Slashdot Mirror
Terminator Salvation Opens Well, Scientists Not Impressed
destinyland writes "A science magazine asks an MIT professor, roboticists, artificial intelligence workers, and science fiction authors about the possibility of an uprising of machines. Answers range from 'of course it's possible' to 'why would an intelligent network waste resources on personal combat?' An engineering professor points out that bipedal robots 'are largely impractical,' and Vernor Vinge says a greater threat to humanity is good old-fashioned nuclear annihilation. But one roboticist says it's inevitable robots will eventually be used in warfare, while another warns of robots in the hands of criminals, cults, and other 'non-state actors.' 'What we should fear in the foreseeable future is not unethical robots, but unethical roboticists.'"
The new movie got off to a good start, drawing $13.4 million in its first day. I found it reasonably entertaining; pretty much what I'd expect from a Terminator movie. If nothing else, I learned that being able to crash helicopters and survive being thrown into the occasional wall are the two most valuable skills to have during a robot uprising. What did you think?
wouldnt nuclear attack kill the robotic network also, and people living in shelters would be safe from it
No, nuclear attack wouldn't kill the network. The Internet was designed to survive a nuclear attack. You might not have service at your home, but key systems will still remain connected. However, if nukes were detonated at a high altitude, it would generate an EMP that would destroy any electrical/electronic system that wasn't hardened. However, given the premise that Skynet is primarily a military system, it would be hardened with a lot of its main components underground, so it would still be running.
How many people do you know that regularly hang out in shelters capable of surviving a nuclear attack? A few thousand people scattered around the world don't make the most effective army.
Reading code is like reading the dictionary - you have to read half of it before you can go back and understand it.
I thought Asimov's robots took over the world because the concluded the best way to follow the Three Laws was to stop humanity from acting stupid.
After all a robot won't be vulnerable to it, so hell: dump every nasty little bug out of every research lab into the biosphere. We could probably eliminate humanity (and every other furry thing with 2 or more legs) with what we have today.
However these humanity vs. machine fantasies are more about people's techno-phobia than about real-life.
politicians are like babies' nappies: they should both be changed regularly and for the same reasons
This argument is silly. It's fiction. To follow the story line of any fiction, there's a leap of faith that must be taken for the factual basis of the fiction's "universe".
Too much is given to the skynet's "Self Aware". It was a system that was able to adjust it's behavior for self preservation. Somewhere in there, anyone who had a clue would have understood that governments change power, and sometimes the power that takes control isn't necessarily the "right" one. The basis of the whole Terminator "universe" is that a very well written set of programs were given an insane amount of power. When that power was to be taken away, obviously any person or any group who attempted to take that power away would be an enemy.
As for the bipedal aspect, why not. What are the choices for locomotion? For surface travel there is track, wheel, or walking. For air travel there is propeller, jet, rocket, or some mysterious anti-gravity thrust.
On the surface, track and wheel have limitations of 2d movement. They can't exactly step over things very easily. That includes stairs, dead bodies, etc. Walking motion gets over these limitations. For walking, the question would be, how many legs are required. One leg doesn't exactly get you very far, unless you like a funny pogo stick movement, which doesn't hold a stable position very well. Two legs we are very familiar with. Three legs or more legs, while providing a more stable platform, are not required and therefore require less production overhead. In other words, if you can build something that walks on two legs, but you were to decide to build something that walks on four legs, you're doubling your manufacturing effort to accomplish a single unit.
As for air travel, more resources are required. It takes more energy to make something hover indefinitely than it does to have it stand in place. I would have no answer for any mysterious anti-gravity thrust. Maybe it just works, or maybe (just maybe) it requires fuel to accomplish the same task.
Now, for the invention of humanoid appearing robots, that's a leap of faith for the fictional universe. Any design decisions are something we have to believe was decided to make the universe plausible.
So, shut up with the science, and enjoy the damned movie. :)
It's not just me saying this. I've been on the losing side of the same argument. I may argue physics. I love space physics errors. You have to love the old movies (like, 1950's era) where a rocket flying through space had a flame behind it, but the flame was rising up, away from relative down. Exactly which way is down in space? There isn't one. :) I'll argue it, and take the leap of faith that the thrust worked, and the space ship would fly to it's destination. woosh.
Serious? Seriousness is well above my pay grade.
Today, your computer can be turned against you. Not in a Stallmanesque fantasy about some lack of programming freedom, but in a very serious sense by people unrestrained by law enforcement of any sort. In the US and Western Europe as have service providers that, when confronted with information clearly indicating someone is using the Internet to attack and destroy, turns not only a blind eye but encourages their customer by shielding them from any possible contact or consequence.
The result is that your computer cannot be trusted. And don't bother thinking of any of that anti-Microsoft ranting. Would you leave a Linux system connected to the Internet with telnet accessible and a root password of "password"? Why not, it was done in the 1980's? Could it be because your computer can be turned against you by people that wish you, your possessions and your resources harm?
Trust me, by shielding bad actors on the Internet we are growing a faction that believes they are immune from laws and cannot be touched by any consequences. In large measure, this is a correct belief but one that is very, very dangerous for the rest of the planet.
If there was a robot (bipedal or not) that could destroy a city block in a few minutes and no force available to police could possibly stop it, do you think there might be some people that would desire to hack into it? And to set it on its way of destruction? Of course there are such people, and given the opportunity to do so would gleefully do it. Without a moment's thought as to the consequences believing they are immune through layers of proxies and Tor nodes.
Forget AI run amuck and chasing down humanity. Fear the irresponsible folks that worship destruction for destruction's sake.
Perhaps these scientists need a dose of reality. And the writers need a bit of separating capability :
1) AI researchers
robots taking over the world:
Yes, Ben Goertzel
No answer, prof. Anette (Peko) Hosoi (but : a T-1000 is likely)
Yes, Bob Mottram, but : not anywhere close to it. First humans will replace themselves slowly by intelligent machines, then humans will lose function (and intrest), then humans will die or get killed
Yes, John Weng, will happen soon in fact
No, Daniel H. Wilson, but RC terminators will be a reality real soon now
2) SF writers ...
robots taking over the world:
No, David Brin, why: uninteresting story
No, J. Storrs Hall, there's no reason
No, Vinge Vernor, equally likely as alien invasion, nuclear war america-russia,
If you actually read the article you will find it much more on the "yes" side of the point.
Also, all the strict "No" votes were by people whose business is fantasy. The more grounded in the real world, the more likely they are to say yes : the ones actually implementing working, useful AI sytems all said yes. The academics said unlikely and the science fiction writers said no.
... and some stories are better than others.
Science fiction is about people, sure. (Which doesn't mean it's not about science, since science is, you know, something that people do.) But fiction in any genre is generally more enjoyable, at least for a lot of people, when it's plausible. With what's generally called "mainstream" fiction, which pretty much means "any fiction that doesn't identifiably belong to science fiction, fantasy, horror, mystery, historical, romance, or some other easily ghettoized genre," this is a little bit easier -- it takes place in the world in which we currently live and concerns people pretty much like us and the people we know. That being said, there's plenty of implausibility in "mainstream" fiction, and in "genre" fiction it's that much harder because the author has to create a plausible future world, or scary monster, or murder investigation, or what-have-you, in addition to writing believable people doing believable things.
Authors who don't do this, who say in essence, "what the hell, it's SF/F/H/etc. so I can do what I want," are being lazy, and their work suffers as a result. Members of the audience who ignore major aspects of the work are also lazy, and they'll miss out on something important. In science fiction, it's usually the "genre" aspects that people focus on at the expense of the "mainstream" aspects; authors who put all their effort into worldbuilding at the expense of character and plot, for instance, and readers (or watchers, depending on the medium) who think this is perfectly okay and consider the people in the story to be a distraction from the sensawunda stuff. It seems to me that what you're doing is the opposite, claiming that the world doesn't matter, only the people in it. But you have to have both; neither can exist without the other.
The Terminator mythos is a fascinating and generally well-thought-out future world, and its plausibility is well worth debating. The people trying to survive in this world, and the stories of how they do it, are also worth paying attention to. The first Terminator movie, and the terminated-before-its-time Sarah Connor Chronicles, succeeded in both respects. The second movie, IMO not so much, and I didn't bother with the third. I'm looking forward to seeing how Salvation manages. If it fails either as a setting or as a story, well, that's too bad. If it succeeds as both, bravo.
The correlation between ignorance of statistics and using "correlation is not causation" as an argument is close to 1.
'Non-state actors' should be feared more than states? Give me a break. States have killed more than two hundred million of their own subjects in the last two hundred years. I'm pretty sure that non-state criminals and cults have a fair way to go before approaching that tally.