AI users prioritize time savings over productivity

As artificial intelligence becomes part of everyday life, expectations are becoming clearer. Rather than seeking ever greater output, many people want to use AI to make daily life easier.

Anthropic, the developer of the Claude large language model, recently released findings from its “80,000 Interviews” project, which sheds light on how people actually experience AI beyond broad debates over artificial general intelligence and automation. Conducted over one week in December last year, the study surveyed 80,508 people in 159 countries across 70 languages. Its method set it apart. Instead of using standard multiple-choice questions, an AI interviewer based on Claude held one-on-one, in-depth conversations and adjusted follow-up questions in real time based on each response. The approach was described as an early effort to solve the long-running tradeoff between scale and depth in qualitative research.

“I used to receive 100 to 150 messages a day from doctors and nurses. I was so tied up with documentation that I barely had time to speak with patients’ families. After adopting AI, that burden disappeared.” That account from a U.S. health care worker reflects a broader pattern. When asked about AI’s role, the most common response was improved workplace productivity, at 18.8%. But as conversations went deeper, a more practical priority emerged. Rather than seeking higher performance, many respondents said they wanted to cut routine tasks such as checking email or handling taxes so they could spend more time with family. That desire to ease everyday burdens accounted for 14.0% of responses.

Fears of a distant future in which machines dominate humanity seemed less immediate for most users. When asked about concerns, respondents most often cited reliability problems, including so-called hallucinations, at 26.7%. That was followed by worries about job losses as outsourced work is replaced by AI, at 22.3%. Concerns about the erosion of human agency, including the possibility of losing the ability to read and think independently, also ranked high at 21.9%.

Perceptions of AI varied widely by region. In emerging economies across Africa, South America and Southeast Asia, respondents tended to see AI as a tool for economic advancement. In contrast, people in advanced economies such as the United States and the European Union were more likely to worry about data privacy, surveillance and gaps in governance.

The findings send a clear message to the AI industry. What people ultimately want is not more output, but a more manageable daily life. Beyond building faster and more capable systems, the tools most likely to win lasting support may be those that make everyday life simpler and more sustainable.


Jae-Hyeng Kim monami@donga.com