4 ms·
As always Amaras law comes to mind: "We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run." https:
by ThomPete 11y ago
As always Amaras law comes to mind:
"We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run."
https://en.wikipedia.org/wiki/Roy_Amara https://en.wikipedia.org/wiki/Roy_Amara
- dredmorbius 11y agoAre there any specific instances which Amara poses? Neither Wikipedia nor its PC Magazine reference list any.
- adventured 11y agoI think it's because the examples are so plentiful. Big leaps forward seem to broadly cause people to either significantly underestimate them (the home computer mid 1980s), or to significantly overestimate their near-term impact (the state of all things stem cells 15 years ago). It occurs almost every time there's a big headline about growing organs, curing HIV/AIDS, curing various forms of cancer, curing alzheimer's, and so on. It occurs with almost everything to do with gene therapy right now. Robotics has been under an almost constant state of overestimation for more than a century, while the industry has still delivered amazing technology in that time. Artificial intelligence has been treated the same way for decades. I think these two instances are heavily caused by fear, people are projecting far beyond what's rational. You see it in batteries a lot. Solar suffered from it for decades. Storage / memory seems to see the effect in action with nearly every breakthrough (the gulf between the date of first public awareness and reaching a commercial economy of scale representing that typically huge gap of time).
- dredmorbius 11y agoI think the "plentiful examples" exhibits very strong confirmation bias. Yes, exponential growth trends result in overestimating short-term, and underestimating long-term, changes. However very few phenomena show persistent long-term exponential growth characteristics. For all the obvious reasons. The growth rapidly becomes ridiculous and/or unsustainable.
- ThomPete 11y agoNot that I know off. But the first internet bubble seems to be a good example to illustrate the point.
- wcarss 11y agoJust to flesh out my parent's point, while the internet existed and was widely available in 1995, its impact over 3 years was far less impressive than expected, but its impact over 20 years is far more impressive than expected. The internet "bubble" is the surest sign of this. Many people thought that everyone would buy groceries via the internet by the late 90s. Grocery delivery via the internet can be a good business, but it wasn't ready 3 years after web browsers got popular. The experience wasn't good enough, the cost of operation was too high, and not enough people were there to use it. Today though, I can speak a few sentences to a wireless black glass box I keep in my pocket, and groceries will arrive at my house in the hour. Beyond that though, I can instantly feel a tingling buzz when someone on the opposite side of the globe wants my attention. Major celebrities have lived their entire careers on the internet. World leaders converse and issue statements via a short-message public-ish broadcast system we can all see instantly. Multiple currencies powered solely by shared public work and cryptography are in wide use. Major mathematical discoveries have been made by free associations of mathematicians in blog discussion threads. That's just the tip of the iceberg, but it's already far more profound than online groceries. The effects of the internet over decades are essentially unfathomable, but we were far too optimistic about them in the short term.
- dredmorbius 11y agoInformation technologies are the exception to many diminishing returns rules. This has several implications. Non-informational technologies often don't see such trends. Where they do they often manifest network effects (telephony, railroads, autos and fuel/road/repair infrastructure), or ... Nominally non-information goods with very high informational content benefit from Moore's Law growth. This can include very high tech manufacturing or control systems (aircraft engines), R&D and process control (pharmaceuticals), an similar goods. Even within high-eech and information goods, where the rubber meets the road, improvements are limited and the Jevons law limits overall efficiencies. Healthcare spending has exploded over the 20th century. Life expectency gains have been modest. Aircraft and ship speeds are lower today than 40 years ago. Greater data access runs into limits of human attention, and the fact that there are only 1,440 minutes in a day. An iPhone is in some regards amazing. In others it merely replaces pen and paper, books, radio, and live performances of entertainment. I suspect tech impacts may be greatly overstated.