Hacker News Reader

"I think by far the most twisted notion in the article is the idea that little changed between 1913 and 1967. I'm pretty sure technological change during that period had more impact on the average Westerner's life than any other 54 year time span in history. It takes a very peculiar world view to completely ignore the invention of television or to consider a biplane as being basically equivalent to a jumbo jet."

That's not what it says at all. He states that the 60s economy is still dominated by technologies that pre-date 1913. He even specifically cites the economically dominant technologies, instead of the vague assertion, "I'm pretty sure ...". He particularly notes aeroplanes and computers as being invented during that period and can be expected to have enormous economic impact. Which they did, but hadn't by that point. The same with television.

Looking at the rest of your comment, I kind of think you should delete and reread the article. Every criticism is incorrect or unsupported - like the MIT crack. He's clearly talking about the "thin client" approach to providing computing - he was wrong, but it was not an unreasonable supposition to make.

> He's clearly talking about the "thin client" approach to providing computing - he was wrong,

Aren't Chromebooks, tablets and smartphones thin clients to Google's mainframe, Facebook's mainframe, Apple's mainframe?

What processes the Siri commands? Who calculates the directions in your navigator?

Yep, sure. The thin client approach has grown substantially recently, mostly I think due to improvements in networking. But the author was talking about "20 years time", so the early 80s. What we actually saw was the arrival of the domestic computer. That was the period in which computing became ubiquitous, and actual physical machines doing the processing locally could be found in every business and home.

So, I would argue, that for the period he is talking about the fat client and stand alone models were much more dominant than the thin client, in contrast to his expectation.

Or to use his reference point, nearly every university had its own computer in the 80s, rather than using a thin client to a host at MIT/Oxford/Harvard etc.

And in general, despite the cloud, I think fat clients are still dominant. I.e. look at gaming, the majority of processing is still done locally, despite MMOs and the internet.

EDIT: I think most people expected thin computing to dominate. But it turns out to be easier to manufacture and ship PCs around than data.

I believe that back then, people thought thin clients would dominate because they didn't expect the price and size of a "real computer" to become so affordable and compact.

So, it wasn't because they thought the thin client approach was better, it was just the only realistic future they could see given the cost and size of computers at the time.

And, in 1963 Moore's Law had yet to be recognised. If it had been, he might have realised that fat clients were going to dominate the future of computing. But at the time, it would have been near impossible to believe how relentless CPU progress was going to be.

> So there were no computers at MIT in 1967

You cpmpletely misunderstood that paragraph. He said that MIT would still have its own computer in 20 years because knowledge is its core business, but that other institutions would, instead of owning their own computers, use rented ones remotely.

He was basically predicting cloud computing.

Strange, I read it and had the opposite reaction. With the exception of the 'steel mill' quote you mentioned, I think he got it mostly right. Even that can be argued to be right to a degree, for a time the mainframe model was in vogue and now we're moving towards the 'cloud'. The little change mentioned between 1913 and 1967 was with reference to the overall trajectory of industrial economies, not technological innovation.

The main thrust was about the increasing importance of knowledge and information, where he was spot on.

Yes but that's a very broad observation that I suspect many people could have arrived at. Generalization is important but if the person doing the generalizing is so ill-informed about the details of their world I think the likelihood of their generalizations being correct is rather low.

If you replace 'steel mill' with 'data center', he was not so far off. At least if you don't count phones, tablets, laptops as computers.