Consciousness. We could and should be robots. There is nothing to reproducing that requires this, which is the survival of the fittest.
Brain teaser: Take absolute nothingness (not the "void" of space, true nothing) then have that create something that allows me to reply to you on HN.
("Philosophical zombie" would be a more precise term than "robot", btw, and might make soi-disant rationalists take you slightly more seriously because it's closer to the sort of language they like to use. I don't know why that would be a desideratum for anyone, but if it is for you, this may be worth considering.)
What do you mean with robots? Describe how our actions would be different if we were robots, compared with our reality where we are not.
Brain teaser answer: Well, that's more of a question of physics, and the origins of the big bang. We are talking about brains here. They came about some billions of years later.
Consciousness is hard to define, but if we examine animal intelligence it unsurprisingly seems to exist at various levels of complexity. So even worms have intelligence (and consciousness IMO) that isn't quite robotic: they react in predictable ways to light, but the reaction is modulated by other factors like temperature, moisture and what they are doing at the time. Their behavior can be described by a simple neural net, which is in fact what they have. (see Darwin 1881)
Larger animals evolved the ability to evaluate multiple such signals simultaneously, and to do so it's necessary that lower level consciousness (like what the worm has) can send signals to higher levels of consciousness for evaluation. So if my foot is cramping a lower level system sends an appropriate pain signal to a higher level system, which can then aggregate that signal with other signals and cross reference it against current active behavioral plans. If I am meditating, I might ignore the pain, but if I'm working I might get up and stretch. This is similar to how the reaction of a worm is modulated, but just filtered up through higher levels of consciousness.
Why would we develop higher levels of consciousness? Because any change to animal brains that allowed them to be aware of sensory input and evaluate it with reference to behavioral goals (instead of just reacting like a robot) would increase that animals fitness, and any marginal increase in fitness will result in those traits being passed on.
The most significant one for human consciousness in particular is that the appearance of language pushed evolution in the direction of favoring a large working memory. If you can tell me "the bananas are down the hill, to the right, across the stream, next to the big stone" and I can remember that information and go find the bananas, then my fitness is improved. Darwin also pointed out that this increased working memory allows people to reason that if I help someone, that other person might help me back, allowing altruism to emerge from the cognitive space allotted to planning, which evolved from the pressures related to holding large chunks of linguistic data in the mind.
So there are two things: first is subjective experience, which to me seems to be easily explainable as being like "system signals" from lower systems to higher systems. It's important to remember here that the appearance of information itself and intelligent responses to the environment is one of the defining features of all life, even the most basic (Adolf Heschel 2002). The smell of rotten eggs (example from the article) is simply a chemical signal that we have, understandably, evolved an aversion to. My reaction to ice-cream is similarly a chemical signal filtered through a lower biological system which interprets it as something REALLY GOOD and sends the appropriate signals to my higher level cognitive systems. The fact that it's a subjective signal and not a "robotic" reaction means that I can respond to it differently depending on other factors, which is clearly advantageous from the POV of natural selection. If I have diabetes I can resist the temptation of "deliciousness" to my great advantage.
The second thing is that humans have a great memory, so we can hold lots of these signals in our minds alongside memories, plans, ideas etc. Since that ability was associated with increased fitness over millions of years, it has increased to the level we see today.
There are plenty of examples of bad design in evolution, human consciousness included (anxiety, depression etc). It's only necessary that the benefits outweigh the drawbacks for the feature to stick around.
Humans are social animals. To interact with other people, it is probably evolutionarily advantageous to have a mental model that can predict the actions of other people. Consciousness might be simply an offshoot of that mental model, in which a person predicts their own actions, too.
I'm not saying that this theory is true, although I do think I heard it somewhere. (Maybe it's in I am a Strange Loop?) But it's just meant to be an example of how consciousness could arise as a side effect of something with an actual reproductive advantage.
Explain to me how a somewhat intelligent creature would experience it's life without consciousness? What's the alternative? How would you, as a living creature, be able to sense your environment, and fight for preserving your survival, if you didn't have a sense of self?
Take the real and complex numbers. The real numbers have the nice property that they are easy to order since they form a line. Given any two you can tell which is larger. When you go up the complex numbers you lose this obvious ordering, you can define a new ordering, but it is no longer trivial since you are comparing points on a plane.
Now, some people see complex numbers as an extension of the real numbers, but it would be more accurate to call them a generalization of the real numbers. The real numbers are a subset of the complex numbers that satisfy the property of having this obvious ordering.
This again extends to the complex numbers and quaternions. A quaternion is a point on a 4D plane that does not necessarily satisfy commutativity, AKA ab != ba. The subset of quaternions that do satisfy commutativity are the complex numbers.
Every time you go up a level you lose an axiom, which is an assumption on how things work for that system. In a sense an axiom is a useful limitation that gives a certain structure to things. So what if there are 0 axioms? This would mean everything is possible, no limitations. 2 + 2 = 5. Obviously everything being possible is not useful.
Now how does this apply to physics? Physical laws are our axioms. But what if there are actually no physical laws, we are just witnessing the subset of "everything" that appears to have structure? If we are only capable of understanding things that are rational, we would be inherently unable to process events that are irrational. We would project this irrational observation down to the rational subset that we can make sense of. Let me finish with an example.
Take a uniform quantum superposition. It has an equally random chance of being measured 0 or 1. There is no hidden information determining this, it is truly random. This is irrational because there seems like there should be a reason for one final outcome being measured over the other due to our familiarity with cause-and-effect, but this ultimately appears to have no cause. We project this phenomena down to the rational and explain it the best we can with quantum states and probability.
People say the same thing the passage of time, too. We live in a block where all events, past, present, and future are fixed and unchanging.
It's essentially denialism: if a phenomenon doesn't fit well into my framework, it must not really exist because framework.
The question still stands. How would a creature experience life without consciousness?
Or let's even say you created a generally intelligent neural net with sufficient sensors, cameras, and actuators for it to live in our physical world, piped into the neural network. It learns to use it's sensors and actuators, much like a growing child would. What is it's experience? How does this neural network experience life as compared to our brains?
To give you a hint, it wouldn't have a Heads-Up-Display with a battery indicator. It would feel tired or hungry. It would experience life immersed in the world to the extent that it's sensors would allow it. It's sensors would be naturally fused in it's neural net in whatever way was most optimal for it's environment. Much in the same way that we use sight, sensation in our feet and muscles, and the liquid in our ears to maintain our balance without ever thinking about the source or distinction in the data our brains are receiving.
It wouldn't. "experience of life" ~= consciousness.
I'm saying consciousness isn't unique to humans.