Where to Find More of Dick Carrigan's Thoughts on Sinister SETI Messages

[beamjockey]

I can't recall in whose Livejournal I read about Richard Carrigan's recent paper. He speculates about the dangers of inimical software possibly buried in SETI signals.

All the discussion on the Web seems to be coming from an article in The Guardian.

You can read more about his ideas in "Do Potential SETI Signals Need to Be Decontaminated?" (it's in Word *.doc format).




At least two scenarios need to be considered in protecting against a malevolent SETI Hacker signal. One is a computer virus in the message that takes over the computer at the receiver. The other is an open message that gives an impenetrable software code or instructions for a hardware translator to handle an opaque message. Both cases are dangerous. The damage may be done before the receiver appreciates that it is under attack. This is the current experience even with earth-based hacker attacks. There may not be an opportunity to pull the signal out of the computer or turn off the power before the intruding signal has taken over.

It is an open question whether an earth-based computer virus can penetrate a computer if it is not familiar with the operating system. The computer and computer security experts I have discussed this with don’t think it is possible. The argument goes that viruses typically enter a computer by exploiting known features in the operating system. Further, experts argue, typical computer operating systems are quite idiosyncratic so that it can be difficult to analyze their structure from a logical point of view.

However, it seems worthwhile to approach the question with an open mind. For example, one could set up a thought or even a practical test with a primitive “toy” computer, perhaps modeled along the lines of the first Illiac and have programmers unfamiliar with the Illiac system try to hack the program. I believe it would also be useful to convene a workshop with diverse participants to discuss the subject in some detail. This might be coupled with broader discussions on the topic of denaturing ETI signals.

Comments

[mmcirvin.livejournal.com]

For example, one could set up a thought or even a practical test with a primitive “toy” computer, perhaps modeled along the lines of the first Illiac and have programmers unfamiliar with the Illiac system try to hack the program.

For this to be a good analogy, they'd have to be able to hack it without ever being allowed to see or interact with the computer. Corrigan's first step—bootstrapping to a working program—already requires the results of his second step, in which the program analyzes the system it's in. Unmotivated biological viruses arise out of a biological environment in which the host organisms are already present.

So it seems to me that this is about as likely a danger as the aliens simply beaming invasion soldiers over physically with a Star Trek transporter. If they're godlike enough to do things that can't be done according to any Earth conception of possibility, how would you possibly plan against it? What if they write a signal so cunning that it is able to reconstruct the missing information after you denature it? What about a signal that turns into a shark and eats you? A signal that makes 2+2 equal 5 so that the laws of physics break and you explode?

Now, the more interesting possibility that I think is a real danger is something like the A for Andromeda scenario, in which the message starts with social engineering assuming that something like an intelligent being is reading it. Begin with schematics for a physical machine, Contact-style, or even an abstract description of a processor with an instruction set; and then once the silly monkeys build that machine, it can do the malicious work. Much more likely than the signal doing it all to the receiver's controller upon receipt.

[mmcirvin.livejournal.com]

...Carrigan's "instant takeover" scenario becomes more likely if computer architecture becomes stable to the point of few security patches over a period comparable to the light speed round-trip time (but the target civilization is still leaking lots of messages that might reveal the machines' internal structure). That's not our world today, though.

[gomeza.livejournal.com]

Hey now, if Jeff Goldblum can inset a virus into an alien network with his PowerBook, I see no reason why aliens can't attack our computers despite never having seen one and with no feedback channel.

This is choice. ^_^

[gomeza.livejournal.com]

oops:
%> s/inset/insert

%> grep coffee kitchen

[ffutures.livejournal.com]

I'm a good deal more worried about what the current implementation of the SETI at Home software might be up to, since it is set up to allow multiple distributed computing tasks, not all of them related to SETI, and you pretty much have to accept that they're benign without much evidence. The old version turned out to be responsible for the apparent failure of the serial ports on my PC, it was stealing interrupts they needed, took me forever to figure it out. Fortunately the new one dosn't want to run on my PC anyway, so I am now out of the SETI business.

As for alien signals taking over our computers over interstellar distances... about as likely as alien diseases being able to attack humans, or human diseases to attack aliens, without physical contact and time to evolve.

[mmcirvin.livejournal.com]

Not to mention that there are well-intentioned people eager to install the client on work machines without permission. In practice it's probably harmless, but if I were a corporate IT guy I would not count on that.

And of course...

[seawasp]

... this is a pretty old idea; Piers Anthony used it as one of the main drivers of the book Macroscope. I'm not sure when it was first used, but before that, I'm pretty sure. (In that book, the "software" was directly educating the living beings observing it; this assumes that there are some basic commonalities in the way living creatures process data, of course, but it's a possibly reasonable assumption)

[minnehaha.livejournal.com]

I blogged this.

B

Dates

[(Anonymous)]

Dick is a friend and he postulated this idea in the 1960's. He and his wife Nancy wrote and published a book about it called "The Siren Stars". He also was, before retirement, a leading particle physicist at Fermilab in Batavia, Illinois. So at the time of this idea, he was about 20 years ahead of the curve and 30 years ahead of IT (although Fermilab was one of the first organizations to have the internet). 3'rd replicator entities would , by nature, operate at a much different level. Could they not gain slight entry and then analyze and improvise their way in through the portals, gaining control?

Re: Dates

[beamjockey.livejournal.com]

I did read their novel The Siren Stars, years before I met the Carrigans. (Even earlier, fear that information from a SETI signal might pose a threat played a big role in Hoyle and Eliot's A for Andromeda.)

(although Fermilab was one of the first organizations to have the internet)

And the World Wide Web in particular-- as far as I can determine, Fermilab was the fourth institution to host a Web site.

Could they not gain slight entry and then analyze and improvise their way in through the portals, gaining control?

Yes, that is the essential question.

It's hard to imagine how this would be possible, starting with reception a passive radio signal.

In numerous SF stories, including the ones mentioned above, trouble starts when someone, using plans provided by aliens, builds a computer he doesn't really understand. Would we really be this dumb? Still a hypothetical question at this point.

Re: Dates

[charlie-meadows.livejournal.com]

*Can* one build a computer one doesn't really understand? How would one know if one had even "followed the instructions" correctly?*

Consider even a situation which doesn't involve a material object: what are the odds that someone correctly solves a problem in, say, mathematical physics if they neither understand the mathematical techniques involved, nor the physical model? (I haven't see a student manage it yet...)

Sure, Cal Meacham built a working interrossiter, but that was a pre-fab kit. The aliens probably aren't sending us packing crates over the radio...


*It occurs to me that the *real* danger is if the aliens include a tech support number...