An interesting piece from 2003 about changing the fundamental object of computation from points to surfaces by Jaron Lanier. IMO what deep learning is semi-blindly doing, but 18 years later we could still be more purposeful.
Are there any specific ideas introduced in this essay that are relevant today?
While essay describes some true algorithms (like wavelet compression), they are not invented by Lanier nor they need any of his "surface" stuff to be used. And Lanier's predictions, like the whole "phenotropic software" ideas, seem completely academic, for example:
> When one module connects to another, it would use the same techniques a biomimetic robot would use to get around in the messy, unpredictable physical world.
I think Daniel Dennet's comment sums it very well:
> he's doing exactly the thing he chastises the early AI community for doing: getting starry-eyed about a toy model that might—might—scale up and might not. There are a few interesting ideas in his ramblings, but it's his job to clean them up and present them in some sort of proper marching order, not ours. Until he does this, there's nothing to reply to.
I agree it isn't a very well supported piece, but I think it's still quite provocative. Here's a comment from Lee Smolin written at the time of the article:
> Reading the critics of Jaron Lanier's essay, in which he speculates about a new form of a computer, based on different principles than those that underlie the standard programmable digital computer, I wonder how people might have reacted, shortly after the invention of the wheel, if some ancestor of Jaron had proposed to invent a new form of transportation that was not a wheel. "Not a wheel!" one can hear them snorting. "Why everyone knows that any device to convey goods must depend on some arrangement of wheels. Not only that, the great thinker van N proved that any arrangement of wheels, whether in parallel or in serial, is equivalent to a single larger wheel, in terms of its ability to move goods."
> "No," said the clearly frustrated proto-Jaron, "What I have in mind does involve lashing some logs together, but instead of rolling them, my idea is to put them into the river and simply put the goods on top and float them down to the next camp. So no wheels, and no need to abide by the great van N's theorem on wheel capacity."
> The answer then must have been, "Well, we've never heard of such a thing, but try it and see if it works." It seems to me that that's what Jaron's critics might be saying to him, instead of arguing that a boat, as a form of transportation, must roll on wheels.
> So it seems to me the question being debated can be framed like this: Is a computer something like a wheel? Is there really only one kind of computer, just like there is really only one kind of wheel? One can arrange them in many ways, in series and in parallel, but in the end, once the wheel or the computer has been invented, they will all work the same way. Even millennia later, wheels are wheels, period. Or, is the computer something more general, like a mode of transportation or a musical instrument. There are many different kinds of musical instruments, which produce sound by means of many different principles. Is it possible that there are actually many different kinds of computers, which will accomplish informational tasks for us by as many different principles as musical instruments produce sounds? In that case, is the problem that the critics are beating their drums, while Jaron is trying to blow the first horn?
This analogy does not work for me, it seems like an strawman. Who are those people who think that wheels are the only form of transportation? I am pretty sure that both sliding and floating were invented before the wheel.
The main Lee Smolin's argument fails the same way: he asks, "is there really only kind of computer" like this is something people never asked before.. but this has question has been asked (and answered) many times. There are many ways to clarify that question, and answers to that are in basic CS courses. For example:
"Is there a different kind of computer hardware, not the same kind as my Windows PC made in 2003" -> Yes, the computational world is wide, and many computing device do not use sequential C-like programming. For example, there (were) analog computers, FPGA, special purpose accelerators. There are even super parallel ones like Green Arrays, which is kinda-sorta similar to what is described in the original article.
"Is there a different data model, other than bits" -> Yes, analog computers were a thing and had no bits. And some algorithms, like machine learning or fuzzy logic, emulate fuzzy values with bits.
"Is there a different kind of computation, different than Turing model" -> Yes, there are uncomputable problems, a halting problem is most famous of them.
Overall, I don't see this as profound, I see it as boring. There may be a genuine, interesting question out there, but this is not what Smolin / Lanier are asking.
> "Computer scientists hate, hate thinking about the loss of idealness that comes with scale."
Unusual interpretation, but if we fail to improve software robustness, reliability, and security, then I interpret this as "don't worry too much about securing your TCP/IP connection with the bank; twenty other forms of information leakage are occurring from the surface contacts involved".
This is unusual interpretation indeed, and I don't think it matches the reality.
Computer scientists don't "hate, hate thinking about the loss of idealness" - there are countless scientists who think about non-ideal cases. In fact, Lanier mentions Shannon in the first paragraph... and his big work was data transmission in non-ideal channels.
Same for your analogy -- I am yet to see anyone who would argue against securing TCP/IP connection with the bank, now or in the future. In fact, there is huge body of work about how to secure it properly, and how to keep it secure in the future (against software exploits, quantum computing etc...).
The only way those sentences make sense to me is if add "some" qualifiers to it: Yes, some scientists hate thinking about non-ideal conditions; and some developers might not worry about securing connections to the bank. But then this is trivially true -- there are lots of people out there, and there are certainly some people with any opinions. This is not worth the whole essay.
7 comments
[ 3.7 ms ] story [ 37.8 ms ] threadWhile essay describes some true algorithms (like wavelet compression), they are not invented by Lanier nor they need any of his "surface" stuff to be used. And Lanier's predictions, like the whole "phenotropic software" ideas, seem completely academic, for example:
> When one module connects to another, it would use the same techniques a biomimetic robot would use to get around in the messy, unpredictable physical world.
I think Daniel Dennet's comment sums it very well:
> he's doing exactly the thing he chastises the early AI community for doing: getting starry-eyed about a toy model that might—might—scale up and might not. There are a few interesting ideas in his ramblings, but it's his job to clean them up and present them in some sort of proper marching order, not ours. Until he does this, there's nothing to reply to.
> Reading the critics of Jaron Lanier's essay, in which he speculates about a new form of a computer, based on different principles than those that underlie the standard programmable digital computer, I wonder how people might have reacted, shortly after the invention of the wheel, if some ancestor of Jaron had proposed to invent a new form of transportation that was not a wheel. "Not a wheel!" one can hear them snorting. "Why everyone knows that any device to convey goods must depend on some arrangement of wheels. Not only that, the great thinker van N proved that any arrangement of wheels, whether in parallel or in serial, is equivalent to a single larger wheel, in terms of its ability to move goods."
> "No," said the clearly frustrated proto-Jaron, "What I have in mind does involve lashing some logs together, but instead of rolling them, my idea is to put them into the river and simply put the goods on top and float them down to the next camp. So no wheels, and no need to abide by the great van N's theorem on wheel capacity."
> The answer then must have been, "Well, we've never heard of such a thing, but try it and see if it works." It seems to me that that's what Jaron's critics might be saying to him, instead of arguing that a boat, as a form of transportation, must roll on wheels.
> So it seems to me the question being debated can be framed like this: Is a computer something like a wheel? Is there really only one kind of computer, just like there is really only one kind of wheel? One can arrange them in many ways, in series and in parallel, but in the end, once the wheel or the computer has been invented, they will all work the same way. Even millennia later, wheels are wheels, period. Or, is the computer something more general, like a mode of transportation or a musical instrument. There are many different kinds of musical instruments, which produce sound by means of many different principles. Is it possible that there are actually many different kinds of computers, which will accomplish informational tasks for us by as many different principles as musical instruments produce sounds? In that case, is the problem that the critics are beating their drums, while Jaron is trying to blow the first horn?
The main Lee Smolin's argument fails the same way: he asks, "is there really only kind of computer" like this is something people never asked before.. but this has question has been asked (and answered) many times. There are many ways to clarify that question, and answers to that are in basic CS courses. For example:
"Is there a different kind of computer hardware, not the same kind as my Windows PC made in 2003" -> Yes, the computational world is wide, and many computing device do not use sequential C-like programming. For example, there (were) analog computers, FPGA, special purpose accelerators. There are even super parallel ones like Green Arrays, which is kinda-sorta similar to what is described in the original article.
"Is there a different data model, other than bits" -> Yes, analog computers were a thing and had no bits. And some algorithms, like machine learning or fuzzy logic, emulate fuzzy values with bits.
"Is there a different kind of computation, different than Turing model" -> Yes, there are uncomputable problems, a halting problem is most famous of them.
Overall, I don't see this as profound, I see it as boring. There may be a genuine, interesting question out there, but this is not what Smolin / Lanier are asking.
> "Computer scientists hate, hate thinking about the loss of idealness that comes with scale."
Unusual interpretation, but if we fail to improve software robustness, reliability, and security, then I interpret this as "don't worry too much about securing your TCP/IP connection with the bank; twenty other forms of information leakage are occurring from the surface contacts involved".
Computer scientists don't "hate, hate thinking about the loss of idealness" - there are countless scientists who think about non-ideal cases. In fact, Lanier mentions Shannon in the first paragraph... and his big work was data transmission in non-ideal channels.
Same for your analogy -- I am yet to see anyone who would argue against securing TCP/IP connection with the bank, now or in the future. In fact, there is huge body of work about how to secure it properly, and how to keep it secure in the future (against software exploits, quantum computing etc...).
The only way those sentences make sense to me is if add "some" qualifiers to it: Yes, some scientists hate thinking about non-ideal conditions; and some developers might not worry about securing connections to the bank. But then this is trivially true -- there are lots of people out there, and there are certainly some people with any opinions. This is not worth the whole essay.
So if you are reading and thinking, "wait, this makes no sense"... scroll down to comments. You will see that your hunch is likely correct.