Asking the wrong questions (2017)

He wants a glider. Not a problem. There are quite good hang gliders. There are also plenty of trolleys around, although most new ones use pantographs instead of trolley poles.

Very few saw a world dominated by giant advertising firms. Or computing becoming a branch of advertising. Even in science fiction. There was Fowler Schocken Associates, in The Space Merchants (1952). The company behind the simulated world in Simulacron-3 (1964) builds it so they can do market testing and opinion polls. As late as "AI" (2001), the tie between search and ads hadn't appeared. In "AI", the "Dr. Know" search service is an expensive pay service.

The lenses that we view the questions also change. In the 1950s, people likely imagined that by the 2020s, roads would be rebuilt with technologies like magnets or rails to support self-driving cars. But they didn’t anticipate the inertia in infrastructure development. Our roads remain largely the same, and this stagnation is what we need to band aid with for autonomous vehicles today.

Everything else aside, I’m impressed his grandfather was born in 1896. My most recent ancestor whose grandparent was born before 1900 was my grandmother, and she died 25 years ago. Some families have much longer generations than mine. I already knew we aren’t a hearty stock, but this difference seems ridiculous.

(2017)

"A Logic Named Joe" is a fascinating story, which has also been discussed on HN occasionally.

This could also be just asking the wrong people.

The predictions we have left are from industry expert or pretty successful people. Fundamentally they fit well in their current world and aren't envisioning social or technical shifts that will completely change the world as they know it.

This is most apparent in the telephone and international fax part, where they see the future of networking through telephone, and not some other technology making it obsolete. We'd have had a different prediction asking AM amateurs how they see the world of telephone communication in 50~100 years (might not have been correct either, but would have been different)

"plenty of people have pointed out that vintage scifi is full of rocketships but all the pilots are men. 1950s scifi shows 1950s society, but with robots. Meanwhile, the interstellar liners have paper tickets, that you queue up to buy. With fundamental technology change, we don't so much get our predictions wrong as make predictions about the wrong things. (And, of course, we now have neither trolleys nor personal gliders.) "

Yes, Asimov's Foundation has people smoking, reading physical newspapers and using physical money, lining up for customs when arriving to Trantor. No women until later on in the series (in his defense, he may have not talked to many women at the age he wrote the first novels).

There was movable sidewalks and other transportation devices though.

Just yesterday I had a "what's coming" discussion with a couple older (non technical) folk. They thought flying cars and fusion power were coming soon.

I contrasted saying that the energy equation for flying cars doesn't work, not to mention the penalty for mechanical failure. (I mentioned helicopters, they mentioned autonomous drones.)

Fusion power is famously "10 years away" but I maintain its simply too capital intensive. If I have 10 billion to invest do I want to make a stunningly complicated fusion power plant, (which will produce power 10 years after the project starts) or do I just buy a bunch of desert, a mountain of solar panels and enough wire to connect it to the grid? Staffed by some cleaners and electricians. Where the worst that can happen is it goes offline. With no moving parts, no sun-like pressures or temperatures.

And yet back in the 50s "free" energy and flying cars were "imminent".

With the increasing adoption of AI into people's workflows I think this is something that's really important to be thinking about. AI is a paradigm shifting technology whether you like it or not.

The author is asking the wrong questions by judging what was present 53 years ago and giving unkind subjective opinions about experts with limited knowledge of the time because their predictions were also true.

Today's expert can't be 100% correct if someone from 2017, 53 years in the future says they are asking the wrong questions.

The world is changing very fast in modern times.

We're obsessed with flying cars, and hang gliders, and drones, and space travel, because gravity sucks.

I often daydream about what life would be like if we could just regulate gravity at will, just at the individual level, so we can modulate our own weight, up to 0 -- or even negative.

>There's no prediction for when everyone on earth would have a pocket computer connected to all the world's knowledge (2020-2025).

It strikes me as tragic that the connection to all of the worlds knowledge has an end date like that, but it seems to be the trajectory we're on. Google's gone to shit, the Internet Archive is going away. Censorship is on the extreme uptick.

I think we should add "Napster - Connected to all of the world's music" (1999-2002)

I do wish the writer would stop justifying the relevance of their experiment by saying “a human would conclude that their time was being wasted long before the LLM”.

This is a fallacy.

A better analogy would be a human who has been forced to answer a series of questions at gunpoint.

At this point it becomes more obvious that the LLM is not “falling short” in some way.

I think this is an important lens to look through especially about predictions of AI. It is likely we will have downstream effects/platforms/interfaces that most of us can't even think about right now.

> And yet, despite predicting half of our world, as a father in the 1950s he could not imagine why his daughter - my mother - wanted to work.

In fairness, typically people imagined a utopian future where nobody worked. The strategic goal, never yet realised although AI might finally manage it, is to push people out of the workforce rather than in to it. Signing your daughter up to be a wage slave may be an improvement on the 1950s it isn't really the sort of thing that makes a good long term goal.

Something I've been thinking about, and then I noticed this essay.

> And yet, despite predicting half of our world, as a father in the 1950s he could not imagine why his daughter - my mother - wanted to work.

As written out, it is suggested that it is a lack of imagination.

I think a better narrative is just that it is work, and who would vulontarily have to work if they don't have to.

This is the more compassionate narrative.

>and computerised taxation (except in the USA).

I don't understand this. Does he not think the IRS uses computers?

Missing 2017 in the title.

Science fiction (novels, short stories, or by the 1950s films, and 1960s telvision) doesn't have a primary goal of predicting the future but rather of selling entertainment. To the extent it is speculative, it's almost always discussing contemporary circumstances sufficiently distant setting (in time and space) to be able to comment on it in a way which both minimises social censure and reaction and gives a potential for a fresh perspective (in the best cases). Of course, much of it is simply, or at least largely, escapist space opera / space westerns (Buck Rogers, Star Wars, Battlestar Galactica (original series). And yes, some of the escapist content does have real science or predictive value.

But in balance it's an exceedingly poor prospecting ground for hard-nosed, realistic, and/or accurate predictions. For much the same reason that most space ships are laid our horizontally, with gravity working on the frontal rather than coronal plane is that sets built on Earth are far more easily built and filmed that way, SF addresses its creations, narrative, and audience-appeal needs over any putative scientific or prophetic accuracy.

And sure, there are notable counterexamples.

E.M. Forster's The Machine Stops is frighteningly accurate in a world much like that of the 2010s / 2020s.

Arthur C. Clark in some works hits on some remarkably accurate depictions of at least parts of a future world. Imperial Earth envisions both handheld computers and a culture obsessed with recording every passing moment in a way that's nearly selfie-culture (though he seems to have missed influencers). 2001: A Space Odyssey predicts tablet computers and video telephony with reasonable accuracy (though all but completely ignores their social implications).

Orson Scott Card's Ender's Game presages blogging to some degree (though generally overstating its influence, as Randall Munroe spoofed: <https://www.xkcd.com/635/>).

And there's a whole slew of dystopian SF which has materialised in some form or another, from Ray Bradbury ("The Veldt"), Philip K. Dick (too many to mention), William Gibson, Neal Stephenson, etc., etc., etc. I and others suspect that's to some extent less prophetic than direct stimulus, with contemporary techbros aping their favourite adolescent sci-fi universes without asking "are we the baddies?" or whether they should.

But if you want hard predictions about the future, it's probably better to look to the literature which specifically and seriously attempts to do this, outside of a fictional context.

One such book is Alvin Toffler's Future Shock, now 54 years old.[1]

I'd read that, for the first time, on its 50th anniversary. I was struck by much, and found it on balance to have stood the test of time quite well, and much better than is typical for the genre. As to accuracy, there seem to be three general cases:

- Specific proponents of specific technologies virtually always overestimated the acceptance and impact of those technologies. The notable exception is, of course, information technology, though even for it the specific ways in which it has and hasn't advanced is worth close study.

- Virtually all of the social dynamic predictions seem laughably modest today --- developments in racial, gender, and sexual equality and acceptance, amongst others. Though on reflection this isn't so much that the predictions were bold, but that they've come to pass. We are on the far side of the singularity for these changes, for the most part. What was written in the context of a world in which these changes lay in the future reads much differently now that the inflection points are in the past. At the same time, it's also clear that such changes need not be permanent, and that perhaps such dynamics tend more towards cyclical patterns or pendulum swings, with greater and lesser liberalisation at different points in time.

- Much of the psychological and sociological concerns over advancing technology, faster paces of change, and an ever-growing onslaught of information seem to me to have been extraordinarily prescient, and largely born out. The disruptive effects, both on a personal psychological level and on a collective sociological one, appear to be profound, and we're still in the midst of discovering just how much so.

In thinking about how technological change manifests, I've come up with an ontology of the types of technological mechanisms which operate: fuels, materials, information (receipt, processing, storage, transmission), networks, systems, process knowledge, causal knowledge, power transmission and transformation, and hygiene (dealing with unintended consequences).[2]

Much of the Industrial Revolution (~1800 -- 1950 or so) was fundamentally grounded in new fuels (coal, petroleum, natural gas) and power transmission and transformation (particularly electricity and magnetism), with strong secondary effects through improved and expanded materials (Bessemer steel, aluminium, plastics), communications (telegraph, telephone, radio, television) and recording (rapid print advancements, photography, phonography, film). Since 1950, it's been information technology which seems to have been in the forefront, making some profound advances (overall processing and storage capacities) whilst remaining stubbornly stagnant in others (forecasting, meaningful automation and controls). Networks and systems have been primary secondary effects.

Hygiene is the ninth factor I'd come upon, and falls out of the recognition that all technologies have both intended and unintended effects. As technologies increase in complexity, I strongly suspect the latter dominate, exacting something of a drag on overall progress.

The element that's missing from my typology is the interaction between technology and society as a whole. I don't have much to say on that at the moment, though I feel it's quite significant. I'm noting that lapse for the moment.

________________________________

Notes:

1. <https://en.wikipedia.org/wiki/Future_Shock> <https://archive.org/details/isbn_0553132644>

2. I've written on this a few times at HN and elsewhere, searching Algolia for "tech ontology" or "technological ontology" should turn up some references. I'm increasingly feeling that the idea probably needs a book-length treatment discussing each mechanism, how it applies (some of the mappings I make may strike some as obscure, e.g., that knowledge is in some ways a network function, as expressed in the phrase "web of knowledge"), and what the capacities and limitations of each mechanism are.

Partially related/"obligatory" xkcd: xkcd.com/1425 (Tasks).

(What appear to be) Hard problems can be very easy, and (what appear to be) easy can be very hard.