It seemed like a bad idea at the time

This past week, a couple of us techheads were shooting the breeze (yes, during midterms, and yes, the night before the algorithms midterm of death) and the turns of the conversation touched briefly on bringing programming to the masses, lingering for a few moments on things like hypercard and The Diamond Age.

The Diamond Age came up specifically because of it’s description of a Primer, a sort of interactive book meant to teach children. When I heard a brief description of it’s operation, I immediately thought “that’s the worst idea I have heard in a while”: trying to teach by making contrived examples which ended up functioning as glorified word problems just seemed like a sure way of turning off readers, no matter how many connections mined from the the reader’s life were included.

However, I was jonsing pretty hard for some good fiction, and there was a large chance that an offhand description of the Primer was much less appealing than the actual description in the book itself. So I entered the university library stacks for the first time, fetched myself The Diamond Age (TDA), sat down, and read the book for 12 hours. Then I noticed it was 4am, so I went to sleep, woke up, and then read the rest of it.

A quick one paragraph review of TDA: I caught some geek references, missed others, discovered I can’t read poetry worth crap, didn’t get many solid ideas for educational devices but saw a plausible mid-game, was challenged by keeping all the story threads in order, and thought the nanotech was cute. Overall, it was a good read, and I might delve back into it, if only to decipher that damn poem.

Okay, now details (general spoiler alert for the rest of the post, although it’s obvious how the book will end from fairly near the beginning, if that’s all you care about).

I picked up TDA primarily to get ideas about how to advance computer aided learning: at the time, I was concerned that adaptive teaching platforms would be able to subsume things like Notesoble, and that I shouldn’t pour effort into it if it was all going to come to naught (hell, I’m still concerned about it). Hence, I was casting around for ideas in the general pedagogical space when this came up. While I planned on focusing my reading to these aspects of the book, and hence hopefully getting some benefits from a focused reading, I eventually just let my fiction-reading engine take over and have some fun.

I should point out that TDA did not try to hold your hand. As a self-proclaimed high caliber geek, I noticed several references going over my head. For instance, I caught the shout-out to Lisp (which in itself is pretty obscure), but wasn’t entirely sure what the castle containing the open market was supposed to represent. The obvious cryptographic system interpretation didn’t fit very well: how was a network of cryptographic signers supposed to be Turing complete? Was this supposed to be a thought on how the internet was Turing complete? Obviously, I have a ways to go before I understand everything in TDA, and I may not even get there, since it probably includes discarded modes of thought and memes that are outside of my future, erm, culture cone.

A particular poem was quite vexacious, and I still don’t have a handle on it. I’m guessing that reading poetry is like reading code, or reading anything: reading natural languages like English tends to be faster than writing only because the read/write ratio of a normal English speaker is enormous. I suspect that if we had similar read/write ratios for things like code and poetry, while keeping similar write amounts, then one could read faster than one writes. It would never be as good, because natural language development starts from the womb, but it’s an interesting thought that I could read code faster than I write it, which is in direct opposition of some schools of thought. Anyways.

TDA was the first book in quite a while that forced me to look up words (EDIT: this is not necessarily a good thing); since Neal Stephenson wrote this, it was somewhat troublesome to sort through which words were actually words and which ones were made up literary cruft, but it was somewhat noticeable which ones were which, and generally made sense even if one didn’t look up the barrage of words. For instance, he definitely had a thing for brocades, which came up every 30 pages or so, but always in a non-essential description, so I only looked it up when I got to page 300 or so.

Coming back to computer-aided learning, I found most of the ideas contained in the book were not very applicable to our situation today. Stab your step-father through the eye? Hardly sound advice coming from an adaptive teaching platform.

However, I vividly remember the description of using the Primer to point at things and then read about them on the spot: something like a generalized Google Goggles with curated information. When I spoke of the mid-game, I was specifically referring to the amount of learning that could happen with this sort of device. Once we lower the cost of asking questions lower than pestering parents with questions, and especially lower than the cost of typing a question into a web browser, then all sorts of half-baked queries can be molded into fully fledged lines of inquiry.

However, the device shouldn’t function as a curiosity stopper, which it could certainly excel at. Socratic dialogs, or a heavily hyperlinked page, or a sort of Notesoble-ish search-and-destroy function might be useful countermeasures. Maybe even *shudder* social media could be used to encourage curiosity (a successful Primer system might even atone for the sins of Farmville).

That last note on social media brings up an observation: it seems that Stephenson is not a reductionist. His ractives (cross between games and movies) were all staffed by humans, without any regard to the fact that things like ELIZA had been written 40 years earlier. With the amount of computing power available in the world, one could emulate humans in any attribute that one cared, including voice synth indistinguishable from real voice actors in a business card, or facial animations that had climbed out of the uncanny valley.

Additionally, he seems to hold human consciousness as not Turing complete (which may well be true), but this is irrelevant as he also believes that human minds have some secret sauce. In TDA, a sect exploits the interface between nanotech and the brain to crack problems in a collective hive mind (where normal nanotech fails), and general AI is not even considered to be an issue. With the amount of computational firepower available, the machines still cannot pass the Turing test. Of course, it would be expedient to the plot to ignore these issues, and I suspect that once you’ve jammed as many things into a book as TDA did, it is time to skip over a digression into moralities of constructed beings and just publish the damn book.

When I mention nanotech as cute, I meant it in a “hey, a fleshed out Drexler-ian vision!” I used to really buy into Drexler’s vision of nanotech, and was only mildly surprised by TDA’s depiction of the sort of excesses possible with nanotech, like the spool that extruded and collected wire on the fly instead of winding and unwinding it, which seems incredibly wasteful to us but would make sense to a practitioner of nanotech. Of course, the depiction of rod logic was somewhat strange, since solid state would probably be better for logic when you’re counting atoms. It does smell like another Drexler-ian construct. I also want to note that Stephenson sidesteps any concerns of a grey goo scenario without fanfare, which isn’t a bad thing (Smalley’s conclusion).

tldr; I would recommend The Diamond Age to geeks of all calibers: if you don’t get every single reference, you are a failure at life.

EDIT: I finally figured out what the poem was saying. It’s pretty meta: she wants to know if he’s a robot, so she asks him.

EDIT 2: While I was cheating by looking on the web (no one cared enough to publish it on the web, so I did have to figure it out in meatspace), I found a discussion of TDA and education in academic settings.