暨南学报(哲学社会科学版)

作者：许雪晨

单位：中国社会科学院数量经济与技术经济研究所、中国数量经济学会。

关键词：大语言模型,生成式人工智能,金融行业,ChatGPT

基金：中国社会科学院经济大数据与政策评估实验室(2024SYZH004);中国社会科学院学科建设登峰战略资助计划(DF2023YS29)。

近几年,以ChatGPT为代表的大语言模型等生成式人工智能发展迅速,并在多领域广泛应用。生成式人工智能除了具备传统人工智能在分析、判断和决策方面的能力,还在模型结构、数据依赖性、应用场景和可解释性方面更具优势,能发挥其创造性特质,推动人工智能行业从决策型、分析型向生成型跨越发展。金融行业因其涉及大量文本数据,需要迅速决策,天然对大语言模型具有较高的需求,而大语言模型对金融业高质量发展也将产生显著影响。因此,需要进一步探讨大模型如何通过改变金融业的业务模式,助力金融业数字化转型升级。本文深入探讨以大语言模型为代表的生成式人工智能的基本原理和应用,同时分析其为金融业所带来的机遇、挑战以及应对策略。最后,对生成式人工智能技术在金融领域的未来发展路径进行规划展望。"/>template{display:none;}.mag-rich-xref-fn { cursor: pointer;}//长视频$(document).ready(function(){if($("#showLongArticleVideo") && $("#showLongArticleVideo").length && $("#showLongArticleVideo").length>0){mag_ajax_update({ele_id:'showLongArticleVideo',url:mag_currentQikanUrl() + "/CN/article/showLongArticleVideo.do?id="+$("#articleId").val()});}});$(document).ready(function(){window.metaData = {"journal":{"issn":"1000-5072","qiKanMingCheng_CN":"暨南学报（哲学社会科学版）","id":2,"qiKanMingCheng_EN":"Jinan Journal"},"fundList_cn":["中国社会科学院经济大数据与政策评估实验室(2024SYZH004);中国社会科学院学科建设登峰战略资助计划(DF2023YS29)。"],"article":{"keywordList_cn":["大语言模型","生成式人工智能","金融行业","ChatGPT"],"juan":"46","zhaiyao_cn":"近几年,以ChatGPT为代表的大语言模型等生成式人工智能发展迅速,并在多领域广泛应用。生成式人工智能除了具备传统人工智能在分析、判断和决策方面的能力,还在模型结构、数据依赖性、应用场景和可解释性方面更具优势,能发挥其创造性特质,推动人工智能行业从决策型、分析型向生成型跨越发展。金融行业因其涉及大量文本数据,需要迅速决策,天然对大语言模型具有较高的需求,而大语言模型对金融业高质量发展也将产生显著影响。因此,需要进一步探讨大模型如何通过改变金融业的业务模式,助力金融业数字化转型升级。本文深入探讨以大语言模型为代表的生成式人工智能的基本原理和应用,同时分析其为金融业所带来的机遇、挑战以及应对策略。最后,对生成式人工智能技术在金融领域的未来发展路径进行规划展望。","endNoteUrl_en":"https://jnxb.jnu.edu.cn/skb/EN/article/getTxtFile.do?fileType=EndNote&id=7040","reference":"","bibtexUrl_cn":"https://jnxb.jnu.edu.cn/skb/CN/article/getTxtFile.do?fileType=BibTeX&id=7040","abstractUrl_en":"https://jnxb.jnu.edu.cn/skb/EN/10.11778/j.jnxb.20240229","qi":"8","id":7040,"nian":2024,"bianHao":"1727145764658-682797959","zuoZheEn_L":"XU Xuechen","juanUrl_en":"https://jnxb.jnu.edu.cn/skb/EN/Y2024","shouCiFaBuRiQi":"2024-09-24","qiShiYe":"108","qiUrl_cn":"https://jnxb.jnu.edu.cn/skb/CN/Y2024/V46/I8","lanMu_cn":"专题·人工智能研究","pdfSize":"1151","zuoZhe_CN":"许雪晨","risUrl_cn":"https://jnxb.jnu.edu.cn/skb/CN/article/getTxtFile.do?fileType=Ris&id=7040","title_cn":"ChatGPT等大语言模型赋能数字时代金融业:基于隐私保护,算法歧视与系统风险","doi":"10.11778/j.jnxb.20240229","jieShuYe":"122","keywordList_en":["large language model","generative AI","financial industry","ChatGPT"],"endNoteUrl_cn":"https://jnxb.jnu.edu.cn/skb/CN/article/getTxtFile.do?fileType=EndNote&id=7040","zhaiyao_en":"In recent years, generative artificial intelligence (AI) such as large language models represented by ChatGPT has developed rapidly and is widely used in many fields. In addition to the capabilities of traditional AI in analysis, judgment, and decision-making, generative AI has advantages in model structure, data dependency, application scenarios, and interpretability. It can give full play to its creative characteristics and promote the AI industry, promoting the decision-making and analytical type to develop leaps and bounds to the generative type. As the financial industry involves a large amount of text data and requires rapid decision-making, it naturally has a high demand for large language models. Large language models will have a significant impact on the high-quality development of the financial industry. Therefore, it is necessary to further explore how large language models can promote the digital transformation and upgrading of the financial industry by changing the business model of the financial industry. However, despite the broad application prospects of large language models in the financial industry, there is still a lack of systematic literature review to discuss in detail the specific applications, challenges, and potential solutions of large language models in the financial industry.This paper systematically reviews the three stages of model development: the nascent stage, the consolidation stage, and the explosive stage, clarifying the evolution of large models. It elaborates on the structure of the Transformer to help readers understand the underlying principles of large language models. Furthermore, this paper summarizes the application of large language models in the financial sector and details how these models are driving a paradigm shift in the industry. This includes enhancing productivity, transforming existing human-computer interaction models, and improving the accuracy of information dissemination and retrieval. Using examples from banking, insurance, financial management, and investment, this paper explores the opportunities that large language models bring to the financial industry. Additionally, it addresses the challenges these models face in the financial sector, such as data privacy and security, accuracy and reliability, legal regulations, technical costs, and implementation. This paper proposes corresponding countermeasures to these challenges. Finally, it offers policy recommendations from the perspectives of advancing legislation, optimizing algorithms, and fostering cross-departmental cooperation to accelerate the implementation and application of large models in the financial industry.This paper is the first to provide a comprehensive overview of the application of large language models in the financial field. Deeply analyzing specific application cases of large language models in the financial field, this article demonstrates the possibilities of large language models in various industries in the financial field. This not only provides valuable reference for practitioners in the financial field but also lays the foundation for subsequent research.Looking ahead, the development of large language models in the financial sector remains full of infinite possibilities. Perhaps they can deeply participate in the R&D of innovative financial products, bringing more personalized and precise services to the financial market. With technological iterations, the data processing capabilities and analytical accuracy of large language models are expected to further improve, better addressing the complex and ever-changing financial environment. Additionally, it is anticipated that more comprehensive industry standards and regulations will be introduced in the future, promoting the compliant and robust application of large language models in the financial sector.","bibtexUrl_en":"https://jnxb.jnu.edu.cn/skb/EN/article/getTxtFile.do?fileType=BibTeX&id=7040","abstractUrl_cn":"https://jnxb.jnu.edu.cn/skb/CN/10.11778/j.jnxb.20240229","zuoZheCn_L":"许雪晨","juanUrl_cn":"https://jnxb.jnu.edu.cn/skb/CN/Y2024","lanMu_en":"","qiUrl_en":"https://jnxb.jnu.edu.cn/skb/EN/Y2024/V46/I8","zuoZhe_EN":"XU Xuechen","risUrl_en":"https://jnxb.jnu.edu.cn/skb/EN/article/getTxtFile.do?fileType=Ris&id=7040","title_en":"Large Language Models Empowering the Financial Industry in the Digital Age: Opportunities, Challenges, and Solutions","hasPdf":"true"},"authorNotes_cn":["许雪晨,中国社会科学院数量经济与技术经济研究所、中国数量经济学会。"]};if(window.metaData && (!window.metaData.authorNotesCommon_cn && !window.metaData.authorNotesCorresp_cn && window.metaData.authorNotes_cn)){window.metaData.authorNotesCommon_cn = window.metaData.authorNotes_cn;}if(window.metaData && (!window.metaData.authorNotesCommon_en && !window.metaData.authorNotesCorresp_en && window.metaData.authorNotes_en)){window.metaData.authorNotesCommon_en = window.metaData.authorNotes_en;}var _nlmdtdXml = $("#article_nlmdtdXml").val();var magDir = _nlmdtdXml.replace(/\/[^\/]+$/,'');window.metaData.magDir = magDir;new mag_vue({ el: '#metaVue', //dataUrl: mag_currentQikanUrl() + '/EN/article/getRichHtmlJson.do?articleId='+$("#articleId").val(), data:window.metaData, loading:true, doTextFun:function(text){ return doDataJsonText(text); }, doAfterFun:function(selector,jsonData){doMagVueAfter(selector,jsonData);}});var pars = {attachType:'PDF',id:$("#articleId").val(),json:true};jQuery.post($("#mag-common-contextPath").val()+"/CN/article/showArticleFile.do?" + new Date().getTime(), pars, function(data){var json=null;if(data.match(/^\[json\].*/g)){json=eval("("+data.replace(/^\[json\]/g,"")+")");}if(json && json.status == 1){//有权限}else{//无权限$("#no-access-div").show();}});mag_ajax({url:mag_currentQikanUrl() + '/CN/article/getRichHtmlJson.do',data:{articleId:$("#articleId").val(),token:'462bca47f16b4fe9984633fadd315fad',referer:document.referrer},dataType:"json",async:"false",success:function(json){if(json.richHtmlJson && json.richHtmlJson.article){if(json.trans){var _nlmdtdXmlCN = $("#article_nlmdtdXmlCN").val();var magDirCN = _nlmdtdXmlCN.replace(/\/[^\/]+$/,'');json.richHtmlJson.magDir=magDirCN;}else{json.richHtmlJson.magDir=magDir;}new mag_vue({ el: '#bodyVue', data:json.richHtmlJson, doTextFun:function(text){ return doDataJsonText(text); }, doAfterFun:function(selector,jsonData){doMagVueAfter(selector,jsonData);}});}}});//altmetricif($.trim($("#article_doi").val())!=''){$.ajax({url:'https://api.altmetric.com/v1/doi/'+$.trim($("#article_doi").val())+'?callback=altmetric_callback',//url:'https://api.altmetric.com/v1/doi/10.1007/s13238-016-0264-7?callback=altmetric_callback',dataType:'jsonp'});}mag_rich_menu(jQuery);$(function(){ $('#autoMenu').autoMenu();})$("#autoMenu").niceScroll({cursorcolor: "#ccc", //滚动条的颜色cursoropacitymax: 0, //滚动条的透明度，从0-1 touchbehavior: false, //使光标拖动滚动像在台式电脑触摸设备 true滚动条拖动不可用 cursorwidth: "6px", //滚动条的宽度 单位默认px cursorborder: "0", // 游标边框css定义 cursorborderradius: "3px", //滚动条两头的圆角 autohidemode: true, //是否隐藏滚动条 true的时候默认不显示滚动条，当鼠标经过的时候显示滚动条 zindex: "auto", //给滚动条设置z-index值 railvalign:'defaul',railpadding: {top: 0,right: -6,left: 0,bottom: 0}, //滚动条的位置});window.autoMenuFlag=true;$("#relatedArticles").niceScroll({cursorcolor: "#ccc", //滚动条的颜色cursoropacitymax: 0, //滚动条的透明度，从0-1 touchbehavior: false, //使光标拖动滚动像在台式电脑触摸设备 true滚动条拖动不可用 cursorwidth: "6px", //滚动条的宽度 单位默认px cursorborder: "0", // 游标边框css定义 cursorborderradius: "3px", //滚动条两头的圆角 autohidemode: true, //是否隐藏滚动条 true的时候默认不显示滚动条，当鼠标经过的时候显示滚动条 zindex: "auto", //给滚动条设置z-index值 railvalign:'defaul',railpadding: {top: 0,right: -6,left: 0,bottom: 0}, //滚动条的位置});//公式序号 居右jQuery(".mag-xml-disp-formula").each(function() {var _this = jQuery(this);var _label= _this.children("label");if(typeof(_label.html())!="undefined"){_this.children("label").remove();_this.wrap("");_this.parent().parent().append(""+_label.html()+"");}});$('body').click(function(){$('.daochu_div').hide();$('.qipao').hide();})$('.qipao').click(function(event){event.stopPropagation();})/*参考文献显示隐藏*/$('.biaotijiantoush1').click(function(event){if($(this).attr('class')=='glyphicon glyphicon-triangle-top biaotijiantoush biaotijiantoush1'){$(this).parents('.mag_main_zhengwen_left_div_ckwx_table_benwenyiny').next('.mag_main_zhengwen_left_div_ckwx_table_ckwx_xiangqing').stop().slideUp();$(this).attr('class','glyphicon glyphicon-triangle-bottom biaotijiantoush biaotijiantoush1');}else if($(this).attr('class')=='glyphicon glyphicon-triangle-bottom biaotijiantoush biaotijiantoush1'){$(this).parents('.mag_main_zhengwen_left_div_ckwx_table_benwenyiny').next('.mag_main_zhengwen_left_div_ckwx_table_ckwx_xiangqing').stop().slideDown();$(this).attr('class','glyphicon glyphicon-triangle-top biaotijiantoush biaotijiantoush1');}event.stopPropagation();});//加载补充材料mag_ajax_update({ele_id:'showArticle_suppls',url:mag_currentQikanUrl() + '/CN/article/showArticleSuppls.do',data:{articleId:$("#articleId").val()}});//加载专题mag_ajax_update({ele_id:'showArticle_subjects',url:mag_currentQikanUrl() + '/CN/subject/showSubjectListByArticle.do',data:{articleId:$("#articleId").val(),forwardJsp:'showArticle'}});//渲染公式mag_mathJax();});function altmetric_callback(json){if(json && json.score){if(json.score - parseInt(json.score) > 0){$("#altmetric-score").html(parseInt(json.score)+1);}else{$("#altmetric-score").html(parseInt(json.score));}$(".main_content_center_right_duoweidu_div").css("width","25%");$("#altmetric-a").show();$("#altmetric-a").attr("href",json.details_url);}}function richClickAuthor(name){var searchSQL = "((("+mag_advSearchValReplace(name)+"[Author]) AND 2J[Journal]) AND year[Order])";mag_request({url:mag_currentQikanUrl() + '/CN/searchresult',data:{searchSQL:searchSQL}});}function richClickKeyword(keyword){var searchSQL = "((("+mag_advSearchValReplace(keyword)+"[Keyword]) AND 2J[Journal]) AND year[Order])";mag_request({url:mag_currentQikanUrl() + '/CN/searchresult',data:{searchSQL:searchSQL}});}function mag_share(){$("#bShare_more_a").click();}function loadRelatedArticles(){if($("#showRelatedArticles") && $("#showRelatedArticles").length>0){//加载过了}else{mag_loading();mag_ajax_update({ele_id:'relatedArticles-content',url:mag_currentQikanUrl() + '/CN/article/relatedArticle.do',data:{id:$("#articleId").val(), forwardJsp:'showArticle'},complete:function(){mag_layer_close();}});}}function showMetrics(){mag_window_location(mag_currentQikanUrl() + '/CN/metrics/'+$("#articleId").val());}function piaofuFn(a,fnId){var content=''+$("#mag-rich-fn-"+fnId).html()+'';piaofu(a,content);}function huoqukuanduimg(){//}function doDataJsonText(text){var str = text;//var _nlmdtdXml = $("#article_nlmdtdXml").val();//var magDir = _nlmdtdXml.replace(/\/[^\/]+$/,'');str = str.replace(/";ss += "";var shuomingTitle = "";var shuomingContent = '';if(fig.titleCn && fig.titleCn!=''){shuomingTitle = fig.titleCn;if(fig.titleContentCn && fig.titleContentCn!=''){shuomingContent += fig.titleContentCn;}}if(shuomingTitle==''){if(fig.titleEn && fig.titleEn!=''){shuomingTitle = fig.titleEn;}if(fig.titleContentEn && fig.titleContentEn!=''){shuomingContent += fig.titleContentEn;}}else{if(fig.titleEn && fig.titleEn!=''){shuomingContent += ""+fig.titleEn+"

来源：2024年第8期

《暨南学报(哲学社会科学版)》期刊编辑部

查看暨南学报(哲学社会科学版)杂志2024年第8期