Krzysztof Węcel, Marcin Sawiński, Milena Stróżyna, Włodzimierz Lewoniewski, Ewelina Księżniak, P. Stolarski, W. Abramowicz
Abstract In this paper the impact of large language models (LLM) on the fake news phenomenon is analysed. On the one hand decent text‐generation capabilities can be misused for mass fake news production. On the other, LLMs trained on huge volumes of text have already accumulated information on many facts thus one may assume they could be used for fact‐checking. Experiments were designed and conducted to verify how much LLM responses are aligned with actual fact‐checking verdicts. The research methodology consists of an experimental dataset preparation and a protocol for interacting with ChatGPT, currently the most sophisticated LLM. A research corpus was explicitly composed for the purpose of this work consisting of several thousand claims randomly selected from claim reviews published by fact‐ checkers. Findings include: it is difficult to align the respons‐ es of ChatGPT with explanations provided by fact‐checkers; prompts have significant impact on the bias of responses. ChatGPT at the current state can be used as a support in fact‐checking but cannot verify claims directly.
{"title":"Artificial intelligence—friend or foe in fake news campaigns","authors":"Krzysztof Węcel, Marcin Sawiński, Milena Stróżyna, Włodzimierz Lewoniewski, Ewelina Księżniak, P. Stolarski, W. Abramowicz","doi":"10.18559/ebr.2023.2.736","DOIUrl":"https://doi.org/10.18559/ebr.2023.2.736","url":null,"abstract":"Abstract In this paper the impact of large language models (LLM) on the fake news phenomenon is analysed. On the one hand decent text‐generation capabilities can be misused for mass fake news production. On the other, LLMs trained on huge volumes of text have already accumulated information on many facts thus one may assume they could be used for fact‐checking. Experiments were designed and conducted to verify how much LLM responses are aligned with actual fact‐checking verdicts. The research methodology consists of an experimental dataset preparation and a protocol for interacting with ChatGPT, currently the most sophisticated LLM. A research corpus was explicitly composed for the purpose of this work consisting of several thousand claims randomly selected from claim reviews published by fact‐ checkers. Findings include: it is difficult to align the respons‐ es of ChatGPT with explanations provided by fact‐checkers; prompts have significant impact on the bias of responses. ChatGPT at the current state can be used as a support in fact‐checking but cannot verify claims directly.","PeriodicalId":41557,"journal":{"name":"Economics and Business Review","volume":null,"pages":null},"PeriodicalIF":0.7,"publicationDate":"2023-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78960462","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Introduction to the thematic issue on digitalisation, big data, and artificial intelligence","authors":"H. Brezinski, W. Jurek","doi":"10.18559/ebr.2023.2.735","DOIUrl":"https://doi.org/10.18559/ebr.2023.2.735","url":null,"abstract":"","PeriodicalId":41557,"journal":{"name":"Economics and Business Review","volume":null,"pages":null},"PeriodicalIF":0.7,"publicationDate":"2023-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83636125","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abstract The article has three aims. The first aim is to develop an improved version of the Keynes-Metzler-Goodwin (the KMG) monetary growth model originally presented and analysed in a series of publications by Carl Chiarella, Peter Flaschel and Willi Semler. The improvement of the model is obtained by modifying some of its equations in a way which ensures that they reflect real macroeconomic dependencies more properly. The equations that have been modified describe final demand expectations, determinants of production decisions, fixed capital accumulation, tax revenues, government budget deficit and money demand. The second aim is to transform the model into an intensive form described by seven non-linear differential equations and determine its unique steady state which shows proportions between variables on the balanced growth path. The third ultimate aim is to present a mathematical proof that the new improved version of the KMG model is locally asymptotically stable.
{"title":"On the stability of a certain Keynes-Metzler-Goodwin monetary growth model","authors":"Damian Sołtysiak","doi":"10.18559/ebr.2023.1.2","DOIUrl":"https://doi.org/10.18559/ebr.2023.1.2","url":null,"abstract":"Abstract The article has three aims. The first aim is to develop an improved version of the Keynes-Metzler-Goodwin (the KMG) monetary growth model originally presented and analysed in a series of publications by Carl Chiarella, Peter Flaschel and Willi Semler. The improvement of the model is obtained by modifying some of its equations in a way which ensures that they reflect real macroeconomic dependencies more properly. The equations that have been modified describe final demand expectations, determinants of production decisions, fixed capital accumulation, tax revenues, government budget deficit and money demand. The second aim is to transform the model into an intensive form described by seven non-linear differential equations and determine its unique steady state which shows proportions between variables on the balanced growth path. The third ultimate aim is to present a mathematical proof that the new improved version of the KMG model is locally asymptotically stable.","PeriodicalId":41557,"journal":{"name":"Economics and Business Review","volume":null,"pages":null},"PeriodicalIF":0.7,"publicationDate":"2023-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73072144","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abstract The widespread use of digital technologies in banking allows banks to obtain and analyse huge amounts of data from different communication channels. While this phenomenon is conducive to improving the quality of services it also increases the risk of privacy breaches. The aim of this study is to identify what factors determine consumer acceptance of banks’ use of public access personal data found on social media accounts. The results indicate the importance of the financial incentive and consumers’ assessment of banks’ information activities regarding the processing of personal data. Determinants relating to the technological sophistication of respondents were also found to be significant, with a particular focus on the ethical evaluation of decisions made by Artificial Intelligence algorithms. The results of the work may be used by banks in practice to adapt the area of personal data management to the requirements of e-privacy and Trustworthy Artificial Intelligence.
{"title":"Privacy frontiers in customers’ relations with banks","authors":"D. Piotrowski","doi":"10.18559/ebr.2023.1.5","DOIUrl":"https://doi.org/10.18559/ebr.2023.1.5","url":null,"abstract":"Abstract The widespread use of digital technologies in banking allows banks to obtain and analyse huge amounts of data from different communication channels. While this phenomenon is conducive to improving the quality of services it also increases the risk of privacy breaches. The aim of this study is to identify what factors determine consumer acceptance of banks’ use of public access personal data found on social media accounts. The results indicate the importance of the financial incentive and consumers’ assessment of banks’ information activities regarding the processing of personal data. Determinants relating to the technological sophistication of respondents were also found to be significant, with a particular focus on the ethical evaluation of decisions made by Artificial Intelligence algorithms. The results of the work may be used by banks in practice to adapt the area of personal data management to the requirements of e-privacy and Trustworthy Artificial Intelligence.","PeriodicalId":41557,"journal":{"name":"Economics and Business Review","volume":null,"pages":null},"PeriodicalIF":0.7,"publicationDate":"2023-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87479749","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abstract Accidentally parallel at the beginning, the transition to value-based pricing and transition to pricing data science have blended harmoniously, changing the pricing landscape. Using the marketing capability approach, I show that the introduction of pricing data science is costly and requires higher management support. Despite its cost, algorithmic price optimisation allows one to react swiftly to changes in demand. The optimisation process is applied to inherently non-linear, multimodal, and right-skewed pricing data. Presenting the interactions between new computational techniques and value-data pricing, I concentrate on altered perceptions of price elasticity, value-driver estimations, and contract opportunity analysis.
{"title":"Pricing and data science: The tale of two accidentally parallel transitions","authors":"J. Wallusch","doi":"10.18559/ebr.2023.2.739","DOIUrl":"https://doi.org/10.18559/ebr.2023.2.739","url":null,"abstract":"Abstract Accidentally parallel at the beginning, the transition to value-based pricing and transition to pricing data science have blended harmoniously, changing the pricing landscape. Using the marketing capability approach, I show that the introduction of pricing data science is costly and requires higher management support. Despite its cost, algorithmic price optimisation allows one to react swiftly to changes in demand. The optimisation process is applied to inherently non-linear, multimodal, and right-skewed pricing data. Presenting the interactions between new computational techniques and value-data pricing, I concentrate on altered perceptions of price elasticity, value-driver estimations, and contract opportunity analysis.","PeriodicalId":41557,"journal":{"name":"Economics and Business Review","volume":null,"pages":null},"PeriodicalIF":0.7,"publicationDate":"2023-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85518334","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abstract Over the last years the use of big data became increasingly relevant also for macroeconomic topics and specifically the conduct and analysis of monetary policy. The aim of this paper is to provide a survey of these applications and the relevant methods. The rationale for doing so is twofold. First, there is no straightforward definition of “big data”. Since macroeconomics and monetary policy analysis has a long tradition in quite sophisticated and data-intensive empirical applications the nature of the innovation big data is indeed bringing to the field is reflected upon. Second, concerning statistical / empirical methods the analysis of big data necessitates the use of different tools relative to traditional empirical macroeconomics which are in some cases a complement to more traditional methods. Hence big data in monetary policy is not just the application of well-established methods to larger data sets.
{"title":"Big data in monetary policy analysis—a critical assessment","authors":"Alexandra Bogner, Jürgen Jerger","doi":"10.18559/ebr.2023.2.733","DOIUrl":"https://doi.org/10.18559/ebr.2023.2.733","url":null,"abstract":"Abstract Over the last years the use of big data became increasingly relevant also for macroeconomic topics and specifically the conduct and analysis of monetary policy. The aim of this paper is to provide a survey of these applications and the relevant methods. The rationale for doing so is twofold. First, there is no straightforward definition of “big data”. Since macroeconomics and monetary policy analysis has a long tradition in quite sophisticated and data-intensive empirical applications the nature of the innovation big data is indeed bringing to the field is reflected upon. Second, concerning statistical / empirical methods the analysis of big data necessitates the use of different tools relative to traditional empirical macroeconomics which are in some cases a complement to more traditional methods. Hence big data in monetary policy is not just the application of well-established methods to larger data sets.","PeriodicalId":41557,"journal":{"name":"Economics and Business Review","volume":null,"pages":null},"PeriodicalIF":0.7,"publicationDate":"2023-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88364857","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abstract The aim of this article is to critically review and evaluate two ESG-based investment strategies—divestment and engagement for alignment of investment portfolios with climate change mitigation goals of the United Nations. The article compares both approaches in terms of their effectiveness of decarbonization, using the case study method. First, the case on fossil fuels divestment by Harvard Management Company is analysed. The second case study discusses shareholder engagement endeavors by Engine No. 1 hedge fund and its investment in ExxonMobil. The findings indicate that divestment may have non-immediate impact on corporate behavior and carries political and legal retribution risks. Engagement, on the other hand, presents itself as a more plausible option as it takes less time to deploy and, therefore, can produce more immediate and impactful results. Nevertheless, both divestment and engagement can play mutually supportive roles in addressing climate change by the investment industry.
{"title":"Divest or engage? Effective paths to net zero from the U.S. perspective","authors":"Andrew Buks, Konrad Sobański","doi":"10.18559/ebr.2023.1.3","DOIUrl":"https://doi.org/10.18559/ebr.2023.1.3","url":null,"abstract":"Abstract The aim of this article is to critically review and evaluate two ESG-based investment strategies—divestment and engagement for alignment of investment portfolios with climate change mitigation goals of the United Nations. The article compares both approaches in terms of their effectiveness of decarbonization, using the case study method. First, the case on fossil fuels divestment by Harvard Management Company is analysed. The second case study discusses shareholder engagement endeavors by Engine No. 1 hedge fund and its investment in ExxonMobil. The findings indicate that divestment may have non-immediate impact on corporate behavior and carries political and legal retribution risks. Engagement, on the other hand, presents itself as a more plausible option as it takes less time to deploy and, therefore, can produce more immediate and impactful results. Nevertheless, both divestment and engagement can play mutually supportive roles in addressing climate change by the investment industry.","PeriodicalId":41557,"journal":{"name":"Economics and Business Review","volume":null,"pages":null},"PeriodicalIF":0.7,"publicationDate":"2023-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74565257","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abstract The aim of the paper is to analyse the likely implications of Generative AI (GAI) on various aspects of business and the economy. Amid the rapid growth and maturing of Generative AI technologies such as Large Language Models (like ChatGPT by OpenAI) a rapid growth of both immediate and potential applications can be seen. The implications for the economy and industries of this technological shift will be discussed. The foreseeable scenarios for the level and types of adoption that GAI might achieve—from useful analytical tool, invaluable assistant to the white-collar workers of the world to being trusted with a wide array of business and life-critical decision making. Both disruptive and premium service opportunities are foreseen. For instance, general purpose models may provide quality service—such as copywriting—to overserved customers leaving human writers as the premium option. In this context, overserved customers would be those who would be satisfied with a non-human, potentially less creative content. On the other hand highly specialized models—specifically trained in a given domain and with access to proprietary knowledge can possibly provide a premium service over that provided by human experts. It is expected that some jobs will be replaced by new AI applications. However, new workplaces will emerge. Not only the obvious expert-level data scientist roles but also low grade, “model supervisors”—people training the models, assessing the quality of responses given and handling escalations. Lastly new cybercrime risks emerging from the rise of GAI are discussed.
{"title":"The rise of Generative AI and possible effects on the economy","authors":"T. Orchard, Leszek Tasiemski","doi":"10.18559/ebr.2023.2.732","DOIUrl":"https://doi.org/10.18559/ebr.2023.2.732","url":null,"abstract":"Abstract The aim of the paper is to analyse the likely implications of Generative AI (GAI) on various aspects of business and the economy. Amid the rapid growth and maturing of Generative AI technologies such as Large Language Models (like ChatGPT by OpenAI) a rapid growth of both immediate and potential applications can be seen. The implications for the economy and industries of this technological shift will be discussed. The foreseeable scenarios for the level and types of adoption that GAI might achieve—from useful analytical tool, invaluable assistant to the white-collar workers of the world to being trusted with a wide array of business and life-critical decision making. Both disruptive and premium service opportunities are foreseen. For instance, general purpose models may provide quality service—such as copywriting—to overserved customers leaving human writers as the premium option. In this context, overserved customers would be those who would be satisfied with a non-human, potentially less creative content. On the other hand highly specialized models—specifically trained in a given domain and with access to proprietary knowledge can possibly provide a premium service over that provided by human experts. It is expected that some jobs will be replaced by new AI applications. However, new workplaces will emerge. Not only the obvious expert-level data scientist roles but also low grade, “model supervisors”—people training the models, assessing the quality of responses given and handling escalations. Lastly new cybercrime risks emerging from the rise of GAI are discussed.","PeriodicalId":41557,"journal":{"name":"Economics and Business Review","volume":null,"pages":null},"PeriodicalIF":0.7,"publicationDate":"2023-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86926805","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abstract This paper introduces and examines a novel realized volatility forecasting model that makes use of Long Short-Term Memory (LSTM) neural networks and the risk metric financial turbulence (FT). The proposed model is compared to five alternative models, of which two incorporate LSTM neural networks and the remaining three include GARCH(1,1), EGARCH(1,1), and HAR models. The results of this paper demonstrate that the proposed model yields statistically significantly more accurate and robust forecasts than all other studied models when applied to stocks with middle-to-high volatility. Yet, considering low-volatility stocks, it can only be confidently affirmed that the proposed model yields statistically significantly more robust forecasts relative to all other models considered.
{"title":"Forecasting realized volatility through financial turbulence and neural networks","authors":"Hugo Gobato Souto, A. Moradi","doi":"10.18559/ebr.2023.2.737","DOIUrl":"https://doi.org/10.18559/ebr.2023.2.737","url":null,"abstract":"Abstract This paper introduces and examines a novel realized volatility forecasting model that makes use of Long Short-Term Memory (LSTM) neural networks and the risk metric financial turbulence (FT). The proposed model is compared to five alternative models, of which two incorporate LSTM neural networks and the remaining three include GARCH(1,1), EGARCH(1,1), and HAR models. The results of this paper demonstrate that the proposed model yields statistically significantly more accurate and robust forecasts than all other studied models when applied to stocks with middle-to-high volatility. Yet, considering low-volatility stocks, it can only be confidently affirmed that the proposed model yields statistically significantly more robust forecasts relative to all other models considered.","PeriodicalId":41557,"journal":{"name":"Economics and Business Review","volume":null,"pages":null},"PeriodicalIF":0.7,"publicationDate":"2023-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76192578","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abstract The aim of this paper is to discuss the role and impact of Generative Artificial Intelligence (AI) systems in higher education. The proliferation of AI models such as GPT-4, Open Assistant and DALL-E presents a paradigm shift in information acquisition and learning. This transformation poses substantial challenges for traditional teaching approaches and the role of educators. The paper explores the advantages and potential threats of using Generative AI in education and necessary changes in curricula. It further discusses the need to foster digital literacy and the ethical use of AI. The paper’s findings are based on a survey conducted among university students exploring their usage and perception of these AI systems. Finally, recommendations for the use of AI in higher education are offered, which emphasize the need to harness AI’s potential while mitigating its risks. This discourse aims at stimulating policy and strategy development to ensure relevant and effective education in the rapidly evolving digital landscape.
{"title":"Challenges for higher education in the era of widespread access to Generative AI","authors":"K. Walczak, W. Cellary","doi":"10.18559/ebr.2023.2.743","DOIUrl":"https://doi.org/10.18559/ebr.2023.2.743","url":null,"abstract":"Abstract The aim of this paper is to discuss the role and impact of Generative Artificial Intelligence (AI) systems in higher education. The proliferation of AI models such as GPT-4, Open Assistant and DALL-E presents a paradigm shift in information acquisition and learning. This transformation poses substantial challenges for traditional teaching approaches and the role of educators. The paper explores the advantages and potential threats of using Generative AI in education and necessary changes in curricula. It further discusses the need to foster digital literacy and the ethical use of AI. The paper’s findings are based on a survey conducted among university students exploring their usage and perception of these AI systems. Finally, recommendations for the use of AI in higher education are offered, which emphasize the need to harness AI’s potential while mitigating its risks. This discourse aims at stimulating policy and strategy development to ensure relevant and effective education in the rapidly evolving digital landscape.","PeriodicalId":41557,"journal":{"name":"Economics and Business Review","volume":null,"pages":null},"PeriodicalIF":0.7,"publicationDate":"2023-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84261407","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：