Monday, June 18, 2007

How Good Are Expert Predictions?

Magazines are ephemeral, but some of my friends compulsively keep stacks of copies of old magazines. I've always wondered what possible use these collections can be, but here's a lovely contribution one of my readers sent, taken from Popular Science of May, 1967, page 93.

"Time sharing, most experts agree, is the key to the computer's future, at least for general use. A few years ago, when people thought about household computers at all, they though of some small, inexpensive, individual unit that would keep track of the family checking account and automatically type of Christmas-card labels. Now we know it won't be like that at all.

"The reason is economic. The bigger and faster the computer, the cheaper it makes each computation. Consequently, it will be far cheaper to build one monster computer with thousands or even millions of customers hooked into it than to have small, individual machines in individual homes."

Now we know that "most experts" were wrong: we know it would be like that, because today, 40 years later, it is like that. I was something of an "expert" in 1967, and I'm proud to say that I wasn't one of those who made such a piss-poor prediction. That's probably because I don't make predictions—except the prediction that almost all of the predictions we make today will turn out to be piss-poor 40 years later.

Why do I make such a meta-prediction? Well, I've researched the past, and, as Patrick Henry said, "I only know the future from the past." But don't take Patrick's or my word for it. Here's how you can find out for yourself. Beg, borrow, or steal a copy of some old computer magazine. Spend as much time reading it as you typically spend on this month's issue of the same publication (or an equivalent one, if the old one is no longer around). I guarantee that the time spent on the old one will be more productive.

Because I was an "expert" in the 1960s, I published a number of articles in the leading computer magazine of the time, Datamation.. I do save my old articles, so I happen to have a copy of Datamation. from September, 1962. My article in that issue is entitled, "How to Automate Demonstrations."

Although the print magazine Datamation. itself shuffled off this mortal coil in 1997, I'm proud to say that my 1962 article would stand up pretty well even today. Perhaps even better today. Now that hardly any part of the computer moves, demonstrations are much more challenging to create. Of course, this was supposed to be a humorous article, though not everyone realized it at the time. I received a dozen requests for the Demonstration Compiler—that is, the compiler that compiled fake demonstrations. (Hmm, is there any other kind?)

On page 79 of that issue of Datamation., there's an advertisement from Computer Dynamics of Silver Spring, Maryland. (What ever happened to them.?
"MEMO Re: COMPUTER TIME
Solve your computer problems efficiently and economically by using our 32K, 10 tape IBM 7090 at $450 per hour." (That's about $5,000 per hour or more in today's dollars.)

Today, 45 years later, I own five computers, each of which is far more powerful than that 7090. As far as their value, I've thrown away a more computing power than that because nobody wanted it. Yes, the ten tape drives would still be a bit expensive today, but why would I want them? I own more than a dozen disk drives, each of which stores far more than those ten tapes.

The list of advertisers from that issue contains many forgotten names of companies selling computers, plus a few companies that are still around but no longer selling computers. Here's some examples:

PHILCO "Philco's on the move."

RCA "What's new at RCA is news in EDP."

GENERAL PRECISION (Surely everyone remembers the RPC-4000.)

ASI "More computation per dollar—on the ASI-210."

GENERAL ELECTRIC "Progress is our most important product."

FRIDEN "This is Practimation."

AUTONETICS "It's called RECOMP III."

TRW "Be operational now with the TRW-130 (AN/UYK-1)"

BENDIX "Is your programming career in a closed loop?"

Bendix didn't actually advertise their machine (no, it wasn't a washing machine), but they were crying out for programmers. And so were most of the others, "from $7,000 on up."

Even IBM (who, at last look, was still around), was desperate for programmers to "shape the future of a new technology." Sound familiar? Although machines are millions of times faster and cheaper, some things—human things, mostly—don't seem to change in 45 years:

"IBM programmers ... are devising programs that in turn use machine capability for formulating new programs. They are creating programs that enable computers to diagnose their own faults through self-checking. And they are helping to design the systems that will let scientists and engineers 'talk' to machines in the everyday language of science and engineering."

Gee, I hope they finish these projects soon. I've been waiting a long time to talk to my computers.

Perhaps, in the end, all this flux of companies and jargon and sales promises is merely an illusion. Perhaps it's what doesn't change that teaches us the most important things about ourselves.

And what is it that doesn't change?

Us.

Oh, the faces change. The names change. But the behavior, the hopes, the visions, the gullibility—they don't change. Maybe that's a prediction you can safely make.

4 comments:

  1. Predictions - I once made one that actually came true. In the early 1980s I visited several "poor third-world countries" in Asia. I had just graduated college where I learned to program in FORTRAN and PL/I (actually used a textbook on PL/I written by some Weinberg guy).

    I looked around at the masses of "poor third-world -ites" who could read, write, and do math. I told my colleagues, "I could teach these guys to write FORTRAN and PL/I programs." What I was really trying to say was that these poor guys could learn to program.

    Well, I was right! Today these "poor third-world -ites" are programming all day long at salaries that are low for Americans and high for Asia.

    The list of my incorrect predictions is much longer. sigh

    Dwayne Phillips

    ReplyDelete
  2. Hi Gerald,A real cool blog...Keep it up!!

    ReplyDelete
  3. Just wanted to let you know that I am an "old" Bendix G15-D computer programmer. They made one of the first office computers - the size of a filing cabinet, vacuum tubes, 8KB of drum memory.

    ReplyDelete
  4. The tangible question is not to challenge the expert predictions. But to assess the background of the expert before challenging the expert predictions. Is he/she a real expert? Does he/she have a broad view that enables them to connect that specific situation to a greater picture? This is the measuring meter stick for good consultants and strategists. ... Most ppl. think that their understanding of a limited situation makes them an expert. ... A Chinese strategy scholar once told me that most people do not knows what is the standard of an expert. If so, we have too many wanna-be experts running around in this world.

    ReplyDelete