\u201cThere\u2019s a lot of very smart people who have concluded that the opportunities for workers outweigh the risks, that it\u2019s about augmenting people\u2019s work, not always just automating it, and so the responsibility that we embrace is to make this work for people, not against them, and make them beneficiaries, not victims of this change, including in the workforce.\u201d<\/p>\n
\n\t\tRelated Article Block Placeholder<\/p>\n
\t\t\tArticle ID: 1215661\n\t<\/p>\n
\t\t\t\t\t\t In the newest Paint by Numbers<\/a>, Crikey breaks down the bewildering statistics behind the impact of AI and the wildest predictions about it \u2014 plus a few of our own.<\/p>\n ***<\/p>\n \tIndependent. Irreverent. In your inbox<\/p>\n \n Get the headlines they don\u2019t want you to read. Sign up to Crikey\u2019s free newsletters for fearless reporting, sharp analysis, and a touch of chaos <\/p>\n By continuing, you agree to our\u00a0Terms & Conditions<\/a>\u00a0and\u00a0Privacy Policy<\/a>.<\/p>\n The amount of money AI could inject into the Australian economy over the next decade, according to today\u2019s Productivity Commission report: $116 billion<\/a><\/p>\n Portion of Australian jobs at risk by 2030, according to the Australian Council of Trade Unions: 1 in 3<\/a><\/p>\n Portion of all energy in the US that currently goes toward AI data centres: 4.4%<\/a><\/p>\n Amount of electricity in the US going to data centres that will be used for AI by 2028: More than half<\/a><\/p>\n Portion of worldwide energy use AI is projected to account for by 2027: 0.5%<\/a><\/p>\n Portion of worldwide energy that the entirety of Argentina accounted for in 2023: 0.426%<\/a><\/p>\n Portion of US households\u2019 electricity use estimated to be used on AI by 2028: 22%<\/a><\/p>\n The amount of water projected to be used by global AI demand in 2027:\u00a04.2\u20136.6 trillion litres<\/a><\/p>\n Amount Apple plans to spend on manufacturing and data centres in the US\u00a0over the next four years: US$500 billion<\/a>\u00a0<\/p>\n Amount Amazon plans to spend to expand data centre infrastructure in Australia:\u00a0$20 billion by 2029<\/a><\/p>\n Number of new data centres likely to be built as part of OpenAI\u2019s US$500 billion Stargate initiative:\u00a0Up to 10<\/a>\u00a0<\/p>\n \n\t\tRelated Article Block Placeholder<\/p>\n \t\t\tArticle ID: 1214425\n\t<\/p>\n \t\t\t\t\t\t Peak of contribution to US GDP growth by spending on AI infrastructure over the past six months:\u00a0Nearly 2.75%<\/a><\/p>\n Consumer spending contribution to US GDP growth over the same period:\u00a0Around 0.75%<\/a><\/p>\n Number of takes it took Jackie Chan to perfect the \u201cshuttlecock<\/a>\u201d scene in 1982\u2019s Dragon Lord: Around 2,900<\/a><\/p>\n Number of classic Kung Fu films \u2014 including Chan\u2019s \u2014 being \u201creimagined\u201d by AI, according to a collection of Chinese studios: Around 100<\/a><\/p>\n Number of \u201csongs\u201d generated by AI start-up Boomy: 14.5 million<\/a><\/p>\n Number of years it would take to listen to those AI-generated songs (assuming the average pop song length of 3 minutes and 30 seconds<\/a>): 96.6<\/p>\n Time it took Michelangelo to paint the Sistine Chapel: Roughly four years<\/p>\n Time it took Crikey to generate the following image of the Sistine Chapel, reworked to feature 1980s pop duo The Proclaimers: Roughly four seconds<\/p>\n \tIs AI usage already getting out of control?<\/p>\n
\n\t\t\t\t![\"Sportsbet\u2019s](\"https:\/\/www.newsbeep.com\/au\/wp-content\/uploads\/2025\/08\/1754698403_93_cornesgpt.png\")
\t\t\t<\/a><\/p>\n
\n\t\t\t\t![\"AI](\"https:\/\/www.newsbeep.com\/au\/wp-content\/uploads\/2025\/08\/AdobeStock_585824945-copy.jpg\")
\t\t\t<\/a><\/p>\n![\"\"](\"https:\/\/www.newsbeep.com\/au\/wp-content\/uploads\/2025\/08\/ChatGPT-Image-Aug-6-2025-10_07_36-AM.png\")
<\/p>\n