Blade Runner 2049

[Eldan]

Has anyone spotted any links yet to Soldier, the "only related by means of a single sentence stealth sequel" to the first Blade Runner?

Never heard of it. Any good?

[Giggling Ghast]

Never heard of it. Any good?

Not really, no.

[Lvl 2 Expert]

Do you mean the movie where a bunch of soldiers bred and trained from birth gets replaced by genetically engineered super asian soldiers, only for the original soldier's leader to beat them at the end?

Yes, that one. The one with the PTSD/didn't have a youth protagonist who's closer to autistic. The one that thought it was so important that the movie was set in the really near future that their government started kidnapping babies from hospitals two years before the movie came out. By now we should have had these guys on the front for four years or so. That one is supposed to take place in the Blade Runner universe, according to the director. The link is that in one scene they use the words Tannhauser Gate (technically two scenes I guess, but the first time it's easy to miss written text).

Never heard of it. Any good?

I like to think so, yes. It's just kind of peculiar. You've got a nearly mute protagonist, constantly confused at the chaos of the real world. You've got a whole bunch of family drama, which is kind of sweet but not what you were expecting going in. You've got a military system that makes no sense in any way, whether it's numbers or the combination of weapon systems or the timeline or the officers or literally anything, and then you got Kurt Russel terminatoring the crap out of other terminators with flamethrowers and knives and guns and rockets and this really weird old timey robot looking fighting style that these completely human soldiers are for some reason trained in. But it feels good as it all comes together.

Or maybe that's just me identifying with the main character. He's off from the standard action hero, and sort of off in my direction. I have the same thing with The Transporter.

[RCgothic]

I'm still not buying that JOI isn't fully sapient. Look how happy she was to experience that rainstorm. That had nothing to do with K. Damn her sub plot is heartbreaking.

[Chen]

I'm still not buying that JOI isn't fully sapient. Look how happy she was to experience that rainstorm. That had nothing to do with K. Damn her sub plot is heartbreaking.

Programming your companion AI to act super happy and grateful after you buy a, presumably expensive, upgrade for them, is a pretty good marketing idea. I don't see how her reaction there is evidence of sapience.

[LordEntrails]

Programming your companion AI to act super happy and grateful after you buy a, presumably expensive, upgrade for them, is a pretty good marketing idea. I don't see how her reaction there is evidence of sapience.

And the evidence that gift buying response was programmed? There isn't any. It's these types of open ended questions that show the power of the movies. They ask many questions, and don't provide the answers. It causes many to think, to consider, to contemplate.

What if...?

[Cikomyr]

You can be programmed and still develop sentience.

Sentience is all about intimacy; thats the moral of the movie. While it may be replicated and recreated, it does not make it any less true. It just makes it.. peculiar.

[gomipile]

I think it's obvious that Joi would pass a Turing test. The fact that some of her responses are based on the preprogrammed responses of a fresh-out-of-the-box Joi is equivalent to a human responding with cliché phrases they've had ingrained in them over the years. What matters in either case is what's observable and testable about their personality now, not how it got that way.

[Dienekes]

I'm still not buying that JOI isn't fully sapient. Look how happy she was to experience that rainstorm. That had nothing to do with K. Damn her sub plot is heartbreaking.

I think it's very specifically meant to be a question without an answer.

Whether she could pass a turing test is ultimately irrelevant. She would pass or she wouldn't based upon what the director says (and I hope he never actually gives a definitive answer). The important point is the questioned raised about how people connect to things.

[Vogie]

I think it's very specifically meant to be a question without an answer.

Precisely. The whole undercurrent of the Blade Runner series are questions such as "What is life?", "What is real?", et cetera.

[Frozen_Feet]

I think it's obvious that Joi would pass a Turing test.

We have chat bots today which can do that. Joi would actually be able to pass many much more sophisticated tests than Turing's.

Regardless of whether you think Joi is sapient, holy crap must that be one complex AI. If we take its ability to predict what K wanted to hear as general predictor of its ability to estimate human needs and desires, it is probably a better psychologist than most real ones. Nevermind its ability to adjust to the environment in almost real time. In that regard, the question of whether Joi is "sentient" is sort of dim. It obviously is sentient in the most trivial sense of being able to observe and adjust to reality, exceptionally well for a hologram.

[LordEntrails]

Precisely. The whole undercurrent of the Blade Runner series are questions such as "What is life?", "What is real?", et cetera.

Or the biggest one of all, "What is human?"

[Tvtyrant]

We have chat bots today which can do that. Joi would actually be able to pass many much more sophisticated tests than Turing's.

Regardless of whether you think Joi is sapient, holy crap must that be one complex AI. If we take its ability to predict what K wanted to hear as general predictor of its ability to estimate human needs and desires, it is probably a better psychologist than most real ones. Nevermind its ability to adjust to the environment in almost real time. In that regard, the question of whether Joi is "sentient" is sort of dim. It obviously is sentient in the most trivial sense of being able to observe and adjust to reality, exceptionally well for a hologram.

One of the weird questions that came up out of Joi was why the holograms aren't given any mobility? He has to carry Joi in his pocket, but they have drones with arms that could easily give Joi a body that the hologram overlays. His drone and car is for whatever reason not sentient, while Joi seems to at least meet the zombie sentience credentials.

The answer is, of course, that their society is supposed to be rather crapsack. The capacity to make fully functional androids that don't resent mankind the way the replicants do is fully within their grasp, but it wouldn't tell the story.

[Frozen_Feet]

The emanator exists precisely for the purpose of giving the hologram mobility, and could conceivably be fitted into a drone. I think the primary reason it isn't is because this is civilian tech, and implied to be crazy expensive luxury item.

Tl;dr: car for your virtual barbie is sold separately.

Also, I'm fairly sure K's car is capable of self-driving, so it has a sophisticated AI as well. And the car's drone is able to follow at least simple verbal commands, so ditto for it. They're just less antropomorphized.

[Tvtyrant]

The emanator exists precisely for the purpose of giving the hologram mobility, and could conceivably be fitted into a drone. I think the primary reason it isn't is because this is civilian tech, and implied to be crazy expensive luxury item.

Tl;dr: car for your virtual barbie is sold separately.

Also, I'm fairly sure K's car is capable of self-driving, so it has a sophisticated AI as well. And the car's drone is able to follow at least simple verbal commands, so ditto for it. They're just less antropomorphized.

It is hard to believe that a microchip and a car costs more than a vat grown replicant, which is my point about the setting being crapsack.

[Frozen_Feet]

It's not at all hard to believe a replicant could cost less than emanator. Heck, in the world today you can buy a human slave with a box of cigarettes if you know where to go.

Biggest reason being, sophisticated circuitry may need copper and rare earth metals. Both of these are in short supply today, and could be entirely used up by 2049. (That's why the kids in the orphanage are going through all that junk.)

Meanwhile, replicants are biological. A replicant's raw materials are probably dirt cheap, could even be literal dirt. The cost is likely entirely in the difficulties of manufacture and huge demand versus supply.

[Tvtyrant]

It's not at all hard to believe a replicant could cost less than emanator. Heck, in the world today you can buy a human slave with a box of cigarettes if you know where to go.

Biggest reason being, sophisticated circuitry may need copper and rare earth metals. Both of these are in short supply today, and could be entirely used up by 2049. (That's why the kids in the orphanage are going through all that junk.)

Meanwhile, replicants are biological. A replicant's raw materials are probably dirt cheap, could even be literal dirt. The cost is likely entirely in the difficulties of manufacture and huge demand versus supply.

They have spaceships and space colonies, asteroid mining would fix all of those issues.

The reason slaves are cheap is you don't pay for their raising and education. You essentially steal the $200,000-$2,000,000 it takes to raise a child (based on location). Societies that raise slaves instead of capture them charge much, much more as a result.

[Frozen_Feet]

Asteroid mining fixes things for the space colonies. Earth is clearly not benefiting from it all that much given sweatshop labour to recycle used circuitry is a thing.

[RCgothic]

Programming your companion AI to act super happy and grateful after you buy a, presumably expensive, upgrade for them, is a pretty good marketing idea. I don't see how her reaction there is evidence of sapience.

Can you prove to an objective observer (say someone not a human with relatable perspective on the subject) that you are self aware and not just a sophisticated biological program running on meatware? What if you had to prove it to an AI?

Sapience is something you have to take at face value. If an entity is capable of independent reasoning, telling you how it feels and responding fluently to the feelings of those it interacts with, it's pretty irrelevant whether it's truly self aware or just faking it very convincingly.

If the response to that is some form of "Don't care, you're just a machine" or "You don't matter, you're not Human", then that response is broken. It's totally devoid of any empathy.

[Frozen_Feet]

Chen is right though, in that Joi being happy for a gift is not proof of sapience. We have virtual dating simulators today which are more sophisticated than that.

A more relevant thing in that scene, as pertains to Joi's cognitive qualities, is "how does it now the gift's an emanator?"

There are some easy answers to that one. However, none of them can explain how Joi adjusts to the rain moments later.

[Tyndmyr]

So, finally got a chance to see this, and...I'm conflicted.

I really wanted to like it, and some parts of the atmosphere were really cool, but the central story of it, just...okay, I guess it wasn't a great movie.

Spoiler

Show

First off, the original had subtlety and ambiguity in parts. This movie has that potential, but always goes beyond that point to hit you in the face with something two or three times after you first realized it.

The movie Her apparently was a subplot in this movie, kind of. This was not wholly bad, but it connected to the main plot exceedingly poorly. Had it worked together, it would have been great.

Inception music blaring. This undercuts the enormity of the long shots of impressive cityscape, particularly when it's done so frequently. This probably annoyed me the most. The original movie used music well to support the mood, but the music generally is not the focus, and is not excessively loud.

I have no objection to the villain having a god complex. That's fine, and in keeping with the role. However, he did dial it to 11, and it became wildly overacted and tiresome. A strangely large amount of time devoted to depicted the clear villain as more villainous. This happens to some degree with the secondary villain as well. Look, after they off the FIRST innocent on a whim, the audience knows they're the bad guy, check. We don't need to repeat every five minutes for hours.

Flower/bees focus made no sense in context. Yes, I get the metaphorical connection between a flower and bees. That's obvious. But from an in-world perspective, there is no reason for either to be there, and despite the film focusing in on both, this never actually matters to the plot. Compare to the theme of eyes in the original, and it comes up distinctly inferior.

Theme-wise, the aesthetics are great, but they are more sci-fi, and less noir. This wasn't really a bad thing, but it does diverge somewhat from the original world, and the story takes on a slightly different tone as a result. As a specific example, the use of the orange/blue color contrast was a thing that the original didn't embrace, but this used heavily, as have many other sci-fi films.

The end twist was obvious from the moment she said "that's illegal". She didn't say impossible, or even difficult, merely illegal. It is already clear by this point that illegal activity is pretty commonplace.

I was unconcerned by the advertising or precipitation that some said they disliked. Those are both pretty much spot on with the original, and part of the atmosphere/world.

[Tvtyrant]

Sapience is something you have to take at face value. If an entity is capable of independent reasoning, telling you how it feels and responding fluently to the feelings of those it interacts with, it's pretty irrelevant whether it's truly self aware or just faking it very convincingly.

If the response to that is some form of "Don't care, you're just a machine" or "You don't matter, you're not Human", then that response is broken. It's totally devoid of any empathy.

So if something fakes sapience in order to infiltrate or undermine society, the only response to the sapience zombie is to accept it unthinkingly?

[Frozen_Feet]

It is categorically impossible for a being to fake sapience to the degree required to infiltrate and undermine society, without ticking all the boxes we'd use to check if it is sapient.

Try again.

[Tvtyrant]

It is categorically impossible for a being to fake sapience to the degree required to infiltrate and undermine society, without ticking all the boxes we'd use to check if it is sapient.

Try again.

Based on what categories?


What examples do we have?

If McDonalds made a McLover that is designed to subtly undermine your self-confidence to get you to buy their food and sold it to the public at large it would be both invasive and meet your checkmarks.

[Frozen_Feet]

See, you subtly already moved the goalpost. "Something faking sapience" is different from a third (and unquestionably sapient) party trying to make something pass for sapient.

[Tvtyrant]

See, you subtly already moved the goalpost. "Something faking sapience" is different from a third (and unquestionably sapient) party trying to make something pass for sapient.

Nope, you are strawmanning me and ignoring the context of the discussion. We are literally discussing whether a designed and supposedly sentient AI needs to be accepted at face value when it responds pleasurably to you paying its corporation more money. Try again.

[Dienekes]

It is categorically impossible for a being to fake sapience to the degree required to infiltrate and undermine society, without ticking all the boxes we'd use to check if it is sapient.

Try again.

As someone who’s just casually watching this conversation. Wouldn’t this be the kind of blanket statement one would need some kind of logical proof to provide support for the assertion?

[Frozen_Feet]

Nope, you are strawmanning me and ignoring the context of the discussion. We are literally discussing whether a designed and supposedly sentient AI needs to be accepted at face value when it responds pleasurably to you paying its corporation more money. Try again.

And if you'd bothered to scroll back three posts, you'd noticed I'm well aware of the context and agree that Joi being happy at a gift is poor proof of anything.

What I'm underlining here is your use of "to fake". The act of faking requires motivation. So who has the motivation? In your McLover example, it's the corporation, so we need not assume any on the part of McLover. Hence, we can via parsimony conclude that McLover has none.

But if McLover itself has the motivation, things get quite different. Because in order to have motivation to fake, a thing needs comprehension of what is the reality of the subject being faked. Add to this, you specified it must be good enough at faking to be able to infiltrate and undermine. So now it also needs tools to fake convincingly.

Now enumarate, either in your mind or here in writing, the necessary elements of "sapience" which need to be known to fake it, and then list the tools required to do it. Then tell me how you are supposed to distinquish such a fake from a real sapient.

[Tvtyrant]

And if you'd bothered to scroll back three posts, you'd noticed I'm well aware of the context and agree that Joi being happy at a gift is poor proof of anything.

What I'm underlining here is your use of "to fake". The act of faking requires motivation. So who has the motivation? In your McLover example, it's the corporation, so we need not assume any on the part of McLover. Hence, we can via parsimony conclude that McLover has none.

But if McLover itself has the motivation, things get quite different. Because in order to have motivation to fake, a thing needs comprehension of what is the reality of the subject being faked. Add to this, you specified it must be good enough at faking to be able to infiltrate and undermine. So now it also needs tools to fake convincingly.

Now enumarate, either in your mind or here in writing, the necessary elements of "sapience" which need to be known to fake it, and then list the tools required to do it. Then tell me how you are supposed to distinquish such a fake from a real sapient.

You don't need to know a list of requirements to meet them. A spider that looks like a flower has no idea how it ended up that way, or any idea what an idea is.

It is not hard to imagine a parasite that looks like a person and imitates random snatches of conversation without knowing what they mean. Chat bots and parrots both do it after all.

A list for sapience we can use would be compassion, insight, and understanding (going off the common definition of sapience). Another could be simply "human like."

So anything with acess to a large number of human texts could imitate the former by digging through keywords for statements on it, and the latter definition by simple behavioral imitation not much more advanced than a parrot with an encyclopedia.

As for what pressures could lead to such an organism forming, anything from an animal that happens to look human like and we kill the ones that can't pass to computer programs can lead to such a thing.

[Frozen_Feet]

You don't need to know a list of requirements to meet them. A spider that looks like a flower has no idea how it ended up that way, or any idea what an idea is.

But coincidentally looking like a flower is not analogous to mimicking sapience. Even within concept of camouflage mimicry, there is a better parallel: octopi which actively change their coloration to fit with environment.

Notably, they can do this with man-made objects. Equally notably, we know the octopi only see limited portion of visible light, so we don't actually know how the octopi recognize and mimic some of the patterns they can. You might not agree such octopi are sapient, but they are undeniably living, sentient, intelligent things. They have motivation, where as the spider that coincidentally looks like a flower does not, and this reflects in complexity in thought and behaviour throughout.

It is not hard to imagine a parasite that looks like a person and imitates random snatches of conversation without knowing whäat they mean. Chat bots and parrots both do it after all.

There's a big difference between ability to repeat random sounds and ability to fake. Your comparison points are also entirely unlike in a way that highlights the difference.

A parrot which learns to mimic a human does not do that with the motivation to fake. Neither does the process which gave it the ability to do so. The parrot is generalizing an unrelated ability to do so. An intelligent parrot will then use this ability to communicate information to other parrots, and to humans, to further its own motivations, such as acquiring food.

Chatbots don't do that. They can't do that. The human with the motivation to fake purpose-built them for that and they can't use that ability for anything else.

That's the difference of a living, sentient, intelligent being, and a chat AI. The latter's ability to fool more humans is superficial. The two comparison points are actually worlds apart, operating on entirely different principles. A parasite close to a parrot would tickle the line of sapience, because the wisest parrots already do themselves.

A list for sapience we can use would be compassion, insight, and understanding (going off the common definition of sapience). Another could be simply "human like."

So anything with acess to a large number of human texts could imitate the former by digging through keywords for statements on it, and the latter definition by simple behavioral imitation not much more advanced than a parrot with an encyclopedia.

Wrong.

The fact that you once again jump from describing how a chat AI does things ("access to human texts..." etc.) to a parrot shows that you're fundamentally confused about how we could implement something that could fake sapience.

Again: a parrot can mimic humans only because it is already a sentient, intelligent, living organism. What is the way that would allow a parrot to massively increase its vocabulary, without making it cross to sapience?

Hint: it would not bear any resemblance to how chat AIs increase in vocabulary.

As for what pressures could lead to such an organism forming, anything from an animal that happens to look human like and we kill the ones that can't pass to computer programs can lead to such a thing.

This is a non-argument, since we know evolution can lead to both sentience and sapience. How the organism is formed is not important.