{
    "componentChunkName": "component---node-modules-gatsby-theme-code-notes-src-templates-note-js",
    "path": "/paper-study-haic-2505142330",
    "result": {"data":{"mdx":{"body":"var _excluded = [\"components\"];\nfunction _extends() { return _extends = Object.assign ? Object.assign.bind() : function (n) { for (var e = 1; e < arguments.length; e++) { var t = arguments[e]; for (var r in t) ({}).hasOwnProperty.call(t, r) && (n[r] = t[r]); } return n; }, _extends.apply(null, arguments); }\nfunction _objectWithoutProperties(e, t) { if (null == e) return {}; var o, r, i = _objectWithoutPropertiesLoose(e, t); if (Object.getOwnPropertySymbols) { var n = Object.getOwnPropertySymbols(e); for (r = 0; r < n.length; r++) o = n[r], -1 === t.indexOf(o) && {}.propertyIsEnumerable.call(e, o) && (i[o] = e[o]); } return i; }\nfunction _objectWithoutPropertiesLoose(r, e) { if (null == r) return {}; var t = {}; for (var n in r) if ({}.hasOwnProperty.call(r, n)) { if (-1 !== e.indexOf(n)) continue; t[n] = r[n]; } return t; }\n/* @jsxRuntime classic */\n/* @jsx mdx */\n\nvar _frontmatter = {\n  \"title\": \"When Are Combinations of Humans and AI Useful? - A Systematic Review and Meta-Analysis\",\n  \"tags\": [\"paper\"],\n  \"emoji\": \"\",\n  \"modified\": \"2025-05-14T00:00:00.000Z\"\n};\nvar layoutProps = {\n  _frontmatter: _frontmatter\n};\nvar MDXLayout = \"wrapper\";\nreturn function MDXContent(_ref) {\n  var components = _ref.components,\n    props = _objectWithoutProperties(_ref, _excluded);\n  return mdx(MDXLayout, _extends({}, layoutProps, props, {\n    components: components,\n    mdxType: \"MDXLayout\"\n  }), mdx(\"h1\", {\n    \"id\": \"when-are-combinations-of-humans-and-ai-useful---a-systematic-review-and-meta-analysis\"\n  }, \"When Are Combinations of Humans and AI Useful? - A Systematic Review and Meta-Analysis\"), mdx(\"h1\", {\n    \"id\": \"abstract\"\n  }, \"Abstract\"), mdx(\"p\", null, mdx(\"strong\", {\n    parentName: \"p\"\n  }, \"(Background)\"), \" Inspired by the increasing use of AI to augment humans, researchers have studied human-AI systems involving different tasks, systems, and populations.\"), mdx(\"p\", null, mdx(\"strong\", {\n    parentName: \"p\"\n  }, \"(Motivation)\"), \" Despite such a large body of work, we lack a broad conceptual understanding of when combinations of humans and AI are better than either alone.\"), mdx(\"p\", null, mdx(\"strong\", {\n    parentName: \"p\"\n  }, \"(Methodology)\"), \" Here, we addressed this question by conducting a pre-registered systematic review and meta-analysis of over 100 recent experimental studies reporting over 300 effect sizes.\"), mdx(\"p\", null, \"'\\uCCB4\\uACC4\\uC801 \\uBB38\\uD5CC \\uAC80\\uD1A0'\\uC640 '\\uBA54\\uD0C0 \\uBD84\\uC11D'\\uC774\\uB77C\\uB294 \\uBC29\\uBC95\\uC744 \\uC0AC\\uC6A9\\uD558\\uC600\\uC2B5\\uB2C8\\uB2E4.\"), mdx(\"p\", null, mdx(\"em\", {\n    parentName: \"p\"\n  }, \"\\uCCB4\\uACC4\\uC801 \\uBB38\\uD5CC \\uAC80\\uD1A0: \\uD2B9\\uC815 \\uC8FC\\uC81C\\uC5D0 \\uB300\\uD55C \\uAE30\\uC874 \\uC5F0\\uAD6C\\uB4E4\\uC744 \\uC5C4\\uACA9\\uD558\\uACE0 \\uD22C\\uBA85\\uD55C \\uC808\\uCC28\\uC5D0 \\uB530\\uB77C \\uBAA8\\uB450 \\uCC3E\\uC544\\uC11C \\uAC80\\uD1A0\\uD558\\uB294 \\uBC29\\uBC95\\uC785\\uB2C8\\uB2E4. \\uC774 \\uB17C\\uBB38\\uC5D0\\uC11C\\uB294 100\\uD3B8 \\uC774\\uC0C1\\uC758 \\uC2E4\\uD5D8 \\uC5F0\\uAD6C\\uB97C \\uBD84\\uC11D \\uB300\\uC0C1\\uC73C\\uB85C \\uC120\\uC815\\uD588\\uC2B5\\uB2C8\\uB2E4. \\uBA54\\uD0C0 \\uBD84\\uC11D: \\uC5EC\\uB7EC \\uAC1C\\uBCC4 \\uC5F0\\uAD6C\\uB4E4\\uC758 \\uACB0\\uACFC\\uB97C \\uD1B5\\uACC4\\uC801\\uC73C\\uB85C \\uD1B5\\uD569\\uD558\\uC5EC \\uC804\\uCCB4\\uC801\\uC778 \\uD6A8\\uACFC \\uD06C\\uAE30\\uB97C \\uCD94\\uC815\\uD558\\uACE0 \\uACB0\\uB860\\uC758 \\uC2E0\\uB8B0\\uB3C4\\uB97C \\uB192\\uC774\\uB294 \\uBC29\\uBC95\\uC785\\uB2C8\\uB2E4. \\uC774 \\uB17C\\uBB38\\uC5D0\\uC11C\\uB294 300\\uAC1C \\uC774\\uC0C1\\uC758 \\uD6A8\\uACFC \\uD06C\\uAE30\\uB97C \\uBD84\\uC11D\\uD588\\uC2B5\\uB2C8\\uB2E4.\")), mdx(\"p\", null, mdx(\"strong\", {\n    parentName: \"p\"\n  }, \"(Results)\"), \" First, we found that, on average, human-AI combinations performed significantly worse than the best of humans or AI alone (Hedges\\u2019 g = \\u22120.23, 95% confidence interval \\u22120.39 to \\u22120.07). Second, we found performance losses in tasks that involved making decisions and significantly greater gains in tasks that involved creating content. Finally, when humans outperformed AI alone, we found performance gains in the combination, but when the AI outperformed humans alone we found losses.\"), mdx(\"p\", null, mdx(\"strong\", {\n    parentName: \"p\"\n  }, \"(Arguments)\"), \" These findings highlight the heterogeneity of the effects of human-AI collaboration and point to promising avenues for improving human-AI systems. \\uC778\\uAC04-AI \\uD611\\uC5C5\\uC758 \\uD6A8\\uACFC\\uAC00 \\uC77C\\uB960\\uC801\\uC774\\uC9C0 \\uC54A\\uACE0, \\uC0C1\\uD669\\uACFC \\uACFC\\uC81C\\uC5D0 \\uB530\\uB77C \\uC5B4\\uB5A4 \\uC870\\uD569\\uC774 \\uC720\\uC6A9\\uD560 \\uC218 \\uC788\\uB294\\uC9C0\\uC5D0 \\uB300\\uD55C \\uAD6C\\uCCB4\\uC801\\uC778 \\uD1B5\\uCC30\\uC744 \\uC81C\\uACF5\\uD558\\uBA70, \\uD5A5\\uD6C4 \\uC778\\uAC04-AI \\uC2DC\\uC2A4\\uD15C \\uC5F0\\uAD6C \\uBC0F \\uC124\\uACC4 \\uBC29\\uD5A5\\uC744 \\uC81C\\uC2DC\\uD558\\uACE0 \\uC788\\uC2B5\\uB2C8\\uB2E4.\"), mdx(\"h1\", {\n    \"id\": \"citation\"\n  }, \"Citation\"), mdx(\"p\", null, \"Michelle Vaccaro, et al. MIT Center for Collective Intelligence, Sloan School of Management Institute for Data, Systems, and Society, Schwarzman College of Computing, MIT\"), mdx(\"p\", null, \"Nature Human Behaviour 8, 2293-2303 (2024)\\nReceived: 06 Apr 2023\\nAccepted: 23 Sep 2024\\nPublished: 28 Oct 2024\"), mdx(\"h1\", {\n    \"id\": \"background\"\n  }, \"Background\"), mdx(\"p\", null, \"\\uB9CE\\uC740 \\uC5F0\\uAD6C \\uACB0\\uACFC\\uC5D0 \\uB530\\uB974\\uBA74 \\uC778\\uAC04\\uC758 \\uCC3D\\uC758\\uC131, \\uC9C1\\uAD00, \\uC0C1\\uD669 \\uC774\\uD574\\uB97C AI\\uC758 \\uC18D\\uB3C4, \\uD655\\uC7A5\\uC131, \\uBD84\\uC11D\\uB825\\uACFC \\uD1B5\\uD569\\uD558\\uBA74 \\uC758\\uB8CC, \\uACE0\\uAC1D\\xA0 \\uC11C\\uBE44\\uC2A4, \\uACFC\\uD559 \\uC5F0\\uAD6C\\uC640 \\uAC19\\uC740 \\uBD84\\uC57C\\uC5D0\\uC11C \\uD601\\uC2E0\\uC801\\uC778 \\uC194\\uB8E8\\uC158\\uACFC \\uD5A5\\uC0C1\\uB41C \\uC758\\uC0AC \\uACB0\\uC815\\uC744 \\uC774\\uB04C\\uC5B4\\uB0BC \\uC218 \\uC788\\uB2E4\\uACE0 \\uD569\\uB2C8\\uB2E4. \\uBC18\\uBA74\\uC5D0, \\uC810\\uC810 \\uB354 \\uB9CE\\uC740 \\uC5F0\\uAD6C\\uC5D0\\uC11C \\uC778\\uAC04-AI \\uC2DC\\uC2A4\\uD15C\\uC774 \\uC778\\uAC04 \\uB610\\uB294 AI \\uB2E8\\uB3C5\\uC73C\\uB85C \\uCD5C\\uACE0\\uC758 \\uC131\\uB2A5\\uC744 \\uBCF4\\uC774\\uB294 \\uACBD\\uC6B0\\uC640 \\uBE44\\uAD50\\uD558\\uC5EC \\uBC18\\uB4DC\\uC2DC \\uB354 \\uB098\\uC740 \\uACB0\\uACFC\\uB97C \\uB2EC\\uC131\\uD558\\uC9C0\\uB294 \\uBABB\\uD55C\\uB2E4\\uB294 \\uC0AC\\uC2E4\\uC774 \\uBC1D\\uD600\\uC9C0\\uACE0 \\uC788\\uC2B5\\uB2C8\\uB2E4. \\uC758\\uC0AC \\uC18C\\uD1B5 \\uC7A5\\uBCBD, \\uC2E0\\uB8B0 \\uBB38\\uC81C, \\uC724\\uB9AC\\uC801 \\uBB38\\uC81C, \\uC778\\uAC04\\uACFC AI \\uC2DC\\uC2A4\\uD15C \\uAC04\\uC758 \\uD6A8\\uACFC\\uC801\\uC778 \\uD611\\uC5C5 \\uD544\\uC694\\uC131\\uACFC \\uAC19\\uC740 \\uC5B4\\uB824\\uC6C0\\uC740 \\uD611\\uC5C5 \\uD504\\uB85C\\uC138\\uC2A4\\uB97C \\uC800\\uD574\\uD560 \\uC218 \\uC788\\uC2B5\\uB2C8\\uB2E4.\"), mdx(\"p\", null, \"\\uC774\\uB7EC\\uD55C \\uBAA8\\uC21C\\uC801\\uC778 \\uACB0\\uACFC\\uB294 \\uC911\\uC694\\uD55C \\uC9C8\\uBB38\\uC744 \\uC81C\\uAE30\\uD569\\uB2C8\\uB2E4. \\uC5B8\\uC81C \\uC778\\uAC04\\uACFC AI\\uAC00 \\uC11C\\uB85C\\uB97C \\uBCF4\\uC644\\uD558\\uB294\\uAC00? \\uADF8\\uB9AC\\uACE0 \\uC5BC\\uB9C8\\uB098 \\uB9CE\\uC774 \\uBCF4\\uC644\\uD558\\uB294\\uAC00? \\uC774\\uB7EC\\uD55C \\uBB38\\uC81C\\uB4E4\\uC744 \\uD574\\uACB0\\uD558\\uAE30 \\uC704\\uD574, \\uC6B0\\uB9AC\\uB294 \\uC778\\uAC04-AI \\uC2DC\\uC2A4\\uD15C\\uC758 \\uC2DC\\uB108\\uC9C0\\uB97C \\uC815\\uB7C9\\uD654\\uD558\\uACE0 \\uB2E4\\uC591\\uD55C \\uD658\\uACBD\\uC5D0\\uC11C \\uC2DC\\uB108\\uC9C0\\uC758 \\uC874\\uC7AC (\\uB610\\uB294 \\uBD80\\uC7AC)\\uB97C \\uC124\\uBA85\\uD558\\uB294 \\uC694\\uC778\\uB4E4\\uC744 \\uC2DD\\uBCC4\\uD558\\uAE30 \\uC704\\uD574 \\uCCB4\\uACC4\\uC801\\uC778 \\uBB38\\uD5CC \\uAC80\\uD1A0 \\uBC0F \\uBA54\\uD0C0 \\uBD84\\uC11D\\uC744 \\uC218\\uD589\\uD588\\uC2B5\\uB2C8\\uB2E4.\"), mdx(\"h1\", {\n    \"id\": \"methodology\"\n  }, \"Methodology\"), mdx(\"p\", null, \"Kitchenham 2004\\uC758 \\uCCB4\\uACC4\\uC801 \\uAC80\\uD1A0 \\uC9C0\\uCE68\\uC5D0 \\uB530\\uB77C \\uC774 \\uBA54\\uD0C0 \\uBD84\\uC11D\\uC744 \\uC218\\uD589\\uD588\\uC73C\\uBA70, Systematic Reviews and Meta-Analyses (PRISMA) \", \"[55]\", \"\\uC5D0 \\uB300\\uD55C Preferred Reporting Items\\uC5D0\\uC11C \\uC81C\\uC2DC\\uD55C \\uD45C\\uC900\\uC744 \\uB530\\uB985\\uB2C8\\uB2E4.\"), mdx(\"p\", null, mdx(\"em\", {\n    parentName: \"p\"\n  }, \"Kitchenham(2004)\\uC758 \\uCCB4\\uACC4\\uC801 \\uAC80\\uD1A0(\\uCCB4\\uACC4\\uC801 \\uBB38\\uD5CC\\uACE0\\uCC30, Systematic Review) \\uC9C0\\uCE68\\uC740 \\uC8FC\\uB85C \\uC18C\\uD504\\uD2B8\\uC6E8\\uC5B4 \\uACF5\\uD559 \\uBD84\\uC57C\\uC5D0\\uC11C \\uB110\\uB9AC \\uC778\\uC6A9\\uB418\\uB294 \\uD45C\\uC900\\uD654\\uB41C \\uBB38\\uD5CC\\uACE0\\uCC30 \\uC808\\uCC28\\uB85C, \\uC138 \\uAC00\\uC9C0 \\uC8FC\\uC694 \\uB2E8\\uACC4(Planning, Conducting, Reporting)\\uB85C \\uAD6C\\uC131\\uB418\\uC5B4 \\uC788\\uC2B5\\uB2C8\\uB2E4.\\uBA54\\uD0C0\\uBD84\\uC11D(meta-analysis)\\uC740 \\uD2B9\\uC815 \\uC8FC\\uC81C\\uC5D0 \\uB300\\uD574 \\uC774\\uBBF8 \\uBC1C\\uD45C\\uB41C \\uC5EC\\uB7EC \\uB3C5\\uB9BD\\uC801\\uC778 \\uC5F0\\uAD6C\\uB4E4\\uC758 \\uACB0\\uACFC\\uB97C \\uCCB4\\uACC4\\uC801\\uC73C\\uB85C \\uC218\\uC9D1\\uD558\\uACE0, \\uD1B5\\uACC4\\uC801\\uC73C\\uB85C \\uD1B5\\uD569\\uD558\\uC5EC \\uD558\\uB098\\uC758 \\uC885\\uD569\\uC801\\uC778 \\uACB0\\uB860\\uC744 \\uB3C4\\uCD9C\\uD558\\uB294 \\uC5F0\\uAD6C \\uBC29\\uBC95\\uC785\\uB2C8\\uB2E4. \\uC989, \\uAC1C\\uBCC4 \\uC5F0\\uAD6C\\uC5D0\\uC11C \\uBCF4\\uACE0\\uB41C \\uD6A8\\uACFC \\uD06C\\uAE30(effect size) \\uB4F1 \\uC8FC\\uC694 \\uD1B5\\uACC4 \\uACB0\\uACFC\\uB97C \\uBAA8\\uC544 \\uD1B5\\uACC4\\uC801\\uC73C\\uB85C \\uACB0\\uD569\\uD574 \\uBCF4\\uB2E4 \\uAC1D\\uAD00\\uC801\\uC774\\uACE0 \\uC2E0\\uB8B0\\uB3C4 \\uB192\\uC740 \\uACB0\\uB860\\uC744 \\uB3C4\\uCD9C\\uD558\\uB294 \\uD1B5\\uACC4\\uC801 \\uC5F0\\uAD6C \\uBC29\\uBC95\\uC785\\uB2C8\\uB2E4. (\\uCCB4\\uACC4\\uC801 \\uBB38\\uD5CC\\uACE0\\uCC30)PRISMA(Preferred Reporting Items for Systematic Reviews and Meta-Analyses)\\uB294 \\uCCB4\\uACC4\\uC801 \\uBB38\\uD5CC\\uACE0\\uCC30(systematic review)\\uACFC \\uBA54\\uD0C0\\uBD84\\uC11D(meta-analysis) \\uC5F0\\uAD6C\\uC758 \\uBCF4\\uACE0 \\uBC29\\uBC95\\uC744 \\uD45C\\uC900\\uD654\\uD558\\uAE30 \\uC704\\uD574 \\uAC1C\\uBC1C\\uB41C \\uAD6D\\uC81C \\uAC00\\uC774\\uB4DC\\uB77C\\uC778\\uC785\\uB2C8\\uB2E4.\")), mdx(\"h2\", {\n    \"id\": \"literature-review-문헌-검토\"\n  }, \"Literature review (\\uBB38\\uD5CC \\uAC80\\uD1A0)\"), mdx(\"p\", null, mdx(\"strong\", {\n    parentName: \"p\"\n  }, \"Eligiblity criteria (\\uC790\\uACA9 \\uAE30\\uC900)\"), \"\\n\\uCCAB\\uC9F8, \\uB17C\\uBB38\\uC740 \\uC778\\uAC04\\uACFC AI \\uC2DC\\uC2A4\\uD15C\\uC774 \\uD611\\uB825\\uD558\\uC5EC \\uC791\\uC5C5\\uC744 \\uC218\\uD589\\uD558\\uB294 \\uD2B9\\uC815 \\uC0AC\\uB840\\uB97C \\uD3C9\\uAC00\\uD558\\uB294 \\uB3C5\\uCC3D\\uC801\\uC778 \\uC2E4\\uD5D8\\uC744 \\uC81C\\uC2DC\\uD574\\uC57C \\uD588\\uC2B5\\uB2C8\\uB2E4.\"), mdx(\"p\", null, \"\\uB458\\uC9F8, \\uC77C\\uBD80 \\uC815\\uB7C9\\uC801 \\uCE21\\uC815 \\uAE30\\uC900\\uC5D0 \\uB530\\uB77C (1) \\uC778\\uAC04 \\uB2E8\\uB3C5, (2) AI \\uB2E8\\uB3C5, (3) \\uC778\\uAC04-AI \\uC2DC\\uC2A4\\uD15C\\uC758 \\uC131\\uB2A5\\uC744 \\uBCF4\\uACE0\\uD574\\uC57C \\uD588\\uC2B5\\uB2C8\\uB2E4. \\uB530\\uB77C\\uC11C \\uC778\\uAC04 \\uB2E8\\uB3C5\\uC758 \\uC131\\uB2A5\\uC740 \\uBCF4\\uACE0\\uD588\\uC9C0\\uB9CC AI \\uB2E8\\uB3C5\\uC758 \\uC131\\uB2A5\\uC740 \\uBCF4\\uACE0\\uD558\\uC9C0 \\uC54A\\uC740 \\uC5F0\\uAD6C\\uB294 \\uC81C\\uC678\\uD588\\uACE0, \\uB9C8\\uCC2C\\uAC00\\uC9C0\\uB85C AI \\uB2E8\\uB3C5\\uC758 \\uC131\\uB2A5\\uC740 \\uBCF4\\uACE0\\uD588\\uC9C0\\uB9CC \\uC778\\uAC04 \\uB2E8\\uB3C5\\uC758 \\uC131\\uB2A5\\uC740 \\uBCF4\\uACE0\\uD558\\uC9C0 \\uC54A\\uC740 \\uC5F0\\uAD6C\\uB3C4 \\uC81C\\uC678\\uD588\\uC2B5\\uB2C8\\uB2E4. \\uC774 \\uC870\\uD56D\\uC5D0 \\uB530\\uB77C \\uC21C\\uC218 \\uBA54\\uD0C0 \\uBD84\\uC11D \\uBC0F \\uBB38\\uD5CC \\uAC80\\uD1A0, \\uC774\\uB860\\uC801 \\uC5F0\\uAD6C, \\uC9C8\\uC801 \\uBD84\\uC11D, \\uB17C\\uD3C9, \\uC758\\uACAC \\uBC0F \\uC2DC\\uBBAC\\uB808\\uC774\\uC158\\uB3C4 \\uC81C\\uC678\\uD588\\uC2B5\\uB2C8\\uB2E4.\"), mdx(\"p\", null, \"\\uC14B\\uC9F8, \\uB17C\\uBB38\\uC5D0\\uB294 \\uC2E4\\uD5D8 \\uC124\\uACC4, \\uAC01 \\uC870\\uAC74\\uC5D0\\uC11C\\uC758 \\uCC38\\uAC00\\uC790 \\uC218, \\uAC01 \\uC870\\uAC74\\uC5D0\\uC11C\\uC758 \\uACB0\\uACFC\\uC758 \\uD45C\\uC900 \\uD3B8\\uCC28 \\uB610\\uB294 \\uB2E4\\uB978 \\uC218\\uB7C9\\uC73C\\uB85C\\uBD80\\uD130 \\uACC4\\uC0B0\\uD558\\uAE30\\uC5D0 \\uCDA9\\uBD84\\uD55C \\uC815\\uBCF4\\uAC00 \\uD3EC\\uD568\\uB418\\uC5B4\\uC57C \\uD588\\uC2B5\\uB2C8\\uB2E4.\"), mdx(\"p\", null, mdx(\"strong\", {\n    parentName: \"p\"\n  }, \"Search strategy\"), \"\\n\\uC778\\uAC04-AI \\uC0C1\\uD638 \\uC791\\uC6A9 \\uC5F0\\uAD6C\\uC758 \\uD559\\uC81C\\uC801 \\uD2B9\\uC131\\uC744 \\uACE0\\uB824\\uD558\\uC5EC \\uCEF4\\uD4E8\\uD130 \\uACFC\\uD559, \\uC815\\uBCF4 \\uACFC\\uD559 \\uBC0F \\uC0AC\\uD68C \\uACFC\\uD559 \\uBD84\\uC57C\\uC758 \\uD559\\uC220 \\uB300\\uD68C \\uBC0F \\uC800\\uB110\\uC744 \\uD3EC\\uD568\\uD558\\uB294 \\uC5EC\\uB7EC \\uB370\\uC774\\uD130\\uBCA0\\uC774\\uC2A4\\uC5D0\\uC11C \\uC774 \\uAC80\\uC0C9\\uC744 \\uC218\\uD589\\uD588\\uC2B5\\uB2C8\\uB2E4. \\uC774\\uB7EC\\uD55C \\uBD84\\uC57C\\uC758 \\uB3C4\\uC11C\\uAD00 \\uC804\\uBB38\\uAC00\\uC640 \\uD611\\uC758\\uD558\\uC5EC \\uAC80\\uD1A0\\uB97C \\uC704\\uD574 Association for Computing Machinery Digital Library (ACM DL), Association for Information Systems eLibrary (AISeL) \\uBC0F Web of Science Core Collection (WoS)\\uC744 \\uB300\\uC0C1\\uC73C\\uB85C \\uD558\\uAE30\\uB85C \\uACB0\\uC815\\uD588\\uC2B5\\uB2C8\\uB2E4. \\uCD5C\\uC2E0 \\uD615\\uD0DC\\uC758 \\uC778\\uACF5 \\uC9C0\\uB2A5\\uC5D0 \\uCD08\\uC810\\uC744 \\uB9DE\\uCD94\\uAE30 \\uC704\\uD574 2020\\uB144 1\\uC6D4 1\\uC77C\\uBD80\\uD130 2023\\uB144 6\\uC6D4 30\\uC77C \\uC0AC\\uC774\\uC5D0 \\uBC1C\\uD45C\\uB41C \\uC5F0\\uAD6C\\uB85C \\uAC80\\uC0C9\\uC744 \\uC81C\\uD55C\\uD588\\uC2B5\\uB2C8\\uB2E4. \\uAC80\\uC0C9 \\uBB38\\uC790\\uC5F4\\uC744 \\uAC1C\\uBC1C\\uD558\\uAE30 \\uC704\\uD574 \\uC778\\uAC04-AI \\uC2DC\\uC2A4\\uD15C\\uC758 \\uC131\\uB2A5\\uC744 \\uD3C9\\uAC00\\uD55C \\uC5F0\\uAD6C\\uC758 \\uCE21\\uBA74\\uC744 \\uCD94\\uCD9C\\uD558\\uB294 \\uAC83\\uC73C\\uB85C \\uC2DC\\uC791\\uD588\\uC2B5\\uB2C8\\uB2E4.\"), mdx(\"ol\", null, mdx(\"li\", {\n    parentName: \"ol\"\n  }, \"Human: human OR expert OR participant OR humans OR experts OR participants\"), mdx(\"li\", {\n    parentName: \"ol\"\n  }, \"AI: AI OR \\\"artificial intelligence\\\" OR ML OR \\\"machine learning\\\" OR \\\"deep learning\\\"\"), mdx(\"li\", {\n    parentName: \"ol\"\n  }, \"Collaboration: collaborate OR assist OR aid OR interact OR help\"), mdx(\"li\", {\n    parentName: \"ol\"\n  }, \"Experiment: \\\"experiment\\\" OR \\\"experiments\\\" OR \\\"user study\\\" OR \\\"user studies\\\" OR \\\"crowdsourced study\\\" OR \\\"crowdsourced studies\\\" OR \\\"laboratory study\\\" OR \\\"laboratory studies\\\"\")), mdx(\"p\", null, mdx(\"strong\", {\n    parentName: \"p\"\n  }, \"Coding\"), \"\\nWe considered and coded for multiple potential moderators of human-AI performance, namely: (1) publication date (2) pre-registration status (3) experimental design (4) data type (5) task type (6) task output (7) AI type (8) AI explanation (9) AI confidence (10) participant type and (11) performance metric\"), mdx(\"h2\", {\n    \"id\": \"data-analysis\"\n  }, \"Data Analysis\"), mdx(\"p\", null, mdx(\"strong\", {\n    parentName: \"p\"\n  }, \"Calculation of Effect Size\"), \"\\nWe computed Hedges\\u2019 g to measure the effect of combining human and artificial intelligence on task performance.\"), mdx(\"p\", null, mdx(\"em\", {\n    parentName: \"p\"\n  }, \"Hedges\\u2019 g\\uB294 \\uB450 \\uC9D1\\uB2E8 \\uAC04 \\uD3C9\\uADE0 \\uCC28\\uC774\\uC758 \\uD6A8\\uACFC \\uD06C\\uAE30\\uB97C \\uB098\\uD0C0\\uB0B4\\uB294 \\uC9C0\\uD45C\\uB85C, Cohen\\u2019s d\\uC758 \\uD45C\\uBCF8 \\uD06C\\uAE30(\\uD2B9\\uD788 \\uC18C\\uD45C\\uBCF8)\\uC5D0\\uC11C \\uBC1C\\uC0DD\\uD558\\uB294 \\uACFC\\uB300\\uCD94\\uC815 \\uBB38\\uC81C\\uB97C \\uAD50\\uC815\\uD55C \\uD45C\\uC900\\uD654\\uB41C \\uD3C9\\uADE0\\uCC28\\uC774(effect size) \\uC9C0\\uC218\\uC785\\uB2C8\\uB2E4\")), mdx(\"p\", null, mdx(\"strong\", {\n    parentName: \"p\"\n  }, \"Meta-Analytic Model\"), \"\\n\\uBD84\\uC11D\\uD55C \\uB17C\\uBB38\\uC758 \\uC2E4\\uD5D8\\uB4E4\\uC740 \\uC0C1\\uB2F9\\uD788 \\uB2E4\\uC591\\uD588\\uC2B5\\uB2C8\\uB2E4. \\uC608\\uB97C \\uB4E4\\uC5B4, \\uC11C\\uB85C \\uB2E4\\uB978 \\uACFC\\uC81C\\uB97C \\uD3C9\\uAC00\\uD558\\uACE0, \\uC11C\\uB85C \\uB2E4\\uB978 \\uBC30\\uACBD\\uC758 \\uCC38\\uAC00\\uC790\\uB97C \\uBAA8\\uC9D1\\uD558\\uACE0, \\uC11C\\uB85C \\uB2E4\\uB978 \\uC2E4\\uD5D8 \\uC124\\uACC4\\uB97C \\uC0AC\\uC6A9\\uD588\\uC2B5\\uB2C8\\uB2E4. \\uC9C4\\uC815\\uD55C effect size\\uC5D0\\uC11C \\uC0C1\\uB2F9\\uD55C \\uC2E4\\uD5D8 \\uAC04 \\uC774\\uC9C8\\uC131\\uC774 \\uC608\\uC0C1\\uB418\\uC5C8\\uC73C\\uBBC0\\uB85C, \\uBD84\\uC11D\\uC744 \\uC704\\uD574 \\uD45C\\uBCF8 \\uCD94\\uCD9C \\uC624\\uB958\\uC640 \\uC2E4\\uD5D8 \\uAC04\\uC758 \\\"\\uC9C4\\uC815\\uD55C\\\" \\uBCC0\\uB3D9\\uC131\\uC5D0\\uC11C \\uBC1C\\uC0DD\\uD558\\uB294 effect size\\uC758 \\uBD84\\uC0B0\\uC744 \\uC124\\uBA85\\uD558\\uB294 random-effects \\uBAA8\\uB378\\uC744 \\uC0AC\\uC6A9\\uD588\\uC2B5\\uB2C8\\uB2E4.\"), mdx(\"p\", null, mdx(\"em\", {\n    parentName: \"p\"\n  }, \"effect size\\uAC00 \\uC2E4\\uD5D8 \\uB0B4\\uC5D0 \\uB0B4\\uD3EC\\uB41C 3\\uB2E8\\uACC4 \\uBA54\\uD0C0 \\uBD84\\uC11D \\uBAA8\\uB378\\uC744 \\uCC44\\uD0DD\\uD558\\uC5EC \\uBAA8\\uB378\\uC5D0\\uC11C \\uB370\\uC774\\uD130\\uC758 \\uC885\\uC18D\\uC131\\uC744 \\uBA85\\uC2DC\\uC801\\uC73C\\uB85C \\uACE0\\uB824\\uD588\\uC2B5\\uB2C8\\uB2E4 \", \"[62, 63]\", \". \\uB610\\uD55C, \\uAC15\\uB825\\uD55C \\uBD84\\uC0B0 \\uCD94\\uC815 (RVE) \\uBC29\\uBC95\\uC744 \\uC0AC\\uC6A9\\uD558\\uC5EC effect size \\uCD94\\uC815\\uCE58\\uC758 \\uBD84\\uC0B0\\uACFC \\uAD00\\uB828\\uD558\\uC5EC \\uD45C\\uC900 \\uC624\\uCC28, \\uC2E0\\uB8B0 \\uAD6C\\uAC04 \\uBC0F \\uD1B5\\uACC4 \\uD14C\\uC2A4\\uD2B8\\uC758 \\uC77C\\uAD00\\uB41C \\uCD94\\uC815\\uCE58\\uB97C \\uACC4\\uC0B0\\uD588\\uB294\\uB370, \\uC774\\uB294 \\uC5EC\\uB7EC treatment \\uADF8\\uB8F9\\uC744 \\uB2E8\\uC77C control \\uADF8\\uB8F9\\uACFC \\uBE44\\uAD50\\uD55C \\uC2E4\\uD5D8\\uC5D0\\uC11C \\uBC1C\\uC0DD\\uD55C \\uC911\\uBCF5 \\uD45C\\uBCF8\\uC73C\\uB85C\\uBD80\\uD130 \\uD45C\\uBCF8 \\uCD94\\uCD9C \\uC624\\uB958\\uC758 \\uC885\\uC18D\\uC131\\uC744 \\uC124\\uBA85\\uD569\\uB2C8\\uB2E4 \", \"[64]\", \". \\uACB0\\uACFC\\uC758 \\uC720\\uC758\\uC131\\uC744 \\uD3C9\\uAC00\\uD560 \\uB54C Knapp \\uBC0F Hartung \\uC870\\uC815\\uC744 \\uC801\\uC6A9\\uD558\\uACE0 \\uC790\\uC720\\uB3C4\\uB97C \\uAC16\\uB294 t-\\uBD84\\uD3EC\\uB97C \\uAE30\\uBC18\\uC73C\\uB85C \\uD14C\\uC2A4\\uD2B8 \\uD1B5\\uACC4\\uB7C9, \\uD45C\\uC900 \\uC624\\uCC28, p-value \\uBC0F \\uC2E0\\uB8B0 \\uAD6C\\uAC04\\uC744 \\uACC4\\uC0B0\\uD588\\uB294\\uB370, \\uC5EC\\uAE30\\uC11C u\\uB294 effect size \\uD074\\uB7EC\\uC2A4\\uD130 \\uC218 (\\uC989, \\uC2E4\\uD5D8 \\uC218)\\uB97C \\uB098\\uD0C0\\uB0B4\\uACE0 k\\uB294 \\uBAA8\\uB378\\uC758 \\uACC4\\uC218 \\uC218\\uB97C \\uB098\\uD0C0\\uB0C5\\uB2C8\\uB2E4. moderator \\uBD84\\uC11D\\uC744 \\uC218\\uD589\\uD558\\uAE30 \\uC704\\uD574 \\uAC01 moderator \\uBCC0\\uC218\\uC5D0 \\uB300\\uD574 \\uBCC4\\uB3C4\\uC758 \\uBA54\\uD0C0 \\uD68C\\uADC0\\uB97C \\uC218\\uD589\\uD588\\uC2B5\\uB2C8\\uB2E4. Effect size\\uC758 \\uC774\\uC9C8\\uC131 \\uC815\\uB3C4\\uB97C \\uD574\\uC11D\\uD558\\uAE30 \\uC704\\uD574 \\uD45C\\uBCF8 \\uCD94\\uCD9C\\uC5D0\\uC11C \\uBE44\\uB86F\\uB418\\uC9C0 \\uC54A\\uC740 effect size\\uC758 \\uBCC0\\uB3D9 \\uBC31\\uBD84\\uC728\\uC744 \\uC815\\uB7C9\\uD654\\uD558\\uB294 \", \"[65]\", \"\\uC5D0 \\uB530\\uB77C \\uB110\\uB9AC \\uC0AC\\uC6A9\\uB418\\uB294 I2 \\uD1B5\\uACC4\\uB7C9\\uC744 \\uACC4\\uC0B0\\uD588\\uC2B5\\uB2C8\\uB2E4.\")), mdx(\"p\", null, mdx(\"strong\", {\n    parentName: \"p\"\n  }, \"Bias Test\"), \"\\n\\uC5F0\\uAD6C\\uC790\\uB4E4\\uC774 \\uC0C1\\uB2F9\\uD55C human-AI synergy\\uC758 \\uC99D\\uAC70\\uB97C \\uBCF4\\uC5EC\\uC8FC\\uB294 \\uC2E4\\uD5D8\\uC744 \\uADF8\\uB807\\uC9C0 \\uC54A\\uC740 \\uC2E4\\uD5D8\\uBCF4\\uB2E4 \\uB354 \\uC790\\uC8FC \\uBC1C\\uD45C\\uD55C\\uB2E4\\uBA74 publication bias\\uAC00 \\uBC1C\\uC0DD\\uD560 \\uC218 \\uC788\\uC2B5\\uB2C8\\uB2E4. \\uC774\\uB7EC\\uD55C \\uC704\\uD5D8\\uC744 \\uD3C9\\uAC00\\uD558\\uAE30 \\uC704\\uD574 Viechtbauer\\uC640 Cheung \", \"[66]\", \"\\uC774 \\uC81C\\uC2DC\\uD55C \\uC9C4\\uB2E8 \\uC808\\uCC28\\uB97C \\uCC44\\uD0DD\\uD588\\uC2B5\\uB2C8\\uB2E4.\"), mdx(\"p\", null, mdx(\"strong\", {\n    parentName: \"p\"\n  }, \"Sensitivity Analysis\"), \"\\n\\uAC15\\uAC74\\uC131 \\uAC80\\uC0AC\\uB85C\\uC11C \\uC2E4\\uD5D8 \\uC218\\uC900 \\uB300\\uC2E0 \\uB17C\\uBB38 \\uC218\\uC900\\uC5D0\\uC11C \\uD074\\uB7EC\\uC2A4\\uD130\\uB9C1\\uD558\\uB294 \\uBBFC\\uAC10\\uB3C4 \\uC7AC\\uBD84\\uC11D\\uC744 \\uC218\\uD589\\uD588\\uC2B5\\uB2C8\\uB2E4. \\uC774 \\uB2E4\\uB2E8\\uACC4 \\uBAA8\\uB378\\uC740 \\uAD00\\uCC30\\uB41C \\uD6A8\\uACFC \\uD06C\\uAE30\\uC758 \\uBD84\\uC0B0(1\\uB2E8\\uACC4), \\uB3D9\\uC77C\\uD55C \\uB17C\\uBB38\\uC758 \\uD6A8\\uACFC \\uD06C\\uAE30 \\uAC04\\uC758 \\uBD84\\uC0B0(2\\uB2E8\\uACC4) \\uBC0F \\uB17C\\uBB38 \\uAC04\\uC758 \\uBD84\\uC0B0(3\\uB2E8\\uACC4)\\uC744 \\uC124\\uBA85\\uD569\\uB2C8\\uB2E4. \\uC6B0\\uB9AC\\uB294 \\uC5EC\\uC804\\uD788 \\uCC38\\uAC00\\uC790 \\uC0D8\\uD50C\\uC774 \\uC2E4\\uD5D8 \\uC218\\uC900\\uC5D0\\uC11C\\uB9CC \\uACB9\\uCCE4\\uAE30 \\uB54C\\uBB38\\uC5D0 \\uC2E4\\uD5D8 \\uC218\\uC900\\uC5D0\\uC11C \\uD074\\uB7EC\\uC2A4\\uD130\\uB97C \\uC815\\uC758\\uD558\\uC5EC \\uD6A8\\uACFC \\uD06C\\uAE30 \\uCD94\\uC815\\uC5D0 \\uB300\\uD55C \\uD074\\uB7EC\\uC2A4\\uD130-\\uAC15\\uAC74 \\uD45C\\uC900 \\uC624\\uCC28, \\uC2E0\\uB8B0 \\uAD6C\\uAC04 \\uBC0F \\uD1B5\\uACC4\\uC801 \\uAC80\\uC815\\uC744 \\uACC4\\uC0B0\\uD588\\uC2B5\\uB2C8\\uB2E4.\"), mdx(\"h1\", {\n    \"id\": \"discussion\"\n  }, \"Discussion\"), mdx(\"p\", null, mdx(\"strong\", {\n    parentName: \"p\"\n  }, \"Performance loses from Human-AI collaboration\"), \"\\n\\uC778\\uAC04-AI \\uADF8\\uB8F9\\uC740 \\uC778\\uAC04 \\uB2E8\\uB3C5 \\uB610\\uB294 AI \\uB2E8\\uB3C5\\uBCF4\\uB2E4 \\uC131\\uB2A5\\uC774 \\uC88B\\uC9C0 \\uC54A\\uC558\\uC2B5\\uB2C8\\uB2E4. \\uC774 \\uACB0\\uACFC\\uB294 \\uC778\\uAC04-AI \\uD611\\uC5C5\\uC5D0 \\uB300\\uD55C \\uC9C8\\uC801 \\uBB38\\uD5CC \\uAC80\\uD1A0\\uB97C \\uBCF4\\uC644\\uD558\\uBA70, \\uC774\\uB294 \\uC778\\uAC04\\uACFC \\uC778\\uACF5 \\uC9C0\\uB2A5\\uC744 \\uD1B5\\uD569\\uD560 \\uB54C \\uBC1C\\uC0DD\\uD558\\uB294 \\uB180\\uB77C\\uC6B4 \\uC5B4\\uB824\\uC6C0 \\uC911 \\uC77C\\uBD80\\uB97C \\uAC15\\uC870\\uD569\\uB2C8\\uB2E4. \\uC608\\uB97C \\uB4E4\\uC5B4, \\uC0AC\\uB78C\\uB4E4\\uC740 \\uC885\\uC885 AI \\uC2DC\\uC2A4\\uD15C\\uC5D0 \\uB108\\uBB34 \\uB9CE\\uC774 \\uC758\\uC874(\\\"\\uACFC\\uC789 \\uC758\\uC874\\\")\\uD558\\uC5EC \\uCD94\\uAC00 \\uC815\\uBCF4\\uB97C \\uCC3E\\uACE0 \\uCC98\\uB9AC\\uD558\\uC9C0 \\uC54A\\uACE0\\uB3C4 AI\\uC758 \\uC81C\\uC548\\uC744 \\uAC15\\uB825\\uD55C \\uC9C0\\uCE68\\uC73C\\uB85C \\uC0AC\\uC6A9\\uD569\\uB2C8\\uB2E4. \\uADF8\\uB7EC\\uB098 \\uB54C\\uB85C\\uB294 \\uC778\\uAC04\\uC774 \\uC790\\uB3D9\\uD654\\uC5D0 \\uB300\\uD55C \\uBD80\\uC815\\uC801\\uC778 \\uD0DC\\uB3C4 \\uB54C\\uBB38\\uC5D0 AI\\uB97C \\uB108\\uBB34 \\uC801\\uAC8C \\uC758\\uC874(\\\"\\uACFC\\uC18C \\uC758\\uC874\\\")\\uD558\\uC5EC AI\\uC758 \\uC81C\\uC548\\uC744 \\uBB34\\uC2DC\\uD569\\uB2C8\\uB2E4.\"), mdx(\"p\", null, mdx(\"strong\", {\n    parentName: \"p\"\n  }, \"Moderating effect(\\uC870\\uC808 \\uD6A8\\uACFC) of task type\"), \"\\n\\uC778\\uAC04-AI \\uD611\\uC5C5\\uC758 \\uD6A8\\uACFC\\uC5D0 \\uC601\\uD5A5\\uC744 \\uBBF8\\uCE58\\uB294 \\uC694\\uC778\\uC5D0 \\uB300\\uD55C \\uCD94\\uAC00 \\uBD84\\uC11D\\uB3C4 \\uC218\\uD589\\uD588\\uC2B5\\uB2C8\\uB2E4. \\uACFC\\uC81C \\uC720\\uD615\\uC774 \\uC778\\uAC04-AI \\uC2DC\\uC2A4\\uD15C\\uC758 \\uC2DC\\uB108\\uC9C0\\uB97C \\uD06C\\uAC8C \\uC870\\uC808\\uD558\\uC5EC \\uC758\\uC0AC \\uACB0\\uC815 \\uACFC\\uC81C\\uB294 \\uC131\\uACFC \\uC190\\uC2E4\\uACFC \\uAD00\\uB828\\uB418\\uACE0 \\uC0DD\\uC131 \\uACFC\\uC81C\\uB294 \\uC131\\uACFC \\uD5A5\\uC0C1\\uACFC \\uAD00\\uB828\\uB41C\\uB2E4\\uB294 \\uAC83\\uC744 \\uBC1C\\uACAC\\uD588\\uC2B5\\uB2C8\\uB2E4\"), mdx(\"p\", null, mdx(\"strong\", {\n    parentName: \"p\"\n  }, \"Moderating Effect of Relative Human/AI Performance\"), \"\\n\\uC778\\uAC04\\uC774 AI\\uBCF4\\uB2E4 \\uB2E8\\uB3C5\\uC73C\\uB85C \\uC131\\uB2A5\\uC774 \\uB6F0\\uC5B4\\uB0A0 \\uB54C, \\uC778\\uAC04-AI \\uC2DC\\uC2A4\\uD15C\\uC5D0\\uC11C \\uC131\\uB2A5 \\uD5A5\\uC0C1\\uC774 \\uBC1C\\uC0DD\\uD588\\uC2B5\\uB2C8\\uB2E4. \\uC774\\uB7EC\\uD55C \\uBC1C\\uACAC\\uC740 \\uC778\\uAC04-AI \\uC131\\uB2A5\\uC774 \\uC778\\uAC04 \\uB2E8\\uB3C5\\uACFC AI \\uB2E8\\uB3C5\\uC758 \\uB2E8\\uC21C \\uD3C9\\uADE0\\uC73C\\uB85C \\uC124\\uBA85\\uB420 \\uC218 \\uC5C6\\uB2E4\\uB294 \\uAC83\\uC744 \\uBCF4\\uC5EC\\uC90D\\uB2C8\\uB2E4\"), mdx(\"p\", null, mdx(\"strong\", {\n    parentName: \"p\"\n  }, \"Surprisingly Insignificant Moderators\"), \"\\n\\uC124\\uBA85\\uC758 \\uC874\\uC7AC \\uC5EC\\uBD80, AI \\uCD9C\\uB825\\uC758 \\uC2E0\\uB8B0\\uB3C4 \\uD3EC\\uD568 \\uC5EC\\uBD80, \\uD3C9\\uAC00\\uB41C \\uCC38\\uAC00\\uC790 \\uC720\\uD615\\uACFC \\uAC19\\uC740 \\uB2E4\\uB978 \\uC870\\uC808 \\uBCC0\\uC218\\uB3C4 \\uC870\\uC0AC\\uD588\\uC2B5\\uB2C8\\uB2E4. \\uC774\\uB7EC\\uD55C \\uC694\\uC18C\\uB4E4\\uC740 \\uCD5C\\uADFC \\uBA87 \\uB144 \\uB3D9\\uC548 \\uB9CE\\uC740 \\uC8FC\\uBAA9\\uC744 \\uBC1B\\uC544\\uC654\\uC2B5\\uB2C8\\uB2E4. \\uC6B0\\uB9AC \\uACB0\\uACFC\\uC5D0 \\uB530\\uB974\\uBA74, 300\\uAC1C \\uC774\\uC0C1\\uC758 \\uD6A8\\uACFC \\uD06C\\uAE30\\uC5D0\\uC11C \\uD3C9\\uADE0\\uC801\\uC73C\\uB85C \\uC774\\uB7EC\\uD55C \\uC694\\uC18C\\uB4E4\\uC774 \\uC778\\uAC04-AI \\uD611\\uC5C5\\uC758 \\uD6A8\\uACFC\\uC5D0 \\uC601\\uD5A5\\uC744 \\uBBF8\\uCE58\\uC9C0 \\uC54A\\uC74C\\uC744 \\uBC1C\\uACAC\\uD558\\uC600\\uB2E4.\"), mdx(\"h2\", {\n    \"id\": \"limitations\"\n  }, \"Limitations\"), mdx(\"p\", null, \"\\uCCAB\\uC9F8, \\uC800\\uD76C\\uC758 \\uC815\\uB7C9\\uC801 \\uACB0\\uACFC\\uB294 \\uCCB4\\uACC4\\uC801\\uC778 \\uBB38\\uD5CC \\uAC80\\uD1A0\\uB97C \\uD1B5\\uD574 \\uC218\\uC9D1\\uD55C \\uC5F0\\uAD6C\\uB4E4\\uC758 \\uBD80\\uBD84\\uC9D1\\uD569\\uC5D0 \\uC801\\uC6A9\\uB429\\uB2C8\\uB2E4. \\uB458\\uC9F8, \\uC800\\uD76C\\uB294 \\uCC38\\uAC00\\uC790 \\uAC04\\uC758 \\uBD84\\uC0B0 \\uD568\\uC218\\uB85C \\uD6A8\\uACFC \\uD06C\\uAE30\\uC5D0 \\uAC00\\uC911\\uCE58\\uB97C \\uBD80\\uC5EC\\uD558\\uBBC0\\uB85C, \\uCE21\\uC815\\uC5D0\\uC11C \\uC774\\uB7EC\\uD55C \\uB2E4\\uB978 \\uBCC0\\uB3D9 \\uC6D0\\uC778\\uC744 \\uACE0\\uB824\\uD558\\uC9C0 \\uC54A\\uC2B5\\uB2C8\\uB2E4. \\uC14B\\uC9F8, \\uC800\\uD76C\\uB294 \\uCD9C\\uD310 \\uD3B8\\uD5A5\\uC758 \\uC99D\\uAC70\\uB97C \\uCC3E\\uC9C0 \\uBABB\\uD588\\uC9C0\\uB9CC, \\uADF8\\uAC83\\uC774 \\uC874\\uC7AC\\uD560 \\uAC00\\uB2A5\\uC131\\uC774 \\uC788\\uC73C\\uBA70, \\uB137\\uC9F8, \\uC5F0\\uAD6C\\uC790\\uB4E4\\uC774 \\uC5F0\\uAD6C\\uD558\\uAE30\\uB85C \\uC120\\uD0DD\\uD55C \\uACFC\\uC5C5, \\uD504\\uB85C\\uC138\\uC2A4 \\uBC0F \\uD53C\\uD5D8\\uC790 \\uD480\\uC5D0\\uB9CC \\uC801\\uC6A9\\uB418\\uBA70, \\uC5F0\\uAD6C \\uC8FC\\uC81C \\uC120\\uD0DD \\uD3B8\\uD5A5\\uC774 \\uC791\\uC6A9\\uD560 \\uC218 \\uC788\\uC2B5\\uB2C8\\uB2E4. \\uB2E4\\uC12F\\uC9F8, \\uC800\\uD76C \\uBD84\\uC11D\\uC758 \\uD488\\uC9C8\\uC740 \\uC800\\uD76C\\uAC00 \\uC885\\uD569\\uD55C \\uC5F0\\uAD6C\\uC758 \\uD488\\uC9C8\\uC5D0 \\uB2EC\\uB824 \\uC788\\uC2B5\\uB2C8\\uB2E4. \\uB9C8\\uC9C0\\uB9C9\\uC73C\\uB85C, \\uC800\\uD76C\\uB294 \\uBD84\\uC11D\\uC5D0\\uC11C \\uD6A8\\uACFC \\uD06C\\uAE30 \\uAC04\\uC758 \\uB192\\uC740 \\uC218\\uC900\\uC758 \\uC774\\uC9C8\\uC131\\uC744 \\uBC1C\\uACAC\\uD569\\uB2C8\\uB2E4.\"), mdx(\"h2\", {\n    \"id\": \"a-roadmap-for-future-work-finding-human-ai-synergy\"\n  }, \"A roadmap for future work: Finding Human-AI synergy\"), mdx(\"p\", null, \"\\uC8FC\\uC694 \\uACB0\\uACFC\\uB294 \\uD3C9\\uADE0\\uC801\\uC73C\\uB85C \\uC778\\uAC04\\uACFC AI\\uB97C \\uACB0\\uD569\\uD558\\uBA74 \\uC131\\uB2A5 \\uC190\\uC2E4\\uC774 \\uBC1C\\uC0DD\\uD55C\\uB2E4\\uB294 \\uAC83\\uC744 \\uC2DC\\uC0AC\\uD558\\uC9C0\\uB9CC, \\uC774\\uAC83\\uC774 \\uC778\\uAC04\\uACFC AI\\uB97C \\uACB0\\uD569\\uD558\\uB294 \\uAC83\\uC774 \\uB098\\uC05C \\uC0DD\\uAC01\\uC774\\uB77C\\uB294 \\uC758\\uBBF8\\uB294 \\uC544\\uB2C8\\uB77C\\uACE0 \\uC0DD\\uAC01\\uD569\\uB2C8\\uB2E4. \\uC624\\uD788\\uB824, \\uC774\\uB294 \\uD5A5\\uD6C4 \\uC5F0\\uAD6C\\uAC00 \\uC778\\uAC04\\uACFC AI\\uB97C \\uD1B5\\uD569\\uD558\\uB294 \\uD6A8\\uACFC\\uC801\\uC778 \\uD504\\uB85C\\uC138\\uC2A4\\uB97C \\uCC3E\\uB294 \\uB370 \\uB354 \\uC9D1\\uC911\\uD574\\uC57C \\uD568\\uC744 \\uC758\\uBBF8\\uD55C\\uB2E4\\uACE0 \\uC0DD\\uAC01\\uD569\\uB2C8\\uB2E4.\\xA0 \\uBE44\\uAD50 \\uAC00\\uB2A5\\uC131 \\uBC0F \\uC5F0\\uAD6C \\uD1B5\\uD569\\uC744 \\uB354\\uC6B1 \\uCD09\\uC9C4\\uD558\\uAE30 \\uC704\\uD574, \\uC6B0\\uB9AC\\uB294 \\uC774 \\uBD84\\uC57C\\uAC00 \\uD2B9\\uD788 human-AI \\uD611\\uC5C5 \\uC2E4\\uD5D8\\uC744 \\uC704\\uD55C \\uD45C\\uC900\\uD654\\uB418\\uACE0 \\uAC1C\\uBC29\\uB41C \\uBCF4\\uACE0 \\uB9AC\\uD3EC\\uC9C0\\uD1A0\\uB9AC\\uB97C \\uAD6C\\uCD95\\uD558\\uB3C4\\uB85D \\uC7A5\\uB824\\uD569\\uB2C8\\uB2E4.\"));\n}\n;\nMDXContent.isMDXComponent = true;","frontmatter":{"title":"When Are Combinations of Humans and AI Useful? - A Systematic Review and Meta-Analysis","tags":["paper"],"emoji":"","link":null,"modified":"May 14, 2025","modifiedTimestamp":"2025-05-14T00:00:00.000Z"},"references":[],"fields":{"slug":"/paper-study-haic-2505142330"},"tableOfContents":{"items":[{"url":"#when-are-combinations-of-humans-and-ai-useful---a-systematic-review-and-meta-analysis","title":"When Are Combinations of Humans and AI Useful? - A Systematic Review and Meta-Analysis"},{"url":"#abstract","title":"Abstract"},{"url":"#citation","title":"Citation"},{"url":"#background","title":"Background"},{"url":"#methodology","title":"Methodology","items":[{"url":"#literature-review-문헌-검토","title":"Literature review (문헌 검토)"},{"url":"#data-analysis","title":"Data Analysis"}]},{"url":"#discussion","title":"Discussion","items":[{"url":"#limitations","title":"Limitations"},{"url":"#a-roadmap-for-future-work-finding-human-ai-synergy","title":"A roadmap for future work: Finding Human-AI synergy"}]}]},"parent":{"relativePath":"paper-study-haic-2505142330.md"}}},"pageContext":{"id":"e9efa302-f99c-5474-bca6-0621d3216b3c","previous":{"id":"a04abe85-1ea2-5e2d-bb77-ff8c8fff4697","frontmatter":{"title":"2026년 5~7월 트레이딩 AI 에이전트 최신 연구 서베이","tags":["paper"]},"fields":{"slug":"/paper-study-haic-2607222200"}},"next":{"id":"2cbfc7aa-a469-5c03-8e62-b6fae2fc170a","frontmatter":{"title":"인간-AI 협업 연구 트렌드 2025; 증강지능 시대의 새로운 패러다임","tags":["paper"]},"fields":{"slug":"/paper-study-haic-2505282230"}},"hasUntagged":false,"basePath":"/","tags":[{"tag":"code","totalCount":1,"slug":"/code","path":"/tag/code"},{"tag":"paper","totalCount":7,"slug":"/paper","path":"/tag/paper"}]}},
    "staticQueryHashes": ["1308211557","2744905544","467212769"]}