As AI becomes more integral to our daily lives and is expected to make significant changes, there are mixed opinions worldwide about embracing it.

To fully tap into the advantages of AI, it’s crucial to grasp the psychological barriers hindering its widespread adoption. Harvard research delves into the conflicting attitudes people have toward a wide range of AI tools, including robots, large language models like ChatGPT, self-driving cars, speech recognition programs, virtual assistants, and recommendation systems.

Sources of resistance

The researchers identified five main ways we don’t like adopting new technologies:

Barrier #1: AI Seems Opaque

Most of us don’t really know how AI algorithms work—it’s like a mysterious “black box” that can make us uneasy and skeptical. This lack of understanding may lead us to stick with human intelligence even when AI is better suited for a task. Overcoming this resistance often involves providing clear but not overly simplified explanations of how AI operates. The key is to show that AI is complex enough to handle its job.

Barrier #2: AI Seems Unfeeling

AI is seen as being made of cold circuits, not warm human emotions. People often underestimate AI’s abilities in tasks that seem to require feelings, like writing music or giving relationship advice. To overcome this resistance, some strategies include making AI seem more human-like by giving it faces, names, or genders. Another approach is to describe emotional tasks in more objective terms, highlighting the data-driven basis for AI-provided advice.

Barrier #3: AI Seems Rigid

Many think of AI as inflexible and incapable of learning, based on outdated views of non-adaptive AI. However, modern AI can learn, adapt, and tailor responses to individual needs. Demonstrating AI’s ability to learn and adapt can reduce resistance. Using terms like “machine learning” instead of “algorithm” can also make AI more approachable.

Barrier #4: AI Seems Too Autonomous

AI’s independence can clash with our desire for control and safety. Movies depicting rogue AIs taking over the world amplify this fear. Overcoming this concern involves ensuring that people have some level of control over AI tools. Designing AI to be predictable and supportive rather than fully autonomous can also ease fears without sacrificing quality.

Barrier #5: AI Seems Non-Human

Our discomfort with AI may stem from its non-human nature. Some people may devalue AI because it doesn’t belong to our species, even if it looks humanoid. Overcoming this resistance is challenging, but encouraging the perception of human-like consciousness in AI may help foster a more positive engagement, especially in cultures that embrace the idea of inanimate objects having a spirit or soul.