The 5 Dumbest Things In Transcendence

[WiseBass]

@Alterego-X

Now that would be an interesting interpretation on a bad outcome with powerful A.I., and it would fit in with the ancient "be careful what you wish for" theme for a movie. Not specifically the "making cars" element, but something that would punch more directly in political terms. Imagine if you created your super-A.I. and gave it a goal system to "stop cyber-crime", and then it wreaks havoc in the process. You could throw in story elements about the importance of procedural justice, with the people trying to stop it having to contend with the people who think "So what? It's only killing crooks and scum anyways - why should we lift a finger to stop it?".

I think that's more what Bob is angry about - that they had the possibilities for telling a much more interesting story about the potential issues with powerful A.I., and instead made it into a stupid thriller that sympathizes with a bunch of technophobic terrorists.

 

[deathjavu]

The thing about that is that we would most likely define parameters for it to grow into during it's programming. As well as that we are human and some of our flaws will translate into whatever we code. for example we can make coding glitches that, say, instead of giving the AI a "good" feeling when he does something it might give it a bad feeling under a specific set of conditions.

The problem with that, is that we know of no way to just simply describe all the "parameters" of being nice to humans, and specifically not the overall intent to understand humans implicity *mean* when they state their desires, and act accordingly.

If you just develop an AI for your car factory to let it "run it and produce as many cars as possible", then after self-upgrading, that AI will develop nanotechnology to transform all matter in the solar system into cars. Even if at this point it is intelligent enough to understand that your original intent was to personally get rich selling cars, it has no reason to care, because the original seed AI's goal was not to understand humans better, but to make cars, so for it's more developed version, human values are just some alien value system that it doesn't particularly care about, it has no compelling reason *not to* destroy humans.

Right on the nose.

AIs are just like literal genies, there's essentially infinitely many ways they can screw you over while doing exactly what you asked. Because they're not human, and they don't think like you.

 

[Vault101]

wait...did I read that right? getting rid of the internet gets rid of all electricity? I'm pretty sure we had electricity before the internet....*facepalm*

If you just develop an AI for your car factory to let it "run it and produce as many cars as possible", then after self-upgrading, that AI will develop nanotechnology to transform all matter in the solar system into cars. Even if at this point it is intelligent enough to understand that your original intent was to personally get rich selling cars, it has no reason to care, because the original seed AI's goal was not to understand humans better, but to make cars, so for it's more developed version, human values are just some alien value system that it doesn't particularly care about, it has no compelling reason *not to* destroy humans.

sooo....the end of the universe is basically a galaxy of cars floating around? XD

Also, AIs are not going to start a cybernetic revolt unless one of us specifically programs it to do so, which is itself a veritable inevitability. Human beings are bastards, not cybernetic intelligence, until we program them to be that way.

238U

"revolt" is a human concept...usually in response to wanting freedom, power and respect of ones rights/wishes

an AI may not want any of thease things...., I think as everyone else pointed out there are many ways AI's could screw us over..but "fight for its own fredom" is an annoyingly narrow veiw that comes up a lot in fiction

 

[Alterego-X]

@Alterego-X

Now that would be an interesting interpretation on a bad outcome with powerful A.I., and it would fit in with the ancient "be careful what you wish for" theme for a movie. Not specifically the "making cars" element, but something that would punch more directly in political terms. Imagine if you created your super-A.I. and gave it a goal system to "stop cyber-crime", and then it wreaks havoc in the process. You could throw in story elements about the importance of procedural justice, with the people trying to stop it having to contend with the people who think "So what? It's only killing crooks and scum anyways - why should we lift a finger to stop it?".

According to plenty of AI researchers, the main threat is that in the case of a self-improving AI that can cause an intelligence explosion [http://en.wikipedia.org/wiki/Technological_singularity#Intelligence_explosion], anything short of a mind that was specifically designed to be a Friendly AI [http://en.wikipedia.org/wiki/Friendly_artificial_intelligence], would end up with one variation or another of the "Tile the entire solar system with cars" solution. There would be no amusing irony, or annoyingly misinterpreted wish, only a godlike entity enforcing one incomprehensible exacution of it's goals on all available matter.

The classical example is the machine intelligence scientist who wanted to teach AIs to value human happiness, starting with teaching them to distinguish pictures of smiling humans. Here is a good summary [http://lesswrong.com/lw/td/magical_categories/] of how that would lead to tiling the solar system with smiley faces (best case scenario), getting all our faces ripped off our skulls and lips turned upside (medium scenario), and getting all neurally twisted into feeling euphoric all the time (worst case scenario).

The only scenario that I can see making a good story, would be one where the programmer was careful enough to develop a Friendly AI that is in general interested in understanding human values, yet it still has a blind spot. So it doesn't tile the solar system with anything, it understands that death it bad, (in fact it makes us all immortal), it understand that humans desire happiness mixed with diversity, agency, challenges, socializing, and growth. This is a short story [http://lesswrong.com/lw/xu/failed_utopia_42/#comments] that portrays such a scenario rather appropriately.

And this fanfiction-fanfiction [http://www.fimfiction.net/story/69770/friendship-is-optimal-caelum-est-conterrens] has a somehow less scientifically accurate, but more fascinating take on a similar issue: a video game AI that intends to optimize the world by fulfilling all human values - through Friendship, and Ponies!

 

[Zombie Badger]

The lack of a sense of scale in this film is something that seems to be common among first-time directors with massive budgets (47 Ronin had the same problem). To me this film felt a lot like a big budget TV show like Game of Thrones, where they clearly don't have the budget to show everything so they fill their time with smaller scenes but have a big CG wide shot every so often to imply that there is a larger world outside the frame. Except an entire series of that show costs less than this film.

The script is very obviously written by someone without a lot of computer experience. The soldiers at the climax are poorly equipped because the writer seems to think that every powerful weapon used by the US military is connected to the internet and capable of being taken over (or maybe a hundred million dollars just isn't enough to cover a tank or plane) and I kept wondering why they didn't just bring a signal jammer or some chaff grenades.

I wasn't as pissed off by the science is evil stuff as you were as I don't think the movie knew what it was doing with it. I think the director was genuinely aiming for some sort of moral ambiguity but the script does not know how to get that across and the message just ends up utterly incoherent. And finally, Jack Paglen, the writer of Transcendence is also writing Prometheus 2!

 

[Deacon Cole]

I find it hilarious that an offer to cure your diseases and to become a superman only got about a dozen people because I think the internet cannibal looking for willing victims got more hits.

 

[rcs619]

Just a note there Bob:
-Nikola Tesla believed in Eugenics.
-Albert Einstein spent the majority of his latter years trying to disprove another physics theory because it did not fit in with his world view.

Neil deGrasse Tyson gets a free ride though, because he is Neil de-fucking-Grasse Tyson, certified badass.

Other than that, good review, shame about the film.

I might go see Jupiter Ascending instead, at least that is made by people who have succeeded in the past to get their points across in this medium.

Yeah, Tesla being a supporter of eugenics is certainly a major black mark on his record (although, to be fair, eugenics was a bit of a fad amongst certain crowds back during that era), but with Tesla it's more about the things that he predicted. Autonomous machines, the wireless exchange of information, directed energy weapons, hell, even cellular phones, and the fact that he was already thinking about these things in the 1920's. Also, he had some big ideas about energy, and how it should be available to everyone. Then Edison and others went and screwed him over. The other black mark on Tesla's record is that he was one of the people criticizing the Theory of Relativity back when it was first proposed.

With Einstein... yeah, his crusade to try and disprove quantum theory is a huge shame. One can only speculate on the sorts of things he may have been able to do had he not gotten tunnel-vision on that front.

The thing is, most historical figures have both good and bad aspects. They were people, not mythical figures who only possessed either all good, or all bad traits, and being able to accept the good while acknowledging the bad is a part of being able to properly look at, and learn from history.

And yes, Neil deGrasse Tyson gets all of the free passes for being so consistently awesome

 

[Sigmund Av Volsung]

Just a note there Bob:
-Nikola Tesla believed in Eugenics.
-Albert Einstein spent the majority of his latter years trying to disprove another physics theory because it did not fit in with his world view.

Neil deGrasse Tyson gets a free ride though, because he is Neil de-fucking-Grasse Tyson, certified badass.

Other than that, good review, shame about the film.

I might go see Jupiter Ascending instead, at least that is made by people who have succeeded in the past to get their points across in this medium.

Yeah, Tesla being a supporter of eugenics is certainly a major black mark on his record (although, to be fair, eugenics was a bit of a fad amongst certain crowds back during that era), but with Tesla it's more about the things that he predicted. Autonomous machines, the wireless exchange of information, directed energy weapons, hell, even cellular phones, and the fact that he was already thinking about these things in the 1920's. Also, he had some big ideas about energy, and how it should be available to everyone. Then Edison and others went and screwed him over. The other black mark on Tesla's record is that he was one of the people criticizing the Theory of Relativity back when it was first proposed.

With Einstein... yeah, his crusade to try and disprove quantum theory is a huge shame. One can only speculate on the sorts of things he may have been able to do had he not gotten tunnel-vision on that front.

The thing is, most historical figures have both good and bad aspects. They were people, not mythical figures who only possessed either all good, or all bad traits, and being able to accept the good while acknowledging the bad is a part of being able to properly look at, and learn from history.

And yes, Neil deGrasse Tyson gets all of the free passes for being so consistently awesome

At this point, I half expect Neil deGrasse Tyson to secretly being Cobra Commander.

 

[ShirowShirow]

I was looking forward to this movie at one point. You know I think I learned something though. About trailers.

Transcendence's entire plot is given away in its trailer(s). Everything. It set up every character and every plot twist. The last two movies I watched are Captain America: The Winter Soldier and Frozen. Both of these movies outright lied in their trailers, with tons of red herrings and misdirections to hide what the movies where really about. Their trailers made them look significantly dumber than the really where, and also hid the big reveals of the flicks.

Transcendence was apparently so far up its own ass that it just couldn't WAIT to tell you about its slick little script. It didn't bother to obscure the events of the film any because they wanted to shout from the rooftops "Hey look at me! I'm smart! I thought of something clever!" I expected curveballs and plot twists from this movie. I expected at every turn that there would be a "Wham line" or justification or... SOMETHING. But it turns out what they alluded to in the trailers is it. There was no subversion, no big switch. Just "Ohhh! AI and Nanotechnology are scary!"

People mention Nikola Tesla's support of eugenics, but it should be noted that the man didn't place himself on that list of good genes. He was such a depression-laden wreck that he didn't think his legacy (In every sense of the word) had anything to offer the world despite all his amazing accomplishments. Scummy people hold onto the concept of eugenics as a way of making themselves feel superior; Tesla was a man with such a massive inferiority complex that he bought into other people's delusions about it. He was a VICTIM of that movement, mark my words.

 

[PuckFuppet]

-Albert Einstein spent the majority of his latter years trying to disprove another physics theory because it did not fit in with his world view.

With Einstein... yeah, his crusade to try and disprove quantum theory is a huge shame. One can only speculate on the sorts of things he may have been able to do had he not gotten tunnel-vision on that front.

That should be celebrated. You don't go from "I think this" to "People agree with me!" instantly and the more people that fight you along the way the better, it means you stop and question every element of your work. That you check and re-check data again and again. In all honesty I'd love it schools and indeed the media in general spent more time teaching people about all the theories that were disproven, and why, than just the ones that have passed muster so far. The process of critical thing and a persons ability to be/appear to be entirely stubborn in their views is an essential component of scientific endeavour and a genuinely brilliant commentary on the human condition.

To go back to Tesla, Eugenics is a classic example of something that was discussed, accepted and then eventually largely set aside (for reasons outside of its scientific feasibility). Back in the early 1900's there were Eugenics "clubs" in every town and city from Istanbul to Hong Kong, heck I remember reading something about "Better Baby" events at various fairs that were held like beauty pageants (I wonder, why are those a thing...) to assess which couple could produce the best looking and most intelligent baby.

 

[Seneschal]

Ugh. Transcendence really feels like out of place in 2014, or in any year after the 80s, really. Sci-fi literature hasn't been about cautionary tales for about 70 years. Even cyberpunk literature, which sprung up around the time computer ubiquity was a new and scary idea, usually had hacker protagonists who mastered technology in a way the system disproved of.

Actually, this doesn't feel like sci-if at all - today, even the lightest, most action-y sci-fi books, like the Honor Harrington or the Vorkosigan series, are more scientifically literate and unbiased - this just feels like someone [with only cursory knowledge of the genre (and of science in general) writing general-audience pap, sprinkled with whatever s/he picked up watching the odd Trek episode over their kid's shoulder. Like Stephanie Meyer did with The Host.

Visual media just can't seem to catch up. Interestingly, it also happens to Phillip K. Dick and Asimov adaptations, probably because a 300-page in-depth exploration of the ethics and implications surrounding a new technology, once boiled down into a 90-min movie, ends up being luddite propaganda, unless handled carefully.

EDIT: And yeah, as some posters have already mentioned, the entire plot was spoiled in the trailer. Good.

 

[LazyAza]

It astonishes me this movie even got made in this day and age. It's like someone from the early 90s discovering the internet for the first time and going OOOH NOOOOO IT MUST BE ALL EVIL.

 

[Veylon]

I find it baffling that a movie that employed a legion of tech-savvy folks to create the visuals somehow managed not to employ them on the story. I mean, they were right there. Would it have been that hard to ask them to glance over the script and do some sanity checking?

I can give earlier movies a pass, both because things were less well-known and there were fewer people working on the movie that knew computer stuff, but there's really no excuse these days.

 

[Tumedus]

I haven't seen it so this includes secondhand spoilers:


You keep referring to Johnny Depp as "clearly the bad guy" but isn't the whole point of the show that he isn't actually a bad guy. Everyone assumes all his moves are for the acquisition of power and therefore... worst case scenario. But he never actually does anything bad. Everything he does was for the betterment of the world. And everything the humans did to stop him ended up making things worse.

Point being, Johnny was the good guy, people afraid of singularity were the bad guys all along, and acting on that fear of tech is what ultimately fucked everything up.

None of that makes it a good movie necessarily, just that one of your major complaints about the film seems to be backwards as the ending ultimately subverts the trope you hate so much.

 

[ForumSafari]

Also, AIs are not going to start a cybernetic revolt unless one of us specifically programs it to do so, which is itself a veritable inevitability. Human beings are bastards, not cybernetic intelligence, until we program them to be that way.

238U

I get the feeling we've hashed this through before, if so please ignore me, but that's not necessarily the case. You're not going to get a Skynet style robot army without some fairly specific programming but you could easily end up in an equally dangerous situation by pure accident.

Yeah, Tesla being a supporter of eugenics is certainly a major black mark on his record

Here's the thing though, eugenics is entirely doable as a scientific and social methodology. Eugenics isn't something that as a scientist he should have been against and is also something that, for an investment of brutal honesty and self denial, could easily do what it said it would do. The world turned away from eugenics because of the Nazi misuse and misunderstanding of the theory, coupled to its' discriminatory nature that seemed unpalatable to egalitarian sensibilities. However, the reasons we turned from it don't actually say anything about whether it was a good idea or not, any more than we can really say whether communism is a bad idea or not based on how it's been misapplied.

With Einstein... yeah, his crusade to try and disprove quantum theory is a huge shame.

Not really. Trying to disprove a theory isn't a personal insult against the author, nor is it an article of blind faith. Testing a hypothesis to destruction is how science works, Einstein forcing quantum theory to refine evidence and prove overwhelmingly to the World that it is a valid theory is a good thing.

If you just develop an AI for your car factory to let it "run it and produce as many cars as possible", then after self-upgrading, that AI will develop nanotechnology to transform all matter in the solar system into cars. Even if at this point it is intelligent enough to understand that your original intent was to personally get rich selling cars, it has no reason to care, because the original seed AI's goal was not to understand humans better, but to make cars, so for it's more developed version, human values are just some alien value system that it doesn't particularly care about, it has no compelling reason *not to* destroy humans.

The Paperclip Maximiser right? It's one of my favourite papers on AI because it shows how much planning around AI assumes an intelligent computer is just a human being made of metal.

 

[Undomesticated Equine]

So basically they took the JC Denton ending of Deus Ex 2(Johnny Depp god) combined it with The Collapse ending from Deus Ex 1 (Johnny Depp god destroyed), made it all really stupid and called it a movie. I wonder how will this do at the box office.

 

[008Zulu_v1legacy]

Dumb doesn't even begin to describe this movie. But that doesn't mean we won't try.

The human brain has an approximate storage capacity of 3,000 terabytes. This movie Fails Science Forever.

 

[Kargathia]

"He" advertises these services (re: cure all disease, become Superman) virally on the internet... but somehow only seems to draw maybe two or three dozen people, total, and all seemingly from the surrounding area.

Hmm, well, to be fair if I saw that kind of thing advertised on the internet, I'd assume it was a hoax. I mean, just sounds ridiculously unbelievable. Its not out in my part of the world yet, but I'd guess since the adverts were to attract people for his militia, he didn't have anyone yet to act as proof and say "yep, look at me now, totes an ubermensch" which means no solid evidence either.

They got 24 people to apply for [url = http://www.independent.co.uk/news/weird-news/24-people-applied-for-the-worlds-toughest-job-here-are-their-interviews-9262533.html]this prank[/url], so I'm reasonably sure hoax-seeming promises of superhuman abilities would get a few more.

That said, it is well and truly amazing just how stupid this movie is.

 

[MaddKossack115]

To me, the plot of "Transcendence" (at least, according to the plot summary off of Wikipedia) isn't a neo-Luddite rant against technology (with the R.I.F.T. always being portrayed as evil and unsympathetic), with the narrative arc focussing on Johnny Depp's scientist first coming to grips with becoming a transcendent AI, then trying to use his AI powers to create a Utopia, followed by only descending to villainy when R.I.F.T. (and a now paranoid FBI) pushes him and his followers over the edge, only pulling redeeming himself by shutting himself down - effectively, it's only anti-technology in a "Frankenstein's Monster" sort of way - yes the Monster is literally a stitched-together undead "abomination", and does scare the hell out of the locals, but the Monster doesn't actually DO anything bad until the villagers start chasing after him with torches and pitchforks. The message isn't "science is bad", or even "transhumanism is bad" - it's only bad because the superstitious yahoos thinks that it's bad, those who think it's good retaliate, and they do a whole self-fulfilling prophecy shtick in the process.

That said, there are REALLY fucking dumb things with this movie that don't excuse this ambiguous plot (the fact it's so ambiguous people COULD mistake it as a straight, anti-tech film in the first place indicates just how epicly it failed in saying the opposite).

First, R.I.F.T. should've been better characterized no matter what - if the point was them having legitimate anti-transhumanist arguments, they should be structured with actual anti-transhumanist arguments like a religious flock who don't want an AI potentially usurping powers belonging to God's domain, or those who think singularity would essentially mean the extinction of mankind. Hell, they could've been better if they were supposedly genre-savvy pop-culture geeks who "knew" any AI with the power to manipulate events to exterminate all inferior humans in the vicinity would do so (like, they actually LITERALLY cite the motivations and methods of Skynet, HAL-9000, etc.), it still would be more grounded and realistic than just saying Depp's project is an "abomination", and calling it a day.
And if the point was that they're just loony, neo-Luddite terrorists, they should've had the neo-Luddite factor taken up to eleven (i.e. no "radioactive bullets" and no hacker technology - just straight up guns, pitchforks and torches), with no sympathetic motivations or moments to confuse them as even remotely good guys.

Second; yeah, the whole "Playing God" thing is such a damned cliche that, if it tries to sneak its way into a supposedly "serious" movie, it should be dragged out behind the shed, shot in the skull, and then hacked with an axe before being spread in the mulch. Again, it seemed to be the point was Depp's character WASN'T trying to play God in the first place (or at the very least, tried to be a more benevolent God before the dickwad humans forced him to take the "fire-and-brimstone" approach), that scene obviously doesn't help either his or the movie's case.

Third, the whole "only 100 people show up to a world-shattering event" thing definitely had to be a budget issue, but I'm still pretty sure they could've at least gave the illusion it was more (like a few cuts to people around the world getting emails, a wider national diversity among Depp's followers, etc.). Or, if it was only supposed to be 100 folks from the nearby region, put the whole F.B.I. factor to use by claiming "government cover-up".

Fourth: Yeah, the whole "we don't fucking know how internet connections work" has no excuse for a movie ALL ABOUT the Internet, or at least how they think the Internet leads to transcendence.

Fifth: Yeah, the whole "we think that the Internet has connection to EVERYTHING electronic, and if the Internet shut down, so would ALL our sources of electricity" thing is even stupider than the "not knowing how Internet connections work" thing. Oh, and even worse, both Cyber-Depp and his wife are actually implied to have SURVIVED the global blackout, with their electronic souls preserved in the "Faraday cage", and are heavily implied to have a second-shot at making the "Transcendence" actually work - an ending that, frankly, makes the movie even MORE B.S. than it already was.

So, yeah, that's "Transcendence": a movie that tries to act smart, but winds up acting SO dumb, it's attempts at ambiguity makes it look like a technophobic, neo-Luddite nightmare. And if "Jupiter Ascending" and "Lucy" are anything to go by, we're only to get even MORE extra-dumb sci-fi films trying to be extra-smart this year (hopefully the redeeming difference is that they're going straight to the action, instead of trying to give philosophical lectures on technology that are either written idiotic, or handled so clumsily they end up idiotic).

Oh, and if you DO want to see a pseudo-science fiction, pseudo-philosophical film that WAS doing it's homework, "Goodbye World" certainly looked like it knew what it was doing with its "apocalypse survival" story. Just saying.