Transcript

Uncanny · Why We Do What We Do

0:03
You're listening to Why we do what we do. Welcome to Why We do
0:16
what we do. I am your slightly off host also three four three guilty
0:20
Spark, Abraham. I don't know what that means. I am your CGI
0:24
host, Shane. That is a very nerdy Halo reference. Okay, So
0:29
this is episode three hundred and forty three, three four three, and I
0:32
was like, hey, three forty three guilty spark, and Hey, we're
0:34
talking about a soda semi related robotics type thing topic today, so I thought
0:39
i'd throw that out there. Welcome to the podcast, everybody. We don't
0:42
usually just ramble at the start. I mean we kind of do, but
0:44
we're psychology podcast, and we talk about the things that people do, and
0:47
sometimes we talk about the effects that things have on us, and then we
0:50
give them names like the one that we're talking about today that I haven't quite
0:54
said. And that's okay because I'm teeing it up for a cool intro full
0:58
of podlight culture references. I can't wait. Thank you for joining us today.
1:02
If you are listening on the day that this is released, it is
1:04
December sixth, and there's several holidays to recognize, so just quickly, let's
1:08
get him out of the way. Happy National Gaspacho Day and National Miners Day.
1:14
I like it. There's also Saint Nicholas Day as we're gearing up to
1:17
Christmas, and it's also put on your Own Shoes Day. I'm tired of
1:19
it. Curious to Saint Nicholas Days all the way. This not even in
1:23
December yet, but all right whatever. It is a National Pawn Broker's Day
1:27
and World Trick Shot Day. I like the second one of those. Yeah,
1:32
And it's also Microwave Oven Day, which I support. But it's also
1:36
Computer Science Education Week, which must have started in the nineties because I remember
1:41
when you computer science was a big deal when we were kids, big STEM
1:45
fans over here. Hm. It is cookie Cutter Week, maybe both metaphorically
1:49
cookie cutter as well as actual tangible cookie cutter. Who knows. And National
1:55
Groundwater Awareness Week, which is very important because there's not very much of it
1:57
left. It's going away rapidly and we needed to survive. So yeah,
2:00
goodbye future top drinking water. It's the advice that we got there. I
2:05
guess, I don't know, may maybe not so much. It's also National
2:07
Paer Month, and it's National Root Vegetables Months. So if you like really
2:12
hard things to eat, those are both good sources of could rotten those two
2:17
things together, Yeah, potato, potato and strawberry. It is Universal Human
2:23
Rights Month, like it, and yeah, keep that one going all the
2:27
time. And AIDS Awareness Month, which is also very important. So yeah,
2:31
absolutely all right, But we just like the Lyssa's holidays. Make sure
2:35
that you know, everybody knows that we see you, we recognize you,
2:37
celebrate all of those or any of those in whatever way you see fit.
2:40
Let's jump into our topic today because it's a it's a fun one. I'm
2:44
gonna start with a nice pop culture reference for any of you out there who
2:46
have been blessed enough to have seen all the Twilight movies, and one of
2:53
them, Bella I think is the main character's name. I'm giving myself away
2:55
on my familiarity with this series already. She gives birth to this baby,
3:00
and upon first look, audience is just recoiled in horror at the distressing and
3:05
menacing child, even creepier than any Chucky doll or exorcist child or Rosemary's baby,
3:12
just revolting to audiences everywhere. M Yes, so this was a CGI
3:16
baby. At the time that CGI just was not up to the task of
3:21
making realistic looking humans. However, when baby jack Jack shows up on the
3:25
screen and the Incredibles with the with his disproportionately large eyes, cartoonishly round face,
3:30
and almost plastic looking skin, most people were perfectly okay with this and
3:35
may have even found him cute. I think he's adorable. Yeah, Jack
3:38
checked is great. Okay. So all of that is to say, most
3:43
of you will have heard of the Uncanny Valley hypothesis or UVH is it sometimes
3:47
referred to in the literature. But what is it? Why is it?
3:52
Why is it called the Uncanny Valley? Why does everyone know how it makes
3:55
them feel but can't explain why. So we're going to try and do our
3:59
best to unpack the Uncanny Valley today. I always like these episodes that are
4:02
like this, like very complicated experience that people have but nobody can explain why.
4:08
I like the Call to the Void episode that we did. I always
4:10
think it's really fascinating. Yeah, yeah, yeah it is. It is
4:13
kind of a weird thing. I think we have, you know, a
4:15
decent We're going to discuss several hypotheses that some have more merit than others,
4:21
and we'll also unpack some of our own. But you're right that not no
4:25
one actually has a clear understanding of what's going on here, but there's also
4:29
some there's some there's some room for debate in here that we will get to.
4:31
But I would like to just say, really quickly, if you like
4:33
the show, if you'd like to support us, join us on Patreon,
4:35
pick up some merch, subscribe, live a rating and review, say something
4:39
nice on YouTube just to throw the world into utter and complete chaos from what
4:43
usually happens on YouTube. Those are all the different ways that you can support
4:46
us. Otherwise, let's go ahead and dive into our topic. So the
4:49
concept was first introduced by Masahiro Mori, a Japanese roboticist in professor born on
4:55
August thirty first, nineteen twenty seven in Japan and more research focused on robotics,
5:00
human robot interaction, and the psychological aspects of robot design, which I
5:05
think is super fascinating. Like I feel like we've kind of talked about this
5:11
before, like around this time and a little bit after, there was this
5:15
obsession with like sci fi and but like hard science fiction, like Isaac ASTHMOV
5:18
stuff. Yeah, so this is like the time for this. Yeah.
5:23
Absolutely, there's probably a very large wormhole, possibly full discussion to go into,
5:30
like what do research or scientists, these engineers, mechanical engineers, computer
5:35
engineers, the people who are making robots, what kind of factors do they
5:41
consider in terms of the psychological interaction with robots. I think that'd be a
5:44
potentially cool discussion, way off the beaten path for where our are expertise lies.
5:48
But that's never stepped us before, so no, never, so maybe
5:53
we'll take that on anyway. Maury held a bachelor's degree in engineering from the
5:57
University of Tokyo and a doctorate in fullosophy from the Tokyo Institute of Technology,
6:02
so he you know, he had his laurels, He knew what he was
6:05
talking about. He was an expert as far as that was concerned in terms
6:10
of roboticists, you know, robotics, not roboticists. Maybe he wasn't roboticists,
6:15
but definitely robotics, right, and thought a lot about these things,
6:17
right, And so before people had any idea that computer generated images could make
6:21
their skin crawl, there was a much more pressing source of weirdness and that
6:27
was robots. I think that there's probably a really interesting history of robots and
6:30
what that looks like to you that we probably need to unpack at some point.
6:33
But yeah, I mean honestly, because they predate a lot of what
6:36
we normally would think of as when the age of robots probably started. People
6:40
have been making mechanical devices that do work for them for a very long time
6:46
time. Yes, but that would be a fun one probably to dig into,
6:49
and that maybe ties directly into that whole side tangent about the psychological aspects
6:54
of our interactions with robots. But all right, well, so Mari here,
6:58
this roboticist. He's putting around the lab. He's trying to figure out
7:00
what makes robots and people tick simultaneously, and he was kind of speculating on
7:04
the future of robotics and he coined this term uncanny Valley in nineteen seventy.
7:10
He did this in an essay published in the journal Energy. This journal was,
7:15
I believe it was like a Japanese technology type of journal, and he
7:18
envisioned our acceptance of robots on a comparative scale, in which he sort of
7:24
looked at as robots become more and more human like people like them more and
7:29
more until they crossed this threshold into very human like but not quite realistic.
7:36
Yeah, so not quite Blade Runner, y'all. So we haven't gotten there
7:41
yet exactly. Now. In two thousand and five, all the way back
7:45
then, we were fresh out of high school, trying to figure out where
7:48
we were going to college. Maybe going to college. I was planning to
7:51
be a dad. At that point in time, McDorman and Takashi Minado translated
7:58
Maury's original paper into England. That increased the concept popularity throughout the literature.
8:03
So like, basically once it was translated, more people found interest in this
8:07
because it was able to get out to a wider audience. Yeah, I
8:09
was reaching a lot more people who are doing work in a similar area,
8:13
who are thinking about these things, who are writing about these things, who
8:16
are building these things, and so Anyway, in his essay, in this
8:18
description of what he did, where he was doing this comparison of as humans
8:22
become more human like, people like them more. He sort of plotted this
8:26
hypothesis on a graph, so you can imagine the Y axis, which is
8:31
the vertical one he has labeled as affinity, and this just means preference for
8:35
something on the X axis, he has human likeness, And so essentially you
8:39
sort of drawing the line. This line's going that, this line goes up
8:43
as it becomes more human like, the line increases, the affinity increases,
8:46
meaning that as it becomes more human like, people like it more and more
8:50
right. However, it reaches this point that we're describing at which the human
8:54
likeness becomes very close but not quite right, and it starts to become ear
9:00
and there's the sudden and immediate dip and the sort of emotional response when encountering
9:05
humanoid robots or computer generated characters, or even like behavior and actions from artificial
9:13
sources that are close to human but not quite realistic enough. And so you
9:18
see, essentially it plummets down so far it basically plummets past baseline. It
9:22
goes from like, oh, toy robots are cool to very clearly I hate
9:28
robots. Like that's essentially what this graph looks like. But then it kind
9:31
of picks back up and it goes up to this thing called a boon Raku
9:35
puppet, which I don't know what that is, and I'm looking it up
9:37
right now. Okay, it looks like something that's used in Japanese puppet theater,
9:41
and the ones that are all on here are painted as geishas it looks
9:46
like kind of like an art. Okay, they're very likable, like it's
9:50
a puppet that's likable. It's not like a Ventroquis dummy, which would fall
9:54
below that likable. You know, ye, no thanks or slappy on this,
10:00
Yeah, no slappy when you read that, I love the goosebumps references
10:03
on here. But then it ends the top of the graph is a healthy
10:05
person. So people apparently really like healthy people. I think they underestimates how
10:11
much people dislike other people. Now, as Abraham mentioned, the drop is
10:15
significant. It's so significant that it passes down through the neutral baseline feeling that
10:18
we might have toward a simple robot such as a timer, all the way
10:22
to causing emotional distress in the viewer. So it kind of goes like,
10:26
oh, like I don't know, oh I do of like this. This
10:28
is really bad, and then it starts picking back up. Yeah, but
10:30
if the human likeness kind of increases too indistinguishable from humans, the affinity increases
10:35
again to maximum levels. So basically there's this one this is weird period where
10:39
it's like it doesn't quite look human. There's something off, something weird about
10:41
it, and it makes me uneasy. Yeah. So the reason it's called
10:45
the Uncanny Valley is just because on the graph it's like sloping upwards and then
10:48
all of a sudden dips down really far and then comes back up again toward
10:52
the end when it reaches like indistinguishable from humans level, and it looks like
10:56
a valley. Like just on the graph, the imaginary hype athetical graph,
11:00
it kind of looks like a valley because it's sort of carved out this deep
11:03
dip in the middle of what was otherwise the sloping surface. Right, that's
11:07
just where it comes from. So yeah, sorry listeners, you can't see
11:09
it. I join us on Patreon. We do post these things there,
11:11
but that's if you google it, you'll also you also find it there.
11:15
Yeah, that's basically what's going on here. And so like a rumba,
11:18
if you will, is a robot that has very little human likeness, almost
11:22
none, but a humanoid tiny toy robot has much more human likeness. And
11:28
if it's still fairly cartoonish and somewhat distinguished from humans, we like it a
11:33
lot. And animatronic robot has even more likeness, to the point that they
11:37
are often seen as creepy. A matter of fact, anything with a high
11:41
human like appearance to that level can be subject to this uncanny value effect,
11:46
but the most examples are things like androids, computer game characters, lifelike dolls,
11:52
and those creepy animatronics that like a check the cheese or five nights at
11:56
Freddy's still gives us that mirrors. That's where it is, where the threshold
12:01
meets the uncanny valley. If you will, There's nothing creepier than Charles Entertainment
12:05
cheese. Such a strange thing, what a weird cultural phenomenon. So to
12:09
be clear, the uncanny value was not something he experimentally validated or tested in
12:13
which he collected real responses and reactions to uncanny imagery, but more of a
12:18
philosophical thought experiment as a consideration for future roboticists that they should be aware of
12:22
as they start to develop robots and kind of move towards more advancements in that
12:26
field. Now, initially, as we said, there was relatively little reaction
12:31
or response to Morey's ideas in this essay that he published. However, this
12:35
was fairly early on, there really wasn't a huge boom and bust going you
12:39
know, or not bust, but huge boom going on in robots or anything
12:41
AI related. But as technology began to improve, as it was translated into
12:46
other languages, Morey's concept of the uncanny valley started to have a much more
12:50
significant, widespread response and recognition and have a significant impact on fields such as
12:58
robotics, computer aim, and the human computer interactions and interfaces that we sort
13:03
of commonly engage with now today. Yeah, the amount of robotics that are
13:07
involved in our life today is kind of absurd when we think about it,
13:11
and part of the reason why we are so comfortable with it is because most
13:16
of the experiences are fairly human Like I mean, think about if Siri's voice
13:20
was more robotic, like maybe Max Headroom, or you start thinking about some
13:26
of the types of talk voices that are very like, you know, stilted
13:30
and the tone is all off, like it's very uncomfortable or weird. But
13:33
Siri can actually have a conversation with you in what seems like a fairly consistent
13:37
affect. And I think that's where it makes it so that we're more likely
13:41
to interact with it. Yeah, it's more human even than Rond de Santis.
13:45
It even smiles better, and Sirih doesn't wear heels as far as we
13:50
know. As far as far as we know, we will eventually have to
13:54
bow to our Ai overlords. But in the meantime we'll say, all hail
13:58
advertisers, all he the advertisers, All hell the advertisers, all he advertisers.
14:15
Okay, well we hailed them. So now we can talk about sort
14:18
of why, or maybe the how is it better way to phrase this?
14:22
What's going on with the uncanny valley? Why does it work? How does
14:26
it work? Blah blah blah blah blah. So it almost seems counterintuitive that
14:31
as something becomes more humanoid that we will suddenly find it disturbing when grotestically disproportionate
14:35
representations of humanoids, such as animated cartoon characters, are charming to the point
14:39
of being lovable. Right, so we see these cartoon characters, we really
14:43
like them. They are really ridiculous. They're very exaggerated. But as something
14:46
like technology in the real world starts to advance, it gets really uncomfortable.
14:50
This, despite clearly wildly inaccurate depictions of what ears, eyes, mouths,
14:54
and noses should look like, or impossibly large arms or impossibly slender waists.
14:58
We tend to have an affection or an affinity as we've been talking about to
15:03
these cartoon characters. So what is it about that point at which things go
15:07
from lovable to unsettling? Popeye should be unsettling, should be very unsettling.
15:13
Those forearms would be a medical emergency, you know. So why does this
15:20
happen? It's really not entirely certain, as we mentioned, but there are
15:22
a few suggestions from psychologists and other armchair experts that will give some platform time
15:26
to for the moment. Let's talk through some of the hypotheses suggestions. They
15:31
call them theories, but a theory is based on data and these are based
15:35
on speculation, So I always call them hypotheses or suggestions for why we get
15:41
that unsettling feeling. Right, So, one of these hypotheses is that there
15:46
is an implicit fear of being replaced by robots or ai as they become more
15:52
human like and difficult to distinguish from real humans. And essentially what we fear
15:56
is they're gonna as they get closer to that, they're going to take over
15:58
right, they're more likely and more capable of taking over as they get more
16:02
advanced. Yeah, they appear more capable of doing this now. Okay,
16:06
so let's focus on this, just this idea for a second. That the
16:08
reason we're creeped out is that we implicitly fear the takeover of robots. Okay,
16:14
this replacement idea doesn't exactly jive with the Uncanny Valley, because if this
16:18
were the case, then we should expect people to start out with the highest
16:21
possible affinity for things that are the least human like, right, and then
16:25
to like them less and less as their human likeness increases. Essentially, this
16:29
is not to say that this theory is incorrect, just to say that it's
16:32
inconsistent with the idea of the Uncanny Valley. So it doesn't make very much
16:36
sense logically as a reason for the Uncanny Valley. Like, it kind of
16:40
makes sense for an opposite hypothesis, the Uncanny Valley, right, right,
16:44
right, Like it should start with we love rumbas and we hate any robots
16:48
with arms, and as it gets, you know, as you get kind
16:51
of further along, we start to hate it less. But that's not really
16:52
what happens because we love toy robots. I have a toy robot collection behind
16:56
me, that's right. It also assumes that humans are the pinnacle of it
16:59
achievement, which we're not, and therefore the thing that is most like us
17:04
is the most threatening. So we certainly do bend nature to our agendas more
17:08
substantially than any of the creature. We've kind of established that, but our
17:11
physical form is not likely the reason for our success. So I'm going to
17:15
go ahead and disagree with this kind of this idea, this hypothesis. It
17:19
also assumes robots would want to take over and replace humans, which, without
17:23
spending a lot of time on this implication, seems a bit premature to be
17:26
concerned about. They're nowhere near that level, and like, I don't know
17:30
that they would ever care, Like right, I don't know why they would
17:33
see us as a source of competition that they would worry about. We would
17:37
have to program them to care. Yeah, I think they would largely ignore
17:40
us and just kind of do whatever it is they do. Right. Anyway,
17:42
let's move on to another suggestion. So we sort of unpacked to that
17:45
one for what it's worth. Maybe it's right, just I don't think it
17:48
makes a lot of sense to me. Another suggestion states that these human like
17:52
representations are inconsistent in the ways that matter to how we perceive them. Their
17:57
movements are not governed by the muscle and skin and nerves that underlie our own
18:00
physiology. They may express little or no emotion in their movements or their communication.
18:07
Their sound may be mismatched from their facial reproduction of those sounds, so
18:11
like their sort of lips move and different sounds come out than what the lips
18:15
movements should be sure, because they're probably coming out of something more like a
18:18
speaker than out of vocal cords, they may be delayed in their their reactions
18:22
in a way that feels a little strange to us. Essentially what people expect
18:26
from something human like starts to violate that expectation and then therefore comes across as
18:32
disturbing. So that's sort of this other hypothesis here. I could see,
18:36
just from like an anecdotal standpoint, the experience of if I'm talking to somebody
18:40
and there's like a two second delay whenever they're having a conversation with me,
18:44
how that would be uncomfortable? Yeah, Or like if I have an expectation
18:48
of what your voice sounds like, and or like if you're looking at me
18:51
and you expect my voice to sound in a certain way, but they sound
18:52
like kurmathy frog, Like you're gonna have a different experience with what that looks
18:56
like, or how would you respond to me? So I understand that kind
19:00
of that subverting an expectation thing. Yeah, but you know, again,
19:07
it's just kind of a thought here might be also sort of weird, jerky
19:10
muscly movements that just don't look like something a normal human would do. Yeah,
19:14
because they're sort of mechanical and not human. That would also be a
19:17
little unsettling. Is that why people are uncomfortable with Michael Jackson when he dances
19:22
the whole thing? Yeah? Yeah, fair enough. Another eyepothesist or another
19:27
theory called configural processing, which sounds so very cognitive psychology, nature does.
19:33
That's cognitive psychology brand theory naming right there. Oh yeah, one hundred percent.
19:37
That's what you learn in your first continent psychology class one on one branding.
19:41
It's a terminology we use. There's also a related eyebotist called a perceptual
19:45
mismatch, and they these suggest that the positioning and size of human features such
19:49
as where the eyes are in relationship knows, or comparison between realistic and unrealistic
19:53
features such as realistic skin with synthetic appearing eyes actually creates an ambiguous incongruent that
20:00
makes us uncomfortable. So, if you've read The Second Dune, the dunken
20:04
Idaho Gola who has steel eyes, probably very uncomfortable. That would be uncomfortable
20:10
for multiple reasons. Yeah, also because he is basically reincarnated from the dead.
20:15
Yeah, yeah, he's a Frankenstein steel eyes. Yeah. We didn't
20:18
mention this when talking about the Uncanny Valley, but some graphics give a little
20:22
a few more heroistics along the curve to sort of give a depiction of like
20:26
what is very unsettling to people, and usually at the bottommost level of the
20:30
uncanny as zombies. I actually I'm not sure I would put them there because
20:33
I feel like they don't create the same unsettling feeling that like really bad CGI
20:40
characters create. No, they don't make me go, oh that's weird.
20:42
They made me go and run exactly. Yeah, they're different kinds of fear,
20:48
you know. Yeah. Some psychologists have posited that we experience cognitive dissonance
20:52
when we encounter these human like creations because we don't know how to behave toward
20:56
them as humans. We don't know how to behave toward them as robots.
20:59
Do we behave towards them as animals, as aliens, as something else?
21:02
We don't really know. We don't have a frame of reference for what we're
21:06
supposed to do in terms of like, I see this thing, and I
21:08
know it's not human, even if it looks very human, I'm sort of
21:11
like, I don't know. I'm not sure what I'm supposed to do here.
21:15
Yeah, And I think in a way that's not an a reasonable suggestion.
21:18
I think there's actually some overlap with what I would take, you know,
21:22
embracing what I understand about the philosophy of human behavior. I would have
21:26
a somewhat similar orientation to hypothesis about what's going on in the situation to what
21:32
they're suggesting here. Sure, because it's kind of a novel stimulus that you
21:36
don't really have any learning history with, so like all of a sudden,
21:40
you're like, what do I do? And because you have no learning history,
21:42
it's by itself kind of unpredictable. So as a result, you're like,
21:47
Okay, well, I'm gonna go ahead and I'm gonna ease into this.
21:49
I'm gonna talk to this robot like it's trying to kill me, and
21:52
then we'll figure out what's going on there, not that the robot's trying to
21:56
kill you. Now, I do appreciate our evolutionary oriented I always love hearing
22:00
from this side of the coin. Sure, I think they have some cool
22:03
ideas. They have suggested that we have evolved to see these kinds of distorted
22:07
features as signs of illness or contagious disease, which we find naturally repulsive.
22:12
Their words, Okay, but that's not This is maybe one of the ones.
22:17
I don't necessarily agree with it. Yeah, I don't really like.
22:19
But still Others in the same camp suggest that disturbing appearances may indicate more characteristics
22:23
for mating, which turns us off. If it's like if things are off,
22:26
are uncomfortable, or perhaps it's because we know implicitly that these animated things
22:32
are not alive and are instead more akin to the undead and cause us to
22:36
think about our own mortality. But by this logic, we should distrust all
22:41
artificially animated things, not just those that are human like. So can't we
22:44
can't just go, oh, this is human like and it's unlive. I
22:47
don't like it. It's anything that is moving that's unlive that should make us
22:52
uncomfortable. Right, there's really no escaping the uncanny valley in that scenario,
22:56
not that it's critically necessary, as this idea he has still just an idea
23:00
and not validated, as we said before, but it's just the evolutionary one,
23:04
like we specifically talked about zombies, doesn't really match up with what Maury
23:07
was trying to argue. Yeah, and also I feel like it doesn't quite
23:11
match up with our actual experiences with this sort of uncanniness that comes with human
23:17
ish but not quite right. I feel like that argument will be made with
23:21
maybe this doesn't account for things like stone carvings or paintings and stuff like that,
23:26
because it has to be moving or kind of like animated. But I
23:29
feel like if you saw certain statues, he'd probably be a little bit uncomfortable
23:33
with how it was designed because it didn't look quite right. But for the
23:36
most part, we kind of generally accept that that's true. Yeah, all
23:40
right, I study in twenty eighteen by David Retajzik, David Ritazwick, There
23:45
you go, Tajzik, There you go, There you go. It was
23:48
one of those. It was somewhere there. I said there you go,
23:51
like three times, but one of them is probably it was better than I
23:55
did. Anyway. Anyway, this guy, I'm just makeing it's a guy.
23:57
This person surveyed hundreds of foul of comments on YouTube videos. Oh my
24:06
god, poor poor fellow. You really took one for the team on this
24:08
one. Yeah. He retired soon after, and I left the human experience
24:12
because I would want to do that. Oh man, chills me to the
24:17
bone. Anyway, So I surveyed all these YouTube comments my videos of robots
24:22
and found that people were most repelled by robots based on their size. As
24:26
one, so, small robots people really didn't find threatening, but larger ones
24:30
they did, but not necessarily their human likeness. Further, the more human
24:33
like their behavior, the more creepy people found them, citing that it made
24:37
them question what it means to be human. Oh, I know, I
24:44
guess you're watching this very human like robot and thinking like, wait a minute,
24:47
am I a robot? You know? Here? They I've seen I
24:53
have seen a lot of weird creepy robots. There's never been a time where
24:57
I've seen a weird creepy robot in God, Am I human? Yeah?
25:03
That's probably not they're they're thinking. But I thought it was a great joke.
25:08
But I think, you know, I could see it being something like,
25:11
wow, if it's humanity, is this easy to replicate using robots?
25:15
Then it is our humanity really that significant or important? Is maybe more along
25:21
the lines of what they were experiencing. But let's go with the am I
25:23
human? Thing? I think that's great? Am I? Wait? Am
25:27
I a robot? Like? I know? But even that, like,
25:30
I don't think there's everything. There's never been anything that's made me And maybe
25:33
I just don't care enough to question that right from a philosophical standpoint, like,
25:37
I've never kind of questioned was it mean to be human? Same and
25:41
seeing a creepy robot would not make me question my own humanity. I think
25:45
what makes me question humanity is like whether or not we treat others well.
25:48
And for the most part, we have found that humanity is failing, failing
25:52
quite a bit, failing so hard in fact, that in the middle of
25:56
a podcast they'll just drop an ad and treat others so poorly with that at
26:07
that was an odd bit of cruelty there. I would actually I tend to
26:11
be more optimistic about humans. I would disagree. I'd say humans generally treat
26:15
each other very well. What happens is that some people who treat people very
26:21
badly get into power, and then they divide people. Sure, those people
26:25
who have the megaphone, who preach fear and hatred and grievances, those are
26:30
the people who I think create sort of a culture of distrust among one another.
26:36
But generally, people I think are actually really good to each other most
26:37
of the time. At the very least, they're very neutral to each other
26:41
a lot of the time. Yeah, I would agree with that. That's
26:44
neither here nor there, except that we are talking about human like things and
26:45
robots are not that. So let's move on to our next thing. We've
26:48
been talking. We've talked about cognitive psychologists, we've talked about evolutionary psychologists,
26:55
we've talked about armchair psychologists. Now let's talk about neuralists. Yes, very
27:00
different group of people. They're the how things work type of people. Yeah,
27:04
or or where they work is often a question? Yeah, where they
27:07
work. So there it is. I see. Ah. Yeah, just
27:11
like going back to our discussion on the extended version of this video episode on
27:15
our Patreon where we talked about the hippocampus, Yeah, that's true. Now,
27:19
some neurologists have conducted brain imaging studies that suggest that when we are interacting
27:23
with robots and other AA like things, we may be using more analytic thinking
27:29
than social interaction oriented thinking. And I think that that's a really important kind
27:33
of way to look at this, that when we are interacting with something that
27:37
we have recognized as either kind of mimicking a live or mimicking a living thing,
27:44
that we're not looking at it from I need to interact this with this
27:47
thing from a social perspective. We're going what is this, why is this
27:49
happening? What does this do? Like, where does this go? Like?
27:53
What like what is this capable of? And start kind of answering different
27:56
questions about that interaction. Not that I would recomon men this matter of fact,
28:00
I recommend not doing this at all. I think it could cause a
28:04
lot of distress for both people on either side of the interaction. I'm about
28:07
to propose, but if you would imagine a situation where you go to interact
28:11
with some stranger and you make yourself believe that that person is actually a robot.
28:17
How differently you would interact with that person versus if you thought that they
28:21
were a real person who was going to go home to their family and have
28:25
dinner and like has plans and a life and hobbies and things like that.
28:29
But instead you might think, like, no, they're just a robot.
28:32
They just turn off at night and you know, recharge, and then they
28:34
turn back on and then they're here the next time. And how different of
28:37
an experience that would be. I can sort of see the rationale in why
28:41
then we would feel very differently interacting with human like things that we know aren't
28:45
humans versus when we are interacting with things that we know are humans. You
28:49
know, even better, though, think of that person as a robot going
28:53
home to the robot family and the robot kids and the robot dog, taking
28:56
a robot shower, sleeping in their robot bed, and then coming to their
29:00
robot work the next day, and then questioning whether or not you are in
29:03
fact a robot in a robot world and you just didn't know until just now.
29:07
That's pretty good that's pretty good. We're gonna get so many people writing
29:11
in just being like you have ruined my life. There was a movie or
29:17
TV show not that long ago where there was some character and it was a
29:19
robot and you didn't know, and they didn't find out until later. It
29:22
wasn't the Blade Runner movies, though that would that sort of fits the bill.
29:26
It was something else. Was it Agents of Shield? Where they had
29:29
the Agents of Shield had the Life model decoys, the lmds, which were
29:33
like, they didn't even know they were robots until they were like disengaged.
29:37
That's horrible. But no, I never saw Agents of Shield, so that
29:40
wasn't the one I was thinking of. But okay, fun it happens they
29:44
replace characters with robots and you don't know until the character gets killed and you're
29:48
like, oh my god, and they're like, oh nope, that was
29:49
an lm D. That was a life model decoy. All right, all
29:52
right, well that's horror sting. Yeah, all right, Well, I
29:59
think it is time for us to sort of unpack what I think would be
30:03
waning to riff on this a little bit, but a behavior analytics sort of
30:06
a very behavioral framework for thinking about the uncanny Valley and the extent to which
30:11
it might occur, what would be an explanation for it, if you will.
30:17
So some of these kind of theories or suggestions get pretty close to what
30:19
feels like a good skeptical, empirical conceptualization of the uncanny Valley. And as
30:26
we develop, we learn to respond to a wide and complex variety of features
30:30
of people around us. So most of us have been exposed to a lot
30:33
of different features and characteristics and what might be described as formal stimulus classes among
30:37
humans. Yeah, the idea is that these are just like formal features the
30:41
look and the movement of eyes, lips, skin, general movements and mannerisms,
30:45
as well as how we look and move ourselves prepare us to engage with
30:48
those features in the real world. And so we have kind of a unique
30:52
learning history with what types of behaviors to expect from other humans, and further,
30:56
we start to learn what not to expect from people around us. You
31:00
can kind of see this too when we see behavior that is quote unquote out
31:03
of the norm. Yeah, you know, where it's like non normal to
31:06
behavior where somebody's doing something that's kind of like not something that's expected from social
31:11
etiquette. We kind of notice that right away. Yeah, you know,
31:14
somebody walking down the street and they just drop their pants. You go,
31:17
Okay, that's something that people don't normally do. Now I'm noticing that,
31:19
right. I think that's important to note. Kind of as our experience with
31:22
engaging with other humans goes, we have kind of an expectation of what humans
31:26
do. Yeah, Like, if you were to go into I don't know,
31:30
some kind of acrobatics training place and see people walking around on their hands
31:34
all over the place, you quickly realize like, okay, that's the thing
31:37
that people do in this context, right, But if you are out walking
31:40
around them all and all of a sudden someone just got on their hands and
31:42
started walking around, that would certainly be like catch your attention, because that's
31:47
not something you tend to see. And I think a lot of people think
31:48
of this as this shocking revelation, as like, oh, look at that
31:52
person stopped and paid attention to someone walking on their hands. I'm like,
31:55
yeah, nobody does that here, Like right, everyone would stop and pay
31:59
attention to that because that's a thing that does not happen here, right,
32:01
big surprise. It stands out in a contextually homogeneous type of behavior. Right.
32:07
I love that. That's beautifully said. When something is out of that,
32:10
you go, oh, that's like if everybody's wearing red shirts and you
32:14
see somebody wearing a blue shirt, you go ah, because it stands out,
32:17
not because not because it's like weird, it's just because it stands out
32:21
compared to a larger group of people. The one person wearing a Joe Biden
32:24
shirt at a Trump rally, and you can spot them very easily, very
32:30
clearly. Yes, they're the one with a book. Yeah, they're the
32:32
one without the swastika. Yeah, got them anyway. It was just the
32:37
blue versus red. That was all they got me thinking about that anyway.
32:39
Still talking about this, So things that are clearly and easily distinguished from humans,
32:45
even when they are cartoonishly human, they don't appear ambiguous compared to what
32:50
we expect because they don't look like how we expect other things to look.
32:54
Right. So that's why when we see a picture or a cartoon or an
32:59
ai jone generrated image of something that doesn't really look human, it looks like
33:02
some very cartoonishly disproportionate figure of a human, we can usually kind of just
33:07
roll with it. It might look weird, it might look even a little
33:09
unsettling in some ways, particularly if it's not disproportioned in the ways that we
33:14
sort of gotten used to. Like all cartoons have these gigantic eyes, and
33:17
so they might look a little different from that. But for the most part,
33:21
we can look at these odd representations and they don't look human enough for
33:23
us to try and match them to what we are used to seeing as human
33:28
And I say, see a lot. It doesn't have to be visual like
33:30
this, you know, blind people could have the experience of Uncanny Valley inness
33:32
too, particularly with sound. Also, the sound is something that we attend
33:37
to. That's part of it in the same way. Right, So the
33:39
more features get close to the features we are used to seeing, that is,
33:43
the more things blur the lines between what we're used to interacting with and
33:47
what is new and different in those AI and robotics. But without exact correspondence,
33:52
it would very predictably feel sort of off. Right. It's the fact
33:55
that those cues, the features, the movements of robots and CGI, they
34:01
would affect us in similar ways to cues, features and movements of normal interactions
34:07
with people that we were around, but they do so incorrectly, ambiguously,
34:12
unpredictably, just like we were just saying that the conversation about what stands out
34:16
in a crowd or in an area where the context says the contextually homogeneous behaviors
34:22
that we see here, this thing is not that. Essentially, we'd have
34:27
almost the same cues, almost the same cues, like they're really close,
34:30
and so there's not a lot of places where we have the lines being so
34:35
closely blurred, where like this thing is just barely off from this thing,
34:37
right right, but those new ambiguous cues would be associated with different contexts and
34:43
features. It's like watching twin peaks where everything just feels wrong the whole time.
34:46
I think that's the way that the one that said it is sort of
34:50
like, you know, it's off putting because it's not what we expect and
34:52
expect, a sort of a fuzzy, nebulous turn that doesn't mean a lot
34:57
except to say that it's derived from the idea that all of our experience and
35:00
so is prepare us to interact with certain stimulus cues in a particular way.
35:02
And you know, little kids are notoriously bad at this, as they're like
35:07
getting used to the world. They say and do bizarre things that are sort
35:09
of like why would you even ask someone that what are you talking about?
35:13
What are you doing? Right? Like why are you like putting that thing
35:15
in your mouth? And it's like they don't have as many features and cues
35:20
and sort of patterns that they've picked up that look like what you might call
35:22
normal in a particularly homogeneous context. Right. Homogeneous context, by the way,
35:28
be like a great super nerdy album name for like yeah, yeah,
35:31
get a really good indie rock record, or like weird like math rock record,
35:36
like you could hear it being like a Minus the Bear offshoot. All
35:38
the members are actually scientists and they thought that was a good idea. That's
35:43
our band now, yeah, contextually homogeneous anyway. All that is to say
35:47
is just like we see examples of this and humanoid things that are not very
35:52
good at the humanoid thing yet, which is kids. Arguably they're great at
35:54
it, but just like they their behavior is not formed really consistent patterns yet
36:00
is not really consistently informed by its context. They're still learning, and the
36:04
same is going to be true for these these humanoid robot or CGI things that
36:07
are like almost there but not quite. And now with kids, we kind
36:12
of expect that again, Like, that's back to the part where as adults,
36:15
as people who have been around long enough, we see kids doing weird
36:17
things. I'm like, yep, that's what kids do. That's totally makes
36:21
sense. And it's actually part of the reason when you see a kid who
36:22
speaks like a very young child who speaks with like perfect articulation and pros and
36:28
says really clever, thoughtful and sightful things, and you're like, oh,
36:30
that child creeps me out, and again it violates the norm, like what
36:35
we expect from them. Yeah, Like, why is this precocious child reciting
36:37
things to me that are more complicated than things I understand? I don't get
36:43
what's going on here, and I don't like it. It's a very similar
36:45
parallel that we can draw. So that's I think that's where I land in
36:49
terms of trying to unpack what might be going on in these situations. And
36:52
I think that's probably where I would land to more than anything. There's always
36:55
a level of discomfort when something subverts your expectation of a scenario or an interaction,
37:00
right, Yeah, and not necessarily always bad, Like it's it's one
37:04
of those things where oh, I had this expectation and now it's changing,
37:07
and I have a skill set to pivot towards that or away from that,
37:10
like I have, I have the behaviors or like the knowledge or the experience
37:15
to be able to handle the situation. But I think what ends up happening
37:17
with the young Kidney Valley stuff is I've never seen a baby that looks like
37:21
an adult. This is weird. I don't like it, you know,
37:24
Like, this is like I have no way to handle, Like what do
37:27
I say to this adult baby? Like what do I say to the parents
37:30
of this adult baby? Like I don't know what to say, like your
37:32
baby is handsome because I can't say cute because it's because it's an adult baby
37:37
and adult babies are not cute. That's why everyone hated Benjamin Button. It
37:42
was weird. Let's let's all admit that's a weird premise for a movie.
37:45
It was a weird premise. It was that was not a bad movie,
37:49
but no, that was great. Yeah, I think anybody hat Probably some
37:52
people hated it, but there was that element of very weirdness and the premise
37:57
was very strange. Somebody was very high and was like, dude, what
38:00
if we were like born old? Yes, h five Edibles later, Benjamin
38:08
Button. Yeah, the script is written and it is a blockbuster. All
38:15
you have to do is cast Brad Pitt and we will make money. That's
38:19
so far has been the case. Name one bad movie that he was in.
38:21
You can't I will think on it, and giving me some space to
38:25
think on it, we will let you think on these ads. There is
38:37
no escaping the ads. But is there escaping the uncanny valley? Oh?
38:43
Oh, what a great lead in I love that. So some do think
38:47
that we cannot escape this uncanny valley. It's more like a wall, like
38:51
we can't ever get past it because there's always going to be that situation that
38:54
kind of toes the line a little bit. It reaches the point that's just
38:58
like you just can't cross it. It's like nope, it's n i'llow has
39:00
been canny? Do you either go more cartoonish pull back from the uncanny,
39:04
or you just stay unkenny right? And maybe I'd like to move beyond uncanny
39:07
because it makes me, it makes me feel weird. But there are some
39:10
more recent examples that have demonstrated that maybe we have already found a way to
39:15
bridge the gap. So, for example, that the aging effects that are
39:17
used in recent movies, such as Samuel L. Jackson's youthful appearance in Captain
39:21
Marvel, or most of the characters in Martin Scorsese's twenty nineteen film In the
39:25
Irishman. The one thing that is kind of an uncanny valley situation in that
39:30
movie is when Robert de Niro, who is deaged, starts beating somebody up
39:34
like he himself as an eighty year old man, Like he's kicking he's swars
39:37
to be twenty and he's kicking somebody like he's eighty, and you're like,
39:39
that doesn't match. That's weird, But the rest of it, that's true,
39:42
not too bad. That's a good point. But there are also several
39:45
entirely CGI characters that most people haven't even noticed for CGI or they just don't
39:50
care, Like Guardians of the Galaxy's a really great example. That's like I
39:53
love watching Rocket Raccoon. He is so like he has done so well and
39:59
he doesn't bother me. Yeah, I think that's very fair. Like sometimes
40:00
they're done with mo cap. I mean, we look at Avatar and Rocket
40:05
raccoon, and sometimes they are just completely just completely completely CGI. And sometimes
40:10
they'll even like CGI a face onto a cgied body, where the CGI body
40:15
was kind of just a dummy, and usually you don't like see it for
40:17
a long time. It's not doing a whole lot of like heavy lifting in
40:22
terms of moving the scene around. But we often sort of just at this
40:25
point, we just accept it. And it's because it's gotten so good that
40:29
we kind of don't notice it. And so it actually kind of gets to
40:30
the point that escaping the Uncanny Valley with CGI seems to mean the fact that
40:35
the images are artificial becomes essentially invisible to us. We don't even notice that
40:39
it's there anymore. So it's almost not that we have an affinity for it,
40:43
it's that we kind of stop noticing that it's not real. Is sort
40:46
of happening, I think, is what's happening with the escaping of the Uncanny
40:50
Valley. Yeah, I think we kind of moved through it and start climbing
40:53
towards that human like the puppet, right, Yeah, and so we kind
40:57
of get towards that. Now in your robotics, they some have managed to
41:00
create essentially indistinguishable robots, but they have extremely limited use, such as pitching
41:05
a single line or a non speaking movement. I mean, Disney has the
41:07
Spider Man that swings from the rope across buildings. Wow, it's pretty impressive
41:14
and you go, whoa. But there was one time where it didn't do
41:16
well and it crashed into a wall and it looked like if it was a
41:19
human it would have died, but it was a robot, so it did
41:21
not die. But it's actually pretty great. But again it's single use.
41:24
It's acrobatics, and it's it's really quickly swinging from a rope across and landing
41:30
somewhere else. You know. Yeah, but most likely people will get used
41:32
to it before the quality of the robotics become so good that they are indistinguishable.
41:37
So like we'll kind of get to a point where we just kind of
41:38
go, like we're starting to notice like the robotics itself, until it eventually
41:43
just becomes invisible to us, like like we see with CGI. Yeah,
41:46
and like Blade Runner. So there is also a point here that maybe some
41:52
of you listening may have been thinking, well, wait a minute, is
41:53
this even really a thing? Like I've heard people say that the Uncanny Valley
41:57
isn't real, and so there's a little for skepticism here before we step into
42:01
that. I think there is I wasn't really sure where to put this idea,
42:06
but there is one solution more epitched, which is to just not make
42:10
robots in a human like they just sort of like, you know, the
42:14
Uncanny Valley is the thing that's going to happen, It'll be really difficult,
42:16
if not impossible, to get out of it. Let's just not even try.
42:20
But this seems pretty unlikely, like an unlikely solution everyone's going to agree
42:23
to. People have already, people have been trying for a long time,
42:27
and they're starting to succeed. So it seems like that's probably not the way
42:30
that we're going to go. And soah, it's funny because the more I
42:35
think about it, it's like, I'm sure there's utility to a human like
42:37
robot, but from what I understand, the robots that have the most utility
42:40
of the ones that are like the industrial robots, or like the ones that
42:44
are doing you know, like the work that is too dangerous for humans to
42:46
do that don't necessarily require human limbs to do. Yeah, you have a
42:52
robot that's going to climb through rubble in like in you know, maybe in
42:54
a natural disaster area. It doesn't need arms and legs. It probably needs
43:00
some way to get through maybe tracks, like maybe like a tank or something
43:02
like that, Like you need something like that, you know. Yeah,
43:05
Honestly, skeptics have pointed to the fact that, like just since we're talking
43:07
about AI a little bit now and we're planning on doing a deep dive on
43:10
AI in the future, but skeptics have talked about the fact that we don't
43:15
really need a general artificial intelligence. Like there's so many places where we can
43:20
use artificial intelligence where it's just going to do the one job we need it
43:22
to do really well, and getting to it the general intelligence is such an
43:27
incredible hurdle to cross that we may not ever need to cross it. Also,
43:31
like we might we might have so many different sort of siloed versions of
43:36
AI, or maybe not even siloed. They're just they're individual They kind of
43:38
do their own thing and then maybe even talk to one another. But they
43:42
don't need to be a general intelligence AI, right, They just need to
43:45
be really good at the one thing that they need to do. So like
43:47
you said, like, if there're one job is to find people who are
43:51
trapped under a fallen building, then like they program the aid to do that.
43:54
It doesn't also need to know how to write poetry and create music,
43:59
you know, and drive a car. Yeah, Like I don't need to
44:00
flip an egg exactly right, Like it doesn't need to invent new dishes.
44:04
All it needs to do is find people who are trapped. So why also
44:07
go out of your way to try and make it capable of learning anything when
44:10
you just need it to be good at the thing that it's going to be
44:14
good at when we don't even know if it's possible to make it general enough
44:16
anyway, So that's Some have pointed to that as being like an unlikely scenario
44:21
that we don't really need to worry about. Maybe why make a screwdriver a
44:23
hammer if a screwdriver works just fine, sure, Or make a screwdriver believe
44:28
that it can be a hammer. Yeah. I don't know if that metaphor
44:30
works, but I feel like it felt good in the moment. Now.
44:32
Some have actually suggested that this might be generational. That is, more and
44:36
more people grow up with AI as a normal part of their daily in air
44:38
actions robots and AI will feel less creepy to those who are more used to
44:44
it, and that kind of aligns pretty well with I think the behavioral expectations
44:47
and like kind of the framework where you know, if it's something that we're
44:51
exposed to and have a pretty decent learning history with and a lot of cues
44:53
with, then it's not going to feel like it's kind of out of our
44:57
normal experiences. I can't speak for everybody, but like, I don't use
45:00
Siri a ton, but I use it, and you know, periodically I
45:04
need it for things. I find it very useful when I'm driving, et
45:07
cetera, and be like, hey, give this person a call or just
45:08
shoot them a text, and I can do I can do that sort of
45:12
thing. But at this point, like it just feels like just a tool.
45:15
Like it doesn't it doesn't feel like anything that's unique or special or weird
45:20
or creepy. It's just like, oh, it's just I'm just talking to
45:23
my phone. Because when I talk to it this way, it'll do stuff
45:27
that I need it to do, right. And I think that when Siri
45:29
first came out, there were a lot of people playing around with all the
45:31
things that Siri could or would do, and all these voice assistants that have
45:35
come out since then, and it feels to me, and maybe I'm completely
45:37
wrong, but it feels like for the most part, we kind of all
45:40
just kind of gotten used to it. It's just the thing that's around.
45:44
And like there are now children who are old enough to sort of you know,
45:47
talk and think of do things independently. Who that is a world they
45:51
have always known as a world with Siri. They've never lived in a world
45:53
that didn't have Siri in it, right, because it's not that new,
45:57
Like that is now at least eleven twelve years old, right, Like these
46:00
things have been around. There's been enough time and it's kind of been so
46:04
normalized that it's not it doesn't stand out as weird or strange or anything like
46:07
that. It's kind of like I feel like, you know, certain cultures
46:10
have that with different experiences, like where you go, okay, well,
46:14
this is what a Canning story looks like, or this is how you like,
46:15
you know, in the South, you pump your own guests, or
46:17
we're not like we're not from New Jersey, so like we're used to kind
46:21
of doing that, and it's just one of those things where it just becomes
46:23
part of the regular routines. Yeah. Another consideration here is that really we've
46:30
been talking about these art of these uncanny valley effects from like looking at CGI
46:35
images or interacting with AI behaviors or robots or anything like that. But honestly,
46:40
any disembodied animated figure is likely to feel a little weird, right,
46:45
Like, imagine you've got a robot hand that's just sitting on the desk in
46:49
front of you. It's not attached to an arm, and it looks one
46:52
hundred percent real, like you're look at it and be like that is a
46:54
person's hand. I'm looking at a hand right now, and it's like moving
46:59
around and doing things completely separate and unattached to any kind of arm. That's
47:02
gonna look weird. Like that's gonna feel weird, right, that's creepy.
47:06
Sure, even though we look it looks so real that we would swear it
47:10
belongs to a person, But it's like marching around our desk doing the thing
47:15
things. Yeah, but it's actually just a rolla and like that that would
47:19
feel word. So really it's just it's any disembodied AMBI like animated thing that
47:24
is doing these these unexpected things going to what we were mentioning before, that
47:28
I think is a reason to think that it's not really this uncanny valley so
47:31
much as it is maybe something else, right, And kind of another point
47:36
to make here is that the highly debated theory that we're talking about this is
47:39
this idea of quote unquote theory. A lot of people are pointing to the
47:43
idea that it has the generational effects that we kind of mentioned, as well
47:45
as studies that have found that varying the circumstances in which robotics or gi are
47:50
presented can significantly impact the extent to which people find it uncanny. That is
47:53
to say, the factors that influence whether or not something appears uncanny in its
47:59
human life is going to depend on its circumstances and its context exactly, and
48:05
therefore those maybe the features that are more important to pinpoint when we're talking about
48:07
what makes something feel uncanny. Right, there's like a movie out right now
48:10
where it's about to come out that's like Badi Alligator or something like that.
48:15
Have you seen the previous this? No? Oh, well, they like
48:19
deliberately inserted, like the effects in the movie actually look pretty good, Like
48:22
it looks like it was a relatively high budget movie, but they intentionally made
48:25
this like horrible looking cgi alligator as the antagonist. I think that was it
48:30
something like that, And it actually looks kind of entertaining. But there's sort
48:35
of like in that aspect, you might totally accept the uncanniness of the alligator
48:39
that like looks like it's trying to be real but it's failing miserably at it,
48:44
and think like, this is funny, this is entertaining. I am
48:46
not creeped out by this, And it's because you went in with the expectation
48:51
of, like, I'm going to see a crappy looking alligator that doesn't look
48:53
that realistic as part of this experience. And so I think again, and
49:00
I said, like, it's something else. When I think that something else
49:01
is a relative history is with these things, what we expect from them,
49:06
what we have learned from them in the past. Absolutely, And I think
49:08
at the end of the day, it's a fairly easy concept to describe,
49:12
and what we kind of look at is intuitively right, Like you know,
49:15
when you experience it yourself, you go, ah, yeah, that's that's
49:17
definitely an uncanny valley situation. It's subjectively a concept that is difficult to research,
49:23
like it is very because of the nature of this type of thing.
49:28
It's very difficult to resoarch, it's very difficult to isolate, and there's just
49:31
not really a lot on it because the theory quote unquote theory itself isn't really
49:36
solid to begin with. It's kind of like it's just an experience that people
49:39
can describe, and that's about as far as it can go right now.
49:43
At least. I think that's actually kind of a beautiful take. Come point
49:45
is like, is this real kind of maybe not in the way that it
49:49
was originally conceived of, maybe not in the way that people think of it.
49:52
It's not just like poorly generated imagery. Context is very, very and
49:58
critically important. The extent to which it does as a cur seems to have
50:00
to do with what our learning history is and our learning history both with the
50:05
image or the behavior that we interact with as well as the context in which
50:08
it appears and sort of what our expectations are going into it, which again
50:13
just goes right back to whatever our learning has been around this thing. Right,
50:15
But yeah, I think that more or less says everything there was to
50:19
say about it, it's like, yeah, it is kind of an effect
50:20
we've certainly had the experience of, like watching Twilight and thinking, yeah,
50:23
and that was you know, and up to and including this cgi baby that
50:29
was on the screens. Everybody shots fired, shots fired. Yeah. But
50:35
yeah, so that's sort of what I think about it. So, yeah,
50:37
I'm piggyback on your on your point, I guess no, I think
50:39
it's good. I think it makes I think it makes sense. So I
50:42
think if you are still if you're walking out this episode and you're unsure of
50:45
what we are still talking about, We're gonna give you a couple examples of
50:49
movies that were made famous for their offensive uncanny Valley experience, the first one
50:54
being the two thousand and three Hulk movie with Eric Banna. True Man looks
51:00
like it's so hit. The Hulk himself stands out so much in that movie.
51:05
It's weird. I've actually never seen this one, but I've seen clips
51:07
of it, and it definitely looks like green Putty that sort of moves through
51:10
the environment in a way that violates all of our expectations of physics. And
51:15
it's not like the Hulk in any subsequent movies has ever like super realistic in
51:20
its movements, but like they look at least okay to us, like they
51:24
did such a good job calibrating the speed and the force at which things move
51:29
that we kind of just accept them, whereas this one, it kind of
51:30
just looks like animated character flying through space at speeds that seem impossible. Yeah,
51:35
it's weird and just don't really jive. Yeah, it's a little weird.
51:37
So another example of this was the twenty nineteen Cats movie, which I
51:43
also did not see. But the images from this are quite distressing. It's
51:46
real bad. It's I haven't seen it, but I've seen enough of it
51:52
to know that I never want to see it. Fair enough snic the Hedgehog
51:54
trailer before it was fixed, before all the fans kind of blew up and
51:58
were like, do not Lisa's movie. Yeah, the joke in the Chip
52:01
and Dale movie, that's right. Yeah, the Ugly Sonic is in that.
52:06
And when you look at Ugly Sonic it is really uncomfortable. Yeah,
52:08
it has like a little too human like teeth and like the mouth was not
52:13
quite right size, like it was a little too realistically human almost, and
52:16
like, yeah, had like the fur around. It was just a little
52:20
too like textured, and it just felt wrong looking at it. And then
52:23
the read done version of it was was totally acceptable. Say shatter features.
52:28
Yeah, go with the cartoon, like, it's a cartoon. It's based
52:30
off of a video game that was basically pixels, so your room to make
52:34
for error is very large, Like, have fun with making it look cartoonish,
52:37
right exactly. There was also the twenty ten Alice in Wonderland Tim Burton
52:43
remake that had just an enormous amount of like cgied human faces on things,
52:51
so they like took a human face and basically reconstructed it to be a CGI
52:54
face on various deformed creatures that again were supposed to look pretty much like people,
53:00
but were rather off putting. Yeah, that's it, that's no thanks.
53:05
There's also the two thousand and seven Beowolf movie, which is real weird.
53:08
I need no thanks, you know, honestly, if you don't look
53:12
at the eyes, if you don't look at the faces, the movie is
53:15
visually. Don't look at the human faces in this or any of the faces
53:20
in it. The problem is they just they could not get any of the
53:22
critical features of the face just right, Like the eyes were very glassy,
53:25
the eyebrows move in a stiff, weird way. The mouth doesn't move quite
53:30
right, the nose doesn't move with the mouth as it's speaking in the way
53:34
that it showed. So it's like they just didn't have the musculature in the
53:37
pole of skin and like how light scatters off skin quite right, So if
53:40
you can, if you don't look at the faces, the rest of the
53:44
movie is very visually presive, but they could not really get the faces down.
53:49
No, not at all. The most offensive. In my opinion,
53:52
I have never seen anything that was more horrifying to me in CGI history.
53:58
Besides maybe Twilight's Baby was the two thousand and four Polar Express. And actually,
54:01
I take it back, the Polar Express is horrifyingly bad in its CGI
54:07
effects. Sorry to all you folks that are watching that for December right now
54:10
for Christmas, because I know you're about to watch it, but just prepare
54:13
yourselves. It is uncomfortable. It is uncomfortable. Yes, maybe if you
54:16
have the expectation going in then it's meant to make you nauseous, then then
54:21
it will not be so bad. Who knows. Yeah, agreed to be
54:22
fair, this is two thousand and four, like this was for that time
54:27
stunning, Like they did a very very good job for two thousand and four,
54:30
which on two thousand and four was not good enough to be in a
54:34
movie, and by twenty twenty three it is alarmingly disgusting. So it is
54:39
a hard to watch movie for the CGI faces and the uncanny Valley effect in
54:45
there, right on top of the fact that it's just incredibly boring and not
54:49
that fun, right, I mean, and arguably, you know, when
54:52
you think about two thousand and four, George W. Bush was elected president
54:54
again, so Polar Express was not the worst thing to happen that year.
54:59
I wish I was not cough laughing through that. Great, Well, that
55:01
was fun. Why don't we let the ai ads do their thing and come
55:07
back with some recommendations. Yay for joining us for the first time again,
55:17
Welcome. We love that you're here. Hopefully you're still here or else you're
55:21
not hearing this, so those of you who are here, thanks. We
55:22
always recommend things. At the end of these episodes, we come up with
55:25
things that we enjoyed over the past week, or even that we just thought
55:29
up off the fly, that we're like Hey, this is a thing I
55:30
would like to recommend, and then we recommend those things. But before we
55:35
do that, we need to thank the important people who help make this show
55:37
possible, which includes my wonderful team of people. Shane and myself do writing,
55:42
in fact checking, audio engineering, and production from Justin Greenhouse. Our
55:45
social media coordinator is Emma Wilson. And without my team, this podcast could
55:51
not happen. It would just be me on my own making things terrible and
55:53
no one would want to listen to them. And I think with my team,
55:55
at least some people do. And that includes an amazing group of people
55:59
who you can be on this list if you wanted to, you could join
56:02
it and it would be good and everyone would be happy. And that is
56:06
the list of people who support us on Patreon, and because they support us
56:08
there, we give them extra goodies because we love them so much. And
56:12
that list of people includes Mike, m Meghan, Leyla, Mike T,
56:15
Justin, Kim, Joshua, Brad, Stephanie, Olivia, Brian, and
56:20
Ashley. Thank you all so much for being amazing. Yes, y'all are
56:22
so wonderful. We appreciate you. I think that's what I have before we
56:25
get to recommendations, ms and anything. No, I think that covers it.
56:29
Groovy. All right, let's recommend some things recommendations. So I recently
56:42
got to go to Epcot. Epcot is a park at Disney. Experimental Prototype
56:47
City of Tomorrow is what it stands for. That's what Epcot gomes for.
56:52
I did not know that. Yeah, So originally it was designed to be
56:53
like this, like live in city that was designed by Disney. They decided
56:57
to go with Celebration, which is a whole different thing. But Epcott is
57:00
theme park and they have this really awesome ride called Guardians of the Galaxy Cosmic
57:07
Rewind. So it's a new ride at Disney. If you've been at Epcot,
57:10
specifically, if you've been to Epcot, you know that most people go
57:14
to Epcot that drink around the world. But they're trying to kind of spruce
57:17
up the park and so they added this ride. The story goes that one
57:21
of the Celestials, it's Marvel, it's a Marvel thing. Sure, one
57:24
of the Celestials essentially tries to rewind time because it basically says humans, We've
57:31
made a mistake with humans and we're just going to erase them, which you
57:35
know fair. I mean, if they were look at today, they'd be
57:38
like, yeah, okay. So, but the whole thing is that you're
57:42
on a roller coaster. It's an indoor roller coaster, and you are supposed
57:45
to stop the celestial from erasing humans and you're with the Guardians, So you're
57:49
flying around with the Guardians in space and it's great. There's like six songs
57:52
that it plays, because the big thing about Guardians is that it's so music
57:55
oriented. We got to go on it and the song that we got to
57:59
listen to was Years for Fears Everybody Wants to Rule the World, which is
58:01
a great song for a roller coaster fun. And it's not just a straight
58:06
line roller coaster. The cars they turn and they shift like kind of like
58:09
you're drifting around corners and stuff, so it's got like a different motion than
58:13
a regular roller coaster. It is one of the coolest rides I have ever
58:15
been on in my life. So if you like theme parks and you like
58:19
stuff like that, it's definitely worth a go. So please go to go
58:22
ride this ride because it's really really great and the music is fun. Yeah,
58:25
if you find yourself in Florida, we're sorry. You can try and
58:30
alleviate that pain by checking out Guardians of the Galaxy Cosmic rewind. That's a
58:35
fun recommendation. Yes, absolutely great. I'm recommending a movie. I spent
58:38
all of Thanksgiving break watching movies. So many movies, just a lot,
58:43
a lot of movies, which is great. I really watching movies, so
58:45
this perfect way for me to spend my time. One of them that I
58:49
watched was this twenty seventeen film called El Camino Christmas. This is actually a
58:52
relatively stacked cast here. We've got Dax Shepherd, Tim Allen, Vincent Dinofrio,
59:00
Luke Grimes, Jessica Alba, kurtwood Smith. There is just a there's
59:02
a lot a lot of people in this and it's kind of it's sort of
59:08
like a comedy crime caper if you will, adventure action sort of thing.
59:15
It might largely takes place in this one setting where there's sort of a standoff.
59:17
But it was fun. It was a good movie. I enjoyed it.
59:21
If you're into Christmas themed action type movies, this is rated R.
59:24
There's like blood and swearing and all that sort of thing, so be wary
59:29
sensitive audiences about that sort of thing. But it was enjoyable and it's available
59:34
on Netflix, So if you're interested in something like that, then you can
59:37
check out El Comuno Christmas and again if you would like to check out a
59:39
groovy new ride and Disney World in Florida, Guardians of the Galaxy, Cosmic
59:44
rewind Yeah, so fun. I love it all right. I think that
59:46
is all that I have to say. As I said, thank you so
59:49
much for recording with me today, Shane. Thank you everybody to who's out
59:52
there listening. Thank you to our Patreon supporters. And you can learn more
59:55
about this episode by going to our website, check us out on or Leavis
59:59
s Lilridging interview all that sort of thing. Is there anything I'm missing before
1:00:01
we wrap this up? Or don't think you'd like dad? I think that
1:00:04
covers everything that we need to today. Cool, go b canny or uncanny
1:00:07
whatever we dig it and you should. If you have something to say about
1:00:12
this, you should email us. You can reach us directly in v at
1:00:14
wwdwwdpodcast dot com or on the social media platform is where we will interact with
1:00:19
you. All right, Cool, that's what I have. So this is
1:00:21
Abraham and it's Shane. We're out see you you've been listening to Why We
1:00:25
Do What We Do. You can learn more about this and other episodes by
1:00:29
going to wwdwwdpodcast dot com. Thanks for listening, and we hope you have
1:00:35
an awesome day.