There are any number of reasons robots will never take over the earth.
Let's take a look at the ways that even the smartest artificial intelligence differs from a typical (that is, awake and non-impaired) human being:
* A human is conscious. This difference is HUGE! When you are conscious, you are able to think and feel, and to know that you are thinking and feeling. Even the smartest AI in the world is not conscious, especially not in the way that you and I are conscious. Consciousness is usually discovered through the "mirror test": when the being in question looks into a mirror, does that being recognize what is seen as an image of itself? Human beings over the age of two pass the mirror test with ease. So do the other great apes (chimps, gorillas, etc.), elephants, rhesus monkeys, bottlenose dolphins, and, believe it or not, octopuses. While it's possible to program a robot to mimic the actions needed to pass the mirror test (for example, doing something at random to see what the image will do), I doubt that we will see conscious robots any time this century, if ever.
* A human is self-aware. That is, you know that you are not me; you are (almost certainly) not the president of the United States; you are not Napoleon, or even Napoleon Solo. When you look at an image of yourself in the mirror, you say, "That's me." While a robot could be programmed to make "I am" statements, no robot can EXPERIENCE an "I am" statement ("I am proud," "I am not an animal," "I am ambitious"), whereas you and I take self-awareness for granted.
* A human has an endocrine system. That is to say, human emotions require glands to create hormones. Adrenaline, for example, is great in times of danger. Acetylcholine helps you sleep. You need serotonin to feel love; you need testosterone and progesterone to get turned on. Every emotion you feel comes about because of a complex dance of hormones. A robot might be programmed to say "I love you," but without hormones, it's just words. Robots cannot FEEL angry or fearful or loving or resentful or envious or lustful. Why should they? What possible benefit to humanity is there in designing a robot that is capable of malevolence?
* A human has a conscience. That is, you know the difference between right and wrong; you feel good when you do right, and you feel sorry, guilty, and ashamed (or ought to) when you deliberately, knowingly do wrong. Since they have no hormones, robots have no feelings; they CAN'T feel proud or ashamed, glad or sorry. A robot will act according to its programming. Period.
* A human has imagination. That is, you can think up new ideas and put them into action without being told to. No robot will ever come up with "I want a hug" all by itself, much less "I want to take over the world."
A robot does not know the difference between awake and asleep — not even between "on" and "off." Unless you programmed it to, a robot would never stop to think about the difference between itself and anything else. A robot can't tell right from wrong. A robot might be programmed to APPEAR to make choices, but without being programmed to do it, even the smartest robot cannot think, "If I do A, then B will happen, but if I do C, D will happen. But I prefer Y to happen, so I'll do Z."
Many decades ago, Isaac Asimov formulated the three laws of robotics:
(1) A robot may not injure a human being or, through inaction, allow a human being to come to harm.
(2) A robot must obey orders given to it by human beings, except where such orders would conflict with the First Law.
(3) A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.
Modern-day scientists are not stupid. If there were even the possibility of a robot becoming intelligent, conscious, self-aware, imaginative, AND malignant, the three laws of robotics would be the first thing they programmed into it. These laws do such a good job that most of Asimov's robot stories centered around what would happen if a robot were reprogrammed so that one or more of the laws did NOT apply.
Asimov's three laws of robotics are essentially variations on the laws of what defines a tool: (1) A tool must be useful. (2) A tool must do what it's designed to do as efficiently as possible without injuring the tool-user. (3) A tool must stay intact while it's being used unless it's SUPPOSED to break while being used, or unless safety requires it to disintegrate.
Robots are tools, not people. What benefit would there be to humanity in designing a hammer that was awake and aware of itself as a unique entity? What benefit would there be to owning a microwave oven with a sense of right and wrong? Who needs a self-driving car with an endocrine system that makes it capable of hate and resentment?
There are robots all over the world. Factory robots build cars and electronics. Robots do jobs that humans don't want to do, like vacuuming the floor or swimming in volcanoes. Robots guided by surgeons do laparoscopic surgery. Robot aircraft spy on enemy nations. None of these robots is conscious or self-aware. None of these robots can think up new ideas. None of these robots NEEDS a conscience. Why should a Roomba suddenly "wake up" and decide, without an endocrine system, that it hates and wants to enslave?
Say for the sake of argument that there is an evil inventor, working in complete secrecy (getting all funding from the tooth fairy), who creates a robot that is conscious, self-aware, and imaginative, and as malignant as its inventor. Absolutely no friend, family member, postal carrier, UPS delivery person, grocer, pharmacist, computer supply house, or competing scientist notices a thing over many decades; the evil scientist is working in a complete economic and interpersonal vacuum. One of two things is going to happen: (1) The evil scientist and the megalomaniac robot come to the notice of the world, which nips their evil plotting in the bud; or (2) the megalomaniac robot starts to build an army of duplicates of itself, with the idea of taking over the world; someone at CompUSA or wherever says, "Hey, who's going to pay for THIS billion dollars' worth of supplies?"; and the evil scientist and megalomaniac robot(s) come to the notice of the world, which nips their evil plotting in the bud.
The one thing the evil scientist and the megalomaniac robot are NOT going to do is persuade the millions of robots that are already working for us today to join their evil plan. You are completely safe from your vacuum cleaner, your thermostat, your refrigerator, and your car.
Humans have been telling stories about "artificial" beings that look and act like humans for thousands of years — zombies, golems, ghouls, Frankenstein's monster. Fear of being replaced or killed seems to drive all these stories. Hundreds of years ago, we were afraid of witches and zombies; today we're afraid of space aliens and insane, malignant robots.
They're just stories, entertainments for people who enjoy being creeped out. Don't be afraid of your singing greeting card; it is NOT plotting to take over the world and enslave you!