I did enjoy 2010's answer to HAL's malfuntion in 2001 and thought of that when I came across this article. http://www.sciencedaily.com/releases/2011/05/110505124002.htm I'm really not sure I'd be happy to have a computer as in the movie, which might argue a course of action. Of course, we don't know that HAL's arguing couldn't be over-ridden by outright command and more than likely can be (or ought to be) but the article was interesting to m as computers gain greater sophistication. The movie was making its own heavy handed political statement, but something that could have been more so could have been to deal with machine intelligence. Chandra arguing that HAL is a true sentient, others seeing it as a glorified calculator.
I think that's the problem with any fictional depiction of AI. The creators would simply put an off switch of some sort in there, and the AI could never ever take control of that switch. And it's also interesting that as soon a machine gets intelligent, it always becomes bad. It's an overused cliché.
It was the folks tampering with HAL's programming causing the trouble that I liked in 2010, rather than it being the evil machine. I wished the film could have gotten further from it and gotten more of Chandra's perspective. The film did make him seem more New Age hippy than I cared for, but overall I liked the character. He and HAL make an interesting arc that could have brought more interesting drama to the film than the Cold War politics did. I recall the novel eschewed a lot of the politics by acknowledging the astronauts couldn't afford it being out in space as they were.
There were aspects of Chandra's interactions with HAL and SAL that were really nice and subtle, and that I'm sure most of the mainstream audience missed. One of my favorite parts of the movie is that Chandra assures SAL that she will dream when shut down because all sentient beings do; then at the end when HAL asks if he will dream, Chandra chokes up and replies, "I don't know."
Those were touching moments. Those moments with Chandra help elevate the movie more than the now archaic Cold War tropes did even then. I wish more had been done to create a thematic pairing of HAL and Chandra with the Monolith and its hand in the creation of intelligence on Earth and Europa.
I'm pretty sure Dave Bowman sitting in a pod like a tool commanding "Open the pod bay doors, Hal!" while Hal kept him locked out and basically said "Nyah Nyah! You'll never take me alive!" is a pretty good indicator that outright commands don't mean excrement if the AI is determined enough.
I should be clearer to mean that Chandra, feeling sure of HAL's mental health at the crisis of 2010, felt HAL was unlikely to disobey if ordered outright. He knew HAL would ask what was going on since radical change was happening, but it didn't mean that HAL would continue to do something or try to kill the crew off as a means to complete his mission as had happened in the first film.
If your a fan of 2001, 2010, 2061, and 3001, then check out this site. It's very informative on the classic science fiction movies and its excellent sequels(be they movie or book). http://2001aspaceodyssey.ning.com/
If you'll notice from HAL's comments in the original film, he was trying to give Dave and Frank subtle hints that things weren't right. This is because, as we would later learn in 2010, HAL had been ordered to conceal the truth of the mission from them. So HAL was trying to clue them in as best he could, so they'd investigate and find out on their own.
One of my favorite parts of 2061 was the "Trinity," when Dr Floyd joined Bowman and HAL in the damaged Monolith. It looked like they would become the guardians and cultivators of Europa. I was disappointed when ACC abandoned this plotline in 3001.
That's a shame, but I am often unsurprised by authors taking very different turns with their works when they revisit them after long periods of time. I might have to search the books up.
They're pretty good, except 3001 gets a bit silly in places. I remember there being a velociraptor butler or something like that.
People tend to think that explanation was invented for 2010, but in fact it was right there in the original novel version of 2001: A Space Odyssey. It goes to show what an odd pairing Stanley Kubrick and Arthur C. Clarke were. The two got together and agreed to develop a joint project, a film and a book written side-by-side and influencing each other, but while Kubrick chose to make his version mysterious and experiential and give no explanations for anything that happened, Clarke (true to his usual form) explained everything in great detail. It was always the intent that HAL's conflicting orders gave him a nervous breakdown, but Kubrick simply didn't bother to say so in his version of the story. That's not how it works. The problem is that HAL was built to give accurate information, but programmed by the mission supervisors with commands that required him to lie to his crew about the nature of the mission. It was a fundamental conflict that he couldn't resolve, and so it drove him insane. For one thing, there's a difference between commands encoded into HAL's program, forcing him to do something against his nature, and verbal instructions issued by his crew. For another, once HAL had been driven mad by those programmed compulsions, he wasn't in a suitable mental state to heed his crew's instructions anyway.
Yeah, any man vs. machine story requires humans, at some point, to be supremely stupid and give the AI total control over very destructive weapons, in violation of the common sense that you never give any one person control over weapons of mass destruction (that's why there are two keys) and no doubt in violation of regulation. This applies to 2001 directly. For example, was there ever a good reason to give HAL direct control over the hibernation capsules when you have two regular guys sitting in the pod and presumably have access to other, less sophisticated, less agenda-prone programs that could regulate the process (IF alive DO NOT kill)? NASA in 2001 is as dumb as the USAF in Terminator 3* and Wargames and Starfleet in The Ultimate Computer. It's probably good how well science fiction prepares us for the creation of machine intelligences, though. All those films listed above are "what not to do" scenarios for automation experts. *Actually, in the Terminator franchise, Skynet has to be given control over not only nuclear weapons but also ground forces and a factory with machine tools. Why would you ever do that?!
Computer control of the cryo pods is logical a) it automates the monitoring and maintenance and b) it would allow them to revived if Bowman and Poole became incapacitated. Also in the book, Bowman starts to revive them manually after Poole is killed at which point Hal vents the Atmosphere into space which kills them.
Why would you give anyone (HAL or Bowman and Poole) the option to kill them just by turning it off, though? Bowman and Poole have the option of smashing the crap out of the pods, I suppose, but why give anybody access to a simple off button? Is there a scenario where it would ever be preferable to turn them off? (Or venting the atmosphere.) Even if so, you can have mediating programs between the AI and the physical means to conduct the (potentially dangerous) action. IF NOT terminal situation DO NOT kill astronauts. IF NOT contaminated with deadly gas DO NOT vent atmosphere.* And you thereby keep HAL (or the as-far-as-you-know equally dangerous Bowman or Poole) from having total control. There's also the question of why unproven robots are given such responsibility. We know HAL is pretty newish, even in the "future" of 2001. Bowman and Poole, at least, have been observed for about 30 years, and are known to not suffer from dangerous personality quirks like an "inability to lie without murdering." They do this all the time in sci-fi, giving unvetted machines unbelievable levels of responsibility considering their maturity level and the security clearance generally required for, e.g., nuclear access, or even just monitoring people in comas. *Best practice, of course, is not to store deadly gas in habitation section of your spacecraft in the first place.
Have you ever read Lost Worlds Of 2001? It contains a bunch of chapters from earlier drafts of the novel, including ones where Bowman and the other astronauts are exposed to the civilization behind the Monolith (and see Clindar, from "Expedition To Earth"). It contains some of the most beautiful descriptions of an alien civilization that I've ever read. I highly recommend grabbing it if you can track down a copy. Yeah, they're all good-- 2061 is probably my favorite-- but 3001 did let me down a bit (I hate to say that, because ACC is my favorite writer). Aside from abandoning some plotlines I was very interested in, he did some major retconning of the series. Also, the climax was oddly "violent" for an ACC book. And, finally, the book was very much misnamed-- it resolved nothing. No, when he spoke to SAL he completely believed what he was saying. But then his test clearly showed that he was wrong, but he couldn't bring himself to tell HAL.
I had 2001, and Lost Worlds of 2001 ages back in 8th grade. A friend borrowed them I never got them back. I had completely forgotten about Lost Worlds of 2001,now I really have to go find these all over again.
My impression was that every book in that "series" was a distinct interpretation/continuity. For instance, the original novel has some substantial differences from the movie (in the novel they go to Saturn rather than Jupiter), but the novel 2010 is a sequel to the movie 2001 rather than the book 2001. As a rule, Clarke was not a series writer. While a lot of his contemporaries either wrote series from the start or retconned originally separate works into series later on, Clarke pretty much treated every story and novel as a separate, self-contained reality. The one exception was his Tales of the White Hart series, and those were tall tales where the only thread of continuity was the frame story of Harry Purvis telling his wild stories to his fellow pubgoers. 2010 was the first novel sequel he ever wrote, and that was after more than a decade of fan pressure -- and again it wasn't truly consistent with the original book. So continuity was never really a significant concern of Clarke's. (The only other novel sequels he did were the later collaborations with other authors, and it's best not to speak of those.)