Your second paragraph seems unpersuasive to me. I would think that designing a program that can wipe out a civilization conditional on that civilization intentionally running it would be many orders of magnitude easier than designing a program that can wipe out a civilization when that civilization tries to analyze it as pure data.
Both things would require that you somehow make your concepts compatible with alien information systems (your first counter-argument), but the second thing additionally requires that you exploit some programming bug (such as a buffer overflow) in a system you have never examined. That seems to me like it would require modeling the unmet aliens in a much higher degree of detail.
Now, you could argue that if an astronomer who is attempting to analyze “gamma ray bursts” accidentally discovers an alien signal, that they are just as likely as SETI to immediately post it to the Internet. But suggesting they would accidentally execute alien code, without realizing that it’s code, seems like a pretty large burdensome detail.
(Contrariwise, conditional on known-alien-source-code being posted to the Internet, I would say the probability of someone trying to run it is close to 1.)
Your second paragraph seems unpersuasive to me. I would think that designing a program that can wipe out a civilization conditional on that civilization intentionally running it would be many orders of magnitude easier than designing a program that can wipe out a civilization when that civilization tries to analyze it as pure data.
Both things would require that you somehow make your concepts compatible with alien information systems (your first counter-argument), but the second thing additionally requires that you exploit some programming bug (such as a buffer overflow) in a system you have never examined. That seems to me like it would require modeling the unmet aliens in a much higher degree of detail.
Now, you could argue that if an astronomer who is attempting to analyze “gamma ray bursts” accidentally discovers an alien signal, that they are just as likely as SETI to immediately post it to the Internet. But suggesting they would accidentally execute alien code, without realizing that it’s code, seems like a pretty large burdensome detail.
(Contrariwise, conditional on known-alien-source-code being posted to the Internet, I would say the probability of someone trying to run it is close to 1.)