Google's Microsoft Complex
James Grimmelman compares Google to Sauron. I'm flabbergasted. I had always thought that Google was the Ringbearer and that Mordor was where Gates and Ballmer hung out.
Was I misinformed? Or am I just confused? If it's the latter, I have a decent excuse. After all, even though Google and Microsoft have very different public personae, it's getting harder and harder to tell them apart as businesses. Both built their empires through the same strategy: gaining control of the economic chokepoint of the prevailing computing model of their day.
Microsoft's dominance in supplying the operating system for the personal computer gave it control over the central money-making engine of the PC economy. In a similar way, Google's dominance over Web search (and search advertising) gives it control over the central money-making engine of the Internet (or "cloud") economy. When you're lucky enough to have that kind of position in an economy, your central business goal becomes expanding the economy. You might not take a cut of every additional dollar that flows through the economy, but you'll take a cut of a lot of them.
That explains why Microsoft expanded aggressively into a vast number of software and other PC-related businesses. And it explains why Google has expanded into so many Internet businesses, from news aggregation to video streaming to web browsers to word processing and spreadsheets. For both, the intent was not necessarily to make money directly from every new product or service but to spur ever greater use of either the PC (for Microsoft) or the Net (for Google), driving more dollars through their chokepoints. An economist would say that all those accessory products and services are simply "complements" to the core business. As the complements expand so does the core.
I think this explains why Randall Stross had such trouble putting his manuscript to bed. It's Google's strategy to be a protean company, constantly changing its shape as it enters new Internet markets. I think it also explains why Grimmelman has the sense of Google appearing at times to be "remarkably aimless." The problem with having complements everywhere you look is that everything looks like a business opportunity. That can lead you to invest in a hodge-podge of activities. Run a social network? Sure. Get into radio advertising? You bet. Shoot a satellite into orbit? Why not?
Frodo or Sauron? When you control the economic chokepoint of a digital economy and have complements everywhere you look, it can be difficult to distinguish between when you're doing good (giving the people what they want) and when you're doing bad (squelching competition). Both Google and Microsoft have a history of explaining their expansion into new business areas by saying that they're just serving the interests of "the users." And there's usually a good deal of truth to that explanation - consumers do often benefit from the companies' investments. But it's rarely the whole truth. As they throw their weight around, they squeeze out the little entrepreneurs who tend to be the most creative actors in a rapidly growing industry.
Go too far with this strategy and, as Microsoft discovered, the government steps in and slaps restraints on you. In recently scuttling Google's proposed search-advertising deal with Yahoo!, the Feds have already sent a warning shot across Google's bow.
Just because Microsoft and Google share a strategy doesn't mean they're similar companies in other respects. Google differs from Microsoft in at least one fundamental way. The ends that Microsoft has pursued are commercial ends. For the most part, it's been in it for the money. Google, as Grimmelman notes, has a strong messianic bent. Google is not just out to make oodles of dough; it's on a crusade - to liberate information for the masses - and is convinced of its righteousness in pursuing its cause. Depending on your point of view as you look forward to the next ten years, or even the next 300 years, you'll find Google's crusade either comforting or frightening. It's worth remembering that Google's chokepoint is not just an economic chokepoint. It's also a cultural chokepoint.
Or, absorb them, especially if they have created something useful. Evolution works that way. Usefulness endures.
And while Google is probably not in the forefront of the actual development, I suspect they will lead the way in the deployment and implementation of artificial intelligence.
January 13, 2009 4:14 PM | Reply | Permalink
Skynet, eh? Colossus? CYC? The awakening Net?
Interesting. A related idea...:
It's more than 20 years since a leading processor design was not itself done using leading-edge computing technology. So in a way, computing systems (hardware then, increasingly software too) are inching toward self replication. Humans still control it but technological advancement requires the input ("cooperation") of existing technology like never before. And there's no question that market forces serve in a role of measuring fitness for selection for continued development.
I enjoyed watching the competition among x86 and 68xxx, transition to x86 vs Power, x86 vs SPARC. Lately the arrival of the GPU (Nvidia sells deskside workstations with near teraFLOP performance based on their graphics processors) is a fascinating advancement.
(One "GPU" now sports up to 240 cores, each with up to 8 threads....)
This stuff is about to get Very Interesting.
January 13, 2009 6:18 PM | Reply | Permalink
I forgot to mention a hypothesis, based on the history of Intel's processor family beginning with the 8008:
8008
8080
8086
80286
80486
Pentium
Pentium II/II/IV
AMD Operon/x86_64
Core
Core I7
It is possible to trace specific processor registers and related facilities all the way through the most recent Core i7.
With that introduction, I think of the Intel processor family the way we see evolution of the natural branches of life forms culminating in the sauria (not Sauron... :-) :-) ). Sauria were a hugely successful branch of advanced life forms, with lasting hundreds of millions of years in their dominant roles. The only thing that toppled them was the event marked by the K-T boundary.
So the hypothetical question... Is the x86 a dinosaur? it certainly has the best record of dominance. Will the x86 lines experience a K-T event?
(Apologies to advocates of the mainframe world! I've been an embedded developer and completely forgot about the long-dominant power of mainframe technology in business.) IBM's mainframe line beginning with the System/360 and compatibles also have a stellar record. One might think of the microprocessor as being the mainframe's "mammalian" competition...
January 13, 2009 6:43 PM | Reply | Permalink
You missed the 386 - the big move to 32bits.
January 14, 2009 4:21 AM | Reply | Permalink
You nailed me totally... The '386 was not only the x86 transition to 32 bits, it was the transition to real support for paging and away from segmentation. The '386 was the first x86 that could run a real UNIX.
The '386 is also the last x86 that had significant human contribution to the component layout.
The irony is that my group at AT&T Bell Labs implemented the first '386 design in the SS7 signalling network to support 56k links, and I wrote its bootloader firmware! How could I forget? (The answer: The following project supported T1 signalling links using twin '486 and twin i960CF processors. I had the first 80486 DX2/66 in the city of Columbus because Intel was only sampling the part then, and I was the only group member geeky enough to have been running UNIX on my fancy '486 workstation...)
January 14, 2009 7:29 AM | Reply | Permalink
"Or, absorb them, especially if they have created something useful. Evolution works that way."
Sorry, but that is totally NOT HOW EVOLUTION works. If bat's learn to fly, rats don't "absorb" that useful ability, each species has to work it out for itself. The only way one species absorbs anything from another is when it digests it through it's stomach. Where evolution might come into this is that the smaller species or "prey" no longer competes against the larger species and the raw ingredients provide some sustenance for the larger species. But none of the "qualities" or "innovations" of the smaller species get picked up by anyone else. Now that I think of it, I think what I have just explained is probably a more accurate representation of how large companies operate in regards to small companies.
But my original point is that you were making an inaccurate claim about how evolution works.
January 14, 2009 1:22 PM | Reply | Permalink
Higher complexity organisms that protect their genetic material through germ cells experience convergent evolution.
At the genetic/molecular level, DNA/plasmid exchange occurs relatively frequently in prokaryotes: Bacteria and viruses exchange genetic material relatively "frequently".
In the world of ideas, "evolution" often involves designers and developers replicating concepts among product families more like prokaryotes forms and less like the more complex forms. And in this context, "absorption" is a reasonably serviceable term. Old Grouch's reference to "evolution" is almost certainly metaphorical anyway.
January 14, 2009 2:26 PM | Reply | Permalink
It is kind of difficult to put Gates in Mordor now given all that money the Gates Foundation has given away. Trying to eradicate malaria is not a Sauron like activity.
January 13, 2009 6:16 PM | Reply | Permalink
you cure a person dying of starvation of malaria they're still going to starve to death.
Gates spends half his life ripping people off then turns around and spends it on self-gratifying "philanthropic" activities to big note himself is, in my opinion, not a very noble cause.
Perhaps if he spent his money on improving the lives of people by helping them help themselves this might be a more worthwhile cause.
Also, don't forget the Gates foundation is into funding some pretty dubious ventures as well as openly peddling Microsoft products in the developing world.
January 15, 2009 2:55 AM | Reply | Permalink
It's important to separate Gates' noble contributions from Microsoft's business practices. I've been watching them for 25 years and been fascinated by their behavior. They used their near monopoly in the OS marketplace to blast away their competition in several major market segments.
Until about 1995, Microsoft was an honest and essentially developer-friendly infrastructure provider, a gentle giant. During the mid to late 1990's however, they became *distinctly* unfriendly to any business whose product line happened to be in their way. And the people connected to those businesses have some very interesting stories to tell.
January 13, 2009 6:27 PM | Reply | Permalink
http://www.htl-steyr.ac.at/~morg/pcinfo/hardware/interrupts/inte1at0.htm
Take a look at which of these functions were "undocumented" - everything to do with extended drives and multitasking. That's why Gate's folks were always able to do more with their software. There was one book for MS guys and another for everyone else. If anyone outside of MS used undocumented functions, they got sued for breach of licensing agreement(no reverse engineering).
Yeah - Gates was a real peach back in the 80's!
January 14, 2009 4:37 AM | Reply | Permalink
I posted this months ago elsewhere. It was obvious then as now.
But Google is far more dangerous than Microsoft could ever be. Microsoft was once at risk of having some of its products put into the public domain. Google itself, in its entirety may have to nationalized, or put into an international public consortium, in the future.January 13, 2009 9:05 PM | Reply | Permalink
I feel a bit out of place - an end user in a room full of developers.
To this end user, Apple is a designer prison. Well laid-out, almost cutting-edge, and a prison nonetheless. It's either their way, or not at all.
That's why I prefer the PC "organ bank" model. I'd rather build than buy any day. And when Windows XP no longer serves my needs I'll find some sort of Linux or something that does, without locking me in. I hear things are getting good over there these days.
January 14, 2009 11:15 AM | Reply | Permalink
Mac OS is only a prison if you choose to live in it. I'm not that interested in computers, but I use Photoshop and Final Cut Pro and the Mac is a smartly designed and stable system in a pleasing package. At some point we may not have a choice but to live in Google's world. That will be a prison.
January 14, 2009 12:37 PM | Reply | Permalink
I've been developing in Android recently, and Google of today is doing all the same things as MS - almost to a T (although in a different order). The first important point is that Android marks a foray from web space to physical space. This article paints them as a cataloger of information, but it seems their plans go well beyond this.
They have (correctly IMO) identified that the next big area for revenue expansion is going to be mobile-based applications. They say the Android Market doesn't make a penny - 30% goes to right to the payment processor. Of course, in order to process the transactions, you have to use Google checkout; so while the "Market" doesn't take in any cash, Google will profit nicely on every sale. This isn't an exorbitant fee, but they will make money from essentially all applications sold.
Apple application growth is limited by the number of devices they can sell. Symbian already has too many devices out to even consider centralizing. RIM and Palm are also manufacturer/software distributors and limited by the number of physical devices they can invent and sell.
In contrast, every Android based device will access the same market no matter who the manufacturer is. Like Microsoft with PCs, as each wave of new devices comes along, a new wave of application purchasers - without Google investing a single dollar in hardware R&D. They take a big chunk off the top and everyone gushes about how "free" everything they do is.
Where the Windows Mobile OS dropped the ball (aside from archaic functionality) was never unifying application space. They, more than anyone, were best poised to do it but either didn't see the need or were too busy trying to dominate the gaming console wars. So there is some irony - because WinMo is likely toast.
In addition to exclusive distribution of all applications, Google also has exclusive access to "private" functions. These are abilities that for marketing/security reasons aren't available to public developers (it's why the G1 won't change orientation and nobody can write an onscreen keyboard). Some make sense, others less so. One way or another, it allows applications made by Google to do stuff that applications made by third-party developers just can't compete with (although open source allows development of bad-ass apps nobody can run).
So to recap: Google has a model likely to make it the dominant platform for mobile/netbook application sales. They have a lock on developing high-end applications third-party developers can't compete with; making them the major provider of applications purchased for distribution on the devices(millions of built-in sales). And finally they take a chunk of all third party sales through Google checkout.
To top it off, they based it all on a community-supported development model and a free Linux backbone. Now with open source, they are getting private programmers to complete the job for free. They control the final release and everyone kisses their ass because it's open source - never mind that open source isn't really a benefit if the end user can't choose what OS to install on their device.
That shit makes Bill Gates look weak! And we haven't even gotten to the obvious benefit of every device coming loaded with applications directly linking to google back-end services like gmail/docs/calender.
I'm just sayin - Google is the fucking Borg. If they are successful, it is an economic chokepoint in the physical space they are moving toward.
January 14, 2009 4:10 AM | Reply | Permalink
Very interesting comment (slashdot points for both "informative" and "insightful"...), and I think your points are worth keeping in mind going forward.
A nit to pick, though: I thought we (developers) could buy a truly unlocked, "you're on your own" G1? I've been tempted to get one if it's truly open.
That is not meant to diminish your point though. If Google controls access to the "official version" of the system, then
January 14, 2009 7:39 AM | Reply | Permalink
The G1 dev phone IS fully unlocked. It's available to registered Android Market developers. It'll take any SIM and it's fully flashable for $399.
http://www.engadget.com/2008/12/05/sim-hardware-unlocked-android-dev-phone-1-surfaces-for-399/
One of the coolest geek things I've seen in a while.
January 14, 2009 3:44 PM | Reply | Permalink
From dataGRIT white paper:
"...nevertheless, regardless of how successful any of these business strategies might be, they all share a fundamental security flaw: by maintaining a monopoly on the technology the company itself becomes a “single point of failure.” Even giant, well‑established corporations go out of business and the executives who run them do make bad decisions. The safety and privacy of our information are too important to be entrusted to just one company and a small group of individuals that may control it.
Transferring the technology into the public domain is the essential step to assure that no intentional or unintentional action of any single company will undermine the safety and security of the data and privacy of the users."
read the rest of the paper on
www.datagrit.com
January 14, 2009 12:00 PM | Reply | Permalink