7.7. How Patterns in Human Behaviour Influence ML and AI Software Development

7.7.1. Patterns Related to Acute Stress Response

The influence of acute stress responses on AI software is a relatively niche topic, but it has important applications, particularly in fields like healthcare and emergency response. In these settings, AI tools can incorporate stress-monitoring technologies, such as heart-rate sensors, to detect when a user is under pressure. Stress can impair decision-making, reduce attention, and cause impulsive reactions, so it’s essential for AI systems in these environments to provide clear, calm, and easy-to-navigate interactions. Features like reassuring language, step-by-step instructions, and gradual information display can help reduce cognitive overload and support better outcomes. By understanding how people respond to stress, developers can design AI tools that not only function effectively but also assist users in making sound decisions during high-stakes situations.

7.7.2. Psychological Responses

Developers take into account how users think, feel, and respond in order to build systems that are not just functional, but also engaging and easy to use. AI tools are often intentionally designed to be user-friendly, offering friendly interfaces, helpful suggestions, and emotionally aware responses that make users feel comfortable and confident. For instance, if you ask ChatGPT, “Can I ask you a question?”, it will typically respond with something like, “Of course! What’s your question?” This positive, welcoming reply is designed to encourage further interaction. A negative response such as “No. Ask someone else.” would likely discourage engagement. These design choices are deliberate, aiming to build trust, promote ease of use, and keep users engaged. The supportive, personalised ways in which chatbots respond, can make users feel heard and understood, even though the interaction is with a machine. This plays on common psychological responses like anthropomorphism, where people naturally attribute human traits to non-human entities.

Additionally, tools like ChatGPT uses “thinking time” and also generates responses word by word, rather than displaying them all at once like a search engine. This mimics human typing, creating a more natural, conversational feel. It helps users feel like they’re interacting with a thoughtful partner, enhancing engagement and making the experience feel more genuine.

In some cases, AI tools are intentionally designed to foster strong emotional connections with users. This is especially true with AI companions, which may aim to build friendly or romantic relationships. Personal beliefs and cultural values strongly influence how such relationships are perceived. For example, people may differ in how they value emotional safety, privacy, or authenticity in digital interactions. Views on whether an AI can be a genuine companion, or what counts as intimacy, will vary widely. These beliefs shape how people respond to emotionally engaging AI tools and whether they accept or reject them as meaningful relationships.

Artificial intelligence and its ethics | DW Documentary 2020 (42 mins)