# Effects of erasing 90% data on humanity

I am playing with idea of Earth being attacked by aliens. Such attack is aimed on our computers and reason of attack is to show to the Federation of planets that we are still weak to be contacted.

The question: How would we handle after such attack?

• The attack is synchronised, covering all the Earth in one shot
• Such shot sets all data stored in RAMs to all logical "1"
• Also, hard drives are given instruction to write to all sectors logical "1"
• Both instructions are on hardware level, so no check of write rights given. Assume the computer under attack will fulfil the request.
• Assuming the aliens know our technology, the write process on HDD starts with OS sectors and then continues to "data" sectors. This applies also to SSD
• Everything else than RAM and HDD/SSDs (namely tapes, DVDs and other "offline" backup systems) remains unharmed
• To be absolutely clear, the attack also targets on routers and modems

While the scope of attack is bit "magic", it obeys the laws of physics:

• The attack is electromagnetic pulse based
• So if a computer is hidden behind something which should prevent electromagnetic pulse, such computer remains fully functional
• Also, the pulse obeys square inverse law, so the more mass is between orbit and the computer, the less effective the attack is.
• My working meta is 10 metres of (any) mass between computer and orbit to render attack fully uneffective.
• And, the attack is only one-off. No repeat of such attack.
• Also, I am aware of one huge flaw: Computer has to be turned on in moment of attack. The attack has no effect on turned off computers. However, as long as the computer is at least in "sleep" mode, the attack has effect.

I know I would not destroy all the computers, but I also know that I would be able to turn off more than 90% of all Earth computers.

Imagine there is alien organisation called "Federation of Planets." They have FTL drives and actively seek other intelligent species to join them. In order to join, you have to pass a mark of setting foot on other planet than your own.

There are "anti immigrant activists" who think, that the acceptance criteria should be raised to "visit other star system than own". These activist know, that "The Auditors" are going to visit Earth soon and check how advanced the Earth is.

And they know that if we are unable to communicate, the Earth is not going to be accepted and next visit is going to happen in 50 years. So while the attack is mainly aimed on our communication, they decided to go over the top and attack us on whole scale.

• "...reason of attack is to show to the Federation of planets that we are still weak to be contacted." Could you clarify what this means, please? – Lostinfrance Oct 11 '15 at 18:18
• Also keep in mind that many large institutions have their critical data secured in storages that are at least partially secure against electromagnetic pulses: thestack.com/data-centre/2015/10/05/… In this way, a lot of the most critical data might be secure from the attack. – Peter S. Oct 12 '15 at 10:48
• I would find a rocket, get into orbit and sucker punch each and every alien for erasing all my game save data... – James Oct 12 '15 at 17:54
• Does it also erase firmware, BIOSes, etc, and at what time does it do so? I'm assuming erase operation are limited to the normal rate of hardware. – timuzhti Oct 13 '15 at 7:45
• For clarity, the effect is not random? It completely wipes the data from every computer that is turned on, but has no effect on the computers that are turned off? (like, say, data stored on backup tapes?) – Cort Ammon Oct 13 '15 at 15:52

Actually, while it would floor things for a while, I don't think it would even be close to long term effect.

While not available to the masses, rolling back to 70's-80's tech would be achievable.

Your EMP doesn't destroy data written to CDs (either mass-produced (printed) or rewritable (heat/melt phase change)) or on things like punch tape or punch cards.

Punch tape could rebuild some 70's type tech which would give you micro controller firmware and so on, and from there something that could read CD ROMs. From the CD-ROMS you can get backups of OS source code and firmware for anything up to probably the 2005 time frame.

There are probably enough people in the world adequately capable of writing straight machine code so they could program basics of firmware, BIOS and bootloaders from scratch.

Because your weapon doesn't destroy the hardware, all it needs is re-initialising.

There'd be some chaos, but if people were organised and had enough time (anything from a week to 3 months depending on how much chaos was going on - and remember old radios, old tvs, old cars and such don't need computerisation, nor do basic electric generators).

Assuming an invasion didn't wipe us all out in the mean time, I'd say given a year everything would be sort of back to some sort of reasonable level. True, we've lost perhaps 10 years of data and tech, financial markets are a mess, no one knows who owes what, but eventually paper records get typed back in and even then not a 100% is lost.

However, that's all assuming 100% currently electro magnetic storage is destroyed.

If even a small percentage is unaffected, things return to normal much faster - it's just the time it takes to slowly rebuild section by section, but it's like the rice on the chess board, it starts off slowly, but as each fixed computer means someone can help fix a couple more, it quickly spreads across the world. There is currently a massive excess of backups, while not everything would get restored, I don't think we'd lose anything truly significant.

• +1 Mind you, I would not enjoy having this done to my civilization moments before an attack, but its the attack that will do the damage. It would not take long at all to bootstrap the world. – Cort Ammon Oct 13 '15 at 15:53
• I think the answer, while feasible, underestimates the damage that the disruption to the status quo will produce. You're talking about reconstructing 40 years' worth of the Information Age from surviving paper documents in the span of a few months, which I would give as the conservative estimate to get even a single computer functional again (keeping in mind that the attack would likely bring down basic infrastructure as well; the electrical and traffic grids are highly computerized, so you'd be running the first computers from generators using fuel pumped manually from gas stations. – KeithS Oct 14 '15 at 18:23
• I honestly think that if such a thing occurred in the real world, it would bring society to its knees; within 24 hours of such an attack you'd have endemic looting in major cities for basic necessities like food, which nobody would be able to buy because they'd have no proof of their bank balance and no way to convert that number into cash or anything else of material value. Once that starts happening there's really no stopping it. – KeithS Oct 14 '15 at 18:27
• And, this is all assuming no yahoo in command of a nuclear arsenal has implemented a "fail-deadly" system that launches the nukes in the event the chain of command is presumed destroyed by loss of contact. There is public knowledge of plans for such systems by the Russian nuclear command, and really all it takes is one ICBM launch and all the nuclear powers will launch what they have at whomever they can in mutual retaliation. – KeithS Oct 14 '15 at 18:31
• Such shot sets all data stored in RAMs to all logical "1"

This would have the same practical effect as the power going out. Restart your computer and everything is fine. The big effect here would be on airplanes and so forth -- I imagine it's difficult to do a hard reset of all the electronics while the aircraft is in-flight. However, this answer suggests it is possible.

• Also, hard drives are given instruction to write to all sectors logical "1"

This would have a much larger effect, since most data is stored on hard drives. Individuals would tend to lose everything (not many people keep backups), but corporations would just lose hours to days of data, and could quickly recover most everything.

Because very little personal data is truly important, most people would be largely unaffected after a few days to weeks, since they could just recover the data from corporations who backed it up.

• And they know that if we are unable to communicate, the Earth is not going to be accepted and next visit is going to happen in 50 years.

We can build radios out of random crap in your garage in like 10 minutes. This attack won't prevent communication with aliens in orbit.

• While the scope of attack is bit "magic", it obeys the laws of physics:

No, it really doesn't.

• The attack is electromagnetic pulse based
• So if a computer is hidden behind something which should prevent electromagnetic pulse, such computer remains fully functional

An EMP is indiscriminate. It will wipe out any and all electronics or none of them. The only question is how sensitive the given electronics are. So if the computer is partially shielded, the most sensitive components will tend to die while the least sensitive will tend to not die. The HDD platters are far from the most sensitive, and anything affecting HDD platters will also affect tape drives and other magnetic media.

What you're thinking of is sending specific signals to a specific part of a hard drive's controller to make it think the OS told it to erase things. But you really can't do that. The normal radio photons are feet to miles across, so there's no way to make sure a photon hits a specific HDD controller, let alone a specific wire on that controller. Even wall-penetrating microwaves are several inches across. By the time you get to EM waves small enough to hit a target wire, we're talking optical light EHF/IR and beyond, which is easily blocked by walls, CPU cases, etc., and isn't readily absorbed into said wires.

Now, sufficiently advanced aliens could hypothetically create an incredibly advanced interference pattern that cleverly hits specific points, even with large wavelength photons. However, this would require billions, if not billions of billions, of photons per bit sent. Because photons have fixed energy amounts, that means such an interference pattern would end up vaporizing everything around the computer, if not the entire planet (conservation of energy says even if there's destructive interference, the energy has to go somewhere).

• Also, the pulse obeys square inverse law, so the more mass is between orbit and the computer, the less effective the attack is.
• The attack is synchronised, covering all the Earth in one shot

The inverse-square laws are about distance, not mass between you and the attacker. This means an attack from 100 km above the planet will have almost zero effect on objects on the other side of the planet (not including the massive planet in the way).

Because everything is synchronized anyways, there are probably dozens to thousands of ships in orbit around the planet. The fewer ships, the further they have to be to get good spread. The more ships, the less distance there is between them. Either way, coverage will be relatively uniform, so the inverse-square law won't be a huge consideration. And the aliens are probably smart enough to concentrate the "bright" spots on the major cities.

• My working meta is 10 metres of (any) mass between computer and orbit to render attack fully uneffective.

Mass between you and the attacker will exponentially decay signal strength, and the coefficient will greatly depend on the material in question, as well as the wavelength of the EM radiation used. The basic form is $I=I_0e^{-\beta D}$ where $I_0$ is intensity at the surface of the material, $D$ is depth into the surface, and $\beta$ is the coefficient for the wavelength and material in question.

Air will have a very low $\beta$ value for most wavelengths, while concrete will have a high $\beta$ value for most wavelengths. Long wavelengths (low frequencies) will tend to have lower $\beta$ values than short wavelengths.

• Also, I am aware of one huge flaw: Computer has to be turned on in moment of attack. The attack has no effect on turned off computers. However, as long as the computer is at least in "sleep" mode, the attack has effect.

This is only a problem if the attack somehow activates the "overwrite" circuits on the HDD controller. If it, realistically, just overloads the physical data structures, it will work on anything whether it's on or off.

• Microwaves are radio waves. So are millimetric waves (tens to hundreds of gigahertz range). The commonly recognized boundary for when we stop talking about "radio" is 300 GHz, which translates to a wavelength of (in meters) 300/300000 = 1 mm (because $\lambda = \frac{v}{f}$, where for radio we can pretty much state $v = c$ and not really need to worry about the exceptions for the purposes of this question). A more pressing concern is that computers are generally built to be RF shielded at the frequencies they are operating on, due to EMI regulations. – a CVn Oct 13 '15 at 7:42
• Also, OP is asking about the effects on something like the described happening, which you only really discuss in the first few paragraphs; a good two thirds or so of your answer is simply stating why the scenario won't happen, which isn't really relevant to discussing what effects it would have. – a CVn Oct 13 '15 at 7:42
• I didn't realize microwaves and radio waves overlapped in definition. Still you'd need 1 mm resolution or better, and wikipedia says the 10-1mm range is blocked by basically anything heavier than clothing, and wouldn't likely make it to the ground from the alien spaceships, which is what I wrote. Also, there may be parts of the computer that are shielded, but the connectors aren't. The only thing between me and the HDD pins on my PC is the plastic holding the pins. So that's going to depend on the PC setup. – MichaelS Oct 13 '15 at 19:45
• And I'm not really saying it can't happen, just describing more realistically how it would happen since the question seemed to want a semi-realistic method of doing it. But yeah, only three of the paragraphs are answers to the main question, which I did answer. – MichaelS Oct 13 '15 at 19:49

Effects concerned (ability to communicate) would be very low.

HOW IT COULD NOT SUCCEED:

1) It would not affect people. Professionals all around the world are keeping up with their respective fields more o less. Technology concerning, only latest research (5 years, 15 max) could be lost. Also keep in mind that research papers are usually literary kept as papers in university libraries.

2) if it would not affect turned off computers, technological impact would be even less. In all filed you have thousands of students caring their laptops filed with latest knowledge. chances that some of them would have their laptops turned off at time of attack are close to certainty.

HOW IT COULD SUCCEED:

1) however, it would greatly impact out 'organization'. Databases of people, logistic databases etc. You would lose all kinds of data from lent books in library to who has access to security facilities. Everybody would know who is president without Wikipedia, but let's take Universities, you know don't know who are students, what marks they had, You would also lose most digital contracts, etc.. Minor a occasionally major social chaos would occur, and that could escalate into conflicts.

2) it would stop us for a while. Top scientist could not continue their researches further, but would need to focus on restoring our knowledge bases from their memories and fragments of surviving data. Put together with 1) it would mean we would be stuck for a while.

SUMMARY:

It has little to no chance to make us unable to communicate technologically, but it could realistically succeed by enduring chaos a making us unable to cooperate on answer.

Two parts to your question, the specified mode of attack and the effect on humanity.

Let's talk about the effects first. Total breakdown of civilization in industrialized areas. Food isn't delivered to supermarkets, trains don't run, power grids fail. Civil defense might have contingency plans for any one failure, but they'll be overwhelmed. Non-industrialized areas might cope a little better, but they're not self-sufficient.

Regarding the attack, you're getting overly specific for something which is magic ("all harddrives are ..."). Some notes:

• The effect of overwriting RAM will conflict with the effect of overwriting harddrives. To do so you need working computers. What if someone pushes the reset button while the overwrite process is running?
• You are talking about "inverse square" and "less effective", but the way you describe the attack there could be no partial effect. Either the harddrive controller is hijacked or not.
• Even if people had offline backups, they'd have to restore the OS, application software, and data. Restoring the OS could run into problems with license codes/DRM.
• How about computers that were switched off during the attack? Either cold backups, or personal computers shut down to save energy/cut the noise.
• Yep. I discovered the same flaw in my attack plan: Everything whats turned off is safe from attack – Pavel Janicek Oct 11 '15 at 15:23
• I agree. The questions initially sounds like a virus, but ends up just being an EMP. – NPSF3000 Oct 12 '15 at 7:27

Assuming that the electronics aren't completely fried, just "code-less" it won't take very long to bring everything back online. An attack of this kind just makes computers into the human equivalent of comatose, not lobotomized.

The OP doesn't mention anything about firmware being '1'-ed out so all the controllers in the hard drives will continue to work. Anyone who has offline storage media (flashdrives, CDs, DVDs) should be able to reinstall the OS and applications.