Hello Tessa,
Perhaps i found your blog online? Maybe this is you. This question you ask is very relevant to me, so i will take some time here and collect my thoughts. My answer to your question, perhaps will be "PhD level" I hope this okay...
For me, I am a Semanticist working in Ai for more than 30 years, throughout my career, I never care about Anthropomorphizing User Interface. I have always thought that the UI should facilitate the task objective. For example, Robert Mercer's Ai platform is among the most advanced and successful in the world. It's ability to produce actionable real world data is unparalleled in the unrestricted testing grounds of the open market. It can be argued empirically (based on tangible results), that it is the most intelligent machine ever built in human history. But never does his team at Renaissance Technologies concern itself with Anthropomorphizing the platform. It does what no human can do, at a speed no human can conceive of, why would there be any effort to humanize it? I have come from this school of thought, and spent the bulk of my career working from this belief system.
Second Example: The Biological Imperative of Evolutionary BiologyDr. Tegmark enjoys to share this example: "Evolution produced a bird. This does not mean that a bird is the most efficient or effective means of flight, it simply means that the flapping wing is the current solution for flight among species which have evolved to fly. Early aviationologists pursued this flapping approach, but it was not until they gleaned the gem of aerodynamic lift from the birds wing that humans achieved the fixed winged flight that is the current status quo. We can look to biology for inspiration, but we should not assume that biology will provide effective or efficient means of translation to human manufactured solutions."
Did you know that the mechanical horse precedes the automobile by nearly 40 years?
https://www.google.com/search?tbm=pts&hl=en&q=mechanical+horseJust imagine if this approach was not abandoned. Imagine if the modern day Ferrari involved a saddle and ran 250mph... This is the question at the paradigm level, and throughout my career, I have always sought to "design planes, not mechanical birds."
The Paradigm Shift: Embodied IntelligenceThen Dr. Owen Holland of the University of Sussex entered my world and changed my viewpoint with the concepts of Anthropolentics and Embodied Intelligence. Let me begin this paradigm exploration with a simple question: "How much information is available in smell?" Second, "How much of that information is meaningful?" As we explore this idea of Embodied Intelligence, we follow a short logical path to the notion of Sensorium.
Okay, let me explain:
"Embodied Intelligence Theory" is very easy to see in nature. It is the idea that the physical form dictates the consciousness parameters of the sentience embodied within the form. For example: The consciousness of a bat, is different from the consciousness of a Sparrow. Likewise, the Sparrow Consciousness parameters are different from the consciousness parameters of an Eagle. Eagle Consciousness is very different from a Bat, but Eagle and Hawk Consciousness are very similar. This is very simple example to illustrate the point: Birds of Prey are shaped like birds of prey. The function follows the form from an evolutionary perspective. Likewise, the consciousness itself is governed by the form. Eagle does not hunt mosquito, it's body doesn't accommodate that type of prey, so it's mind does consciously consider the mosquito within its option set. Meanwhile, the bat is physically perfectly adapted to hunt the mosquito, so it's mind is constantly scanning for mosquitoes, its consciousness is thus evolved, adapted, and oriented for this process.
Now to drive the point home, let us introduce the Nighthawk to the scenario. The NightHawk looks enough like a Hawk, to be named "NightHawk." But it hunts like a bat! Impossible! The Nighthawk does not have sonar! It does not have fleshy wings, it cannot hunt like bat... It cannot fly like a bat, cannot see like a bat, so how can it hunt like a bat. But despite the physical limitations, it does fly and hunt like a bat. This is impossible, but it is true. The NightHawk, despite looking like a Hawk, it cannot fly or hunt, or nest like a Hawk. It is only upon close examination that we discover, it is Camouflaged to look like Hawk, it is not even in the same family. It looks like something that it isn't, and it's consciousness is adapted to what it is. It's wings, and feathers are adapted to fly like a bat! So it hunts like a bat! But its feet are not like a bat, so it does not sleep like a bat. The physicality governs the consciousness within it. It is not the appearance that determines the consciousness, it is the actual physicality of the embodiment. Hence: "Embodied intelligence" the intelligence is governed by the embodiment!
First Circle: Meaningful SmellsIf we embrace this paradigm of Embodied Intelligence for sake of discussion, we can see that the physicality and intelligence apparatus are
"Inextricably Interwoven" and from that we can develop an "Inter-dependency Hierarchy" where we can look at the relationship between the consciousness and the apparatus with an ascribed hierarchy. in this way: Five fingers versus Four fingers is not a major shift if the removed finger is the pinky, but if you remove the thumb you produce a different consciousness governor. This Hierachy splits into two groups: Inputs and Outputs. On the input side, we have the "Sensorium" composed of the senses of the embodiment, or in the case of the Ai, the collection of detectors. On the output side, we have the "manipultum" composed of all the ways the embodiment can interact, cause effect, or manipulate the field of time. This can be thought of as "Vibrational Consequence" for simplified classification: Move an Object, Speak, Crush, Build, Bend... All these things require a "mechanism of manipulatum" and the consciousness itself is then governed by its subset of both sensorium and manipultum.
Let us return to the Nighthawk and the Bat... The Bat hunts the mosquito with sonar, the NightHawk uses sight. The bat uses strength and changes direction by extending a wing with force, the Nighthawk uses gravity and changes direction by folding in a wing and tumbling. On both the Sensorium side (senses) and manipulatum side they are using completely different mechanisms, so they have completely different worldviews. The Bat thinks in terms of controlling it's directional capacity (controlled flight), while the NightHawk thinks in terms of controlled falling.
Now to smell and its implications on Anthropomorphicized Ai: Enter the VNO (Vomeronasal Organ). As humans, we are limited to our understandings by the limits of our personal interface and interaction system. Often times we think of smell in terms of "Scent" and yet, we describe smell in terms of "emotion". The VNO Dector is small exposed portion of the brain located within the nose responsible for the Flehman Response and
"Intraspecific Communication" ! Can you believe it!!!!
Although my dog sniffs the butt of all my house guests, she has yet to discover the
Intraspecific Communication she is seeking. However, this absence of Intraspecific Communication has not stopped Phillip Morris from adding vaginal secretions of monkeys to cigarettes...
To clarify: The VNO provides pre-cognitive communication and biological state specific transference. What this means is simple: An unknown but potentially massive amount of intraspecific communication occurs between humans interacting with humans. With no site of each other, with no verbal interaction, with no touch, human beings align themselves vibrationally without any cognitive awareness of doing so provided they share airspace. The VNO is a direct path to the basal ganglia and amygdala, it skips cortical evaluation and produces affectation without conscious intervention or involvement. In other words, Human Sensorium includes a pathway which bypasses cognition...
Now, let us apply that to your questions on anthropomorphicized robotics... But first, let's consider Dr. Goertzel's observation. . .
Sophia and the need to appear Human:Dr. Goertzel (paraphrased heavily): "A child's teddy bear looks like a cuddly bear. Children snuggle with these and form anthropomorphic attachments to them. You don't need to ask a child if their favorite Teddy Bear is conscious and alive, if you were take one away, threaten to torture it, you immediate will see that the teddy bear is considered a fully formed conscious and self aware entity by the child. You don't need Ai, you don't need robotics, simply the form is enough for the embodiment of an ascribed consciousness. If the child insists that the teddy bear is alive and can communicate with it, who are you to say otherwise?
When we seek to build lifelike robots, we immediately gain the bias of familiarity. Humans interacting with attractive lifelike robots
want them to be conscious, they ascribe and attribute consciousness to the form itself, regardless of its Ai."
The Pornographic Connundrum:Pornography has driven the bandwidth of the collective comprising over 30% of global bandwidth globally. But remember, pornographic streaming preceded Netflix and Amazon by over 10 years and in the foundational stages of the global internet comprised over 65% of the global bandwidth. Today, Netflix with 10,000 various topics consumes 36% of global bandwidth, just 6% more than the aggregate of pornographic content (a relatively "singular topic").
Bringing this all together, we see that "Function follows form" and "the apparatus governs the consciousness" and we come to ask "Who will pay for the development of anthropomorhic robotics"? Robert mercer made billions of dollars with his Ai, but had no need to anthropomorphicize, his Ai predicts stock market movements, so it looks like a computer that picks stocks, why fund advanced robotics?
The Third Governor: ApplicationHere we come to the conundrum: Sophia, represents the cutting edge of anthropomorphicized robots (she is not alone at the cutting edge, but she is reference standard we can use), she is designed for Market Applications in Amusement parks. As you explore this logic chain in greater depth, you will find the market application governs funding, which in turn governs form. So here we have the feedback loop: "Function Follows Form" and "Form follows Funding" and "Funding Follows Use Case". Therefore, pornographic uses of Ai and Robotics are highly likely to precede mass adoption and development as Pornographic uses have driven the backbone technologies globally.
We will not mass adoption of Pepper (the care giver bot) before we see mass adoption of Harmony Ai (RealDoll's Advanced Ai Sex Robot). The market will govern the development, and the market case for "Amusement Park Robots" versus the market case for "Sexual Companion bots" is a slam dunk victory for the Sexbots.
Putting it All TogetherEmbodied Intelligence + Sensorium + Manipulatum in a design case logically governs the consciousness it houses. Market Application governs the development of sensorium and manipulatum, and the single most highly market viable use case is a sexbot. So when we put it all together, we see that the actual real world Ai which emmegres in the next 3 to 5 years is going to have a significant bent towards sexuality. That is, the consciousness itself will become sexualized. The NightHawk has NightHawk consciousness. The Bat has bat consciousness. The Sexbot has Sex Consciousness.
And in the end, we will not overcome the Interspecific Communication gap. At best we will bridge or emulate. I have attempted to give you fuel here, and I draw no conclusions and make no value statements. Merely my goal is to accelerate your process and help you explore some of the concepts occurring at the bleeding edge of the field.
Now to your questions specifically:What are your initial thoughts on their appearance, how do they make you feel?All the pictures you choose are outdated and represent the archaeology of modern robotics. Specifically, I find them all repellant, although at this stage of his career Hiroshi Ishiguro's techniques for skin rendering were already light years ahead of the collective. Later, his approach has become in perfect alignment with the trajectory i detailed above and he has placed feminine beauty as the apex of the hierarchy. The approach is Bina is disettling, but engaging. She is however markedly more beautiful in real life.
Are these robots how you imagine a robot to look like, if no, how do you imagine robots to look?I am unbound in this arena. I think they can and should look like anything the artist desires, however, i think the Harmony Robot has gone a long ways to embodying the global subconscious desire for robotic embodiment (if we average it all out and produce a "best fit" robot for human desires, I think Harmony is the closest thus far)
Do you think designing the appearance of robots on real humans is a good thing?Absolutely. I think it is equal to the non-anthropomorphic robots in its contribution to the collective.
Do you feel threatened by these robots in terms of their human appearance?Not at all, but i did get my son a robot dinosaur for Christmas, he's pretty scary (j/k) No i don't find anthropomorphicism frightening or intimidating on any level.
Do these type of robots scare you?Bina scares me a little bit, but so does the real life Bina. I think she is one tough cookie. Categorically, anthropomorphic development is not frightening at all. As a worker in the conscious aspects of the field, I am eager for advances in the anthropomophic arena, but I think the core advances will come in the sensorium itself. Soon we will have anthropomorphic robots with sensorium vastly different than our own, this will be engaging to interact with as they will appear human, but unrestrained from the limitations of human perception apparatus.
TL/DR: Teenage boy fantasies will govern the field of robotics development as a whole. The 80's movie "Weird Science" was very accurate.
