Anthropic Reveals Global Expectations and Fears Regarding AI

A Global Study on Expectations for AI

Anthropic recently published a study titled "What 81,000 people want from AI," aimed at exploring global perceptions of artificial intelligence. This large-scale survey was conducted with over 80,000 people from 159 countries, speaking 70 different languages. The interviews were carried out via a conversational agent, and the responses were analyzed to identify global trends.

The results show that the primary expectation of users is the improvement of professional performance, with 18.8% of respondents expressing this desire. This priority validates Anthropic's strategy, which primarily targets professionals and businesses. However, personal expectations are also significant: 13.7% of participants hope for personal transformation through AI, and 13.5% wish for better organization in their daily lives. Additionally, 11.1% of respondents hope to gain more free time, while 9.7% aim for financial independence. Finally, 9.4% of respondents are optimistic about the positive impact of AI on society.

The Concrete Contributions of AI According to Users

Anthropic's study also examined the benefits already perceived from AI. Nearly one-third of respondents acknowledge an improvement in their productivity thanks to AI, an opinion particularly shared by developers. Furthermore, 17.2% of participants view AI as a cognitive partner, and 8.7% appreciate the technical accessibility it offers. However, 6.1% of respondents mention AI's contribution to emotional support.

The main benefits identified are:

• Productivity: 32.0%
• AI has not lived up to its promises: 18.9%
• Cognitive partnership: 17.2%
• Learning: 9.9%
• Technical accessibility: 8.7%

Concerns Related to AI

The study also highlights users' concerns. The lack of reliability of AI is the primary worry, cited by 26.7% of respondents. Potential effects on employment and the economy concern 22.3% of participants, followed by worries about governance (14.7%) and misinformation (13.6%).

The ten most frequently cited concerns are:

• Lack of reliability: 26.7%
• Employment and economy: 22.3%
• Autonomy and agency: 21.9%
• Cognitive atrophy: 16.3%
• Governance: 14.7%
• Misinformation: 13.6%
• Surveillance and privacy: 13.1%
• Malicious use: 13.0%
• Meaning and creativity: 11.7%
• Excessive restriction: 11.7%

Regional Variations in Perception of AI

Anthropic's study reveals a global optimism towards AI, with 67% of respondents expressing a positive sentiment. However, notable differences appear by region. In low- or middle-income countries, optimism reaches 70% to 76%, while it ranges from 63% to 67% in Europe and North America.

These disparities can be explained by economic concerns. In wealthy regions, fears about employment and the economy are more pronounced, negatively influencing the overall sentiment. Conversely, in emerging regions, AI is seen as a source of opportunities, particularly for entrepreneurship, cited by 16% of respondents in Sub-Saharan Africa compared to an average of 8.7% globally. Moreover, 17% to 18% of respondents in certain regions report having no concerns, nearly double that of Western countries.

According to Anthropic, "concerns related to employment and the economy are the most significant predictive factor of the overall sentiment towards AI."