Industry question of the day

[yesthattom]

Intel x86 (Pentium, etc.) CPUs became very cheap because of the consumer market gave it economies of scale, and the people using it for niche things (like servers, web servers, embedded systems, etc.) benefitted.

In the The Economist article about Google Apps for your Domain, there is a throw-away line

In the past, innovation was driven by the military or corporate markets. But now the consumer market, with its vast economies of scale and appetite for novelty, leads the way.

which makes me think: is this going to be as big a shift as the consumer use of Pentiums?

Question for the group: What will be the trickle down benefactors of this change?

Update: Ok, Ok, it was the business use of Pentiums that gave consumers a cheaper chip. I reversed it. But more importantly, I don't mean just Google Apps, I mean all Web-API, Mashup-able, web-based applications.

Comments

[niherlas.livejournal.com]

Observations:

Innovation driven for government/military/corporate often results in a laserlike focus to generate a specific high-cost must-have technology for single/few uses (ie, Apollo rocket technology).

Innovation driven for a consumer market requires a business case, cost-benefit analysis, and market study proving that the new idea will sell enough units quickly enought to boost the corporate stock in time for the year-over-year comparison.

That isn't innovation. For the most part, it's refinement. Take a successful widget and make it smaller, faster, longer-batter-life, larger screen, add a camera, MP3 player, HD video, iPod integration, etc.

Should the consumer market - dominated by shortsighted stockholder capitalism - become the cradle of technological innovation, then innovation will likely become as effective as our government (which is dominated by the shortsighted need to generate $$ for re-election).

No long view. No "pay now for big results in a decade".

[xthread.livejournal.com]

Interesting. What would you describe as the last three successful 'pay now for big results in a decade' projects?

[niherlas.livejournal.com]

Darpanet
Shuttle program
Bell Labs/PARC

Forgive my lack of clarity - in the past ten days I've been through the PacNW windstorms, christmas, and a very sick household. But one of the things common to "pay now for big results later" projects is that they create science and technologies which - at their onset - often seem to lack consumer application. But the innovations that these "money sinks" (which is what your quarterly-report stockprice-watchers would call them) generate have a slow but inexorable effect on many surrounding industries.

Of course, it doesn't just require large sacks of cash thrown into a think tank. Focus, intellgent management, an environment of exploration, and careful management of resultant IP are all key.

These are the things that create the future. The insatiable demand of the consumer market is what refines it and hones it to a sharp edge.

[xthread.livejournal.com]

Hmmm.. I don't think I agree with your description. Darpanet was useful to the folks who were building it while they were building it, and things that weren't didn't get funded or built. Everyone building it wanted it to be much, much bigger than it was, but the funded projects were funded because they did something then that was useful. Bell Labs was a similar story, as was IBM Research - while both divisions were chartered to perform 'blue sky' research as well as solve concrete, ongoing operational problems, a whole lot more of their concrete solutions made it into the world.

The one clear exception is the Shuttle program, which was definitely a 'pay now for returns later' sort of project.

[gulfie.livejournal.com]

The shuttle was a program forced thorugh so that the spy satalite flyboys could get there packages to and from space without anyone else knowing what they looked like. The big shuttle volume was based on that requirement, everything else was basicly a cover story and buy off of the technocracy to keep them STFU.

[kimuchi.livejournal.com]

Even if the project as a whole isn't successful, or isn't successful in a way that brings the primary object to consumers, sometimes the side inventions become hugely successful. C.f. teflon and the nuclear bomb program.

[xthread.livejournal.com]

Your premise is broken.
The Intel CPU line became very cheap because the business market gave it economies of scale. Once it was cheap, the mass consumer market came into existence.

[gizbot.livejournal.com]

(Ignoring the man with the tail until signal-to-noise improves...)

I'll bite on your premise: the consumer market rarely creates new technology from scratch but it does develop and fund new technologies when they can be integrated into an existing product line.

The current explosions are in cellular phones, automobile crash avoidance, and networked video. Once one can talk into a little box, every new feature on top of that can be leveraged into market share for the "talking boxes" market. Cameras, advanced sensors, biometric security, walking credit cards, wireless music services, etc., do not improve cellular phone service, but they do affect market share. Similarly, funky safety devices in cars and all sorts of entertainment-on-demand are in competitive markets. These markets are technology competitive primarily because none *really* does what the consumers want.

Coffee shops, retailers, toiletries, airlines, and other large markets all work and are competitive over price and execution. One coffee shop might be better than another, but they all offer free-wifi, espressso, sugar, etc. A technical breakthrough ("fresh rasberries mixed into the whip creme" or "high resolution latte art") will not dramatically change market share in these markets.

Consider what the consumer really wants in these cases. They want everything they could do on a laptop available, one-handed, on a cell-phone. They want cars that simply do not crash. They want whatever game, video, or music that pops into their head to play whenever and wherever they think of it.

[xthread.livejournal.com]

I'll bite on your premise: the consumer market rarely creates new technology from scratch but it does develop and fund new technologies when they can be integrated into an existing product line.

How do you explain the existence of cellular phones, television, or telephones?

[tim1965.livejournal.com]

Bullshit. Innovation was driven by government investment, not by the military. And hardly any modern innovations were driven by the corporate market! The history of technological innovation is one of government intervention, not one of the free market bringing us closer to consumerist bliss.

An economy of scale can be achieved only when the marginal cost of each additional unit is very low. The consumer market has nothing to do with it.

I'm not entirely clear what the "this" in the sentence "is this going to be as big a shift as the consumer use of Pentiums?" means. If you mean that Google apps for everyone is going to be a big shift, I'd say: Duh. Talk about the obvious! "The Economist" ignores the fact that software development is a classic case of economy of scale. Creating one application on one computer is very costly. Creating one application and then getting 1 billion people to use it is not. That's a model which was developed by the book printing industry (talk about old skool technology!) more than 500 years ago.

Trickle-down benefits? It depends. What advantage does Google derive from having a half-billion people use its apps? Does Google merely expand its advertising empire (e.g., forcing users to pass a Google ad through their Web site)? Or is this an example of loss-leading, where Google sells the product for free and then requires you to buy it later on? One way or the other, consumers will pay. Google is not a charity.

[xthread.livejournal.com]

Innovation was driven by government investment, not by the military. And hardly any modern innovations were driven by the corporate market!

Like, say, the semiconductor?

Creating one application on one computer is very costly. Creating one application and then getting 1 billion people to use it is not.

Do you actually mean that? I think it would come as a great surprise to pretty much the entire industry...

[tim1965.livejournal.com]

You mean Fairchild Semiconductor wasn't financed by government money in the form of military contracts?? You mean Texas Instruments wasn't financed by government money, either??? Or that Bill Shockley, Walter Brattain, and John Bardeen weren't all educated with government money, didn't work in companies (like Bell Labs) financed largely with government money or educated in universities primarily financed with government money?

That is an astonishing claim.

As for your second point, think about it. Creating one computer program may cost $10,000. And the one user pays the $10,000. But say you can get 100 users to buy the same program. The cost of copying each disc might be 5 cents per disc. Now total costs are $10,005. Per-unit costs drop to $100.05 per customer. That's called an economy of scale, because the marginal cost of producing an additional item is miniscule compared to the cost of producing one (or just a few) items. This should surprise no one in the industry.

[xthread.livejournal.com]

I think you're burning down a straw man.

You mean Fairchild Semiconductor wasn't financed by government money in the form of military contracts??

Bell Labs invented the semiconductor in pursuit of solving AT&T's problems; Yes, Fairchild and TI went on and did really valuable things with it, but the innovation of creating the semiconductor in the first place went on at Bell. But what leads you to describe the research arm of the largest regulated monopoly in the world as 'financed largely with government money?'

As for your second point, think about it. Creating one computer program may cost $10,000. And the one user pays the $10,000. But say you can get 100 users to buy the same program. The cost of copying each disc might be 5 cents per disc. Now total costs are $10,005. Per-unit costs drop to $100.05 per customer.

However, that is unfortunately far and away not what goes into, as you described it, 'Creating one application and then getting 1 billion people to use it' - getting even one hundred customers to use something costs rafts more than it costs simply to make a hundred copies of the product, in general.

[yesthattom.livejournal.com]

Actually, when I worked at Bell Labs I was shocked to learn the dirty little secret of Bell Labs was that a big part of their $ came from the government project to build the missile systems of the 1950s, 1960s, and beyond.

Much of the electronics invented at Bell Labs was in a push to develop the guidence systems for missiles. If you think about the technology you had in, say, a Commodore PET in 1980 might have been enough to guide a missile... and then remember that they need that 20-30 years earlier. (I'm making a bad analogy... you are technical enough to know the history of electronics).

Anyway...

I was quite shocked to learn how much $ came to Bell Labs to fight the cold war. And it haunts me to think that the end of the cold war may have contributed to the end of Bell Labs.

[gulfie.livejournal.com]

Long term thought.

10 years of exponential growth can mean nothing, then 'suddenly' it takes off as far as the common understanding, even if it's just continuing it's exponential growth.

(as to others comments)
There are _many_ cases of earth shattering (literally) innovations that were caused by individuals or small companies coming up with new ideas. A plow made out of steel, this is why everyone knows who makes green tractors. The cotton gin, one guy. The flowbe, you name it.


The differences are of required capital investment. A Plow or a flowbe, may be invented in the spare time and capital of one or a small set of people. A new jet plane or trans continental switching network... maybe not so much.


As for the 'new consumer based novelty' crap... bullshit. Horse shit. Pig shit even. Every 'novel' thing that comes out of these large hierarchical structures (government, businesses whatever), has most of the novelty pounded out of it before it gets to market, and the customers didn't ask for it. The marketing departments and visionaries did.

Effectively the corporations are offering behavior, in the behavioral training sense of the phrase, hoping they'll get a positive response. Or so goes there argument... but with multi billion dollar advertising and marketing strategies... well the little hick field mouse has no hope of surviving in the big city. When a multi billion dollar well crafted message of want crosses your mind... well many people fall for it, in the best of cases they even think it's themselves who came up with the idea of wanting it. The customer isn't driving, the companies are driving the customer to want it. This novelty argument, as the french might say, 'eis shit'.


If you need any more proof go grab a number two pencil, and write down on a piece of paper 'carbon is what De Beers is selling'.



Now to a different tack on the question. What say you about the English / Chinese opium wars of the 1800's? You could make a valid argument that opium, or tea and the sterling silver they demanded was all consumer goods. Then why were so many people fighting, and dieing? Corporate (aka consumer's often time non consentual top from the bottom) and military (aka government, typically working on behalf of the corporations), are not so stricly deliniated.


Second, Google apps, unless they are including some sort of games package, may have nothing to do with novelty and everything to do with all the other solutions (a.k.a. Lotus Notes, or MS Exchange) suck more, and often require vast quantities of money, time and energy to get working right. Where as an outsourced solution nicely avoids all that cost and dissapears most of the risk to the purchaser.


I would argue that the primary effect that he's talking about is really the 'corporate leaders' trying to justify there new products and corporate tactics. Be they SUV's with $10k worth of electronics and flat screens that always seem to be in the shop or cellphones that have no battery life because they are attempting to be and do things that a fair segment of the market doesn't want. It's a simple fact of the current economy that if you make a good simple tool that works and works well, you go bankrupt. There are rare exceptions, but this is basicly the rule. So companies atempt to 'differentiate them selves in the market', remember the unix diaspora? and why everyone had to be different? Because you can get the same for free (a.k.a. no profit margine to speak of).


Half the reason Google is where it is, is precisely because it's not following the novelty feature rich crazy complexity bandwagon of insanity. Sure they are slurping in companies and offering other tools, but they are to some extent keeping the tools simple, not bells and whistles novelties. And most importantly it found a way to make money at it. It found an externality to the use of a simple tool that now flows billions of dollars.


4300 character limit... my ... something something...

And another 510 chars! too!

[gulfie.livejournal.com]

Search is search, maps is maps, email is just some well UI'd and most importantly _functional_ (unlike previously mentioned products).


It's not the novelty of going to amazon and buying a book, it's the simplicity and ease. Starbucks may sell complex and novel drinks, but that's not why people buy the majority of them. It's the simplicity. It's the effectivity of the simplicity to the customer.



Yes I have had a tremendous amount of coffee today.... why would you ask that?

Désirent ardemment de phase la révolution !

Re: And another 510 chars! too!

[xthread.livejournal.com]

It's a simple fact of the current economy that if you make a good simple tool that works and works well, you go bankrupt.

I'm curious - who and/or what do you offer as your example?
And do you consider the iPod to be a counter-example?

Re: And another 510 chars! too!

[gulfie.livejournal.com]

Examples of 'non-inovating' companies or products leaving the sphere of buisness.

0) The main thrust of it was mostly the things we don't make anymore, or only made some of for a while. There are lots of examples of buisnesses that are running on equipment that is now nolonger produceed. Be they vaxes from a by gone era' ( now done in emulation ), or some wierd wire twisting machine that's not been orginaly manufactured for decades. These are mostly durable goods that are durable, and simply fixable, and you don't need that many of them. Which is exactly the buisness problem. Good product, never needs replaceing... poof. out of buisness.


1) AMPS.
The advanced Mobile phone service.
It just worked (tm)... well after they got the kinks worked out. It went away because more calls and 'services' could be handled on the 'newer' and 'better' digital systems.

2) The bag phone.

3) The panasonic duramax.

4) Cell Phones without color screens and wizz bang 'graphical' interfaces.

5) basicly every non microsoft computer or software company, any number of imbeded solutions, or market specific products that died during the 80s and 90s, primarily because of some marketing lies about 'the new system being so great and maintinance free'...

6) Any number of industrial equipment man


If you'd like an oposite case, look for anything that is designed to fail. Engeneered obsilecence. Sure, cars, roofs and expendable items are a good exampls. My favorite is the standard light bulb.

To some extent this has been changing for the past decade and a half but hear out the argument.

The cost to produce an Edison based incandecent light is basicly free. There is very littel material in it, the machinery that builds them has been paied off for decades and the process is rather tuned in.

Almost no one buys lamps based on light / power ratings. People just go out and get another 100 watt lamp, when the previous one dies. If you wish you can purchase extended life bulbs, they run at a slightly lower light/power rating but at a much higher cost, primarily due to there application in hard to reach or expensive to swap out fixtures. Increasing the MTBF for normal bulbs would not cost that much, but the savings would go to the consumer and why would a company do that?

Re: And another 510 chars! too!

[gulfie.livejournal.com]

Sorry, Ipod.

The Ipod isn't a durable good. An Ipod that lasts 2 years is an amazing thing. Maybe the shuffles and the newer more flash based things will last longer.

The ipod also has a nice healthy profit margine on it, and most importantly a vector for reocurring income. A.k.a. the itunes store.

Re: And another 510 chars! too!

[xthread.livejournal.com]

The Ipod isn't a durable good.

Absolutely, sold.
Should it be?

[awfief.livejournal.com]

I think there's evidence that this kind of big shift has happened again -- look at the ipod. That's not military-driven or for corporate use. And it's definitely an innovation.

As for Google Apps, I wouldn't say they're an innovation. There have been online document collaboration and sharing tools around for a long time. What Google is doing is what everyone wanted at the turn of the millenium -- a "portal".

It's hard to think of the benefactors, because you have to think of what it will be used for that it's not used for now. Sure, companies and institutions will (and probably do) use Google's portal -- for e-mail, documents, etc.

What would the niches be? Perhaps some kind of system that needs to get information from a centralized source will use an API into a Google App instead of some kind of centralized company database? Perhaps there will be some kind of Google DB available, such that applications similar to each other can share data? ie, if I want to move all of my journal posts from livejournal to my personal wordpress, if there's a "Google DB" backend, it might not be so difficult...who knows?

Perhaps it will lead to actual real standards with things like newspaper articles, recipes, etc? Maybe it will drive XML and XHTML forward?

[yesthattom.livejournal.com]

The iPod wasn't military driven, but the drive for small, fast, portable hard drives was. The iPod benefitted. However, now I think iPods and laptops are the driver more-so than the military.

I didn't mean "just google apps", I meant all the various web-based systems... Amazon, Yahoo, Microsoft all are doing these apps.

(The Google DB is called base.google.com, but obviously it isn't a database as the queen of MySQL would want.) There is gData, which is a read-write version of RSS (actually, RDA, I think. I can't remember).

Thanks for your thoughts. I like the ideas of an XML scheme for recipes.

[xthread.livejournal.com]

Time-shared computing is, by definition, not new. Google has re-invented it. This is cool. But it's not new - go look at discussions from the period 1968-1974 about the future of technology, and in particular the future of time-shared computing, and you see the world of web-based applications.