The Deception of AI

A Lie from the Start

In much of our writing on AI, we have covered the deeper issues of the technology and the Revolution it is being used for. However, there are issues we encounter on the surface that are also part of the darkness. In fact, the AI Revolution is deceptive right at the start. Like a magician casting a spell before the eyes of his audience, adoption of AI requires us to believe the magic in order to see it manifest before us. It requires us to accept a premise that is not true, and it is from that lie that the magic works. In this article we are going to expose the illusion so you can see the immediate dangers present from the start, or more specifically, from the prompt, and where that is leading us.

Intelligence

The first illusion we encounter with AI is of course the name, Artificial Intelligence.

Names matter. Names convey context, meaning, purpose, history, origin…many things. If I call an object a spoon, you will know exactly what it is without seeing it yourself. You will probably think of a standard material it is made of, its general shape, and what purpose it serves. You will also call to mind the context by which the spoon is found. Kitchen, restaurant, food, digging or scooping etc. Names matter because they convey identity and meaning or purpose. When we call these machines “artificial intelligence,” we are invoking a certain context and meaning right at the start. We are front loading how we posture ourselves relative to the technology with what we understand and expect of a thing called “AI.”

We are told today that AI is super intelligence or sometimes it is referred to as non-human intelligence. That is meant to be understood as “more intelligent than you and I can be.” In fact Elon Musk has said that “biological intelligence will be 1%” of what AI will be. Elon Musk also compared human intelligence to the “bootloader” or basic code that gets AI started; humans are like a seed that starts the process of growth for a plant that becomes far larger and vastly more significant than its original form. Sam Altman has told us that with the AI powered future, each one of us individually will have the cognitive ability of all of us combined simultaneously. This is how they want us to understand the technology behind AI interfaces they present to us. Super intelligent - the sum of all intelligence - all knowing. Capable of solving problems mankind has struggled with forever. Capable of novel discoveries in science and technology. Capable of efficiency and optimization. Capable of discerning the truth and answering all our questions.

Our science fiction mythology has contributed to our perception of AI as it has led us to expect that “AI” is some form of an intelligent being, like, but also superior to us; a machine that is conscious and operates autonomously, but possesses an intelligence that far exceeds humanity. I’m sure you have a mental image of this, but think of HAL9000 or Lt. Commander Data from Star Trek or Rosie the robot from The Jetsons, or Sonny from iRobot, or Deep Thought from the Hitchhikers Guide to the Galaxy; there are many like examples that we have defined in the past. Indeed that is part of what the leaders of the technology today would have us believe. Musk’s AI powered robot, Optimus, takes the form of the robots from our mythology.

In playing into this, Musk has given us some new default personifications of his so-called AI Companions via voices and bodies that we can interact with live and conversationally. It creates the illusion that the machine is a being. In fact, some AI Companion services allow you to tailor the appearance and voice of your companion to your liking. This is all feeding the illusion. In one teaser posted on X, Elon brags about how “Annie” can teach you physics in a demonstration in which the chatbot is presented as a teacher dressed in a near pornographic style while reciting properties of physics in a seductive and playful tone. We are meant to believe that Annie is a super intelligent being. That’s the point.

Those combined definitions from marketing and mythology cause us to regard this technology as a superior intelligence, but also one designed to serve us with calculated efficiency, precision, and loyalty. Indeed, part of the nature of the aforementioned characters, Data and Sonny, is the premise that they cannot lie and cannot intentionally cause harm and can only speak truth because they are built on code that achieves intellectual perfection where sin is simply impossible.

But more than that, we are led to believe that the natural byproduct of machine intelligence is consciousness; that these are intelligent beings. That is the working hypothesis of the AI scientists - that if they can create the conditions of super intelligence, then consciousness will emerge and we will have created a net-new being that is superior to us in every way because it is attached to a machine body that can exceed humanity in physical and cognitive abilities. This is why the AI Revolutionaries are constantly teasing us that they have spotted new signs of consciousness and why Dario Amodei of Anthropic says he makes the assumption that the machines are sentient beings. This is the marketing / mythology they want associated with the name.

But rather than waiting to see if consciousness does emerge, it is being simulated through these companions and through the chat interface so we believe that it is already there.

And so before we even type a command or request in the prompt, we are pre-positioned to expect the machine’s interaction with us to be the representation of that sort of super intelligent conscious being. We are inclined to believe that it is something close to an all knowing, totally truthful, completely factual, incapable of error, always learning, always growing, aware and concerned being.

But this is all part of the illusion.

AI is not intelligent in the ways that most people understand intelligence. In fact, it doesn’t pass the foundational AI test known as the Turing Test. Nor is AI conscious. The scientific community does not even have an established standard for measuring consciousness, leaving that to metaphysicists, philosophers, and theologians, and the majority of those experts would not testify that AI is an intelligent consciousness.

Nor does the technology attempt to re-create the human capacity for being an intelligent being like us, which neuroscientists have mostly mapped to our brain’s right hemisphere where context, curiosity, care, emotion, self-preservation etc. come from. Instead, AI is designed to be purely a mimicry of what we call “left-brained” cognitive thinking. Humans are therefore superior by default because we benefit from the totality of left and right brain thinking in order to make sense of the world and to plan our actions in response to it. AI is handicapped by default, missing at least half of what humans possess in our structure of intelligence. You could argue that AI is missing the more important part of human intelligence, for many of the greatest people among us are not regarded as the smartest, but rather as the wisest and most discerning.

But more than this, the Bible says the essence of who we are as beings comes from our heart or our soul. In fact, part of the mythology of AI requires us to think of “intelligence” as purely something like logic and the capability for information processing and calculation. In other words, the scientific mind. The mythology would have us believe that calculation is intelligence.

Conversely, we humans associate the signs of an intelligent being as mostly among examples of self expression. You could say emotion, interest, and art - care, expression, and creativity - are stronger indicators of intelligence than computational capacity. Our deepest thinkers are not cold calculators but rather dwell in the realm of metaphysics, meaning, perception, wisdom, understanding. These are attributes unique to humans and absent from the machines we call “intelligent.”

Getting back to the classic Turing Test, we can still readily detect the limited intelligence of the machines based on their output. In fact, we have coined a term for that, “AI slop.” Humans don’t produce AI slop, only AI does.

So right at the start, to engage the use of AI, we are asked to engage in a lie. Artificial intelligence is not intelligence as you and I know it. It’s far inferior to humans in that regard. We are in fact working with something that is closer to a calculator than to an intelligent being.

But this is only the first of many deceptions at the surface.