Alan Kay: "We're running on fumes technologically today." | Hacker News Reader

Alan Kay: "We're running on fumes technologically today."

windley.com

63 pointsasciilifeform20 comments

I think that the reason for this is that businesses are very reluctant to adopt revolutionary technologies. They are much more comfortable with technologies that offer a small and defined improvement over what they are using (eg. typewriter to word processor).

If you pitch something completely new, many people will look at you like you are from Mars. My startup's product has the potential to drastically improve the profitability of a lot of businesses. However, when we pitch it for what it is, people tend to react fairly violently, because it runs counter to a lot of their intuitions.

In order to get any traction, we have had to pitch it as an incremental improvement, with the real game changers hidden behind the scenes.

They had to do that with Javascript, too.

I think you are right on. This is how fiber optics gained their foothold in the communication space. Coax was reaching its limits in terms of transmission amounts and cost of improvements, but people were skeptical of switching over to an unproven medium due to the many different technologies that had to come together to make it work (fibers, lasers, semiconductors). It took many iterations of fiber technology and many years, starting with "large" graded-index fibers with laser-based repeaters to get to tiny single-mode fibers with optical amplifiers.

I think that the reason for this is that businesses are very reluctant to adopt revolutionary technologies.

It's no surprise that the revolutionary technologies of the internet were developed and deployed by (gasp) the Government. It wasn't until the internet was already well-established that commercial interests started to pay attention.

More academia than government, but mostly on government's money.

Actually, the initial development was done almost entirely by think tanks (Rand Corp., BBN, etc.) and academics, with the crucial piece being supplied by the concept of packet switching, invented by Paul Baran at Rand Corp.

The essence of government is centralization of power over individuals. The essence of the internet is decentralization of communication among individuals.

The internet was a revolutionary invention because it solved a revolutionary problem: how to maintain communications after a massive nuclear attack. Pentagon (DARPA) money was behind most of this because it was the Pentagon that was responsible for solving the problem. More to the point, the key technologies that the internet depended on were the transistor (AT&T/Bell Labs) and the integrated circuit/microprocessor (Intel).

Transistors are an interesting case. AT&T was a government-supported monopoly, allowing it to pursue expensive blue-sky research. And since transistors were expensive, the government purchased 100% of the advanced transistors until they were viable on the market. http://tech.mit.edu/V128/N51/chomsky.html

The New York Times discusses how technology flows from federally-funded research to industry. http://www.nytimes.com/2005/04/02/technology/02darpa.html

I agree that national states are centralized institutions, but corporations are another type of hierarchical power structure with their own set of disadvantages.

I think the only way to get businesses to adopt radically new technologies is to start new businesses. This is one advantage small companies/startups have... they can leverage radically new technology as an advantage.

Eventually these new companies become the 'big companies'... but by then they are also hopelessly out of date because processes and tools have become ingrained and can only change incrementally. The cycle begins anew.

"people tend to react fairly violently, because it runs counter to a lot of their intuitions."

When will they ever learn? Wasn't it Thom Watson himself who said the world only needs about five computers? (Sorry, no record can be found.) Or Gates who said 640K is enough RAM for anybody? (Sorry, no record can be found.)

Many great things are counter-intuitive. Like, beating the House in Vegas. But, then, when they react violently, can't you show them a simulation that demonstrates that their intuitions can be beat?

The semiconductor industry, both in the design and manufacturing processes, has invested substantially in revolutionary technologies for several decades to deliver what is know as Moore's Law.Processes and methodologies in use today are drastically different than what was used even 8-10 years ago.

But profitability is a function of competitive alternatives. Many technologies act like electricity, they dramatically improve productivity without necessarily leading to higher levels of profitability because of adoption by competitors.

Interesting side note: Alan Kay challenged the Stack Overflow community several months back asking for "Significant new inventions in computing since 1980" (http://stackoverflow.com/questions/432922/significant-new-in...).

Sounds like Alan Kay is looking at the world of business through a pinhole in the floor of his car when he'd be better off looking through the windshield.

Anybody can automate payroll, inventory, order processing, etc. In fact, almost everybody has. I call that "striking out the pitcher".

But only some people have figured out how to use sophisticated software to run simulations, mimic real world processes, and as Michael Gerber puts it, "Work on the business instead of working in the business."

Apparently Alan Kay doesn't see many of these businesses. He should. Because they're the ones who'll be left standing while the "automators" wonder what happened.

I think this migth be some confusion between Mr Alan Kaye (http://people.forbes.com/profile/alan-kaye/51664) and Alan Kay of SmallTalk & Object Oriented fame.

"Kaye" appears to be a typo. "Alan Kay" was interviewd by Fortune Magazine's "Fast Forward" column in the July 12, 2004 issue: http://slashdot.org/articles/04/07/13/1226206.shtml

I would speculate that innovation in this domain effectively stopped when the complexity of computer applications surpassed what the layperson was willing or able to wrap their head around. This happened some time just after the spreadsheet.

At this point, businesses became highly dependent on skilled IT staff to hand-hold them through everything. IT staff are expensive, scarce and difficult to work with. They are also utterly preoccupied with rejiggering old ideas to work with flashy new technology, devoting very little time to truly new ideas.

The problem concerns all of computing, both home and office. A vast knowledge gap leaves most people helpless while the experts are buried in architectural cruft, and often unaware of it.

Maybe it shouldn't be that surprising that most of the big ideas seem to be behind us in computing. Automobiles have been fundamentally the same for decades (speaking broadly). It could be that new areas of technology are more like explorers finding new lands. The newer finds don't tend to be as big within that area after a while.

Coincidentally, Freeman Dyson recently said that physics has been undergoing such a slowing down. (His daughter was mentioned in the article). He also mentioned that the field of biology is speeding up. (Link to his Charlie Rose interview http://www.charlierose.com/view/interview/10560 (doesn't go into much detail concerning the given point, but it is Freeman Dyson.)

> Think about it this way: if you're in business, then it's likely that the most important process you have is your "order to cash" process wherein an order is received and, eventually, turned into cash. Who owns that process in your business? How was it created? When is it reviewed? Did someone build a model for it that can be simulated to test various scenarios?

We don't need to simulate this in many cases. We can just try different things and measure the results. This beats analysis and simulation.

You cannot extrapolate from running a small business 'by the seat of your pants' to the level of analysis that's applied by large corporations.

Large businesses do an enormous amount of simulation of their "quote to cash" process. There are complex demand forecasting, supply chain management, and inventory management software platforms offered by the likes of SAP and Oracle that are used for this. There are a number of decision tree tools used to simulate and analyze billion dollar investment decisions used by pharmaceutical companies, oil and gas firms, and other corporations that need to make large investments. Airline companies run "yield management" software to adjust pricing frequently based on estimates of demand for a particular seat on a particular flight. Wal-Mart adjusts pricing and offers specials based on complex models of how stimulate demand. There are many other examples of business using sophisticated simulation and analysis tools.

Article is from 2004.

Yes, but it may as well have been written yesterday

Reply on news.ycombinator.com