The Speculist

Of course, didn't you see 'Independence Day'?

I received this CERT-like advisory and shot coffee out my nose:
http://www.f-secure.se/virus/hoax/hoaxinfo.asp?Id=82

Tob

I argue that a "perfect", system-hopping, all-destructive virus would be impossible to program for anyone having only a black-box view of a system.

We must not forget that any piece of software is only a means to switch some transistors in a meaningful fashion. Any abstraction of code that goes beyond this principle (i.e. higher-level programming languages) is based on a seperate piece of software (a compiler/interpreter) hiding the actual hardware from us humans (or aliens, for that matter). I deem it impossible to write any purposeful piece of software lacking any knowledge of the underlying hardware, which would be the situation an alien programmer would find itself in.

The most pseudo-plausible (and far-fetched) approach would probably be to implement an advanced AI software that is somehow "aware" of the system environment it is placed in, and could self-modify to do something harmful (possibly by recognizing some kind of bit patterns that identify certain parts of the system, i.e. network functionality). This still implies that alien computer technology rests on the same pillars as ours, namely transistors and the binary system, which is wildly unlikely if you ask me. But then again, this is just the kind of anthropomorphic (wishful) thinking the SETI people seem to caught up in.

I think that eisendorn is correct. Computer code requires a knowledge of the the instruction set. A virus written for x86 hardware would not run on a mainframe.

I seem to remember an 'Outer Limits' show that had a more interesting take. If memory serves, a teenager astronomer records transmissions from deep space that sound addictively like new-age music. As people listen to the 'songs', biochemical changes occurred that caused them to mutate. As far fetched as that sounds, I would think that more likely than a super-JAVA (write-once-run-anywhere) virus.

Tob

Eisendorn:

I agree that a standard virus couldn't work.

Even a learn-as-you-go approach would need to have some knowledge of the system environment.

Perhaps if you had many digital "spores" one might work welll enough to evolve into something. But the hacker alien civ couldn't predetermine what it would evolve into - It might turn out to be a great recipee for oatmeal raison cookies.

Toby:

Eisendorn has convinced me too, but the example you gave - a biological hack - is the exact same problem on a different substrate. You still have to know how to address the system, right?

The only thing that would make this (slightly) more likely than the SETI hack is the fact that our biology has changed little in millions of years. This would give ET's more time to learn our biology, write the hack, and deliver.

Sounds like more trouble than it would be worth to a highly advanced civilization.

Stephan, I agree that both scenarios are unbelievable remote. I guess I was just suggesting that it was more likely for us to have similar biology with ET than for us to independently derive the same binary instruction set for our computers and that is key. Even a self-modifying program has to get loaded from a PSW and then understood by the chipset to get executed. Its not 'alive' to be modified until its running. I'm doubtfull that an alien tech would even use binary systems. There is a good argument that analog systems (yes/no/maybe/some/alot) will be the preferred computer eventually. I've been in the computer industry for almost 30 years and IMHO the chance for this alien virus scenario approaches zero.

Tob

The 419 letters is one thing. I just don't want any pseudopod enlargement emails.

Alien hacking is not a problem if you have Mac OS -- the Aliens have been Wintel since the mid 1980s.

Ok, since the hacker ET side is losing, let me present a "Fire in the Deep" scenario. An alien civilization at A sends elaborate nested plans for building nanotech machines. They don't even have to be close enough to otherwise communicate. The plans describe how to build machines which build machines, etc. At some point down the road when the builders at B no longer have the intellectual capacity to keep track of what's going on, the nanotech goes "grey goo" using secret codes hidden in the complex plans, terraforms planet B, and builds replicas of the aliens from A. Ie, the aliens from A have colonized a planet just by broadcasting information.