Why are we afraid of AI?

While most people know that artificial intelligence (AI) is here to stay, it scares us that intelligent machines are taking over more and more human domains. At the same time, AI is the driving force behind technologies like augmented reality (AR) that are far from scary. Knowing what’s going on under the hood will help you better understand these trends and give you the opportunity to become a true tech optimist.

In 1950, the famous British computer scientist Alan Turing wrote a paper called Computing Machinery and Intelligence, which essentially asked: Can machines think? To answer this question, he invented the so-called imitation game, which became famous as the Turing test. It involves a human judge who interacts with a machine and another human through text notes. The judge tries to determine which of the notes originated from the machine. If the machine can fool the human judge, it passes the test.

In 2018, Google took a similar approach. At their Google I/O developer conference, they showed a demo video of an AI technology called Duplex. At the time, Duplex was the latest version of Google’s voice assistant, like Amazon’s Alexa or Apple’s Siri. The difference was that in this demo, Duplex was able to call a hair salon and make an appointment with a real person over the phone, all by itself, without that person even realizing that they were talking to a computer.

The result of this presentation was a public outcry about the dangers and ethics of AI, and Google recently shut down parts of the project. How could this happen? The answer is simple. Duplex not only passed the Turing test by interacting with another human via text, but it also dared to imitate the human voice in a way that could fool real humans.

Human, all too human

Since then, AI technology has certainly evolved with the likes of ChatGPT, Bard, and other new text-based AI tools. These technologies are amazing in what they can do, but they’re not pretending to be human or have human characteristics.

Rather, they present themselves as useful tools that are powerful but not human-like. No one is currently trying to stage another Turing test. Why not? The results would most likely irritate people once again and do more harm than good for the advancement of the technology. For example, Italy’s national privacy watchdog has already ordered an effective ban on the AI chatbot ChatGPT. But why are people so afraid? Let’s take a step back and look under the hood at what AI really is.

How to teach a machine

Although artificial intelligence is often thought of as a system in itself, it’s actually a set of technologies that enable a system to reason, learn, and solve complex problems. One such technology is machine learning (ML). While AI encompasses the idea of a machine that can mimic human intelligence, ML doesn’t. The goal of ML is to teach a machine how to perform a specific task and provide accurate results by identifying patterns.

ML automatically enables a machine or system to learn and improve from experience. Instead of explicit programming, machine learning uses algorithms to analyze large amounts of data, learn from the insights, and then make informed decisions. The performance of ML algorithms improves over time as they are continuously trained — i.e., exposed to more data.

For example, you can train algorithms to analyze a two-dimensional camera image to create a three-dimensional model of the scene based on data in the image, such as depth and color information, as well as camera movements. This three-dimensional model acts as an invisible grid on which augmented reality objects such as arrows, lines, or other information, can be placed as if they were actually sitting on the real objects captured by the camera lens. This is how machine learning enables augmented reality solutions.

AI is a car, machine learning is the engine

Machine learning and its applications sound a lot less scary than the dystopian image of human-like artificial intelligence, right? That’s not because the concept of ML is less powerful than that of AI. You could say that if AI is a car, then ML is its engine. And although only few people are emotionally attached to engines (except for the good old V8), many people have strong opinions about cars in general, car brands, car design, or whether cars should exist at all in a modern society facing the threat of climate change.

And while it’s possible to imagine a society without cars, it’s quite impossible to imagine a society without engines or motors that power all sorts of things we need every day, like trains, elevators, automatic doors, machines, etc. The same goes for AI and ML: if these technologies are used in a way that seems useful to us, such as in chat bots, games, search engines, navigation software, writing tools, smart glasses, image and video editing, or in professional augmented reality solutions, then our acceptance is almost a given.

Augmented reality is fueled by AI

Speaking of augmented reality (AR), this technology is so successful in recent years because AI is integrated into its core functionality. AR, as experienced through smart glasses and mobile devices, is fueled by AI and ML. These technologies work in harmony to analyze data from hundreds of sensors, bridging the digital and physical worlds.

Smart glasses and mobile devices are equipped with various sensors that collect data about our surroundings. AI takes this raw sensor data and transforms it into a digital representation of the environment, a process often referred to as mapping. This mapping is the foundation that allows AR annotations to be linked to the real world. It’s what makes digital objects appear seamlessly integrated into your physical environment.

One of AI’s standout capabilities is its ability to identify patterns in vast amounts of data. In an industrial setting, this means that AI can help analyze worker behavior and activities. For example, shop floor workers often spend unnecessary time searching for specific items. AI can track these patterns and suggest more efficient item locations by considering the entire warehouse stock, reducing the need for trial and error.

AI friend, not AI foe

By now, you should no longer be afraid of AI. As always, it’s the way technologies are communicated and marketed to the public that makes them seem desirable or scary to us. Understanding what’s going on underneath the flashy surface will help you evaluate technologies the right way — and maybe even help you become a tech optimist.

Most new technologies have been met with skepticism in the past. Even electricity and X-rays. The next time you’re skeptical about a new kind of technology without knowing much about it, ask yourself: Would you want your doctor to use technology from 200 years ago?