Hacker News new | past | comments | ask | show | jobs | submit login
How to destroy the Earth (2006) (qntm.org)
92 points by Smaug123 on May 31, 2016 | hide | past | favorite | 44 comments



This reminds me of what the Web used to be: amusing entries by interesting people, devoid of advertising, marketing or JavaScript.


In the same vein I've got to recommend

* Norman Yavin's Usenet archive of interesting technical and science related posts (http://yarchive.net/)

* Prof James Calvert's pages on about everything worth knowing (http://mysite.du.edu/~jcalvert/)


Prof Calvert's page reminds me of John Savard's home page: http://www.quadibloc.com/ -- I ran across this back when I first got into slide rules, but there is a lot of other good stuff there too.


It's still there, the signal-to-noise radio is much, much worse though.


Your comment suggested a new tool/feature to me: what if we had something like Google, except it added the additional constraint where it only led you to sites/pages that were in that old retro lean static simple HTML style. The graph of text pages, perhaps with embedded images, version of the web.

Ideally it would be just an add-on feature to an existing web search engine like Google. But if you wanted to boot it up faster you could perhaps meta-spider/index on top of them, where your wrapper site added this constraint as a filter.

I'd love to be able to explore a curated web subset that was just text and images, no bandwidth bloat, no obvious SEO abusers, no pirated content farms, no ads, no spyware, no audio, no (automatically streaming/playing) video, etc.


>Your comment suggested a new tool/feature to me: what if we had something like Google, except it added the additional constraint where it only led you to sites/pages that were in that old retro lean static simple HTML style. The graph of text pages, perhaps with embedded images, version of the web.

Sometimes when looking for knowledge on Google and can only find blogspam, if the information isn't particularly volatile over time, e.g. information about effective exercises for various purposes, then I've got a trick that sometimes helps:

Use google's search tools to limit the results to a custom range of time, and only specify the end date as something like 2002. You end up getting just pre-web-2.0 stuff.


nice suggestion! I'll try that


Yeah, article like this one would exist, but with a caption "You wouldn't believe 10 ways this guy invented to destroy the earth! Number 6 will blow your mind!" and then the article is split in 100 pages just to get more clicks.


There were far more interesting USENET articles than this op in rec.arts.sf.science. There were back of the envelope calcs by researchers involving hypothetical attempts by humanity to boil the biosphere using decades worth of global industrial output to seen the entire planet with a grid of hydrogen bombs. This one is a 1+ steps removed rip-off of those discussions.


Unfortunately, the increasing noise is rapidly pushing interesting information to below Shannon limit.


Hmmm.

I wonder how difficult it'd be to spider the web and index all the pages with a certain density of useful information.

There's ole entropy, but we also have a priori knowledge of certain bits of information that suggest that useful information in the content is nil. E.g., anything that is a listicle.


Method 1: Annihilated by an equivalent quantity of antimatter

I don't think this would work as well as one might assume. When Earth and anti-Earth collide, the vast majority of the mass will be blown into space by the initial explosion, and will never actually get a chance to annihilate. There might even be some coherent chunks left over.

If you can manage to transform every other atom from matter to antimatter, evenly distributed through the body of the planet, then you've got yourself a good kaboom.

Method 11: Ripped apart by tidal forces.

The idea here is to move Earth within the Roche limit of either the Sun or Jupiter. The problem is that the Roche limit depends on density, and Earth is substantially denser than either the Sun or Jupiter (Earth 5.51, Sun 1.41, Jupiter 1.33). I haven't done the math, but it's possible that the relevant Roche limit in both cases would be below the surface -- i.e., the Earth would (barely) survive the tidal stress even in a surface-hugging orbit. And once you've done that, you might as well just drop it a little lower; falling into either the Sun or Jupiter would do the trick quite effectively.


> I don't think this would work as well as one might assume. When Earth and anti-Earth collide, the vast majority of the mass will be blown into space by the initial explosion, and will never actually get a chance to annihilate. There might even be some coherent chunks left over.

The author mentioned an Earth made from antimatter. How about a spherical shell of gaseous antimatter, whose total mass is equal to Earth? Make Earth the center of this sphere, then wait.


The matter-antimatter annihilation near the boundary would generate enough energy to blow the antimatter shell into deep space (where it would cause some serious problems for the rest of the Solar System).

The released energy would likely (I haven't done the math) be enough to vaporize the Earth, which would meet the original requirements, but I don't think it would annihilate a large percentage of the Earth's mass.


I read this site and am reminded of EE Doc Smith's Lensman series. He has many of the methods described, including hurling an antimatter planet at another, relativistic speed planets, and even a sun beam though that was used to vaporize a fleet.

Great imagination with a few throw backs to the sensibilities of his age but he did go beyond them too


The author, qntm, recently posted a follow-up article entitled "To Destroy the Earth", which looks at the problem from a different perspective: https://qntm.org/destro


> Antimatter... it may be possible to find or scrape together an approximately Earth-sized chunk of rock and simply to "flip" it all through a fourth spacial dimension, turning it all to antimatter at once.

Obviously this idea is entirely speculative, but I haven't heard of it before and was wondering what the theoretical foundation for it is.

I think it's alluding to a 4D rotation causing a change in parity, which seems plausible. Physics is very nearly symmetric under a conjugation of parity and charge (the CP symmetry, which is violated only by the weak nuclear force); so the idea would be that a change in parity through this 4D rotation causes a charge conjugation, maintaining CP symmetry.

I don't think this would work. Symmetries in Physics imply a conserved quantity, through Noether's theorem. The problem is that the conserved quantity associated with a charge-conjugation symmetry is not charge itself, but 'charge parity', a multiplicative quantum number. Under CP-symmetry, the product 'charge parity' * 'spatial parity' is the conserved quantity, so a change in spatial parity from the 4D rotation would cause an associated change in charge parity (assuming CP symmetry), which merely means the system would behave differently under a charge conjugation operation, not that the charge is actually changed.

Charge conservation itself originates from the gauge invariance of electromagnetism, and I'm not sure how that would be affected by this scenario.

I'm not saying the idea is necessarily wrong, but I think it's likely to have originated from a misunderstading.

For further information:

- https://en.wikipedia.org/wiki/Symmetry_%28physics%29#Conserv...


Man, I feel old. I remember this when it was the alt.destroy.the.earth FAQ...

https://jult.net/adte.htm


This comes up on HN every year but I re-read it every time.

My idea is just figure out a way to nudge earth towards the sun and inertia will do the rest, slowly but surely, so a vote for their #10

Or settle for wiping out all human life on earth and earth will survive just fine without us. Picking a US President that will use nuclear weapons might be a dangerous starting point to investigate.


> My idea is just figure out a way to nudge earth towards the sun and inertia will do the rest, slowly but surely, so a vote for their #10

I don't think you can. Even if you could give the whole Earth a push toward the sun, it has enough orbital velocity (~30,000 m/sec) that it'll miss and keep on going around. You might turn its orbit into a slightly different ellipse.

Say we want to straight up (down?) launch a probe into the sun. It's a hard thing to do because it has to shed all of that orbital speed to get its velocity vector pointed in the right direction (at which point it can fall in on its own). If you only get rid of most of it, you'll fall into a long elliptical orbit.

You could potentially dump some energy with a gravitational assist off of Venus (transfer some of our momentum into speeding up its orbit while slowing ours down), but there's still the matter of getting the Earth to Venus for that to work. And at that point maybe you just want to stick to #9.

I haven't played enough KSP to be sure of all that, but that's my understanding of the orbital mechanics.


> You could potentially dump some energy with a gravitational assist off of Venus (transfer some of our momentum into speeding up its orbit while slowing ours down)

I've heard that a hard turn around Jupiter is even better. I remember Scott Manley dropping a probe into Kerbol via Jool gravity assist in one video[0].

Actually, now I want to make a simulation with Earth doing Jupiter flybys. What's good for that nowadays? Universe Sandbox?

[0] - https://www.youtube.com/watch?v=uNS6VKNXY6s


Scott Manley also made a trip to Minmus using only the starting Career parts. Just because he did a thing doesn't mean it's the optimal way to do that thing :-)

(In fact, it's probably a better indicator against, heh.)


It's straightforward: throw the moon off in the direction of Earth's motion around the sun. The Earth has 81 moon-masses, so you'd have to throw the moon 81 times faster: still less than 1% of the speed of light. That'd take considerable energy, but the sun outputs 3.846×10^26 W, so we'd only need 17 years of total solar output. It'd be a lot easier to just roast the Earth's surface, it's true.

(I haven't read the original post in a long time, but had thought about it myself before.)


If you wait for a while (8 billion years, give or take) the Sun will expand enough to meet the Earth, and presumably that will happen long before any drag effects will cause the reverse to happen.


Small amounts of force applied in any direction will not set the Earth into a collision course with the Sun. An intuitive explanation for why can be found in the game Osmos.

www.osmos-game.com

The short reason is that inertia is the entire reason the Earth is orbiting the Sun right now, you have to overcome that inertia rather than generate more of it. When you're playing Osmos during the orbital sections, a little bit of playing around reveals that the best way to reduce your orbit is to thrust in the same direction you're currently moving in. To increase your orbit, you thrust in the opposite direction. Thrusting towards or away from the 'star' alters your orbit somewhat but doesn't really change how far away from the star you are on average.

After awhile you figure out the right way to finesse your way into orbits that carry you on a trajectory to eat smaller bodies. With that skill comes the dying off of the fantasy of being able to easily dispatch the Earth in this fashion. Inertia doesn't work that way.


I didn't bother to search HN for the title, just the link, and this is the first time this particular variant of the link (https) has been posted. Sorry; I wouldn't have posted it if I'd realised it had appeared multiple times before.


Nothing to be sorry about: a lot of HN readers (invluding me) are seeing it for the first time. And I will upvote this next time I see it, too.


> Or settle for wiping out all human life on earth and earth will survive just fine without us. Picking a US President that will use nuclear weapons might be a dangerous starting point to investigate.

This may be very dangerous. When you take antibiotics, you need to do the full course, 'lest some bacteria survive. Similarly, if the nuclear destruction isn't complete enough, the surviving humans will repopulate the planet and continue to trouble it for millennia to come. When planning an apocalypse, one needs to be sure it can be carried out to the end.


That's an excellent point. If you think humans are bad, nuclear-resistant humans are worse.


My idea is just figure out a way to nudge earth towards the sun and inertia will do the rest

ARRGH! This is about as big a misunderstanding as not knowing the difference between sending a rocket to space vs. getting to orbit:

https://www.youtube.com/watch?v=uNS6VKNXY6s

tl;dr -- the Earth is whizzing around the Sun at a really high speed. You'd have to cancel most of that velocity to get a spacecraft to get anywhere near the Sun. It would actually be less energy to send the Earth to Jupiter than to drop it into the Sun.

You've repeatedly re-read this article, but never delved into basic orbital mechanics all that time!? I'm going to have to remember this in my chronicles of "Stuff I've actually seen posted in comments on HN."


I choose a "meticulously deconstructed" variant. I call it the "onion method".

First, invent a self-replicating robot. It does not need to be miniaturized. It would actually probably work better if they were all stadium-sized. Allow replication, using solar energy and near-surface resources, until the entire planet is covered by a uniform layer one robot deep.

The robots then securely connect to their nearest neighbors. You now have a uniform robot tensegrity shell around the planet. The linkages expand until the shell rises to a uniform altitude above the surface. This likely requires a material with tensile strength higher than any currently known, and a material with compression strength higher than any currently known. I haven't done the math. Reasonable strengths may require that the shell robots run on the surface a bit to start the whole thing spinning faster than the planet.

A circular section of the shell directly facing the sun de-links, falls back to the surface, and begins replicating as before. The remainder of the robot shell opens like a jellyfish mantle, to catch the solar wind and reflect additional solar energy to the dark side of the planet. The shell sails to a different orbit, so as to not re-collide with the planet.

Repeat until all shells have sailed away.


Solar wind is not strong enough to put something in orbit.


I dunno. Given a self-supporting spherical structure around the planet, you could raise it above the atmosphere and set it up to experience a net torque from solar radiation pressure. Just add "sails" that are black on one side and mirrored on the other. Then sit back as it very very slowly spins itself up to orbital velocity.

Of course, there would be plenty of practical difficulties. (For one thing, radiation pressure is a hideously inefficient use of energy; for another, your spherical shells will be unstable and require constant maneuvering to keep them centered.) So if you have an army of self-replicating robots, it's probably easier to just have them build millions of solar-powered mass drivers and fire rocks directly into space.


The center of gravity of the shell starts coincident with the center of gravity of Earth. So as long as the shell keeps its integrity, and the center of mass of the Earth is inside the spherical shell, it is effectively unattached to the Earth by gravity.

Let's also assume that the first round of robot-building made the Earth a perfectly smooth ellipsoid. So now you have a spherical shell on a spherical planet, standing on trillions of robot wheels. All the robots sync up, and do some math, and at the same millisecond mark, start turning their wheels.

So the shell starts spinning faster than Earth. Its increase in angular momentum is subtracted from Earth's angular momentum. The robots now have to link arms pretty tightly to keep their wheels in contact with the surface. They connect to each other with cables of adjustable length.

Escape velocity at Earth's surface is about 11.2 km/s. That's not practical due to atmospheric drag. Even a stable orbit at 200 km altitude requires 7.8 km/s. So the shell runs as fast as it can in air. Then all the robots simultaneously jettison their legs and spool out their cables, and the whole thing jumps in altitude.

Now they start sacrificing. The robots nearest the spin axis twist as hard as they can on their own cables, slowing their own spin and increasing that of the remaining robots in the shell. Then they let go and drop back to Earth. The next opposing rings of robots do the same. Eventually, you reach a point where a spherical ring section of the shell is spinning fast enough to keep the linkages under a reasonable amount of tension, entirely outside the atmosphere.

Now each robot deploys a solar sail, and angles it to accelerate the ring towards the winter end of Earth's axis, until Earth gravity starts pulling it back. Then the sails tack, and start pushing in the opposite direction. It's like pumping your legs on a playground swing. Eventually, the entire ring achieves Earth-escape velocity and flies off. Then you repeat with a new shell.

[Edit:] Actually, let's make the shell an ellipsoid with greater eccentricity than the Earth during the spin-up phase. The robots nearest the poles can therefore keep running with wheels on the surface until their arms get torn off. Then the polar discs get sacrificed.


This is from 2003, in fact.


Here's my way: An enormous particle accelerator.

Shoot electrons into space in vast quantities until the whole earth becomes so highly charged it blows itself to pieces.


Shooting electrons out gets harder and harder as you go since you're counteracting an ever increasing positive charge from the Earth.

Besides, on a uniformly conductive sphere (which admittedly the Earth is not at the crust but probably is near the mantle and core), excess charge distributes over the surface, so only the top layer of the Earth would get blown off (along with your accelerator).


> Shooting electrons out gets harder and harder as you go since you're counteracting an ever increasing positive charge from the Earth.

Sure. This is fantasy of course - my accelerator is powerful enough to handle it.

> excess charge distributes over the surface

Yes. The plan is to put the accelerator in a hole so it can tap into all the electrons that are hiding below the surface (since the surface would be depleted of electrons). As the surface is blown away it would dig itself deeper.

It wouldn't blow itself to pieces like a bomb in a move, but rather as giant chunks fling themself into space chasing the electrons I've shot there.

I like my plan because I think it's the most energy efficient one out there that is actually theoretically possible today (i.e. doesn't require new physics).


hmm...the singularity not the same as event horizon. The Schwarzschild radius of the earth is 9mm. The singularity is infinitely small


You talk about "the singularity" as if it exists. For all we know, that could be a mathematical fiction due to our limited understanding. The Event Horizon as we currently understand it might as well be the demarcation to a separate universe. We have no access to any information from the events beyond it.

https://www.youtube.com/watch?v=vNaEBbFbvcY


Fools. The way to destroy the Earth is for me to remove 1% of the electrons from my motorcycle.


plan B: wait another 10 billion years, and the sun will destroy the earth when it becomes a giant.


hasthelargehadroncolliderdestroyedtheworldyet.com


I'm sure electing Trump will get us most of the way there.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: