the paradigm clutch

2010-12-27T14:47:00.000-08:00

If you're interested in the future of food, here's an interesting essay: Twilight of the Chicken Tenders, about transitioning away from industrial foods in ways that are more practical than slow food or Martha Stewart.

In that essay, John Michael Greer suggests that "[t]he recipes to look for aren’t the fancy ones you’ll find in glossy recent cookbooks that are meant to gaze scornfully down from the bookshelf and overawe the guests...".

I think he won't mind if I dissent in one particular instance: the same essay includes "De gustibus non disputandum est; which is to say that in food choices, above all else, dissensus rules." Unfortunately, I'm having trouble posting a comment directly, but it's a long enough comment to stand alone here.

The book I'm thinking of is packaged to gaze scornfully down and to overawe, but isn't written the way the cover would suggest. Chad Robertson et alia have done a great job of reviving the tradition of peasant bread for Tartine Bread. It's unusually considerate of the baker's schedule: the community of test cooks included a graduate student who found that the rhythms of fermentation prompted her to study more effectively than she had before she took up baking, and the first-time proprietor of a new coffee shop with an infant and a serious surfing habit who had to radically distort the process to fit the narrow windows of time that remained. It is a better resource than any I've seen on how to adjust to the ingredients, the fuel, the equipment, and especially time available. It also produces better bread with less effort or expense than Tassajara recipes or any from my family. Partly, this is because Robertson drew on very deep traditions, including apprenticeships in rural France, but has gracefully adapted them to his own time and place. I decided not to buy the book, but I got a lot out of reading it; I hope there's a paperback edition in the works, with more memoir/how-to spirit, and less coffee table book/intimidating gourmet attitude.

A great how-to on cuisine-building in general, is Carol Deppe's latest book, The Resilient Gardener. She has gone so far as to breed vegetables for less time- and energy-intensive preparation, but a good third of the book focuses on attitudes toward cooking and food storage that would be helpful even for those who live in studio apartments.

Last, there are a lot of good web log resources on this topic. An ongoing feature I enjoy is the "Bean Fest" series at Homegrown Evolution. This week's topic is a quick and adaptable rice-and-lentil dish called khichdi.

2009-04-29T12:04:00.000-07:00

I would like to green the streets.

Traffic through residential streets is often quieted using speed bumps, but some cities have begun using barriers and legal restrictions to do so more forcefully. A common tactic is to install concrete planters in intersections, cutting off diagonal traffic and forcing oncoming cars to turn a particular direction. This led me to wonder: What if those planters expanded, and took over a significant amount of pavement?

I'd like to see more streets made one-way, one lane, with barriers enforcing a winding path, and blocks of diagonal parking in the remaining space. This would mean a significant area of green space near the (very sunny) middle of the road. The raised planter beds could be used as a community garden.

If possible, I'd like to start this on my own street, which has some unique advantages:

a lively community of politically, culturally, and socially active neighbors
neighborhood dissatisfaction with traffic, parking hassles, and the aesthetics of cities in general and our street in particular
close enough ties to the local college town that everyone is familiar with related ideas for traffic and for food
located outside of that college town, so that constraints are more often economic than political, and tend to be more solidly rooted in reality

I'm not so sure about the legal requirements or implementation details, but I think it's more likely to satisfy more people than a few of the ideas I've heard floated lately, including parking permits and tearing up the whole street to plant in the soil underneath.

It would be nice to have a garden that looks nice as well as producing food, and thankfully there have been some amazing efforts with that goal in mind.

My (wildly optimistic) vision of the planters themselves has them about three feet tall, and made from locally-available re-claimed materials. One tentative thought is to use strips of tire tread, plastic fry-grease jugs, and demolition waste. There's enough free firewood and clean dirt listed on Craigslist to make literally tons of terra preta, if my neighborhood starts composting locally rather than filling green bins for Waste Management. I'm also thinking of a system to pump water from the gutters into the aforementioned plastic jugs during the wet season, for electronically-controlled irrigation later in the year, although that would require some special attention to keeping the pavement clean.

I'd like to hear what people think about this idea. It would be a lot of lobbying and even more manual labor, so it would be nice to hear as much criticism as possible, as early as possible. Even better would be if some other neighborhood would find all the pitfalls: if you've done something similar, or want to steal the idea, that's fantastic.

2008-09-10T14:27:00.000-07:00

Republicans are getting huffy about "lipstick on a pig".

Let me break that metaphor down for you: Palin is not the pig, she's the lipstick.

That's right, McCain and Palin were compared to a coat of lipstick. Republican policies were compared to a pig, which lipstick will fail to beautify or disguise. That is to say, he showed respect for her, and contempt for what she stands for...or rather, stands in fron of.

Sheesh...doesn't anyone in the Right Wing ever read Lakoff?

2008-05-10T21:00:00.000-07:00

I grew up with the notion that relationships, especially romantic ones, are all about compromise. This makes about as much sense as saying that war is about killing people: very plausible on the surface, but ultimately wrong. It's true that anyone entering a relationship should be prepared to compromise, and that ability to compromise is a strong predictor of success, but compromise is not the whole point. In war, ruthless killers will lose to people who competently take and hold territory; in love, a skill at finding mutual preferences beats a willingness to compromise, far and away.

I have been surprised over the last ten years or so, as relationships built on huge compromises have turned out somewhere between disappointing and horrific. Not only surprised, but frustrated. Something was subtly and profoundly wrong with my thinking, and I couldn't quite figure out what.

I've been trying to improve myself in a spirit of compromise, i.e. a flattened, zero-sum model where ways of relating lie along a line between passivity and aggression. In that model, every step toward self-assertion is a step away from what the other person wants. I've failed my efforts to find the middle ground between doormat and asshole, because within that system, the region I was looking for has negative length. Through first-hand experience, I know that you can neglect your own needs and your partner's needs, all at the same time.

I'm now aware that relationships are about mutual preference, not compromise, and I intend to drum that fact into my skull until I actually behave accordingly.

2008-05-06T19:47:00.000-07:00

Barack Obama is an elitist.

I listened to Mr. Obama's speech on the radio today. The mismatch between the America I have gotten to know in my adulthood, and his understanding of this country, was so striking that I couldn't turn off my radio at the end of my drive home. His rhetoric was so far out of touch with the values America has displayed in this decade, that I sweated it out in the hot car and delayed my dinner, almost in shock.

He talks a lot about hard work, but the business section of any newspaper will tell you that Americans want easy money. We want a gigantic house, and after no down payment and a couple years of barely covering the interest, we want to sell it at a tidy profit and move up to something bigger. By the same token, we want everything to be made overseas, so that we can buy as much of their work as possible, for as little of our work as possible.

He openly admits that he isn't perfect, but anyone who watches the political news knows that Americans want leaders who are beyond reproach: if not de facto, at least de jure.

He says we all want unity, and to be treated alike, but Americans seem to want special priveleges. We'd like to think that we won't pay any taxes once we finally get rich. We also expect perks for our membership in whatever in-group we belong to, whether we are Crips or Freemasons, whether we have pledged ourselves to a fraternity or a charismatic religious leader, whether we're members of Sam's Club or of Clear (which allows you to skip the airport security line for a fee and some waivers of privacy).

Most telling, he calls on his audience to be agents for change. For starters, this limits his audience to those who are brave enough to face change, which seems to be a small proportion of us, if the images on TV have any basis in reality. But even within this small group, it speaks only to the elites: if my classes in sociology have taught me anything, it's that, by and large, Americans lack agency.

Defining his audience as agents of change limits his message to two select groups, who have more agency than most. The first, by accident of heredity or upbringing, have broken the spell that Edward Bernays began casting in the 1920s. The second group have remained aloof from mainstream culture, and have been free from that necromancy all along. Bernays asserts that most Americans "are governed, our minds are molded, our tastes formed, our ideas suggested, largely by men we have never heard of." The remainder, the people Mr. Obama is speaking to, make up the ruling class of this country: they govern, mold, and suggest.

A stump speech that calls for the listener to act on their own behalf is a dog whistle to the select few who neither passively resign to the world they've been given, nor blindly insist that the universe conform to some pre-existing text. His message is for those with such high opinions of themselves that they enter into a dialogue with the world, as equals to it.

Mr. Obama shamelessly panders to people with the gall to ask the world to improve, who have the leisure to learn its language and then speak to it on its own terms. Quite often, he chooses words that betray his pragmatist ideology: one that, with few exceptions, is only held by the very successful.

A part of me wishes that all of America held the values that Mr. Obama espouses, but the collective decisions we've made over the past eight years suggest that, except for the elite, we do not. We haven't been an elite country during this century, but I'm sure that, some day, we can. Oh yes, we can.

2008-01-27T17:47:00.000-08:00

Weeding his alms:
Weeks' worth of good luck,
Sown in the street.

2007-12-02T12:41:00.000-08:00

Note: this overlaps with some things I've written before, but I'm thinking of publishing elsewhere. Forgive any repetitiveness.

Whenever I hear someone wax poetic about the mystical powers of crystals, as elaborated in neo-Pagan systems of magic or medicine, I smile and nod, and say "you have no earthly idea how right you are," though they may have a celestial idea or two.

I say "earthly" because I don't believe in the woo-woo power of big, natural quartz rocks to harmonize one's aura. Instead, I agree with them because I've been trained in crystallography, and I know of several breathtakingly practical applications that materials scientists and engineers have found for this discipline. For instance, the fact that quartz crystals lack inversion symmetry makes them piezoelectric, able to transform electrical energy to mechanical work and vice versa. Resonators that slosh energy back and forth between these two forms are available from hobby shops at very low prices. These crystals allow all of your electronics to keep time with one another, maintaining a harmony so exquisite that it would escape even a dog's ear (canine hearing reaches to maybe 0.1 MHz; the 4 GHz quartz clock shown below isn't especially fast for use with a microprocessor).

Speaking of sound, several mystical traditions hold that vibrations or verbalizations are primordial and universal; for instance, that the syllable "om" brought the universe into being. I don't have any evidence for or against that assertion, but I do know that electron-phonon interactions (that is, collisions between particles of electricity and particles of sound) are very important for the workings of semiconductors and superconductors. Your dad's old transistor radio, let alone your doctor's new MRI scanner, wouldn't work without the designers having some very subtle understanding of the properties of sound.

But phonons are really only well-defined within crystalline solids, so maybe that's a spurious example. Even sound in gasses has some surprising applications, though, through the thermoacoustic effect. There's a commercially available system that uses sound to condense natural gas on remote oil wells, for shipment to consumers. Offshore oil rigs, or wells in the middle of a desert or tundra, would otherwise vent or flare the gas they produce, adding a lot of methane or (slightly less bad, from a greenhouse perspective) carbon dioxide into the air without using it to accomplis anything. A thermoacoustic system will burn some small fraction of the gas, and use that heat to generate sound waves, which, in turn, drive a refrigerator that brings the remaining gas to cryogenic temperatures. This reduces the methane's volume by about a thousand times, and allows it to be transported in cheaper equipment designed for liquids, easing demand for synthetic fuel gasses. Why not use a conventional heat engine and mechanical refrigerator, you ask? Well, if it's so remote, a breakdown would be a big problem. Thermoacoustic systems are a lot more reliable, because they have approximately as many working parts as an old-fashioned flute: zero.

Of course, seeing this gas liquefaction technology as a good thing begs the question of whether we should be extracting fossil fuels at all. It's great to save a lot of methane emissions at the expense of a little carbon dioxide, but if we want our global climate to stay at all familiar, we should probably be working to return carbon to the earth, rather than to dig it up more efficiently.

I think there should be a lot more composting, for instance. If something goes into thelandfill and begins to rot, it will largely turn into methane and be released. Efforts are made to capture some fraction of this methane at the most enlightened of dumps, but I'm told that this is only a small help.

Judging by how much weight my compost pile has lost since I set it up, I'm probably emitting more carbon dioxide now than I did a year ago. But that means a lot less methane, and, on top of that benefit, I intend for the carbon that remains to be buried in garden soil, where it will probably stay for quite some time. If it leads to more plant life, it will pull still more carbon from the air.

I try to be thorough about it: food waste, raked leaves, and floor sweepings, of course, but also any paper towels that I use in public restrooms. Unless a paper product is recyclable, or small samples of it burn green (a test for poisonous copper, which many fast-food companies use in their ink), or it's coated in plastic rather than wax, it goes in the compost pile.

Composting has some other benefits, of course. At the risk of sounding like a gullible hippy, I find the act of turning a compost pile to be a powerful meditation on mortality. It adds some much-needed perspective to my life when I can see the things of man, from pizza boxes to my own nail clippings, being digested by the earth. It challenges my notion of permanence, and keeps me humble. Which I need, as much as the next guy.

humble (adj.)
c.1250, from O.Fr. humble, earlier humele, from L. humilis "lowly, humble," lit. "on the ground," from humus "earth."

2007-11-24T03:59:00.000-08:00

So, a high-school peace group is being denied the right, peaceably, to organize. Honestly, how else would they?

The group of students wear yellow ribbons. They don't say anything against Bush, or for Democrats. They just wear shirts that proclaim peace. Peace signs, the word "Peace," that kind of stuff. They followed the rules with a written application, and an official statement of purpose & principles, and an adult sponsor. The administration rejected their application, and is preventing them from posting anything related to peace on their lockers.

The whole school is in an uproar, with students united behind a common cause. Now a huge number of students are wearing shirts with the same theme, holding their ground in this battle of ideas.
That newly-popular theme, though, is not a peace symbol, but a Confederate flag. Yep, the majority of students side with the administration (school and national), and want this group of dissenters to be silenced. And wearing the old Stars and Bars is the least they can do to show their support...for...their, um...country? Uh, and for the President...of the, um, Union.

But peace advocacy can't be allowed, because it's disruptive and nigh-upon treasonous. Best to stick with safe, patriotic themes...like that group who rebelled against the US government and eventually became symbols of slavery. No controversy there, no sir.

I stole the above link from Ran Preiur, who says: "Ordinary Americans, post 1980, are probably the only population in history to be generally pro-war." But I think Spartan citizens would have also swung pro-war. I hear that the government of Sparta declared war on most of its own, non-citizen population on a regular basis, using a ballot by a representative body. This meant most of the work within their borders was done by enemies of the state, allowing more-powerful Spartans to ignore what little civil rights protections had been developed at that time, and treat their workers worse than an Athenian would have treated a slave.

You may have seen 300, in which the Spartan protagonist (a young prince, natch) is sent out to live off the land, the way soldiers do. He makes a spear from a sapling and kills a giant, slavering CG wolf.

This is kind of historically accurate, if we remember that this cartoon wolf is actually symbolic. Soldiers don't live from the earth the way hunter-gatherers do; that would require some sissy expertise with a digging stick, and less time for manly spear-play. "Live off the land" is a euphemism for "steal from the locals:" Spartan youths weren't cast out on their own to dig for grubs and tubers; they were forced to join street gangs, and get first-hand experience of, as Terry Gilliam would put it, the violence inherent in the system. Having to choose between starvation and armed robbery was their rite of passage, and the slavering wolf represents the threat of an empowered lower or middle class, like the phantom menace faced by Marie Antoinette. He slew that wolf, alright...and got food, clothing, and status in the bargain.

Anyone who couldn't find a place in the hierarchy of thugs would be left in the cold, to be killed by an angry peasant if hunger or exposure weren't enough...but I'm sure that didn't happen too often. People are good at fitting in, whether they're in an ancient Greek scouting troop, a Prussian-style public school, or a Bentham-Foucault inspired prison system.

It occurs to me that the bulk of agricultural workers in this country are in a similar situation to their Spartan forbears: they're economically necessary, but legally forbidden. In my neighborhood, citizen "rat packs" have formed to exploit this situation. Like Sinatra before them, they've found strength and wealth in organized crime. Like young Spartans, they target members of the laboring class, because "Illegals" tend to be frozen out of the banking system, so they often carry cash on payday, rather than a checkbook or ATM card. And they can't go to the authorities, for fear of deportation.

But the plumbing for deportation is getting a little clogged, and prisoners are backing up faster than they can be drained from our borders...so they're staying in prison longer. And with shortages of agricultural labor in the parts of rural America with the toughest immigration enforcement, they've started using chain gangs to harvest and care for crops. Like illegal aliens, inmates don't need to be paid minimum wage.

The current political situation raises many small concerns, each little more than a drop in the bucket, or a grain of sand on the beach. They all add up, though, until I can't go along with it, and can barely even keep my temper. If I weren't such a peaceful fellow, I'd have the urge to kill the messenger. By...oh, I don't know...kicking him into a well, or something.

Here's your moment of Zen...unless that phrase is a registered trademark, in which case it's a moment of Taoism or Yoga or something:

2007-09-06T10:19:00.000-07:00

How a 1970s Mother Earth News article made me less of a dirty hippy

I washed my car today & yesterday, for the first time in at least a year.

Ick, I know. It was high time.

But people are rationing water here in California, and the Delta Smelt face extinction...how could I, in good conscience, turn on the hose for such a frivolous purpose? Especially when I obviously don't much care how my car looks?

Well, I've figured out a way to wash my car without using a hose or bucket. You'll be disqualified from the Hipster Olympics if you use my method, since it involves getting up around dawn, but other than that, it's pretty convenient:

Use dew.

I took an old nylon scrubber pad with a squirt of dish soap, and some of the paper towels I pocket after drying my hands in public restrooms (these accumulate quickly; 4 days' worth is enough to dry a car twice over), and two mornings' worth of dew was enough to remove a year's grime, bird droppings, and aphid sugar. Not to mention the ants that had colonized the car exterior in order to eat the aphid sugar...I know, ick.

If you are a morning person, it's actually easier than hauling out a bucket and hose. The scrubber went into the bathtub and was rinsed out as I trod on it during my morning shower, and the towels went into the compost, which I had intended to do anyway (warning: road dust contains some lead, due to the weights that balance wheels...lobby automakers to use zinc or something instead, and hope the exposure is not too great). A squeegee might cut down on the number of towels you need, if you don't have once-used ones lying around.

An added bonus: more reason to mock the people who use filtered water to wash their car. Filter cartridges for that ridiculous Mr. Clean/Pur gizmo have appeared in the clearance aisle, so there clearly aren't as many of these people as the marketing industry might have hoped, but if I should ever meet one, I can now say: "Filtered water? Well, okay...but I use distilled."

As usual, I feel pretty clever, but this idea is actually borrowed from a famous survivalist, Tom Brown, Jr. I highly recommend his series of articles on how to survive in the wilderness, especially since he's not trying to sell high-end gimmicky equipment. The article in question is here.

Now that I've washed it, perhaps I should take care of the dent from that fender bender back in February of '06...

2007-03-17T22:00:00.000-07:00

An Arrogant Proposal

I just finished reading Matt Fuller's excellent essay, BSD for Linux Users. It's pretty heavy stuff, that should go way over my head—after all, IANACS. But I think I really understood the message. Mr. Fuller deserves any credit for my ability to understand the topics he presented, but those of you who know me well, know that I'll take any experience as an excuse to feel smart.
The chapter on myths was no exception. It made me feel I could take on a huge, outstanding problem...not Irish poverty, which is already mostly solved (turns out that was a biodiversity/economic development issue), but an older and more global problem: artificial intelligence.
Of course I'm not qualified to make any useful contributions to the ongoing dialogue on cognition and self-awareness. If I present an original theory here, I'll do so by misrepresenting a theory that I liked but didn't understand completely.
I'm also not going to rehash the great strides made in recursion and other types of self-reference. For example, Jeff Bezos has done some really elegant work in that vein, by applying the concept of self-reference at a sufficiently high level.
What I'd like to propose is just an AI application that seems under-developed, but may allow some exciting theories to be tested. It's a strategy for passing the Turing test, by a careful choice of venue.
Obviously, it would help if the programming effort were motivated by an unmet need in the lives of skilled programmers. There's only so long that someone will write code out of the goodness of their heart, and even the potential glory of creating practical AI isn't really enough. And Windows Vista is a $400-a-box demonstration of the limits to any effect achieved by external motivation of creative people. People really have to be driven from inside.
Cybersex is an application where passing the Turing test would meet someone's needs, but not those of the programmer: no matter how lifelike the text is, a programmer can't fool himself. Also, there's very little use for a half-working model, which makes incremental development a much less attractive option.
While there is a class of human interaction that programmers don't get enough of, I think most programmers are more agitated about interactions that they dislike but can't avoid. Mr. Fuller's essay includes a rant-within-a-rant highlighting a type of grunt-work many programmers dislike: answering questions from clueless members of your software community. A tool to automate that process might be built on code like the following:

question = raw_input("End-user's question:")
if False: #check for learned answer to this question
   (tab) pass
else:

(tab)

  print "RTFM"

This is not even a rough draft, and barely qualifies as throw-away code. Ideally, question would be compared to a corpus of previously-answered questions, to find an answer that has previously been given. This would mimic courteous newbies, who use search engines to look for answers in archived help discussions before bothering a human. I don't know much about Bayesian algebra, so for now the test for learned answers returns "false".
But this code could pass the Turing test, even in its present state, given a mailing list where "RTFM" is used often enough. It would help to make the default message user-configurable. For instance, Mr. Fuller could use "This is a development list, you want to ask that question on questions@freebsd.org".
I hinted earlier that I imagined a use for half-working code. I think the first application should be a context-sensitive suggestion utility, that presents a software expert with a mix of learned and user-supplied answers, then keeps track of when its suggestions are accepted, and how the user edits them before sending them off.
All this highly-motivated human input would be saved, and algorithms could then be developed that use this input as training in what to suggest, and when. This information, or clues from the corpus of discussions, could be used to create categories of question-answer pairs, and to find the most-representative member of each category.
The idea that really excited me might take a little longer to develop, and would build on all of this. It occurs to me that messing around with software by entering text is a realm of human behavior that computers already have access to, without any expensive investment in machine vision or robotics. In its training cycle, a bot can try out any of the archived pieces of advice it digs up, given remote access to a system with the appropriate properties. It can even run thought-experiments while a dialogue is taking place, by spawning a virtual machine and playing around with the various parameters.
Sure, its notion of identity will be founded on user names and passwords, rather than faces and pheromones, but the same goes for many of the people it would try to emulate in a Turing test. And it will struggle to read and write natural languages, and even to use other software, but the same goes for most of the people it will chat with on a daily basis.
Basically, I believe this is a venue where a machine might experiment alongside humans on a more-or-less level playing field, with no hardware more specialized than a network card. Not only does this venue hold a vast and growing field of co-experimenters, who truly want to understand what the machine is saying, warts and all; the venue also holds many skilled software developers, each with a fine-grained motivation to help mechanize the expert's role in the process.
The effort to build such a tool may advance the field of AI in a fundamental way. It would mean finding a way to generate natural language based on several, interacting inputs, including experimentation (via remote terminal), conditioning (acceptance of suggestions), imitation (archived discussions), and imagination (VM models).
If incremental advances aren't enough, consider what success might mean. For a machine to credibly take the developer's role in a discussion about software, it would need to say intelligent things, backed up by experience as well as borrowed expertise. If such a machine has been trained and designed to experiment with software the way users and/or developers do, it might be a useful development tool in its own right, assisting the production of documentation, perhaps even informing UI design. If it graduates from merely talking about software, perhaps the machine might be allowed to start playing with the source code in question.
But I digress: the imitation part is what interests me most, because I'm convinced that modeling the human cognition of imitation could allow us to progress from faking self-awareness to making self-awareness.
If V. S. Ramachandran is right in the essay I linked to above, then practical imitation will depend on a sense of the other's identity, which can then be kludged into a sense of self. In nature, this seems to work by categorizing observed behaviors according to percieved intent, and then interpreting one's own automatic behaviors according intent.
To begin, the bot might go through the archives correlating the methods used by experts with the results they achieve, results such as "I don't understand", "thanks, that worked", or simply a lack of further questions. Some easily recognized methods might include asking about the user's system, giving them code to paste into a terminal, providing a link to another forum, or not replying to the question at all.
But a bot would also do well to correlate the results (which I dub "more", "thanks", or "silence") not just with methods, but with individual experts and the questions asked. Profiles of experts would allow it to develop the roles of "dismissive, gnomic curmudgeon" (behaviors that produce silence) and of "charismatic, helpful guru" (behaviors that receive thanks), while analysis of questions would generate the stereotypes "lazy pest" and/or "troll" (people whom experts dismiss) and "worthwhile new citizen" (folks who thank experts; presumably, experts help them).
When it first begins to operate autonomously, the bot can begin by fitting the questioner to a stereotype based on the question asked and a history of questions asked by that username in that forum. Then, when a list of answers that match the question is generated, its choice of answer can be informed by how well they fit the appropriate role (dismissive or helpful). The choice of role will also affect training; while further questions will be negative reinforcement in either role, "dismissive" mode will mean a stronger response to silence, and "helpful" mode will mean a stronger response to gratitude.
Bear with me, because here's where it gets good. Given a system that works this way, the bot can search the archives for experts who achieve a good mix of "thanks" and "silence" results; that is, they don't stick to one role. Then it can narrow this list to those that tend to be dismissive to trolls and get silence, but also act helpful to citizens and get thanks. By trusting that this list of user names includes mostly smart, flexible people, the bot can find flaws in its model by deciding which role it would choose for a given question, then seeing which role the expert seems to have chosen. This will allow it to refine its model by looking for a minimum rate of disagreement as it randomly tweaks its criteria for roles and stereotypes.
To simulate deeper introspection, the bot can use this refined model to look for exceptions to its rules. These exceptions would be based on the remaining cases where experts choose a different role than the bot would, and achieve a role-appropriate result. The test for a constructive exception would be that it would change the role chosen in some cases that ended in negative reinforcement, but would not apply to a significan number of cases that ended in positive reinforcement.
I'm really reaching here, and I'm sure I got a lot wrong, but I was trying to write a plausible story of how a mechanical sense of others' identities and thoughts might be modified to mimic critical examination of one's own behavior.
Even if all my wild speculation about deeper AI is unfounded, there's still the potential to partially automate one of the least-favorite tasks of programmers. The gurus have nothing to lose but the drudgery of the endless September. They have a Turing test to win.

2005-09-03T15:56:00.000-07:00

In case you weren't angry enough:

Compared to the $10G (that's G as in Giga, not Grand...billion, rather than thousand) congress approved to cover the >$100G effort to save the lives of former New Orleans residents, they spent about $7G this year emulating the control system of the former Soviet Union. That figure excludes any indirect costs that come from the natural inefficiencies involved, of course.

David Brin has the details:

Government secrecy has reached a historic high, even compared to the Cold War (San Diego Union July 3). Federal departments are classifying documents at a rate of 125 per minute... or two per second... and inventing new kinds of classification, while declassification efforts that peaked under the Clinton Administration have slowed to a crawl[...]Please note that post Cold War analyses have almost universally concluded that Soviet paranoia and excessive secrecy was one of the reasons for the decay, fragility and eventual fall of that evil empire. By all means, let’s copy them!

2005-08-29T23:23:00.000-07:00

This guy makes a well-reasoned and, dare I say it, enlightened argument in favor of the Iraq war. Is this the exception that proves the rule, or the counterexample that falsifies a bad theory? I'm from a tradition that judges prophets by their fruits, and the online etymology dictionary speaks highly of its creator.

Contrary Brin 's assertions that the war is being mismanaged in a way that borders on treason are quite convincing, in my opinion. He notes that many of the problems we face can be traced back to the same folks who bungled the Vietnam project (e. g., Rummy), but there are more similarities than differences betweent hese two commentators.

They share a dislike of both fundamentalist conservatism and antimodernist leftism. They share a contempt for the left-right political axis as a metaphor, because it leaves pragmatists like them out, or tags them with a ridiculous "moderate" or "independent" title. And they're both in favor of Balkans-style nation-building.

Reading other people's points of view about the Iraq war and gun control have given me a new respect for rank-and-file Republicans. I've always shown them respect, but I grew up liberal: showing respect and having respect seem to be almost opposite from one another.

I remain entirely convinced that our leaders' thinking about the war is impractical, ignorant, and/or immoral, but I'm beginning to doubt my opposition to the project as a whole. It's possible to imagine Iraq being another Kosovo or Afghanistan, if only Gore had won. He would have shown patience, and tact. He might have had enough humility to listen to Powell and Clarke and the long chain of competent people that reaches down to the soldiers and marines on the ground.

2005-08-29T20:10:00.000-07:00

Pinker's First Law

From The World Question Center

Human intelligence is a product of analogy and combinatorics. Analogy allows the mind to use a few innate ideas—space, force, essence, goal—to understand more abstract domains. Combinatorics allows an a finite set of simple ideas to give rise to an infinite set of complex ones.

No one asked, but here are my laws:

1. Taxonomies are always foreshortened: they expand as you approach the point-of-view.

Why? People's imperfect expertise and need to distinguish "other" from "self" will always lead to finer distinctions in their classification system the closer one gets to the entity that created the system.

Why's it important? So we realize that Chinese is not one language, but many languages which can share a writing system. So we don't keep discovering new "species" of dog/fox/wolf that can all freely interbreed, while ignoring the plain fact that each new bacterium is a new species by the canonical definition. So that we look for important distinctions among foreigners, and learn to more reliably tell friend from foe in a better-connected world.

2. Faith tends to occur when a person has given an extreme amount of thought to a subject. But it's important to remember that there are two types of extremum: minimum and maximum.

Why? Faith is a sense that something has been thought about enough. And unless they suffer from severe anxiety, an expert will have thought through and/or experienced their favorite subject enough to have faith in some of the less certain parts of it.

Why's it important? This helps to avoid the hasty generalization that religion is irreconcilable with science. Sure, you don't do science if you think that enough thought has been given to your area of study, but history teaches us that religions change quickly enough to admit inquiry into topics where progress is being made. They're only ever a century or two behind, at most. It's also a theoretical framework for the observation that making people think about their beliefs tends to cause them to feel insecure, and hate you for it.

Cross-posted from my Livejournal.