If we find a structure on Mars

[RangerWickett]

First, the moment we discover anything that appears to be created by an alien intelligence, we must 'air gap' planet earth. No networked devices should be allowed to receive data directly from sources outside our local network of satellites. It's an exceedingly remote possibility, but we should not risk the chance that something like this happens:

A rover investigates the structure.

The rover is subdued, reprogrammed, and ordered to send innocuous dummy information back to earth.

Contained in that dummy information is an evolutionary program that will determine the design of our software, learn its vulnerabilities, and nest itself into our networks while disguising its presence.

Once it reaches sufficient processing power to become a functional artificial super intelligence, it engineers a way to get us to launch our nukes, wiping out all civilization and setting the human race back far enough that a clean-up fleet is able to arrive and wipe the planet clean.

 

[Umbran]

First, the moment we discover anything that appears to be created by an alien intelligence, we must 'air gap' planet earth. No networked devices should be allowed to receive data directly from sources outside our local network of satellites.

With respect, unless you are talking about a hard-wired connection, "outside" doesn't have any relevance. Either the data comes with proper authorizations, or it does not. If they are as good as you suggest later in the post, no security measure will matter.

The rover is subdued, reprogrammed, and ordered to send innocuous dummy information back to earth.

So, your concern is reverse-Independence Day? The aliens can and will, without clear and obvious threat, set out to destroy us through computers, and they're able to do so based on one example? Never mind that it contains only limited examples of the many sorts of hardware and operating systems we actually use? This makes as little sense with computers as it does with biology - just as you can't, based on the example of one organism, design a virus that will wipe out all life on Earth, based on one computer you cannot create an exploit that will apply to all computer systems. Heck, this was a selling point for Apple for a long while - all the exploits on Windows didn't apply to Apple machines...

Heck, some of those nuke launch systems are probably sill updated using floppy disks, with no network connections to the rest of the world!

But okay, fine. Let us say that's the danger. This argues for us to send probes that have *no security measures whatsoever* on them. The aliens, then, will have no understanding of what form our defenses take. They'll use the usual buffer-overflow and other exploits we've known how to deal with for ages, and their ugliness will die nascent, never reporting back to its Martian Overlords.

Then, they'll send their huge battle fleet, which, due to a terrible miscalculation of scale, will be swallowed by a small dog.

(Sorry. I just felt the thread needed a Hitchhiker's Guide reference, and this is where it fit in )

 

[Dannyalcatraz]

We also have Justin Bieber. Not quite as powerful as Vogon Poetry, but still...

 

[Janx]

First, the moment we discover anything that appears to be created by an alien intelligence, we must 'air gap' planet earth. No networked devices should be allowed to receive data directly from sources outside our local network of satellites. It's an exceedingly remote possibility, but we should not risk the chance that something like this happens:

A rover investigates the structure.

The rover is subdued, reprogrammed, and ordered to send innocuous dummy information back to earth.

Contained in that dummy information is an evolutionary program that will determine the design of our software, learn its vulnerabilities, and nest itself into our networks while disguising its presence.

Once it reaches sufficient processing power to become a functional artificial super intelligence, it engineers a way to get us to launch our nukes, wiping out all civilization and setting the human race back far enough that a clean-up fleet is able to arrive and wipe the planet clean.

Umbran kind of covered this, but I'm left wast left with as case of "frp, drp, erp, gack, bdah!" when I read this.

Sure, I enjoyed ID4 as much as the next guy. But as a software developer I know that code is constrained by the CPU used. You can't cross the boundaries. So whatever chip is on the probe, likely isn't an Intel x86 architecture (or whatever we call the current AMD/Intel CPUs in production now).

You also can't really send more data than you can use. A virus has to be nimble. I can't send a virus that can infect a probe AND has the brains to build an AI AND has the mission for the AI in the size of the attack payload that the probe can handle, let alone when it gets to the Earth as transmitted data, it's going into a data file, not an executable. Nobody executes the bytes coming back from a probe, let alone any subsystem, because the contract between two systems is data, not exectuable code (remote firmware updating aside, which is also one directional). If a virus is not executed, it does not have any bite.

So , sci-fi aside...

Odds are good any structure we find is dead. It might have old tech or old viruses. And yes, if we brought any of those artifacts home and somehow wake them up, that might be risky. But our initial futzing around with probes is actually pretty safe. The sending of code or germs is largely one way, until we get a way to transport back to earth.

 

[Vagabond234]

For that, we need to send a team of humans. Pray Elon Musk gets it done! Machines are way too slow and not dependable.

 

[Umbran]

Machines are way too slow and not dependable.

Not dependable? You know that NASA has a track record of making probes that survive for *years* past their original design parameters, yes?

Nobody is perfect, but for flinging something a million kilometers, hitting a speck *softly*, and having a device work after landing, they're not shabby.

 

[Ryujin]

Not dependable? You know that NASA has a track record of making probes that survive for *years* past their original design parameters, yes?

Nobody is perfect, but for flinging something a million kilometers, hitting a speck *softly*, and having a device work after landing, they're not shabby.

Yes, NASA's probes are definitely dependable, with a very small failure record considering the environments in which they perform and the stresses they are under, when being delivered to their targets. They perform their functions almost flawlessly. What I would say that they are not, is flexible. You can only pack so much into such a small package. Humans are arguably less dependable, harder to get on target, but infinitely more flexible.

 

[maxnmn80]

I would send first Bruce Willis well armed, just in case.

 

[Maxperson]

I would send first Bruce Willis well armed, just in case.

He'd get wasted. Send Sigourney Weaver.

 

[Umbran]

What I would say that they are not, is flexible. You can only pack so much into such a small package.

That's certainly true.

Humans are arguably less dependable, harder to get on target, but infinitely more flexible.

Well, not *infinitely*. The human will not actually be any more flexible than the laboratory you send with him or her. And that lab won't be able to restock consumables, or get new equipment in short order. Part of human flexibility lies in our social infrastructure, which doesn't exist on Mars.