Star Trek (TOS): The Ultimate Computer

[ScreamingDoom]

Couldn't we run a bunch of simulations for this computer to make sure it wouldn't murder anybody first? It is not like the computer can tell the difference between a simulation and reality any better than a Matrixed human could.

Didn't Spock mention in the episode that the M5 went through tons of simulations and passed them all with flying colors? The wargames with actual ships was a final test, I thought.

[rickgriffin]

When I watched this episode during my binge of TOS, I did not like it. Although my dislike, I think, was due to the relatively small point that Kirk is whinging that a future of automated ships means he doesn't get to be captain. The issue I had is that this kind of problem is usually framed in terms of blue collar workers having their factory jobs taken away . . . but Kirk isn't a blue collar worker, he's a captain of a dang starship.

I did get, though, that Kirk's position was tempered with a lot of other positions so it didn't actually come across as one-sided as it did in my head. It's just that between Kirk on the edge of not getting to play captain no more, and him being vindicated in the end when the computer goes crazy, made me feel like it was attempting to justify Kirk's petulance.

Then again, that's probably because I also don't really like the super old school computer-goes-crazy plots. Which don't get me wrong, I love mad AI, but the old school stuff feels like an inevitability--of COURSE the AI can't actually be all it's cracked up to be or else our heroes wouldn't have anything to do. It feels anti-technology in the way that only old school sci-fi did with its Scientific Hubris setups (even when the inevitable disaster had absolutely nothing to do with the science in question)

Despite that, I would admit that the characterization in the episode was on point. My own personal misgivings with the setup would only likely have me deduct a point or two.

[Rocketboy1313]

Didn't Spock mention in the episode that the M5 went through tons of simulations and passed them all with flying colors? The wargames with actual ships was a final test, I thought.

That is what I don't get. To a computer, all reality is virtual. So why did it pick this instance and not any other to do all of this?
How was the computer able to tell the difference between a computer simulation and reality? If it performed well in sims then it shouldn't have performed differently here.

[Crowley]

Thinking of this episode next to Search for Spock, it would make a lot of sense if Scotty took some pointers from M5 for how to rig up the Enterprise to operate with virtually no crew.

[ChiggyvonRichthofen]

I think it comes back to the way M-5 was created, with the impressing of Daystrom's memory engrams. The M-5 may well have seen a difference between simulations and reality, because that's how Daystrom saw it. Daystrom points out that it is learning and growing by being in control of the Enterprise as well. It wouldn't shock me that the effect of all that power wouldn't have been perfectly anticipated by simulations.

One of the ironies of the episode is that the M-5 may have passed its test if it had used the engrams of someone other than the man who created it in the first place. I think the disastrous results here probably do a lot to explain why automation in ships hasn't advanced further by the TNG era.

I did get, though, that Kirk's position was tempered with a lot of other positions so it didn't actually come across as one-sided as it did in my head. It's just that between Kirk on the edge of not getting to play captain no more, and him being vindicated in the end when the computer goes crazy, made me feel like it was attempting to justify Kirk's petulance.

Then again, that's probably because I also don't really like the super old school computer-goes-crazy plots. Which don't get me wrong, I love mad AI, but the old school stuff feels like an inevitability--of COURSE the AI can't actually be all it's cracked up to be or else our heroes wouldn't have anything to do. It feels anti-technology in the way that only old school sci-fi did with its Scientific Hubris setups (even when the inevitable disaster had absolutely nothing to do with the science in question)

Yeah, that's why I think this episode doesn't always get the reception it deserves (its well-liked, but I don't see it on a lot of "best of" lists). There had been a lot of computer-go-crazy plots before this, and it seems like just another one at first glance, but there's a lot more nuance to this one.

I don't think the episode is trying to justify Kirk. Kirk has two conflicting instincts- one that the computer is bad news, which proves to be correct. The other instinct is a sort of nostalgia for captaining a starship, which he expresses in that great scene with Bones in his quarters. The latter is pure sentimentality, and it's really only a (happy for him) coincidence that that instinct is vindicated.

[FaxModem1]

I think it comes back to the way M-5 was created, with the impressing of Daystrom's memory engrams. The M-5 may well have seen a difference between simulations and reality, because that's how Daystrom saw it. Daystrom points out that it is learning and growing by being in control of the Enterprise as well. It wouldn't shock me that the effect of all that power wouldn't have been perfectly anticipated by simulations.

One of the ironies of the episode is that the M-5 may have passed its test if it had used the engrams of someone other than the man who created it in the first place. I think the disastrous results here probably do a lot to explain why automation in ships hasn't advanced further by the TNG era.

I'm not sure about that. The Enterprise-D and E are much larger than the original, and while the crew of the Enterprise-D is a 1000, most of those are specialists in one field or another, botany, archeology, astrophysicist, etc. or civilian crewmembers/families. It's quite possible that they used the automation possible by the M-5 for a lot of running of the ship, with crew members there in case things go wrong, if repairs are needed, or there are modifications needed.

Of course, we haven't seen dreadnoughts crewed with only 20 people or anything like that, but the Starfleet ships do seem to have a better ratio of ship to crew than they did in the 23rd century.

[ChiggyvonRichthofen]

I think it comes back to the way M-5 was created, with the impressing of Daystrom's memory engrams. The M-5 may well have seen a difference between simulations and reality, because that's how Daystrom saw it. Daystrom points out that it is learning and growing by being in control of the Enterprise as well. It wouldn't shock me that the effect of all that power wouldn't have been perfectly anticipated by simulations.

One of the ironies of the episode is that the M-5 may have passed its test if it had used the engrams of someone other than the man who created it in the first place. I think the disastrous results here probably do a lot to explain why automation in ships hasn't advanced further by the TNG era.

I'm not sure about that. The Enterprise-D and E are much larger than the original, and while the crew of the Enterprise-D is a 1000, most of those are specialists in one field or another, botany, archeology, astrophysicist, etc. or civilian crewmembers/families. It's quite possible that they used the automation possible by the M-5 for a lot of running of the ship, with crew members there in case things go wrong, if repairs are needed, or there are modifications needed.

Of course, we haven't seen dreadnoughts crewed with only 20 people or anything like that, but the Starfleet ships do seem to have a better ratio of ship to crew than they did in the 23rd century.

Even in The Ultimate Computer we see a robot ship, which naturally raises the question of what exactly the M-5 does differently. I think the difference we see in the episode is M-5's sapience and its ability to make command decisions- what course to take, attack patterns, landing party configuration, and so on. All those tasks are still performed by organics in the TNG era. On the other hand, it's true that we don't see the nitty gritty details of what exactly the 400 people who were no longer needed actually did, so it's hard to say whether any jobs there were replaced.

One example is how they use the computer for troubleshooting in TNG. In bottle shows in particular, the computer often "knows" the solution to the problem. And yet someone, usually Geordi, still has to ask the computer to scan for something, ask if this or that technobabble solution will work, and tell the computer to enact the solution. In a universe where artificial intelligence can be created on accident, the Enterprise-D's computer seems to be carefully crafted so that it doesn't have any sapience or agency.

[Beastro]

Liked this episode and Chuck's review of it. This episode seems more relevant today than when it came out. Driving around in the Philadelphia area, yesterday, I couldn't help but think "self driving cars can't get here nearly fast enough."

You will until you realize how much it'll cut into your autonomy.

Am I the only person who doesn't anthropomorphize AI?
"Would you kill a child?" Is a massive begging of the question as far as "Is it alive?" and "If alive is it worth moral consideration?" or "If it is alive and worth moral consideration, can we still remove it from torpedo control until we are sure it won't lose its shit and murder a bunch of people?"

We boxed up Lore and it was a humanoid android.

Couldn't we run a bunch of simulations for this computer to make sure it wouldn't murder anybody first? It is not like the computer can tell the difference between a simulation and reality any better than a Matrixed human could.

Same.

For me any sort of advanced, autonomous AI is a sociopath lacking God knows how many years of social evolution that has produced the make up of our.

People need to stop looking on machines as just different kinds of humans and more like a strange animal, and like any sort of animal when it threatens human life the appropriate response is to destroy it.

[TrueMetis]

Liked this episode and Chuck's review of it. This episode seems more relevant today than when it came out. Driving around in the Philadelphia area, yesterday, I couldn't help but think "self driving cars can't get here nearly fast enough."

You will until you realize how much it'll cut into your autonomy.

How will it cut into people's autonomy?

[Durandal_1707]

Okay, here's a devil's-advocate counterpoint to the classic Sci-Fi "AI goes crazy and kills us all!!!" trope: what if AI should replace us?

Some time in the early 00s—I don't remember when it was, but it was definitely when Pluto was still a planet—I found a webcomic called Nine Planets Without Intelligent Life, in which humanity had died out and AI had risen up to take its place. But, it wasn't because the AIs had murdered them, it was because humanity had died out on its own, and its robot servants had simply kept on going in their absence. Now, granted, the method by which humanity died out in the comic was very silly—they'd discovered holodeck technology and subsequently lost the interest for actual reproduction—but between climate change and the ever-present threat of someone setting off World War III over North Korea or something similar, perhaps it's just the case that we are living in the twilight of our species. Perhaps we should have some kind of being to replace us ready to go, that could live in a radioactive and/or climate-altered environment, so that when the inevitable eventually happened, our culture and legacy would have a way to live on.

Plus, of course, androids could be adapted to live on planets like Mars that don't have oxygenated atmospheres, which let's face it, is just cool.

Discuss!