Possible, Probable, Personal: Arguing against Castles in the Sky

Introduction:

 

I labeled this article with the heading ‘Possible, Probable, Personal’ because I think that a lot of failures in qualitative forecasting and putting the boundaries on quantitative forecasting result from an inability to differentiate what category new technologies and forecasts about technologies fall into. Like the flying car mentioned in the last post, people who were enthusiastic about it jumped straight from it being possible to it being personal, instead of an intermediary category of ‘Pragmatically Improbably’.
It is my hope that this framework or one that will evolve from it will help people understand why some technologies take a long time to make it to market, some are adopted immediately, and some never see the light of day at all (failing other interests, which are beyond the scope of this blog).

 

Is a technology impossible?

Flying cities--depending on how you count them, improbable or impossible. Credit to http://solartistic.deviantart.com/art/Laputa-The-Flying-City-385811018

Flying cities–depending on how you count them, improbable or impossible. Credit to http://solartistic.deviantart.com/art/Laputa-The-Flying-City-385811018

The first question to ask when evaluating the future with regards to a new technology or scientific discovery is to simply ask if it’s possible. Rather than establish all the different ways something could be determined to be possible, we can establish the ways that technologies could be easily determined to be impossible. Now, there could be specifics to certain scientific and technological fields that I don’t cover here, but I believe that these are the major ones.

  • Axiomatic impossibility: This is a scientific or technological discovery that violates the absolute most fundamental principles that we understand about the universe. This might include things like the values attached to fundamental forces, entropy, or the fact that there are no integers between 3 and 4.
  • Physical impossibility: This is a scientific or technological discovery that is completely un-grounded by what we understand about the universe currently, but doesn’t violate an axiomatic statement. This might include FTL travel, vacuum energy systems, anti-gravity, etc.
  • Conditional impossibility: This is a scientific or technological discovery that is impossible but only in a relational sense. In other words, describing one technology as being required to have higher capabilities than a different technology that it has never been shown to have an advantage over.

Now, as is the case with both of the previous articles in this series, none of these are absolute statements. As someone who studied Physics in my undergraduate years and still follows discoveries with the eye of a keen hobbyist, I’m well aware that there are still a number of interesting inconsistencies. All the same, just because we don’t know what the answer is doesn’t mean we can arbitrarily say that some result is likely–something that you can’t even comprehend or guess is just as likely, which is to say completely made up. It’s even worse in cases like FTL, where relativity is one of the most confirmed results in all of physics and no amount of wishing will get us around that. Choosing one outcome over another for no other reason than you like what it might mean is intellectually dishonest.

 

Is a technology improbable?

This question is much harder than the impossibility of a technology–and rightly so, as it relies significantly more on opinion than anything else. Well, perhaps  not opinion, but arguments are likely to be driven b opinion which will lead to cherry picked facts. All the same, we can still attempt to try to break it down further anyway.

  • Theoretically improbable: the most strict form of the term, a theoretically improbable technology is unlikely to be seen in even a laboratory or purely research sense, due to some restriction on the part of funding or even ethics. An example of the first would be experimenting with large equipment made out of rare earth metals (arbitrary-I do not know at the time of this writing if there’s any POINT to making large objects out of rare earth metals) and an example of the second might include human cloning or a number of experiments likely to leave their subject maimed, dead, or out of their mind.
  • Pragmatically improbable: while not burdened by some form of practical hard cut-off like theoretically improbably technologies, a pragmatically improbably technology is one that could theoretically be constructed unburdened by any realistic concerns, but is unlikely to be implemented on any larger scale. This relates back to the discussion of flying cars in the previous post, in that I’d consider flying cars to fall into this category. Notably, not even the military (which is well known for spending significant quantities of money on devices that wouldn’t necessarily be worth it outside of that context) has used ‘flying cars’ (when technologies that might be considered whackier, such as the Osprey, have been the focus of significant amounts of development effort because they were pragmatic).
  • Individually improbable: The last filter establishes that while a technology may see some form of large scale use (on an industrial, military, organizational, or government level–ie supporting a large number of individuals or requiring the support of a large group for the usage of one individual on behalf of that organization) it is unlikely to ever reach the hands of one person fro their own sake, either via purchasing or as an individual’s piece of equipment they used on their own behalf as opposed to the behalf of an organization.

Much of this blog will be about discussing the probability/improbability of various technologies and their implementations, though that specific discussion is likely to be far in the future considering the material yet to be discussed.

So, we’ve established impossible and improbable categories of technology. Many technologies will sit in one of these categories forever (in the case of impossible technologies) or for a very long time (in the case of improbable technologies). As our capacities advance, though, the scope of what is pragmatic is pushed back in some cases (transistors), though not all (flying cars), and things move to ‘personal’.

 

Is a technology personal?

So if a technology isn’t impossible, and it isn’t improbable–even on an individual scale, we can say that it may become a personal technology. In my opinion, technologies that are individually improbable (ie the loosest class of improbable technology implementation) are the most likely to eventually become personal, as they are often simply limited by results of scaling or development (mainframes, genetic sequencing, etc.).

Personal technologies are not necessarily technologies for private use. They may still be restricted to certain organizations–an example might be a certain firearm. The distinguishing feature here is that they are being used by an individual on behalf of themselves (and possible additional individuals as a side effect, but not as the point). Exoskeletons in warfare might go from ‘individually improbable’ to ‘personal’ when they become standard issue as opposed to being issued to squads (which itself is a forecast, though one I hope to cover in depth eventually).

 

Using this Framework

To conclude, this gives us a rough framework in which we can place technologies to evaluate their likelihood, at least at a very high level sketch. Axiomatic, physical, and conditional impossibilities can be examined first. If none of those prevent a technology, then theoretic, pragmatic, and personal restrictions on implementation can be examined. If the forecasted technology isn’t restricted by any of those reasons, then it likely is (or will be) a personal technology.

The “Flying Car Fallacy” and Why It’s Wrong

Motivation

Where is my flying car? Popular science cover.

Where is my flying car? Popular science cover.

One of my least favorite responses to hear when talking with someone about the future is what I call the “flying car fallacy”. While the precise phrasing differs from person to person, the most general form looks something like: ‘They promised us flying cars’ or ‘Where are our flying cars?’. The implication, of course, is that any form of techno-optimism or forecasting should be regarded with extreme skepticism due to the failure of flying car predictions to come true.

 

Honestly? The concept that you should regard what people tell you about the future with skepticism isn’t a bad one. However, while it’s not a precise corollary, I think Scott Alexander’s Innoculation Effect comes into play here. Because there’s this very basic, obviously failed prediction about the future, people applying the fallacy are able to dismiss all other predictions as wishful thinking or, if proven correct, lucky guesses. And as someone who likes to think that he can make useful predictions about technologies, that’s hardly ideal.

 

While the correct usage parameters for technological forecasting and trend tracking aren’t 100% established (and I hope to get a better handle on those exact things in this blog), there’s a large amount of data that shows that there is -SOME- predictive value. I think it’s worth it, from a professional or even just hobbyist perspective, to examine this misconception and take it apart. So, what powers the flying car fallacy?

Dissection

“They promised flying cars. There are no flying cars.” Perhaps a bit more verbose than the normal phrasing, but it serves my needs. I’m perfectly comfortable admitting that I’m deconstructing a particular phrasing of this claim that I myself constructed, but I feel the above version adequately represents the concepts and thoughts typically being expressed.

 

To examine this more closely, it needs to be decomposed. Let’s look at the first sentence: “They promised flying cars.” This is more interesting than it might appear at first glance. There’s a lot of meaning contained in the three concepts in this sentence. First, “they” implies some form of authority or expert. Honestly, while I can find plenty of general references to a ‘they’ or ‘scientists’ or ‘futurists’ discussing flying cars, most of these references fail to cite a specific source. Other controversial topics in futurism (such as AI) often have specific names claiming specific dates, a much more falsifiable claim.

 

An article on Livescience discusses broken science promises, flying cars among them, and mentions Popular Mechanics and Popular Science. Further investigation reveals that discussion of personalized flying transportation (and railway cars, something that did in some ways come true) go all the way back to the 1930s and 1940s. But still…these were popular magazines. By those years, they had already transitioned into appealing to the public mind and eye as opposed to serious, calm discussion.

 

This isn’t to say that experts haven’t forecast flying cars–but that allows us to take a look at the second word in the first phrase, “promised”. What does “promised” mean? Well, it might be semantics, but to my mind that means that people are taking a personal guarantee that flying cars will exist rather than a prediction, projection, or forecast. Forecasters who “promise” something about the future are generally not experts but popularizers–experts will generally couch statements carefully.

 

What does Google say? Well, if you search explicitly for the string “promised flying cars”, it returns about 239k results. If you search explicitly for the string “predicted flying cars”, it returns only 19.5k results. Other configurations return even less. Because it’s a personal guarantee, a promise, people feel betrayed. And that betrayal sticks in people’s minds as a failure for an interesting future to come about, disregarding all the other things that HAVE happened. Things that have happened become natural–things that don’t happen are failures.

 

Finally, the term “flying cars” itself is a bit of a misnomer–or at the very least unspecific, and looking at exactly what it means lets us address whether or not flying cars ‘exist’ or not. The off the cuff thought is simply a car that is unconstrained by the rules of the two dimensional road. If you think about it a bit more, you conceive of one of two things. The unrealistic/idealistic version is trivially having access to flight to ease up traffic without requiring building additional expensive infrastructure. The more realistic/pragmatic example, however, is a car that also works as a plane, and is subject to all the restrictions that entails (something that has been shown in prototype multiple times, with the most recent being the Parajet Skyrunner, which works off a parachute and a very light body, and the Terrafugia Transition, which is more technically a roadable airplane).

 

 

Why Didn’t They Happen?

Okay, so we’ve established that actually using the flying car fallacy generally has to do with a failure to live up to a flight of fantasy that appealed to the imaginations of readers and fans of technologies, rather than a serious, pragmatic estimation of the future. But why DIDN’T they happen? Well, to go over it very briefly (as there are more issues holding back flying cars than can possibly be dealt with in a competent manner here, but they have been discussed in depth elsewhere, we’ll keep it brief):

  • Safety: Three dimensions and velocities make things much more dangerous, meaning that there will necessarily be quite a bit tighter…
  • Regulation: Pilot’s licenses are more difficult to get, violations would be punished more harshly, variations would take longer to be approved.
  • Cost to consumer: Flight takes more fuel than driving, not to mention the fact that the additional engineering required would mean that vehicles would be drastically more expensive–in turn, leading to the fact that a comparatively small portion of the population would use the needed infrastructure. Speaking of…
  • Cost of infrastructure: New structures to support new dimensions of travel, new ways to deal with cars taking off and landing, etc. As mentioned above, this would be required for a comparatively small portion of the population, unlike roads which (with the exception of some private toll roads) are used by both expensive and economic cars.

Arguing Against It

To argue against the flying car fallacy, you need to clearly delineate between different types of technological predictions (or promises). Making sure that wild unsupported flights of fancy aren’t put forward as representative of your beliefs is ideal. In fact, one of the major things you should take away from this is that you SHOULDN’T let your hopes rise when someone claims something about the future, but can’t present rigorous evidence as to why or how it will happen. The next blog post will go into how to counter poorly framed techno-optimism and promises.

 

Citing the lack of flying cars isn’t the only pithy response people make in response to positive predictions about the future, but it is most certainly a common one. However, it’s founded on a bitterness on the lack of follow through from visionaries more than any widespread failure on the part of people who make rigorous evaluations of future possibilities. I am in no way trying to say that people should by default accept optimistic predictions about the future, any more than they should reflexively dismiss them.

 

In the end, if someone wants to believe that a failure of techno-optimism to bring about their hopes means that all forecasts fail, you can’t convince them otherwise. The best thing you can do is try to show them amazing things that WERE predicted ahead of time.

 

Thanks to Paul Bragulla for Proofreading and Discussion

Status Quo Fallacy

It’s almost inevitable that if you openly speculate about new technologies that have any impact on society that change the status quo, you’ll be met with a pithy response of ‘Previous technologies didn’t change (Relevant aspect of society), what makes you think this one will?’ It might not always be phrased that way, but it’s certainly a common sentiment that I expect many readers will have heard.

Of course, this pops up most commonly when you’re discussing automation. As an example, if you discuss the displacement of workers in various industry segments and skill levels, you’ll be met with responses ranging from ‘technology opens up more new jobs than it replaces’ to ‘the luddites originally complained about being put out of work too’. Incidentally, for that last example–turns out many luddites WERE put out of work.http://books.google.com/books/about/Writings_of_the_Luddites.html?id=NG6ABlDQ10MC That’s not to say I agree with them-history bore out that mechanization of industry improved the lot of the working class in Britain tremendously…but wage suppression still hurt in the short term.

One of the easiest ways to model the future is to assume something won’t change. This isn’t accurate, but it’s easy. People like thinking that what they are familiar with will always be around. Sometimes it’s selective–people will isolate one or two issues they have with the current dynamic of society and humanity, and look at how they  might be affected by technology, but assume the context is the same.

More often, they might pick one or two coming technologies and see how they cause a systematic impact, but not take into account momentum of existing social structures and whatever underlying human motivations cause them (unless the technology is one that directly alters the human condition, such as various neuroprosthetics or nootropics). This is particularly common, and problematic, with people whose futurism is driven by a particular political ideology they wish to see come to pass (or not come to pass, depending on if they are seeking utopia or avoiding dystopia).

The pithy way to phrase this belief, what I call the ‘Status Quo Fallacy’, might be the statement that “Quantitative change will not lead to qualitative change”. Can we argue this point? Is it possible to point to any sort of historical popular statements or predictions that stated that quantitative change wouldn’t lead to qualitative change?

Conveniently, yes! Professor Donald Simanek of Lockhaven University of Pennsylvania has conveniently collated a list of some of the best overly negative/pessimistic quotes about the future of science and technology throughout history, as well as a few overly optimistic ones (which we will revisit in a later article).

I strongly advise reading it yourself for some degree of amusement, especially one of the earliest quotes:

I also lay aside all ideas of any new works or engines of war, the invention of which long-ago reached its limit, and in which I see no hope for further improvement…- Sextus Julius Frontinus, governor of Britania, 84 C.E.

Regardless of ancient humor, there are some more relevant quotes. As the topic of this blog concerns science and technology, and as I’d like to avoid stepping into politics and theology, I’ll disregard those quotes. Furthermore, I think the point is best illustrated in a case where a technology had already been demonstrated as achievable and practical, if not quite ready for mass production. The final requirement is that the quote comes from someone with a degree of authority on the topic or at the very least understanding–luckily, nearly all of the quotes fulfill that requirement.

What does that leave us? Erasmus Wilson claimed that the electric light would not surface again after the Paris Exhibition closed. Lord Kelvin, a mathemetician and physicist, claimed that radio had no future outright. The inventor of the vacuum tube, Lee DeForest, said television had no future. Even the head of 20th Century Fox, who had made his fortune on the movie boom, claimed that television would never gain market share. These are all from the telecom industry, as they in many ways are the predecessor of today’s computer and data industries.

What’s the commonality between these? Individuals who knew systems intricately, who helped develop them or had helped develop their predecessor, were blinded by the status quo. When you are living in a system and much of your life revolves around that system, you either don’t want it to change or you can’t let yourself expect that it will change.

Does this still apply today? Absolutely. People have continued to make predictions, especially about computers. But every time they can’t show, quantitatively or qualitatively, why their prediction holds true for the future instead of simply being a hold-over from the past, you have to treat it with immense skepticism.

This is not to say all negative predictions should be ignored, however. Many have completely valid reasons, and should be used to construct a rigorous model of the future.

In further articles in this series, I hope to examine precisely when various failed predictions about technology came true, and see if there is any sort of consistency between when something was getting enough attention to make a bold statement about it, and how long it took to come true. I will also discuss the value of tempering enthusiasm (lowering the ceiling as opposed to raising the floor, so to speak), and being able to differentiate between different types of failures in technological development.