When Extrapolation Goes Wrong

[JGittes]

Yep. AFAICT the risks associated with current "AI" technologies are the usual job-disruption ones created by any new tech, plus "information on the Web gets less reliable, betcha didn't even realise that was possible", plus "grandpa just gave his bank account login details to that nice robot on the phone". Harder to make an exciting movie out of that though.

I suspect at least some of the current techbros talking about the menace of smart AIs are doing it less because they believe in it, more as a way to distract from the real problems and limitations of the "AI" products they're trying to sell us right now.

I completely agree with you. There is an underappreciated movie, 'Blackhat,' that explores exactly those themes in an exciting and entertaining way.

 

[PennameWombat]

<snip>
Aside from blue collar spacemen. What other tropes from the past have failed to materialize? Let's hear them.

Has it failed, or is the timing just off?

"The Expanse" novels were written within the last few years and they also posit a large blue-collar or working class presence in the eventual colonization of space.

The reason for government involvement in space flight is that it is ungodly expensive, and only governments have access to the financial resources to fund it. It's the same issue that railroads had in the 19th century.

There is a significant difference between railroads and space travel, which will continually limit access to space. It's ungodly expensive. For practical purposes, it's infinitely expensive to lift mass into space. Yes, SpaceX has lowered that cost by taking risks that the government cannot take, but half of infinitely expensive is still infinitely expensive.

This.

The failed trope isn't really "blue collar spacemen." It's "space travel will be cheap and easy and will provide economic returns that outweigh the costs." The first part gets large numbers of people of all types into space, the second part gives them reason to be there.

That's ALWAYS been the driver behind exploration and expansion. Sure, "what's over there?" is a factor, but it's only when there is positive economic return that it goes beyond "huh, funky animal." If they can't eat it and its fur isn't useful and valuable, it's boring. There needs to be a driver (overpopulation, war, loss of arable land, drought, etc.)

Now that the cost and difficulty of exploration is significantly higher than putting sails on a ship and pointing west, the economic potentials will need to likewise be as high.

 

[ronde]

I have my serious doubts about time travel, but also being an engineer, I say just give the rest some time to come to fruition. Technology just takes time, though technology is increasing at an exponential rate. My grandfather courted my grandmother in a buggy pulled by a horse and farmed with horses, but when he retired, he farmed with a tractor and picked his grain with a gasoline powered combine. He had a party-line phone for a long time instead of a cell phone, and he did his taxes on an adding machine instead of a computer. He lived long enough to see the birth of commercial aviation and to see a man step out onto the moon from the spacecraft that took him there. It took less than 70 years for that to happen.

Just because it hasn't happened yet only means that those early authors just miscalculated a little. We're probably a ways away from "Space Truckers" hauling "square hogs", but I think we'll get there, probably within the next hundred years or so. In the meantime, I love writing about the possible future.

 

[Bramblethorn]

In the event, banks were hit hardest, in terms of their databases, and any system that ran on people's birth dates. Did 01 mean 1901 or 2001? Who knew?

Wiki lists some of the Y2K issues that weren't caught. As you say, plenty of banking/payroll glitches, but also stuff like radiation monitoring equipment at a nuclear plant failing, day-long heating outages, dialysis machines briefly failing, miscalculation of Down Syndrome risks (resulting in two abortions for pregnancies incorrectly assessed as high-risk, and four "low-risk" babies born with DS), traffic light outages at major intersections, "an unspecified malfunction in a system for determining the weight and composition of nuclear substances at a nuclear weapons plant", and train outages. Some of the glitches showed up as much as 20 years later because of people applying a half-arsed patch that punted the problem down the road instead of properly fixing it.

Aside from the pregnancy one, mostly reversible, and none likely to be a catastrophe if it's the only thing you have to deal with. If your dialysis machines aren't working, you can drive your patients to another hospital with a different model until you get them fixed. If the heating fails in one apartment block in mid-winter, people can go somewhere else that's heated. But if you can't drive your patients to another hospital in reasonable time because the traffic lights are down, and everybody else's heating is out too, and the power plant shut down because they can't guarantee safe operation, and you can't buy emergency supplies because the payment systems are down, that's when systems can get overwhelmed and things get very bad.

I feel for the poor bastards who put so much effort into patching stuff, to ensure that it was just rare failures here and there, only to have their competence taken as evidence that the work wasn't important.

 

[Bramblethorn]

The failed trope isn't really "blue collar spacemen." It's "space travel will be cheap and easy and will provide economic returns that outweigh the costs." The first part gets large numbers of people of all types into space, the second part gives them reason to be there.

Picking a small nit there: it's not just economic returns that outweigh the costs, it's economic returns that outweigh the alternative options.

Sending canned primates into space is hideously expensive. You need to carry food, water, air, temperature control systems, radiation protection for any long journeys, in most circumstances you need to allow for a return trip, and you need to engineer to a very high level of reliability.

Maybe there's something out in space so valuable that it'd be profitable to get humans there. But if that application exists, it's probably going to be much more profitable to send expendable drones/remotes to do the job.

I think this is a large part of where sci-fi prognostication went wrong on this issue - yes, SF was generally over-optimistic on the question of "how much will it cost to send a human into space?", but also it was very reluctant to ask "and why wouldn't it be cheaper to send something else?"

That's ALWAYS been the driver behind exploration and expansion. Sure, "what's over there?" is a factor, but it's only when there is positive economic return that it goes beyond "huh, funky animal." If they can't eat it and its fur isn't useful and valuable, it's boring. There needs to be a driver (overpopulation, war, loss of arable land, drought, etc.)

Now that the cost and difficulty of exploration is significantly higher than putting sails on a ship and pointing west, the economic potentials will need to likewise be as high.

Yup. In my ancestors' time, migration was a way for poor people to escape bad situations. But if you have the money for an interplanetary space trip and enough to keep you alive at the other end, you have some pretty comfortable options here on Earth.

 

[JGittes]

........

I feel for the poor bastards who put so much effort into patching stuff, to ensure that it was just rare failures here and there, only to have their competence taken as evidence that the work wasn't important.

This is the problem with successful vaccination programs, disaster preparedness efforts, and international diplomacy; the results aren't visible as anything other than life goes on as usual, and the efforts expended in making that so are unrecognized and frequently derided as unnecessarily expensive.

 

[nice90sguy]

The thing about SF is that it's above all a commentary on the present, and puts out alternatives, extrapolating to the future, either hopefully, or as a cautionary tale.

Pohl and Kornbluth's "The Space Merchants" was really prescient of the monopolising forces of advertising, but was written at the height of the "Madmen", in the early 1960s.

"1984" was completely wrong (certainly as to the date), but was a warning against the spread of Stalinism, not a prediction.

The far future predicted in "The Time Machine" was a commentary on the class struggle of Wells' time.

So "getting it right" is not the aim of futuristic writing.

 

[NoTalentHack]

I can remember the great "Y2K" scare that preceded the turn of the millennium. Some feared that planes would fall from the sky. Lawyers built up specialties in the expectation of massive damages and liability and lawsuits.

The late 60s and early 70s were full of predictions that by the end of the century we'd be running out of everything and there would be massive global starvation.

Catastrophism goes back a long way and there's a seemingly endless appetite for it.

As someone who was in the middle of the Y2K thing (tech support at Dell and the son of a programmer feverishly working to get a bunch of systems upgraded), that absolutely would have been a catastrophe if not for everyone worrying about it being a catastrophe. And that's kind of the problem; after one of those types of things is averted (Y2K, ozone hole, a bunch of other stuff), there's an attitude that crops up later of, "well, those other things didn't happen, so what's the big deal?" The climate change scientists have had to deal with the aftereffects of the successful fight against CFCs to restore the ozone layer, for example.

 

[NoTalentHack]

Yep. AFAICT the risks associated with current "AI" technologies are the usual job-disruption ones created by any new tech, plus "information on the Web gets less reliable, betcha didn't even realise that was possible", plus "grandpa just gave his bank account login details to that nice robot on the phone". Harder to make an exciting movie out of that though.

I suspect at least some of the current techbros talking about the menace of smart AIs are doing it less because they believe in it, more as a way to distract from the real problems and limitations of the "AI" products they're trying to sell us right now.

I believe some of them are sincere. However, in the case of a number of the ones that are, like Musk, it's because they don't actually understand the technology and wrongly think the current LLM stuff is a strong step towards general AI when it's not.

 

[M_K_Babalon]

Has it failed, or is the timing just off?

"The Expanse" novels were written within the last few years and they also posit a large blue-collar or working class presence in the eventual colonization of space.


This.

The failed trope isn't really "blue collar spacemen." It's "space travel will be cheap and easy and will provide economic returns that outweigh the costs." The first part gets large numbers of people of all types into space, the second part gives them reason to be there.

That's ALWAYS been the driver behind exploration and expansion. Sure, "what's over there?" is a factor, but it's only when there is positive economic return that it goes beyond "huh, funky animal." If they can't eat it and its fur isn't useful and valuable, it's boring. There needs to be a driver (overpopulation, war, loss of arable land, drought, etc.)

Now that the cost and difficulty of exploration is significantly higher than putting sails on a ship and pointing west, the economic potentials will need to likewise be as high.

I think the real question is; will there be a reason for space travel and widely available ships, to the point it's no different than being a airplane pilot or cdl holder? The reality would be people on moons, planets with land, which there are few, and space stations. That we know of, it's not going to be like Star Wars, Cowboy Bebop, or Outlaw Star, where there's all these planets and races. Funny thing is a lot of ships in Star Wars are pretty cheap, given a credit is equal to the us dollar, if I recall correctly, a Cerulian Corvette is 40,000cr and an X-Wing is like 10,000/15,000. You'd save money getting that Corvette over a Chevy one.

 

[MrPixel]

Y2K? Oh, just wait for the year 2038 problem.

One example: for folks like my wife and I who tend to keep vehicles a long time (20 and 15 years currently), given the total reliance of today’s cars on computers running complex software for anything and everything, this date problem will simply cause most functions to roll over and die. Instant heaps of scrap metal too costly to upgrade, everywhere.

 

[lc69hunter]

It has been over 50 years since humans last walked on the moon, and we haven't returned. We won't have bases on Mars doing what you suggest, probably ever.

Without a vision, the people perish. It has always been the risk takers who change the world, and the course of human events. We have one living among us now. Elon Musk. "Experts tell him, "you can't do that". How well has that worked out, so far?

 

[alohadave]

It has been over 50 years since humans last walked on the moon, and we haven't returned. We won't have bases on Mars doing what you suggest, probably ever.

Unless there is a good political or economic reason to go, we won't.

We stopped going to the moon because we beat the Soviets. After that, there was no political will to go.

Now that commercial space is opening up, there is more pressure to go back, hence the Artemis missions.

 

[djrip]

Without a vision, the people perish. It has always been the risk takers who change the world, and the course of human events. We have one living among us now. Elon Musk. "Experts tell him, "you can't do that". How well has that worked out, so far?

Boy that guy really used to be cool. Before he turned into Howard Hughes. Well at least Space X is still cool.

 

[Gamblnluck]

I remember one from the early years of my youth, 'Skylark series written in the early 1900's as a pulp story and then published as a book later.
What I loved about it was not just the fantastic crap they made, but the characters were developed. You had 'Blackie" the evil scientist who had few if any morals. You had the main character Dick who was a standup guy and scientist.
The story dropped in use of a heavy metal as a power source Way before uranium.
Sometimes it got lost a bit explaining how things might work (gravity or lack thereof). But the story was fun and written by an actual scientist.

 

[NoTalentHack]

Boy that guy really used to be cool. Before he turned into Howard Hughes. Well at least Space X is still cool.

Everything cool he’s ever been associated with flourished in spite of him, not because of him. The engineers at every one of “his” companies had to develop strategies for dealing with him that essentially boiled down to “pat his head and tell him that’s an awesome idea, then do the opposite.” The places where they couldn’t are where you get things like Cybertruck, the Twitter meltdown, and an attempt at SpaceX to literally use epoxy to hold a rocket together.

Musk is good at one thing: generating publicity.

 

[Kelliezgirl]

Everything cool he’s ever been associated with flourished in spite of him, not because of him. The engineers at every one of “his” companies had to develop strategies for dealing with him that essentially boiled down to “pat his head and tell him that’s an awesome idea, then do the opposite.” The places where they couldn’t are where you get things like Cybertruck, the Twitter meltdown, and an attempt at SpaceX to literally use epoxy to hold a rocket together.

Musk is good at one thing: generating publicity.

Kind of hard to believe that could be true. You really believe that he became the world's richest man in spite of being wrong about everything except publicity? Perhaps the truth is closer to the middle? He's hit some homeruns, and he's struck out, he's had some great ideas and put the right people in the position to be successful, and he's made some bad judgement calls. Just like everyone else, but on a far, far larger scale.

 

[djrip]

Everything cool he’s ever been associated with flourished in spite of him, not because of him. The engineers at every one of “his” companies had to develop strategies for dealing with him that essentially boiled down to “pat his head and tell him that’s an awesome idea, then do the opposite.” The places where they couldn’t are where you get things like Cybertruck, the Twitter meltdown, and an attempt at SpaceX to literally use epoxy to hold a rocket together.

Musk is good at one thing: generating publicity.

Well I don't know that Space X would exist at all without him. Tesla he bought, Space X he built. I wouldn't think he was important to the engineering, but without a boss saying, let's do this in new ways and let's get to vertical landing and reusability, I don't know that it would have happened anytime soon. Boeing certainly wasn't even trying. Bezos is slowly working towards the capability, but that's in a world where Space X was working on it too, so a race. Bezos isn't even trying to do functional deliver stuff and people to orbit type launches, just tourism. Space X gets to orbit for a tenth of the cost of previous solutions. It's important what they have done. It's certainly debatable how much Musk had to do with it, but I don't see any other space companies doing anything like it. Clearly he did something. I suspect what he did was put a company together for idealistic reasons and then take risks on longshot strategies that companies motivated only by money would not have taken. If those risks had not worked out, Space X would have been a footnote no one would know about. But they did.
But that doesn't mean what he's doing now is any good.

 

[Kelliezgirl]

But that doesn't mean what he's doing now is any good.

What is he doing now that isn't any good?
Starship seems to be making genuine progress, they are kicking SLS's ass.
Starlink is pretty awesome.
Tesla is innovating and just LOWERED prices, which is pretty awesome.

In general, a certain segment of the population seems upset about his decisions regarding Twitter/X and are letting that color their opinion of everything else he's done.

 

[djrip]

What is he doing now that isn't any good?
Starship seems to be making genuine progress, they are kicking SLS's ass.
Starlink is pretty awesome.
Tesla is innovating and just LOWERED prices, which is pretty awesome.

In general, a certain segment of the population seems upset about his decisions regarding Twitter/X and are letting that color their opinion of everything else he's done.

What he's doing now is Twitter. He already made Space X and at this point I'm sure it can keep going just fine with or without him. I'll give him a high five if he lands something on Mars anytime soon though.

 

[Kelliezgirl]

What he's doing now is Twitter. He already made Space X and at this point I'm sure it can keep going just fine with or without him. I'll give him a high five if he lands something on Mars anytime soon though.

He's probably more involved getting Starship ironed out than he is day to day at Twitter.
And what is so terrible about what he's done at Twitter?
Advocating for free speech is a good thing.

 

[NoTalentHack]

In general, a certain segment of the population seems upset about his decisions regarding Twitter/X and are letting that color their opinion of everything else he's done.

That's sort of true, but not in the way you mean. Basically, when it came to automotive engineering, rocketry, etc., there wasn't a large enough swathe of trusted people who both got to see what was going on in the company and could intelligently critique it.

Then the Twitter debacle happened, in full view of the world's software engineers. Elon's excuses for why things weren't working made absolutely no sense, and, in fact, often the opposite of sense. The ways he talked about addressing issues also didn't jibe with reality.

The source code got released, and it showed places where Musk had recently asked for things to be done that could only be called narcissistic. Stories from former SpaceX and Tesla employees came out of the woodwork for how he'd done the same thing at those companies, but not as publicly. Then people started paying attention to journalists that had been ringing the alarm bells for a while, but that they'd been ignoring in favor of the image Musk projected, because people liked the idea of this real-life Tony Stark.

Basically, Twitter was the cause of his problems, but only because it meant so much light got cast onto him, in the same way that Trump and family probably would have never seen the inside of a courtroom for longer than it took to agree to a settlement if they'd been content running their scams in NYC and the surrounding area, instead of going for the brass ring.

If you want more documentation, here:

https://www.reuters.com/investigates/special-report/spacex-musk-safety/

Musk himself at times appeared cavalier about safety on visits to SpaceX sites: Four employees said he sometimes played with a novelty flamethrower and discouraged workers from wearing safety yellow because he dislikes bright colors.

The lax safety culture, more than a dozen current and former employees said, stems in part from Musk’s disdain for perceived bureaucracy and a belief inside SpaceX that it’s leading an urgent quest to create a refuge in space from a dying Earth.

https://arstechnica.com/science/202...-his-spacex-employees-to-work-thanksgiving/2/

Relating an anecdote from 2003:

It was near Christmas, and Musk thought he had a solution for the chambers, which would crack a few seconds into firing. Musk believed an epoxy-like material could be applied to the ablative chamber, which would then seep into the cracks and fill them. Dressed in expensive shoes and designer jeans, having foregone a swanky Christmas party in Los Angeles, Musk worked with the propulsion team late into the night. He ruined the $2,000 shoes with the sticky epoxy. His idea was a spectacular catastrophe, but eventually, through hard work, long nights, and ingenuity, he and the company found a solution to the Merlin engine issues.

That's a good example of the spin that Musk got before about 2022. "Was this a terrible idea? Yes. But these kinds of terrible ideas are why SpaceX succeeded." Instead, it's more likely that SpaceX succeeds largely ins pite of him, instead of because of him.

https://www.nytimes.com/2022/11/17/business/spacex-workers-elon-musk.html

The day before, the company had moved to fire five employees who had written a letter calling on SpaceX to condemn the “harmful Twitter behavior” of Mr. Musk, who had used the social network to make light of a news report that SpaceX had settled a sexual harassment claim against him. Several of the engineers filed into the meeting expecting a sympathetic ear, as some managers and executives had indicated that they did not condone Mr. Musk’s behavior.

But the meeting, which has not been previously reported, quickly became heated, according to two SpaceX employees in attendance.

They said Jon Edwards, the vice president leading the meeting, had characterized the letter as an extremist act and declared that the writers had been fired for distracting the company and taking on Mr. Musk. When asked whether the chief executive could sexually harass his workers with impunity, Mr. Edwards did not appear to answer, the two employees said. But they said the meeting had a recurring theme — that Mr. Musk could do whatever he wanted at the company.

https://finance.yahoo.com/news/elon-musk-fired-dozens-more-172906651.html

Thousands of Twitter employees have already been laid off, resigned, or fired by Musk, leaving critical workers stretched thin. Twitter teams are already so short-handed that many employees have had to work on Thanksgiving, the people said.

The firings are thought to be, at least in part, a result of Musk's new policy requiring Twitter engineers and other staff working on code or technical projects to update him every week about what they worked on, what they hoped to complete, and specific lines of code they wrote, as Insider previously reported.

As a software dev, that's fucking insane.

https://insideevs.com/news/658439/elon-musk-overruled-tesla-autopilot-engineers-radar-removal/

Back in May 2021, Tesla announced that it was eliminating radar on its new cars in a push to switch to the so-called Tesla Vision approach, which only uses cameras to “see” the road and neural networks that are supposed to mimic the way a human brain works.

Before the announcement, however, several Tesla engineers were worried that this move had a high risk of causing problems, including increasing the risk of accidents if the cameras were obscured by raindrops or even bright sunlight.

“Some Tesla engineers were aghast, said former employees with knowledge of his reaction, speaking on the condition of anonymity for fear of retribution. They contacted a trusted former executive for advice on how to talk Musk out of it, in previously unreported pushback,” The Washington Post story says.

“Musk was unconvinced and overruled his engineers. In May 2021 Tesla announced it was eliminating radar on new cars. Soon after, the company began disabling radar in cars already on the road. The result, according to interviews with nearly a dozen former employees and test drivers, safety officials and other experts, was an uptick in crashes, near misses and other embarrassing mistakes by Tesla vehicles suddenly deprived of a critical sensor,” the report adds.

According to the story, Musk’s erratic leadership played a role in the unpolished releases of its Autopilot and so-called Full Self-Driving features, with engineers forced to work at a breakneck pace to develop software and push it to the public before it was ready. More worryingly, some former Tesla employees say that even today, the software isn’t safe for public road use, with a former test operator going on record saying that internally, the company is “nowhere close” to having a finished product. The person in question was fired by Tesla in February 2022.

There are tons more articles like this, some even worse: how he pushed the Tesla founders out, his labor practices, anger management issues, willingness to push untested features to the public (and then turn around and rely on settlements and NDAs to absorb bad press), and on and on. And that's the thing: he did a great job keeping all of this under wraps for a long time because his biggest strength is his ability in PR. But then, once it started to unravel (especially on Twitter), his response was "well, I'll buy Twitter and fix the 'issues' with it," which seems, so far, to be his Waterloo.

I believed his hype for a long time, too. And I think there are places where he's been sincere, passionate, and even occasionally effective. But I also think that, at some point, he started believing his own press.

 

[Bramblethorn]

He's probably more involved getting Starship ironed out than he is day to day at Twitter.
And what is so terrible about what he's done at Twitter?
Advocating for free speech is a good thing.

Musk hasn't been pro-free-speech, though. He's been pro-speech-that-he-agrees-with, while repeatedly curtailing other speech.

Under Musk, Twitter has become much more willing to accept government requests to censor and/or surveil its users.

Aircraft tracking info is publicly available by law, but Musk suspended an account that posted this info for his personal jet for "doxxing" (after having previously talked about how he wouldn't do this exact thing, because he was such a big believer in free speech.)

He then suspended a bunch of journalists who reported on that suspension, falsely claiming that they'd doxxed him. Around the same time, Twitter also suspended Linette Lopez, a journalist who hadn't been discussing the jet tracker - but who had previously reported information critical of Tesla.

He announced that the word "cis" would be considered a slur, grounds for suspension as harassment.

(Context: "cis" is a neutral Latin-derived prefix that's the antonym to "trans". It's been around since the Roman Empire. Today it's used in geography, chemistry, and rocket science - for instance, SpaceX's "Starship User's Guide" advertises Starship's capability to deliver payload to "Earth, cislunar, and interplanetary trajectories". But because it's also used to mean "not transgender", it's a popular bugbear for people who want to portray themselves as anti-woke/anti-trans. Musk started making a big deal of being anti-woke not long after his girlfriend left him for a trans woman and one of his own adult kids who's alienated from him came out as trans.)

After Twitter introduced the paid "verification" system, Twitter suspended users who advocated for blocking people who bought verification.

Soon after Substack introduced a product similar to Twitter, Twitter blocked its users from linking to Substack and apparently even from searching for the word "Substack".

Twitter shut down its newsletter platform Revue immediately after Jack Dorsey (co-founder) posted a long essay critical of Musk.

Most recently, Twitter/X has sued MediaMatters for showing advertisers how their ads were appearing next to neo-Nazi/white nationalist content.

Probably a few others I've missed, but that'll do for starters.

 

[NoTalentHack]

Oh, wait, there’s also how he offered Starlink to Ukraine, then shut it down (or at least refused to let it be used) when it was going to be used to coordinate an attack on Russia once Ukraine was dependent on it.

 

[Cagivagurl]

The most important part of reading science fiction is....
It's fiction... Written for entertainment.
They were not predictions of future events.
Merely writers exploring their imaginations.
Some of the things imagined in those works, may or may not come to pass.
Does that diminish the writing?
I don't think so. I loved reading them. Fueled my own imagination.

Cagivagurl