Why Google’s Troubles Run Deep

This may be the only thing (archive.org) I’ve ever read that made me actually want to work at Google.  First off, ignore the stupid framing and scroll down to the actual content.
Let me pick some quotes.

Writing about Amazon: “Organizing into services taught teams not to trust each other in most of the same ways they’re not supposed to trust external developers”

Or writing about a fundamental truth: “Accessibility has an evil twin who, jilted by the unbalanced affection displayed by their parents in their youth, has grown into an equally powerful Arch-Nemesis (yes, there’s more than one nemesis to accessibility) named Security. And boy howdy are the two ever at odds.”

But then getting to Google’s deepest fundamental tragic flaw: “one last thing that Google doesn’t do well is Platforms” … “product is useless without a platform”  …  “.Google+ team took a look at the aftermarket and said: “Gosh, it looks like we need some … Let’s go contract someone to, um, write some… for us.”  … “Any teams that have successfully internalized the notion that they should be externally programmable platforms from the ground up are underdogs” … “The problem is that we’re a Product Company…”

ok, nevermind.  But,  I think it’s shockingly exactly right and in the long run it’s not a bad reason to short Google.

If I’ve peak’d your interest read the whole thing (see archive.org now)

I suspect this says something about  Hal Varian’s status in the senior leadership.  I presume he understands this.  But maybe not how fundamental it is.

The Wegman Family

Much excitement, Wegman’s is opening a store in Northboro Massachusetts.  Wegman’s the archetype for the modern grocery store chain: gigantic, exhausting to browse (like a Home Depot or a Walmart), lots of signal value, lots of experience value; now quite as pretentious as WholeFoods, pricey.  It’s privately held, so the wealth of it’s owners is obscured.

“Wegmans has paid the lobbying firm of Patton Boggs $570,000 in fees to lobby on the estate tax in every year since 1998. In six of these years, the estate tax has been the sole issue on which Patton Boggs has lobbied on Wegmans’ behalf.” –  Spending Millions to Save Billions

That was written in 2006, so – at minimum – in those eight years they spend 4.5 millions dollars.

This store will bring the Wegmans into contact with another famous regional Grocery store: Market Basket/DeMoulas’.  The DeMoulas’ spend their free time  suing each other.

Surely somebody has repurposed that song from Camelot; “What do the simple folk do?”  for the 1%.

The view from no where

Here’s one of those b-school charts, this one about media reporting:

I the short days between when AI finally works and the Singularity clicks in we can imagine that metadata will tag all articles with this information.

I’ve only recently started reading Jay Rosen’s blog “Quote and Comment,” and I’m surprised that I like it. I generally find all the arguments about old/new media to be strangely pointless and hopelessly ill informed.

He draws our attention this wonderful Marshall McLuhan quote:

To start announcing your own preferences for old values when your world is collapsing and everything is changing at a furious pitch: this is not the act of a serious person. It is frivolous, fatuous. If you were to knock on the door of one of these critics and say “Sir, there are flames leaping out of your roof, your house is burning,” under these conditions he would then say to you, “That’s a very interesting point of view. Personally, I couldn’t disagree with you more.”

Exactly right.  The arguments in the old media about old and new media are exactly of that kind.  It must be exhausting to be modern media critic; to spending your days standing on that doorway having that conversation.

I haven’t read that much of Jay’s stuff yet.  But one idea he toys with is that the media must to stop pretending to be dispassionate observers.  They need to let go of the what I think he calls “the view from nowhere.”  My imaginary tagging AI might do this for you – it might tag a report so you can know the presumptions and constraints of the reporter.  That, for example he has deep intimate relationships with the institution he is reporting on and that these constrain his ability to be entirely transparent about what he knows; or – on the contrary – that he is an outsider likely to lack appreciation of the complex trade-offs the institution’s members are living within and thus likely to take cheap shots at their necessary hypocrisies.    Or that the reporter is deeply invested in the current group consensus and often acts as a guardian of that consensus; e.g. that his primary loyalty is to his role and stature within that dominate consensus.

Of course we could imagine that something else, something other than my imaginary AI, might take responsibility for this tagging.  Obviously there are lots of possibilities: ratings agency, a media critic (like Jay), the reporter himself, the blog sphere, … the possibilities are endless.

There is money in this.  Everybody is slowly waking up to the commercial value in having increasingly accurate profiles of your counter parties.  I have no doubt PR firms would pay a pretty penny for accurate profiles of the various entities in the media they must handle.  And our AI tech is good enough already  that it can do a pretty good job of this.  Sadly, and typically, it is probably easier to make money on this by computing the profiles and selling them to a limited audience v.s. – say – Google doing it and then tagging everything.

This reminds me of an incident from back in the 80s.  At the time there was an active small catalog industry.  Once or twice a week the mailman would bring around another catalog full of stuff my household might want.  Google used to have an awesomely fun search engine for scanned versions of these (toy) catalogs, they shut it down a decade ago; I assume  small catalog industry shrank and moved online.

The mid-sized businesses that sent out these catalogs purchased mailing lists from vendors who had built models of each household/zipcode/region/etc.  At some point a company, I think it was Lotus, brought out a product who’s intent was to empower smaller businesses to do this.  The product came with a CD, and the CD had a simple profile for lots and lots of addresses.  The idea being that the small business man would filter out a small targeted mailing list and send his catalog to those folks.

This was perfectly straight forward evolution in the technology of targeted marketing.  It’s an information industry, so costs plummet.  In information industries, what a big firm can afford usually quickly becomes something a small firm can afford.

What happened to this product?  Outrage!  People had no idea that this was going on, and they got very very angry.  There was a little feeding frenzy in the press, and the vendor quickly withdrew the product.  But of course, that did not stop the practice.  It only left it in the shadows.  To this day the targeted marketing industry still demands a veil of secrecy.  The functional benefits of the “view from nowhere” voice as used in the media seems quite analogous; it provides a veil.  I am not unsympathetic to the need for such compartmentalization.

a story of money

This is a fascinating interview with David Graeber, an anthropologist, about the origins of money.

For years now I’ve been convinced that where is something curiously wrong with the presumption that “the books balance.”  One way I talk about that is to ask: “Wouldyou rather die with people owing you, or with you owing them?”  Graeber gets into this in the interview; arguing that exchange exists against a deep background of existing relationships; that the entire myth about the market arises from a simplifying assumption of early economists where they start from a surely impossible initial state where the participants have no existing relationships; i.e. the books begin balanced and we continuously strive to get back into that lonely atomized state.

Graeber outlines a sharply different story about money from the usual just-so-storywe have all been told.  The fairy tale he is dismisses goes like so: barter comes first with individuals exchanging A for B.Over time traders recognize that some commodities are money like – their value is durable, they store easily, their value is easy to decern – and these commoditiesthen become units of exchange.  Later tokens appear that represent these commodities,and at that point it’s off to the races.  Graeber highlights a problem with this story, e.g. that anthropologists haven’t found any examples of this in practice.  I am surely exaggerating, but I come away wondering if barter actually exists.  He says the only place it appears to happen is when an existing currency regime collapses.

Graeber says nope.  He argues, and I think this is exactly right, that what happensfirst is that exchange takes place in the form of “I’ll give you A, and you’ll owe me.”  The “owe me” clause is part of the social sphere.  It’s in the set of books that balance only very roughly.  I’d argue the entire idea of such books coming into balance is fairly toxic to the social sphere.

In this telling the emergence of money is largely as a means to keep a measure of the extent of outstanding debt.  And curiously one place that debt appears is betweeninstitutions (clubs) and thier members – in that context the club members draw uponthe benefits of the club and in turn they accumulate a debt to that club.  In time,as the institution becomes more formal and it’s members more alienated from it themanagement of this debt becomes more formalized.  The club then introduces membershipfees, or taxes.  Money becomes that which the King will accept to pay your taxes.

It’s worth reading the entire interview, because money (reified obligation) has a design flaw; the debt trap.  And there are complex currents about it’s interplay with a societies presumptions about what is moral; and thus a societies morality.  He has fascinating things to say about slavery; debt forgiven, cycles between money as commodity (for example gold) and money as virtual, and on and on.  And possibly most importantly the balance a society based on virtual money most find between protections for debtors v.s. creditors.

Theorizing about money is fun, but try not to forget; studies have shown that 97.3% of those who do it are nuts.

Speaking of clubs … I’m now 18th in line at the public library to borrow their single copy of Graeber’s book: Debt: The First 5,000 Years.

Selling handmade tires and cigarettes

The distribution of wealth creates a curious dynamic, a few producers garner all the profits but often a long tail of small producers continues to exist.  Denialists like to jump up and down and point as this long tail of marginal small producers.  An interesting example of this is the craft beer movement.  In this  interesting essay the author states that the entire craft brewing industry accounts for only 4% of beer sales.  So the brewery industry is extremely concentrated, with 96% of the sales in going to only ten producers, and 92% to the top five.

But here is an amusing thought.  This chart taken from that essay.  It shows (on the right) the industries with the highest concentration.

So.  Could eBay, etsy, Amazon, etc.  target these industries – where presumably there are monopoly profits to be captured?    Home vulcanization!

FYI – The best essay ever written on the topic of craft breweries is Daniel Davis’ “The  Christmas Sermon“.

Migration Assistant Leopard -> Lion

Just an FYI, since apparently the Internet doesn’t know this yet.

You currently can not use Migration Assistant to export your setup from Mac OS X 10.5.8 (Leopard) to OS X Lion (10.8),  as soon as the assistant on Lion makes contact with the one on Leopard it announces that “You need to update Migration Assistant on you other Mac.”    Meanwhile, back on Leopard, software update confidently reports that no updates are available.

Apple phone support finally reported that the Leopard update to Migration Assistant is TBD, and they are snail mailing me a Snow Leopard upgrade.