107159 research outputs found
Sort by
The levers of political persuasion with conversational artificial intelligence
Get PDFThere are widespread fears that conversational artificial intelligence (AI) could soon exert unprecedented influence over human beliefs. In this work, in three large-scale experiments (N = 76,977 participants), we deployed 19 large language models (LLMs)—including some post-trained explicitly for persuasion—to evaluate their persuasiveness on 707 political issues. We then checked the factual accuracy of 466,769 resulting LLM claims. We show that the persuasive power of current and near-future AI is likely to stem more from post-training and prompting methods—which boosted persuasiveness by as much as 51 and 27%, respectively—than from personalization or increasing model scale, which had smaller effects. We further show that these methods increased persuasion by exploiting LLMs’ ability to rapidly access and strategically deploy information and that, notably, where they increased AI persuasiveness, they also systematically decreased factual accuracy
User preferences for large language model refusals: implications for moderation and market structure
No full textLarge language models (LLMs) differ in their moderation and content policies, which determine which prompts these models refuse to answer. These refusals can affect user decisions of which models to use and whether to make safe or risky prompts. Using data from LMArena, where users select preferred responses to their prompts from paired LLM comparisons, we estimate a discrete choice model that captures user preferences for making risky prompts and their choice of which LLM provides the best response quality given the possibility of refusals. We leverage this model to analyze how moderation policies affect market shares across proprietary and opensource LLMs. Our findings reveal that proprietary LLMs provide higher quality responses and maintain larger market shares, but implement stricter moderation policies with higher refusal rates compared to open-source alternatives. This stricter moderation by proprietary LLMs reduces market concentration by allowing lower-quality open-source LLMs to compete effectively in the risky prompt segment. Mandating uniform moderation policies across all LLMs could increase market concentration favoring proprietary LLMs, potentially hampering competition. Our framework characterizes the efficient frontier of moderation policies that balance market concentration and safet
Global youth re-create symbols from past popular culture: discussion about the popularity of retro culture on TikTok
Get PDFIn the era of globalization, the recreation of retro cultural symbols by youth groups through social media platforms (especially TikTok) has become a remarkable phenomenon. Through focusing on the re-creation of retro culture, this paper explores how youth develop novel and unique cultural expressions by deconstructing, reorganizing and reinterpreting historical symbols. The paper also argues that this symbolic practice represents a form of cultural resistance through carnivalization and emotional deconstruction. Youth utilize visual, verbal, and behavioral symbols to construct a comprehensive system of expression. This system serves as a medium to convey their identity, emotional needs, and resistance to mainstream societal norms. The re-creation of retro culture is not only a kind of nostalgic behavior, but also an important way for youth to establish meaning, realize self-expression and emotional belonging in the digital space
Redundancy analysis using lcm-filtrations: networks, system signature and sensitivity evaluation
No full textWe introduce the lcm-filtration and stepwise filtration, comparing their performance across various scenarios in terms of computational complexity, efficiency, and redundancy. The lcm-filtration often involves identical steps or ideals, leading to unnecessary computations. To address this, we analyse how stepwise filtration can effectively compute only the non-identical steps, offering a more efficient approach. We compare these filtrations in applications to networks, system signatures, and sensitivity analysis
Nature degradation and the macroeconomy: insights from Latin America and the Caribbean
No full textAiming to examine the economic and financial dimensions of environmental degradation in Latin America and the Caribbean, policymakers, central banks, financial supervisors, academics and practitioners convened at the Nature and the Economy: Environmental Change, Economic Adjustment, and Policy Challenges conference in Mexico City on 2-3 October 2025. Elena Almeida and Rob Patalano detail three key insights that emerged from the discussions
Are debt sustainability frameworks compatible with climate and nature action?: findings from a new dataset of the IMF’s Debt Sustainability Analyses
Get PDFDebt sustainability analyses (DSAs) face numerous challenges. Highly technical methodologically, their outcomes bear enormous political significance as not only do they govern the operational lending of the World Bank-International Monetary Fund (IMF) grant-loan mix for low-income countries (LICs) but they also affect the risk perception of the private sector. In the context of heightened debt vulnerability, which a growing proportion of countries currently face, DSAs indicate the susceptibility of debt levels to various types of shocks and how different policy scenarios impact debt sustainability. Importantly, these assessments determine distributional impacts of the cost of a debt crisis – and whether and what size of debt restructuring is needed to bring a country’s debt back to sustainable levels. They provide the envelope for negotiations regarding the amount of debt relief the borrower will seek to negotiate with its creditors. The macro-critical impacts of climate change and nature loss represent key additional concerns on top of the longstanding challenges inherent to these exercises
Bridging the $1.3 trillion climate finance gap: a pragmatic agenda for emerging markets
No full textLuiz Awazu Pereira da Silva and Fernanda Gimenes propose a sequenced and pragmatic reform agenda to help close the US$1.3 trillion annual climate finance gap for emerging markets and developing economies by 2035. The agenda connects regulatory, institutional and technological reforms across the financial system, showing how each component can lower risk premiums, unlock investment and make climate finance both viable and scalable. It is an old debate in international macroeconomics: global savings do not naturally flow to emerging markets and developing economies (EMDEs), where returns to investment are higher and the need for capital is greater than in advanced economies. EMDEs are constrained by perceived risks, persistent home bias and deep structural asymmetries in the international financial system. This creates a double paradox that has evolved since the global financial crisis
Online risks and benefits – the 4C’s framework
No full textDigital technologies are increasingly embedded in all aspects of children’s lives from an early age, affording important opportunities for learning, participation, creativity, entertainment and belonging. Since today’s children and young people are often at the forefront of new media adoption, they are also exposed to a range of ever-evolving, risky or negative experiences for which they may be unprepared
How to strengthen UK business and public sector against cyber-threats
No full textCyber risk is increasing in the UK – costing companies and citizens billions. Notwithstanding robust efforts by the government, Alexander Evans argues another step is needed to bolster individual responsibility at company Board level