{ "docId": "019dd923-5e88-73ef-bd5e-0fcffb6515fe", "docSlug": "483646da9fd73f55", "documentTitle": "2026 University of Sydney The 2026 Skills Horizon", "authorId": "USydney", "authorName": "University of Sydney", "documentKindSlug": "consulting-deck", "documentKindLabel": "Consulting deck", "sourceTypeSlug": "academic_think_tank", "sourceTypeLabel": "Academic / think tank", "presentationDate": null, "orientation": "landscape", "aspectRatio": 1.414, "pageNumber": 50, "pageCount": 98, "prevPage": 49, "nextPage": 51, "slideType": "appendix_methodology", "function": "summarize", "density": "dense", "nDataPoints": 0, "notes": "The slide includes a quote from Dan Clancy (CEO, Twitch) and details the use of TopicGPT for analysis.", "elementsJson": [ "headline_text", "quote_block", "paragraph" ], "metadataConfidence": 1, "imagePath": null, "slideHref": "/slides/019dd923-5e88-73ef-bd5e-0fcffb6515fe/50", "deckHref": "/decks/019dd923-5e88-73ef-bd5e-0fcffb6515fe", "deckJsonHref": "/decks/019dd923-5e88-73ef-bd5e-0fcffb6515fe.json", "deckAnchorHref": "/decks/019dd923-5e88-73ef-bd5e-0fcffb6515fe#slide-50", "components": [ { "bbox": null, "kind": "callout", "text": "For the analysis, we did as we teach—we employed AI. We used large language models (LLMs) for topic modelling across a sample of these five outlets.", "attrs": null, "subkind": null, "toolName": "Visual emphasis", "toolSlug": "visual-emphasis", "confidence": null, "componentId": "019dd952-5644-756a-a85e-ef61c75bc8b5", "frameworkName": null, "frameworkSlug": null }, { "bbox": { "h": 0.085, "w": 0.285, "x": 0.362, "y": 0.431 }, "kind": "paragraph", "text": "This is a behind-the-scenes look at how we identified and analysed the public conversation. We built a corpus that includes the metadata for a number of future-focused outlets (for the period 2020-2025), which allowed us to answer two big questions:", "attrs": null, "subkind": "paragraph", "toolName": null, "toolSlug": null, "confidence": null, "componentId": "4d8bdf14-5f05-4a2e-848c-2697d9ad54d0", "frameworkName": null, "frameworkSlug": null }, { "bbox": { "h": 0.385, "w": 0.285, "x": 0.652, "y": 0.431 }, "kind": "paragraph", "text": "WHAT'S EVERYBODY TALKING ABOUT? If you want to get a feel for the future, there are two big conferences that help. World leaders meet in Davos every year for the World Economic Forum to discuss big changes ahead-as they happen and as they make them. And every March, the who's who of tech, media, and the arts flock to Austin (TX) for SXSW to reveal the future, as driven by technology, culture, and media. For the analysis, we did as we teach-we employed AI. We used large language models (LLMs) for topic modelling across a sample of these five outlets. We used TopicGPT, a prompt-based, topic modelling framework to identify the topics and sub-topics that characterise the conversation. Frequency analysis gave us percentages. We present the top topics for each outlet, and some of the standout insights.", "attrs": null, "subkind": "paragraph", "toolName": null, "toolSlug": null, "confidence": null, "componentId": "7b4572e3-cded-4f2e-93f8-7824f9fdb287", "frameworkName": null, "frameworkSlug": null }, { "bbox": { "h": 0.225, "w": 0.285, "x": 0.362, "y": 0.608 }, "kind": "paragraph", "text": "WHAT'S EVERYBODY READING? Best Business Books of the Year lists help us to see what's exciting everybody. Books take a while to write and publish, but they can have a big impact on the public conversation. Best books lists give us a good sense of where the conversation is coming from and where it might be going. Harvard Business Review and The World Ahead (from The Economist) are faster than a book, but slower than the news cycle. They give us a sense of what's happening when you've got a finger on the economy's pulse.", "attrs": null, "subkind": "paragraph", "toolName": null, "toolSlug": null, "confidence": null, "componentId": "9b85661f-7783-4f41-947e-bdc6d1c863bd", "frameworkName": null, "frameworkSlug": null }, { "bbox": { "h": 0.146, "w": 0.214, "x": 0.071, "y": 0.431 }, "kind": "quote", "text": "The skill that is so difficult, that so many people don't have, is that -when data works, just use data, get more data, do experiments. But for some questions, there are too many relevant factors to collect data. Then use judgement.", "attrs": null, "subkind": "pull-quote", "toolName": null, "toolSlug": null, "confidence": null, "componentId": "dc441eb4-9a63-477d-8fbd-df30887b3e13", "frameworkName": null, "frameworkSlug": null }, { "bbox": null, "kind": "quote", "text": "The skill that is so difficult, that so many people don’t have, is that -when data works, just use data, get more data, do experiments. But for some questions, there are too many relevant factors to collect data. Then use judgement. — Dan Clancy, CEO, Twitch", "attrs": null, "subkind": null, "toolName": "Authority citation", "toolSlug": "authority-citation", "confidence": null, "componentId": "019dd952-5644-756a-a85e-f19fa625ab90", "frameworkName": null, "frameworkSlug": null }, { "bbox": { "h": 0.153, "w": 0.285, "x": 0.358, "y": 0.083 }, "kind": "title", "text": "The world in words", "attrs": null, "subkind": "headline", "toolName": null, "toolSlug": null, "confidence": null, "componentId": "2ec5824a-f488-4e8f-a3a8-63f4f3b31efb", "frameworkName": null, "frameworkSlug": null } ], "metrics": [], "tools": [], "frameworks": [ { "name": "TopicGPT", "slug": null, "matchId": "966c485a-8409-4d29-abb3-6ef726ad8872", "evidence": "Mentioned as the prompt-based topic modelling framework used for analysis.", "confidence": 0.9 } ], "arcBeats": [ { "to": 90, "from": 43, "beatId": "b85339a3-59a4-4909-b322-a7960cca828b", "arcName": "Monroe's Motivated Sequence", "arcSlug": "monroes-sequence", "beatName": "Satisfaction", "beatSlug": "monroes-sequence-satisfaction", "evidence": "The document provides satisfaction by outlining key skills and trends for future leadership.", "position": 2, "confidence": 0.8, "parentBeatName": "Turn", "parentBeatSlug": "turn" } ], "loops": [], "imagePathAlt": null, "thumbSrc": null, "thumbSrcAlt": null, "locked": true }