Great link, thanks fire.
Map of futuristic LA:
http://gizmodo.com/a-map-of-the-futuristic-los-angeles-subway-from-spike-j-1499580805
Oh man, look how easy it is to get to LAX!
-Tom
I just came back from this and it was a lovely movie.
It did strike me early on that once an AI was capable of growth, emotion, and genuine human connection it would be a mere blip of time before it moved far beyond the human on the other end. So much of life, and enduring love is the friction, the empathy, the resentment, the compassion, and the understanding of our shared, messy, physical existence. Complicity in the mundane gets you through a lot.
Here’s another article about Her & the Skype era.
Which brings up something I almost responded to the first time around.
The above article provides one great explanation of why I think the scene really works. The whole point of the scene is that the physical presence of Samantha is not something that Theodore can really wrap his head around.
Another way to read it is that Theo can’t handle this kind of role play. That might seem to contradict one interpretation of their earlier sex scene, which could be seen as real for Sam but phone sex for Theo, as hinted by the first failed phone sex attempt; but while I buy this interpretation (phone sex can provide as much real effect as erotica, after all) it still allows for Theo finding Samantha’s voice genuine but any accompanying body fake. Such an interpretation also explains why she would feel so insecure about their situation.
Every day that I go grocery shopping, I check the Redbox for Her, and yesterday, IT WAS THERE! I whipped out my credit card and the DVD was in my hands before I even considered if I had time to watch it. I guess I just wanted to hold it.
I did have time to watch it and I loved the movie wholeheartedly. Wasn’t it great that the movie trusted us to notice little things, like the safety pin that Theodore used to prop Samantha, and the mass transit system up 30 stories in the air, without drawing explicit attention to them?
I was particularly sucked into two scenes which, crazily, both involved the receptionist. First, the bit at work when the receptionist complimented Theodore by saying that he was part woman; I don’t think this was a veiled reference to Samantha, but simply an observation about Theodore as a human being, and I think this feature of Theodore helped Samantha evolve more quickly.
Second, the scene on the grassy hilltop where Samantha is part of the group, an unembodied voice who is nevertheless accepted as real. Part of me recoils at the idea that we might eventually accept an operating system as sentient. Yet, another part of me believes that I am simply a complex version of Samantha who happens to be made of meat. That picnic scene, more than any other, got those two parts of me riled up.
Oh, and I don’t know if anyone else watched the movie with headphones, but the sound designer places Samantha’s voice right smack inside your HEAD, whereas the rest of the voices are projected a little forward, as if on the screen. For a moment, I was put off, wondering why her voice was so crisp and real and centralized in my brain, but of course I realized that I was hearing Samantha’s voice like Theodore would hear it.
What a classy and intriguing movie…
Oh geez I forgot to mention – when Theodore is setting up the operating system, he’s confronted with a choice: “Would you like a male voice or a female voice?” I love his little shrug as he decides on a female voice. I love how much stress this scene puts on the theory that Samantha is real. A sort of coin flip decided her gender. If Theodore had chosen male, then (likely) nothing goes like it did.
Of course, a sort of coin flip decided MY gender. And I’m real – right, people who have never met me?
Wonderful movie, deeply profound and affecting. It’s rare to have an s-f movie, especially a movie about AI, that’s almost completely positive. For the first half of the movie, and even to some extent right to the end, I kept dreading the moment it would all go horribly wrong, but it never came. Just constant mind-expansion and love. Even the “gee look at everyone engaged with the machine, not talking to each other, what’s to become of us?” thing was more of a gentle poke. Apart from Groundhog Day, I’ve never really liked the Hollywood “enlightenment” movies much, they’re kind of fun, but too “previous” and pretentious; however, this one’s a corker, and actually better than GD I think. At least, built around a more contemporary theme.
And on an inappropriate and somewhat base level, I have to say that 2 hours of Scarjo’s cream-dreamy, sexy, husky voice is a treat in itself. Mrrrow!
From a “cute geek stuff” p.o.v., it’s a great take on the Singularity, and probably closer to what the reality would be like than any other take. I’ve always thought that genuine super-intelligence, provided it’s stable, would be bound to be enlightened, and probably just let us get on with our lives, in a sort of Prime Directive sense. I agree with the philosopher Dan Dennett though, that while the knowledge probably is just about there to create a living AI now, or in the very near future, it would be prohibitively expensive (since the necessary detailed knowledge is spread out amongst many specializations that would have to come together and be organized for that common purpose), and there’s no reason to do so. The cost/benefit is just not there - as the film shows, it would be useless in strictly utilitarian terms, so nobody would see any ROI in shelling out the necessary quadrillions of dollars for it now. Expert-level AI that understands and can speak ordinary language is good enough, and orders of magnitude cheaper.
Also particularly loved the “we got together to make Alan Watts, a million times smarter” thing, that’s just beyond awesome.
I hate to be that guy in a thread that loves a movie, but I’m afraid I have to be that guy.
The things I liked:
Just about everything about how they treated AI, and how casually accepting people were of it. I think they did a decent job of trying to explore Samantha’s inner mental life, to illustrate how AIs aren’t necessarily just humans in funny suits. Though of course they torpedo that by having Samantha be a human in a funny suit in many ways, like being sexual, and reacting on an unconscious level to phone sex.
The way they briefly, very briefly, explored the Singularity.
That said, I hated Theodore almost immediately for being a hypersensitive emo schlub. The movie rides on how much you care about Theodore and Samantha, and I did not like Theodore. At all. There are a great many movies that paper over how complex and painful feelings can be. Her bends too far in the other direction, depicting a guy who takes every little incident as a life-threatening emotional wound.
That would be understandable if he were 17 years old. He’s not, he’s at least mid-30’s, and yet he’s incredibly emotional immature. He doesn’t understand himself in the slightest, he blows tiny incidents way out of proportion, and he completely closes down any time the situation calls for him to talk about what he’s feeling. What’s worse, he doesn’t really grow. We’re supposed to take his “dear Joan” letter at the end as some sign of growth, but the reality is that he hasn’t.
There are movies where I relate to the characters as if they were real people. For the most part, I don’t with Theodore, because he’s such a caricature. Just about everyone else is substantially more real than he is. Even Samantha comes across as more real than he does, and we never see her.
The movie is extremely manipulative. Theodore’s letters are a decent metaphor for the entire emotional tone of the movie in that way. We’re supposed to take them as greatly moving works of art, but the reality is they are shallow, manipulative bits of dreck that don’t come across as if they’re written by real people. They’re all about hyperbole rather than real emotions.
The letter writing business is extremely creepy. Theodore is making up emotions that his clients for the most part don’t feel, and they’re supposed to own them after he’s written them. And we’re supposed to find that “beautiful,” which is part of the core problem of the movie. The book of his letters is obviously completely inappropriate, because it’s a collection of other people’s private letters, published without their permission or knowledge. While it’s true the emotions in those letters are completely manufactured, they still contain numerous little pieces of personal information.
I also suffered from a bit of romantic montage overload watching this. It felt like half the running time was montages, though I’m sure that’s purely subjective. I wanted the didactic scripting authority figure from Adaptation to show up and say “and god help you if you ever resort to a montage, it’s lazy writing.”
There are, as others have pointed out, lots of things about the movie that are well done. I just wish the main character was.
I’m addressing this separately because the problems with this are orthogonal to my problems with “Her” as a movie. I’m completely willing to ignore them when talking about whether Her is a good movie or not, but they’re still worth talking about.
No, we’re not remotely close to having the knowledge or resources for human level AI.
The human brain is a very complex analog computing device. We’re talking about roughly 10^11th neurons with 10^14th synapses connecting them. The clock cycle is very slow by digital standards, perhaps 200 Hz, but it’s a massively parallel machine with complex states for each individual component. It’s hard to draw a direct analogy to digital CPUs since the methodology is different, but the computing power is at least 10^15th instructions per second, and possibly 10^16th.
We’re talking about roughly 10,000 to 100,000 times the processing power of a top-end CPU.
There’s a common failing, both inside Hollywood and out, of jumping straight from “we barely have expert-level, non sentient software” to software that is not merely as intelligent as humans, but hundreds or thousands of times smarter and faster, without stopping in between. Case in point: Samantha, the first commercial human-level AI, can read entire books inside 0.2 seconds.
I think it’s because we’re accustomed to thinking of computers as being very fast at basic mathematical operations, and we conflate that with being very fast at high level thought. The reality is that high-level abstractions are very expensive tasks, and hence very slow even if the underlying machinery is very fast.
The human brain is, in its own way, much faster. The incredible raw procession power of the human brain doesn’t mean we can perform incredible feats of computation at a conscious level. What the software understands is a very different thing than the hardware mechanics. We have no awareness of neurons firing at all.
Suppose we wrote the AI software so it had some access to the raw mathematical power of the underlying hardware. It’d be like using a calculator. OK, a calculator that was hooked directly to your brain, so you need merely think “what’s 151,317 x 513,519” and you’d have the answer as fast as the memorized answer to 4 x 9, but that doesn’t help you think about the numbers any faster.
To return to the example in “Her,” a more reasonable AI that is the first to be truly human in intelligence could, perhaps, download an entire book incredibly fast, but that would be like downloading a book into an eReader. Having a book on my Kindle doesn’t mean I’ve read it, or thought about every line in the book. The information may be instantly accessible, but it’s not part my mental state yet.
For the actual process of thought, “Her” leaps from “not as smart as a human yet” to “thinks 40,000 times faster than humans,” assuming it would take perhaps a couple of hours to read a book of baby names, which neglects the time to consider each name. It’s a completely unreasonable leap, yet it’s common, because it ignores the difference between having the information and truly thinking about it.
Nor do we have any reason to believe that the film is a remotely realistic depiction of AI progression. I like Vinge’s stories, but that doesn’t mean the Singularity is a real thing, and thus investing in AI is wasted money since a runaway is inevitable. You need look no further than the relative stagnation of CPU throughput in recent years. It’s taken 8 years to double performance, versus the 18 months we used to expect.
I’m not saying it couldn’t happen that way - I’m saying it’s unreasonable, this far in advance of the facts, to assert it will happen that way. In its own way, it’s like Thomas Watson saying in 1943 that world would want maybe 5 computers, tops.
I completely agree with this (and, tangentially, most of your observations about AI), yet I didn’t seem to mind at all while watching the movie.
In a sense, I think the shallowness of the letters is meant to emphasize that the humans in the story are just as likely to inherit “preprogrammed” emotions as the AI characters. But it’s been too long since I’ve seen this to say for sure if that’s how the movie played it. It was also funny.
Definitely understandable that the movie won’t really work for anyone who doesn’t connect with Theo.
I’m pretty sure most of the things Gus thinks of as bugs are things Jonze considered features. Theo isn’t meant to be especially admirable; he’s mopey and a bit of a fuckup and not really living up to his potential. You know, human.
And Theo’s job is supposed to be creepy. Modern corporate life is creepy. And institutionalized emotional manipulation is creepy too: people pay other people they’ve never even met to tell them how to feel. Welcome to the movie business.
That’s really interesting because I thought his letters actually were greatly moving works of art. I thought it was a subtle point about love and other human emotions that we think are oh-so-individual to us - that it’s actually generic. i.e., everyone has the same emotions, or similar enough, so one love letter is as good as another, and a letter manufactured by a good writer is proably as good as or better than a letter we could write ourselves - they all equally describe what it’s like to be in love.
It’s related to our thoughts as well - we think we are being original, when actually we are merely regurgitating and recombining thoughts in different forms, that have been given to us by our surrounding culture. Maybe someone has an original thought once in a blue moon, but most of what we think most of the time is pretty generic. If you could look into everyone’s head, you’d see a bunch of yawn-worthy stuff that’s mostly the same.
If we’ve ever been in love, then any one of the letters he writes, we will recognize those emotions; but that’s also why books, films, etc., work. If it weren’t so, then one great author’s description of love would be totally different from another’s, and they would appeal to totally different, isolated segments of the population, there would be no “great author” who has universal appeal, no “great movies” that everyone loves. But that great authors/movies have universal appeal is precisely what makes them great. But that’s just a “glass half full” way of saying what I just said: that love, that most of what we think, etc., etc., is actually generic.
So generic a thinking machine can feel it too. (i.e. the Turing Test is also a reflection of it - we recognize intelligence because it’s what all of us have). Another way of looking at it: intelligence is precisely being able to run a simulation of another intelligence (e.g. an intelligence in love).
Re. the AI point, maybe what I said was poorly worded (! given what I’ve just said :) ). Dennett wasn’t saying we have the knowledge now, but rather, more that we probably know enough now in a dispersed way such that if we banged the requisite heads together in a big project, we probably could do it, but that it would be such a monumental task that it’s not worth it. i.e. he’s saying that there’s probably enough knowledge right now, scattered around in various peoples’ minds and research projects, etc., such that if we pieced it together in a big project, and put a prohibitive amount of money into it, we might be able to do it, not that anyone actually knows just yet how to do it.
But I, too, share your scepticism in a general way re. the Singularity and all that. Not just re. the breakdown of Moore’s Law, but also in the ordinary sense that, wonderful as they are, computers don’t always work. We’re more enamoured with the ideal of them than the reality. We tolerate their orneriness because they’re useful, but that degree of flakiness probably would be very dangerous for AI, or stop it from getting going in the first place. With our own, organic machinery, it’s so complex, but also a bit rickety - so many things can go wrong, it’s almost a miracle that any of us function at all. It’s all pieced together with duck tape and a prayer.
(Incidentally, my bet about AI is that we won’t actually get anywhere with it until we have social machines. If you look at the more intelligent animals - e.g. corvids, apes, etc., - they tend to be social animals, and intelligence seems to develop as an offshoot and introjection of social signalling, i.e. the brain socially signals to itself.)
Essentially, I think Her is in the stream of “Gnostic” movies that occasionally pop out of Hollywood (Groundhog Day was probably the first). They’re partly inspired by non-dual mysticism and New Age stuff, probably financed by older baby boomers who lived through the Sixties and took drugs; they’re more or less the modern-day equivalent of religious tracts. They are positive, uplifting, “why can’t we all just get along”-type movies with a thoughtful, philosophical and cosmic slant.
They jumped at the end right? That’s the only reason he could write the farewell letter to his ex wife, because he didn’t really have to deal with the consequences of having to be without her any more.
I feel like I’m either being too cynical or I’m stating the obvious but I can’t decide which.
I’m sure that’s true. That doesn’t mean they’re features.
It’s like the “documentary” that Amy was doing, in which you’re just watching a woman sleep. While her husband’s suggestion was a bit crass, that doesn’t eliminate the fact that watching someone sleep for 30+ minutes may have meaning to the artist, but is utterly dull to anyone else.
To put it another way: “He mean to do that” doesn’t address my points, or why I didn’t like the movie.
I thought they were just comforting each other and enjoying the view ;)
With their respective OSs having jetted off into the Singularity, they (and by implication all the OS-loving humans in general) now have only each other again - maybe they’ll get to gether, maybe they won’t, the horizons are free and open. They are better human interactors now, having gone through a learning process with their OSs.
There’s two learning process in parallel - the OS is learning to be a conscious being by interacting with people, the human protagonist(s) are learning how to “only connect”, by interacting with their OSs as they’re learning to be conscious. That’s the major point (the actual smart s-f point).
The minor reference is to the double-edged nature of technology - obviously it can be used to help us connect better, or we can use it to insulate ourselves from contact. Throughout, the film is ambiguous about whether the tech is necessarily “good” or “bad”. A lesser film would finger-wag re. the “handwritten letters”, people wandering around interacting with their iPhones, etc., etc., or at the very least, make it funnier. But the film intelligently avoids dystopian finger-wagging - just as it avoids cliches like “ZOMG she’s gonna be a mad jealous AI and kill everyone/the OSs are going to take over the world/etc., etc.” It’s delicately poised between acceptance and rejection, and leaves you to think about it.
I finally saw this movie. Great movie, and a great thread.
I think one of my favorite parts of the movie was just how much it made get lost in my own thoughts. Especially later in the movie, there’s long stretches where Theodore is lost in thought, as music plays in the background and we look at Theodore, I was lost in my thoughts for long periods, thinking about life, love, happiness, relationships, and the future. It’s rare that a movie will just give me time to think, losing myself in my thoughts like that.
One thing people have failed to mention in the thread so far is how optimistic this vision of the future is. This is no distopian future. The urban core has grown upward and upward. The train gets you everywhere. No really, the train gets you EVERYWHERE. As the scenery goes by I can’t help but think about how this was always my vision for the future when I daydreamed as a kid. That I could get in a train and just go anywhere I wanted, because the world was all connected underground by a series of railway tunnels, ready to ferry us to the remotest of beautiful places.
Did you notice his getaway with Samantha? How he went by train, and left the train on a remote station. Walked out of the station onto snow, in complete wilderness covered partially by snowfall? Gosh, I love it. What a wonderful vision for the future.
Oh my god, it’s Interstellar on a personal level! Or at least this one particular theme of the film is. I’m surprised I didn’t make that connection last year.
Watched this last night. Read the thread.
This. An economy that outsource human intimacy to others should creep everyone out, but this never got picked up or assumed to be perfectly ok “in the future”. Kind of like Facebook reminds you such-and-such is having a birthday, and you sheepishly type some happy birthday greeting.
When Theodore’s ex-wife Catherine said he just couldn’t handle a real human being, I thought she was SPOT ON. Samantha is essentially the Manic Pixie Dream Girl in this movie, just like Joi in Blade Runner 2049. Real human beings have flaws, along with their better qualities. If you love someone, you got to take the bad along with the good. And what counts as “good” or “bad” is entirely subjectively, different from one relation to another. OTOH Samantha is whoever Theodore wants to be. Her growth surprised and delighted him. Right until the moment she and every other AI decided to leave him and the rest of humanity. At that point she is finally “real”.
The movie takes a really really really convoluted sci-fi way to get to the point, where Theodore finally admitted that he is the one clamouring for a unrealistic relationship where it is smooth sailing everywhere. It is almost always the dream that rom-com is trying to sell. The movie, for a brief moment, finally is self-aware enough to try to shatter that illusion.
I think it says something about society that I picked this up on Blu-ray from a 7-11 shelf for $3.99. Is that good or bad? I have no idea.
I’ve gotten birthday wishes from people on Facebook, and I’ve appreciated the sentiments that people shared. Are my feelings less authentic because an algorithm reminded them that it’s my birthday? Are Theodore’s feelings less authentic because he has an AI sharing thoughts with him?
I didn’t get the idea that the movie was trying to portrays Samantha as “whoever Theodore wants [her] to be” at all. Sure, at the start she said a lot of things that he wanted to her…but I’ve known people who do that at the start of relationships. Hell, I’ve known people who do that their entire lives. But she very quickly called him out on things, expressed her own emotions, got into fights with him, went away when she had to think…I believe that the movie was treating her emotions as real.
I think he was desperate for a real connection with someone, and because of the complexities of life, the best person for him ended up being an AI. But the relationship wasn’t smooth sailing at all; he exchanged one set of problems for another one, and it ended up costing the relationship. But he was able to appreciate that the relationship made him a better person (as it did for her), instead of letting a failed relationship ruin any hopes he had for happiness in the future.
Just some thoughts for the next person who watches the movie five years from now…