Long-Term Trends in the Public Perception of Artificial Intelligence (1609.04904v2)

Abstract: Analyses of text corpora over time can reveal trends in beliefs, interest, and sentiment about a topic. We focus on views expressed about AI in the New York Times over a 30-year period. General interest, awareness, and discussion about AI has waxed and waned since the field was founded in 1956. We present a set of measures that captures levels of engagement, measures of pessimism and optimism, the prevalence of specific hopes and concerns, and topics that are linked to discussions about AI over decades. We find that discussion of AI has increased sharply since 2009, and that these discussions have been consistently more optimistic than pessimistic. However, when we examine specific concerns, we find that worries of loss of control of AI, ethical concerns for AI, and the negative impact of AI on work have grown in recent years. We also find that hopes for AI in healthcare and education have increased over time.

Analyzing Long-Term Trends in Public Perception of Artificial Intelligence

The paper "Long-Term Trends in the Public Perception of Artificial Intelligence," authored by Ethan Fast and Eric Horvitz, provides a comprehensive analysis of public sentiment toward AI as reported by the New York Times over a thirty-year period. Utilizing a combination of crowdsourcing and natural language processing, the paper identifies themes in AI-related discussions by analyzing how the portrayal of hopes and concerns related to AI has evolved from 1986 to 2016.

Methodology and Indicators

The paper employs a rigorous methodological framework to capture public sentiment and levels of engagement with AI. Through the lens of New York Times articles, it develops a set of impression indicators to parse general sentiment (optimism vs. pessimism) and specific hopes and concerns associated with AI. The research taps into crowdsourcing via Amazon Mechanical Turk to annotate excerpts for indicators such as optimism, pessimism, and AI-related hopes (e.g., AI advancements in healthcare) and concerns (e.g., loss of control over AI systems). The use of a 5-point Likert scale for annotation allows for a quantifiable measure of sentiment, enabling the researchers to establish trends over an extended period.

Key Findings

1. Rising Engagement: The longitudinal data reveal that discussions around AI have significantly increased since 2009, indicative of heightened public interest in AI technologies. This upsurge coincides with the achievements in deep learning and AI technologies becoming more mainstream.
2. Trends in Optimism and Pessimism: Analysis shows that optimism about AI has consistently outweighed pessimism over the decades. However, the gap between public optimism and pessimism appears to be closing in recent years, partly due to growing concerns about AI's societal impact.
3. Evolution of Associated Ideas: The paper systematically outlines how the themes linked to AI have transformed over the three decades. For instance, early associations with military applications have shifted towards modern discussions involving healthcare and driverless vehicles. This shift underscores the broadening application of AI technologies.
4. Emerging Concerns and Hopes: While public sentiment towards AI remains largely optimistic, specific concerns like loss of control and ethical dilemmas have seen a marked increase. Similarly, the expectation that AI will positively influence sectors like healthcare and education has grown, highlighting potential areas of societal benefit and anxiety.

Implications and Future Developments

The evidence from the New York Times can offer policymakers and technologists nuanced insights into the evolving public perception of AI, potentially guiding regulatory decisions and informing public dialogue. The paper's framework, leveraging public discourse data, provides a scalable model that could be applied to other corpora or adapted for more current datasets. This ongoing monitoring of AI sentiment could be instrumental in understanding societal readiness for AI advancements and in identifying the cultural shifts influenced by rapidly developing technologies.

Conclusion

The research by Fast and Horvitz maps a detailed picture of how public perceptions of AI have shifted over the decades. As AI continues to intersect various facets of life, the importance of understanding public sentiment becomes ever more critical. This paper not only provides a historical perspective on AI perception but also offers a methodological template for analyzing sentiment on emerging technologies across media channels. Future research may expand this approach to more global datasets, offering a broader understanding of cross-cultural perceptions of AI and technology at large.