 80's I had a cartoon hanging on my bulletin board which IIRC originally ran in Isaac Asimov's Science Fiction Magazine. It showed several very concerned and professional people standing behind a strange contraption which was aimed at a painting on the wall. An ad-like blurb read: IS IT ART? The Art Detector Can Tell!. The humor was obvious, and this piece of software doesn't impress me.
Cohen's 1.5mloc obviously implement a model of the world from which it can select various groupings of objects, allowed to interact in various ways, and portray them in a way which has been come to be considered "artistic." But that doesn't make it art.
Consider the picture displayed with the story; it shows three women apparently getting ready to play tennis. What is the emotional hook here? What are the women thinking? What is the higher reality behind their situation? The image looks "artsy" but elicits no emotional involvement. It is what some hoity-toity art types would pass off as an illustration, and a pretty murky one at that.
Now I am about the furthest thing from an Art Snob you will ever find, but I will always remember the time a high-roller friend took me to dinner at Melange, at the time the restaurant at the Mirage where Steve Wynn had his personal collection of original Picassos on display. I barely noticed the food for the power of the art; we circled the restaurant as our food was prepared, examining each one. The images were bold and powerful and filled with energy; they almost seemed to leap off the canvas. It was not always obvious what Picasso's abstractions represented but one could sense in every one of them a concentration of power, or sexuality, or awe; some of them left you thinking for hours, teasing multiple meanings out of the symbology used. Picasso was also fond of piling on the paint, and in person his works have a three-dimensionality which does not come across even in lithographically reproduced art books. That evening, for the first time, I realized what art was about.
I look again at the picture of the tennis players, and nope, it doesn't do it. The program does not really have emotions to attempt to represent; it merely throws things together and illustrates them, without any sense of a dynamic behind the scene. An artist would have made something of this image -- you would look at it and know who hated whom, who was worried about their performance, maybe who was having an affair with the others' man. But there is nothing like that. Ho hum.
I would suspect the primary enforcement is that, if you break the agreement, you won't get invited to any more early screenings. If a critic were to piss off enough studios in this manner he might soon find himself hard-pressed to find material to review.
Nice starting point, actually everyone's starting point because it's the most easily studied via ablation and point-scan type studies.
When you were born as a blank IBM RAM chip, who taught your eyes to differeniate a luminosity function and form egdes?
You get to the gist: I didn't.. I learned those things in the first hours/days after my eyes opened, or (alternately) after the wiring to my cerebral cortex became myelinized enough to permit the learning. How did I learn it? By looking at stuff..
We are not born knowing how to, for example, detect short line segments or connect them into shapes or detect luminosity variations and do derivatives on them. This is the fundamental error of current AI research. How do I (O great swami, I hear you say)) arrive at this startling conclusion?
It's simple, but important. THERE IS NO PLACE IN THE GENETIC CODE FOR THE INFORMATION TO GUIDE SUCH WIRING.
Please go back to CSCI 100, re-read Knuth or whatever, until you grok that last statement, because it is really, really important.
OK, we were discussing this at work Friday but it's a major miss. Imagine what the 50's droids would have thought of an actual working aphrodisiac that is covered by medical insurance.
It is interesting to see the perspective of people who were not around at the time.
I was around at the time, and actually used all the computers I mentioned. Your comments about IBM teaching uSoft its trader are on-target. Development in C was unheard of in the microcomputer industry, which is what we were talking about. Yes, there were some advanced compilers (for their time) running on mainframes. This had no relevance to the microcomputer industry, which grew out of developments that could be made on microcomputers.
You're incorrect, Daisy. The brain is enormously pre-wired.
Since you are so knowledgeable on the subject, perhaps you can supply me with the wiring diagram?
I thought not. Your attitude is symptomatic of why no progress is being made in the field. While there are a few tendencies which may be pre-programmed, which act as relatively poor predictors of future behavior, the vast array of data structures which make up our personalities are acquired through life. It's the only place they can come from -- regardless of your religious views on the subject (and IME most believers in social darwinism have a nearly religious fixation on the idea), there is simply no room in the genetic code for those properties to be coded in a non-emergent way.
(P.S. Emergent means they emerge like the features of a fractal; small changes in the code result in large, often crippling, changes in the result, making them non-evolvable.)
The problem with "signal-based systems" like the brain is their inherent chaos -- which I mean in the strict mathematical sense, as in Mandelbrot et al. While this can make them very useful it also makes them fundamentally unpredictable, a characteristic which bean counters don't like.
One reason AI research isn't going anywhere is that we are failing to face up to an important truth about how brains develop. Since they program themselves, starting with really very little seed information, most of their observable properties are emergent. The same would be true of any artificial system that really mimics the brain. The reason we don't have good AI isn't that the hardware isn't good enough -- I think it is, at this point -- it's that nobody wants such a system. Imagine educating your self-driving car for six years only to find out it's become a chance-taking rebellious delinquent!
You mention "message-based communication between objects" as if those objects will somehow know how to talk to one another. Clue time: They don't know how to talk to one another until they learn. And they learn through experience. Sometimes their learning is imperfect, and it can be very difficult to recognize the holes in that learning. We don't even know how to reliably program our own children, much less an air traffic control system that will differ significantly from all known types of brain, animal and human.
The newly formed brain is every bit as blank as the newly powered-up dynamic RAM chip -- anybody with an ounce of objectivity can look at the 7 Gb genome vs. the 10 ex 14 connectivity of the cerebral cortex and figure that out. Do you really want to go into the cyber day-care business, teaching your machine to speak english and habla the espanol and so forth the same way human babies learn it? Of course not. Since that's what it takes to do it the way you mention, it won't be done that way.
Don't laugh, this was a big deal in 1981. At the time outfitting a box with the theoretical maximum of 640K would set you back several thousand bucks and an outboard power supply. At the time 64Kx1 RAM chips had just been introduced and ran IIRC somewhere around $200 per set of 9.
2. It was not a Zilog chip.
Zilog made the Z-80, used by then-big competitor Tandy in their machines. IBM did not want their product confused with that of a major competitor.
3. It was not a Motorola chip.
Motorola made the 6502, used by then-big competitor Apple in their machines. IBM did not want their product confused with that of a major competitor.
4. Rich instruction set
In those days before the religion of RISC had been formed, much important code was written directly in assembly language. The luxury of writing the operating system in C was not practical on a 4 MHz processor with an 8-bit bus to memory. The x86 set was designed to be used by humans, not compilers. In these days of fast CPUs and optimized computers this is considered a Bad Thing, but in those days with a rich lode of 8080 assembly source code waiting to be ported and legions of programmers familiar with the techniques it was considered a Good Thing.
In many ways the 8088 was ahead of its time. Early MS-DOS code was notably bloated compared to code written for 8-bit platforms like the Z-80, and the first wave of PC's were terribly underpowered to be doing anything useful. The PC succeeded in those early days only because of its IBM brand; it was at least 1984 before the XT with its hard drive and AT with some actual CPU performance became serious competitors to low-end machines like the C64, TRS-80, and Apple II. Businesses kept the standard going until it became commodified, the price of useful systems came down, and its ability to do personal type applications (read: games) began to actually approach that of its competitors.
It's worth noting that what Rambus got slammed for was lying to the JEDEC committee about whether it planned to patent the technologies it was proposing for the standard. Unfortunately, this does not touch on their right to sue everyone who ever manufactured a RAM chip for patent infringement.
I personally see patents as being a useful and even necessary evil in a capitalist world, but our own patent office has gone completely off its rocker in the last couple of decades. This bitchslap would have been much more satisfying if it had been for the fraudulent act of suing everyone in sight over something that obviously should never have been patentable in the first place since it was clearly described in prior art.
Slashdot Mirror
I can see the events reconstructed by one of these watches...
- 7:30 PM radio WRNO (car)
- 8:15 PM disc jockey at Club Vinyl
- 8:47 PM conversation with female acquaintance
- 9:05 PM radio WRNO (car) and conversation
- 10:15 PM Bolero from CD
- 10:27 PM watch taken off
Hmmmm, I wonder if it keeps listening after you take it off too...Whether the owner of a trademark can stop others from using it depends on such factors as:
Any more questions?
Enter Yrs. Truly. I can't get one working at all. Critical files are missing from the flash; they are not present on the CD, not available on the Web, and in general there seems to be no way to recover them. I carefully reconstitute the second unit, get it set up again to access his internet account, etc. It sits up for a week while I wait for him to come get it, during which time the primary battery dies. When I power it up, it crashes, and (get this) when I go to set up the internet account again, the user interfaces are different and no longer conform to the instructions I got from the Web for setting it up. I have yet to figure that one out.
In short, WinCE manages to bring to the palm everything we've come to love in Windows desktop apps: their bloat, their instability, and their occasional need for a fresh reinstall from scratch. Except that, unlike the desktop, you don't have a standard way of doing the reinstall.
Meanwhile, the fridge, stove, and toaster never crash on me...
Yes, that is the essence of a trademark. You can enforce that trademark when it is used with regard to a particular service or business. And it develops that Aimster is a service very similar in some respects (even compatible with) AOL's AIM. It's not just about "confusion," but about implied endorsement and the value of the brand.
I am thinking A1 steak sauce should now sue AOL.
Why? AOL doesn't make steak sauce.
And how exactly would this work? I would agree that we have too many lawyers in America today, and the law is too byzantine and obscure so that they are more necessary, but do you really think it is possible for human beings to implement a legal system of any kind with nobody at all performing such a function?
No matter how clear and simple the law is, if there is a procedure for administering it (such as our network of hearings and trials) then someone will have to advise plaintiffs and defendants on how to adhere to that procedure, since most of us won't know all the ins and outs.
And even the clearest law may not be understood by everybody. Lots of people seem inclined to interpret the law one way when they are the aggrieved party but another way when they are being sued. It's the job of a lawyer to explain these folks' true position to them, so they can avoid going to trial and getting whacked because they didn't understand the situation (or, alternately, so they can sue for just compensation which they might not have realized could be theirs).
This doesn't mean we need a system like we have, but try to be realistic. When Shakespeare wrote "first, we kill all the lawyers" those characters were planning a revolution; this was recognition that lawyers are a linchpin in the rule of law, and the first step in subverting that rule of law would have to be to get rid of them.
Of course, I suppose we could try anarchy, but I don't think that would work too well with our population and our level of technology. One day all those wunnerful feedback mechanisms that drive the invisible hand of the market would overshoot and we'd end up living in caves again. So, have you some other idea? Just wondering...
And of course we all know that (as they actually argued in court) the chain restaurant Hooters refers to the sound made by their owl mascot.
My favorite line is in the list of reflections he is making as he tries to decide whether to let himself sink further into the fantasy/coma or struggle out into the daylight world and the pain of his inevitable recuperation. He begins thinking of all the things he hasn't done, and finishes off with going to bed with three women at once!
It is distinctly implied that this thought pushes him over the edge into choosing life, heh.
The Wasp Factory -- how can you resist something billed by one reviewer as "the most perverse book ever written?"
Or is it redirected, possibly to some place in Louisiana like in David Brin's Earth?
Homeschooling. It's the only answer.
Thank Eris I prudently decided not to have kids myself. I thought my childhood was a living hell, but I had no idea. Sheesh.
Actually Qbasic is native to Dos and breaks in significant ways under W9x, unless you boot in DOS mode. And some of the functions (serial ports, anyone) don't work at all in NT.
Incidentally, my experience with QuickBasic (the upscale non-bundled superset of Qbasic, with compiler) is that it generally makes faster code than C++ (no dereferencing to support OO).
Oh, and you forgot to work threading and XMP into your screed.
And most importantly: QBasic and QuickBasic are the development environments of choice for anyone who is hardware hacking bits out of the parallel port. (I am, in fact, embarking on such a project soon myself, to control a stepper motor which will drive my rock saw. An old 386SX16, dedumpsterized flat-panel VGA display with touchscreen, and and a little solder and WHEEE I won't have to run the damn thing by hand any more hehehe.)
while pattern_not_entirely_covered
WendBack before computers got fast enough to do 3D screensavers quite a few screensavers were written this way. I did one which is quite fun to watch while drinking beer that drew random lines on the text display, in random ASCII characters; it would draw a random number (20 to 50) in each character before cycling, so that characters would "splat" the screen. Originally wrote it to test my Bresenham's Algorithm code without switching into graphics mode. Was very engaging, and ran on a 8088. You can waste beaucoup time playing with the parameters of an algorithm like this to get it to look "right." :-)
No, what I mean is that there is no spark, no apparent purpose in the interaction of the women. It is quite possible to encode such meaning into line drawings like these, and such meaning is absent. Aaron gave the three human characters essentially random expressions, and it shows; they are not neutral enough to be passed off as an illustration, such as one might find in an advertisement; but neither do they make any sense.
Anyway, even if the argument can be made that the pantings elicit "no emotional involvement", that just makes it bad art, but art nonetheless.
That makes it illustration, as considered distinct from art within the art world. A good example of a human artist on the edge of the distinction is Andrew Wyeth. While there isn't general agreement on what specific works constitute art there is a pattern of agreement in the art community that a work must affect somebody in a meaningful way, even if you don't personally share their feelings, in order to qualify as art.
The problem with Aaron is that, while it may in fact have a great deal of "knowledge" about how to place objects (including human figures) in its "imaginative" world it obviously does not have any concept of how they interact on anything except the bare physical level.
It doesn't matter how well it is done, how much work went into it, or whatever; once it is a matter of law it is the State's duty to make it as freely available as possible. Otherwise one has the Kafka-esque situation of being required to comply with laws which will only be revealed to you for a price.
Cohen's 1.5mloc obviously implement a model of the world from which it can select various groupings of objects, allowed to interact in various ways, and portray them in a way which has been come to be considered "artistic." But that doesn't make it art.
Consider the picture displayed with the story; it shows three women apparently getting ready to play tennis. What is the emotional hook here? What are the women thinking? What is the higher reality behind their situation? The image looks "artsy" but elicits no emotional involvement. It is what some hoity-toity art types would pass off as an illustration, and a pretty murky one at that.
Now I am about the furthest thing from an Art Snob you will ever find, but I will always remember the time a high-roller friend took me to dinner at Melange, at the time the restaurant at the Mirage where Steve Wynn had his personal collection of original Picassos on display. I barely noticed the food for the power of the art; we circled the restaurant as our food was prepared, examining each one. The images were bold and powerful and filled with energy; they almost seemed to leap off the canvas. It was not always obvious what Picasso's abstractions represented but one could sense in every one of them a concentration of power, or sexuality, or awe; some of them left you thinking for hours, teasing multiple meanings out of the symbology used. Picasso was also fond of piling on the paint, and in person his works have a three-dimensionality which does not come across even in lithographically reproduced art books. That evening, for the first time, I realized what art was about.
I look again at the picture of the tennis players, and nope, it doesn't do it. The program does not really have emotions to attempt to represent; it merely throws things together and illustrates them, without any sense of a dynamic behind the scene. An artist would have made something of this image -- you would look at it and know who hated whom, who was worried about their performance, maybe who was having an affair with the others' man. But there is nothing like that. Ho hum.
One day we will have AI, but this ain't it.
I would suspect the primary enforcement is that, if you break the agreement, you won't get invited to any more early screenings. If a critic were to piss off enough studios in this manner he might soon find himself hard-pressed to find material to review.
Nice starting point, actually everyone's starting point because it's the most easily studied via ablation and point-scan type studies.
When you were born as a blank IBM RAM chip, who taught your eyes to differeniate a luminosity function and form egdes?
You get to the gist: I didn't.. I learned those things in the first hours/days after my eyes opened, or (alternately) after the wiring to my cerebral cortex became myelinized enough to permit the learning. How did I learn it? By looking at stuff..
We are not born knowing how to, for example, detect short line segments or connect them into shapes or detect luminosity variations and do derivatives on them. This is the fundamental error of current AI research. How do I (O great swami, I hear you say)) arrive at this startling conclusion?
It's simple, but important. THERE IS NO PLACE IN THE GENETIC CODE FOR THE INFORMATION TO GUIDE SUCH WIRING.
Please go back to CSCI 100, re-read Knuth or whatever, until you grok that last statement, because it is really, really important.
OK, we were discussing this at work Friday but it's a major miss. Imagine what the 50's droids would have thought of an actual working aphrodisiac that is covered by medical insurance.
...without exception. Especially loved the TWADDLE acronym.
I was around at the time, and actually used all the computers I mentioned. Your comments about IBM teaching uSoft its trader are on-target. Development in C was unheard of in the microcomputer industry, which is what we were talking about. Yes, there were some advanced compilers (for their time) running on mainframes. This had no relevance to the microcomputer industry, which grew out of developments that could be made on microcomputers.
Since you are so knowledgeable on the subject, perhaps you can supply me with the wiring diagram?
I thought not. Your attitude is symptomatic of why no progress is being made in the field. While there are a few tendencies which may be pre-programmed, which act as relatively poor predictors of future behavior, the vast array of data structures which make up our personalities are acquired through life. It's the only place they can come from -- regardless of your religious views on the subject (and IME most believers in social darwinism have a nearly religious fixation on the idea), there is simply no room in the genetic code for those properties to be coded in a non-emergent way.
(P.S. Emergent means they emerge like the features of a fractal; small changes in the code result in large, often crippling, changes in the result, making them non-evolvable.)
One reason AI research isn't going anywhere is that we are failing to face up to an important truth about how brains develop. Since they program themselves, starting with really very little seed information, most of their observable properties are emergent. The same would be true of any artificial system that really mimics the brain. The reason we don't have good AI isn't that the hardware isn't good enough -- I think it is, at this point -- it's that nobody wants such a system. Imagine educating your self-driving car for six years only to find out it's become a chance-taking rebellious delinquent!
You mention "message-based communication between objects" as if those objects will somehow know how to talk to one another. Clue time: They don't know how to talk to one another until they learn. And they learn through experience. Sometimes their learning is imperfect, and it can be very difficult to recognize the holes in that learning. We don't even know how to reliably program our own children, much less an air traffic control system that will differ significantly from all known types of brain, animal and human.
The newly formed brain is every bit as blank as the newly powered-up dynamic RAM chip -- anybody with an ounce of objectivity can look at the 7 Gb genome vs. the 10 ex 14 connectivity of the cerebral cortex and figure that out. Do you really want to go into the cyber day-care business, teaching your machine to speak english and habla the espanol and so forth the same way human babies learn it? Of course not. Since that's what it takes to do it the way you mention, it won't be done that way.
At least, not by most people (sly grin).
1. 20-bit addressing.
Don't laugh, this was a big deal in 1981. At the time outfitting a box with the theoretical maximum of 640K would set you back several thousand bucks and an outboard power supply. At the time 64Kx1 RAM chips had just been introduced and ran IIRC somewhere around $200 per set of 9.
2. It was not a Zilog chip.
Zilog made the Z-80, used by then-big competitor Tandy in their machines. IBM did not want their product confused with that of a major competitor.
3. It was not a Motorola chip.
Motorola made the 6502, used by then-big competitor Apple in their machines. IBM did not want their product confused with that of a major competitor.
4. Rich instruction set
In those days before the religion of RISC had been formed, much important code was written directly in assembly language. The luxury of writing the operating system in C was not practical on a 4 MHz processor with an 8-bit bus to memory. The x86 set was designed to be used by humans, not compilers. In these days of fast CPUs and optimized computers this is considered a Bad Thing, but in those days with a rich lode of 8080 assembly source code waiting to be ported and legions of programmers familiar with the techniques it was considered a Good Thing.
In many ways the 8088 was ahead of its time. Early MS-DOS code was notably bloated compared to code written for 8-bit platforms like the Z-80, and the first wave of PC's were terribly underpowered to be doing anything useful. The PC succeeded in those early days only because of its IBM brand; it was at least 1984 before the XT with its hard drive and AT with some actual CPU performance became serious competitors to low-end machines like the C64, TRS-80, and Apple II. Businesses kept the standard going until it became commodified, the price of useful systems came down, and its ability to do personal type applications (read: games) began to actually approach that of its competitors.
TSIA. It's achievable if you work at it -- though why anyone would try is a mystery to me.
I personally see patents as being a useful and even necessary evil in a capitalist world, but our own patent office has gone completely off its rocker in the last couple of decades. This bitchslap would have been much more satisfying if it had been for the fraudulent act of suing everyone in sight over something that obviously should never have been patentable in the first place since it was clearly described in prior art.