AI Summit_Sept. 13 2024

How Social Media Algorithms Assist in Spreading MDM 8THNFQ RJINF FQLTWNYMRX \JFUTSN_J MZRFS UX^HMTQTL^ G^ UWNTWNYN_NSL HTSYJSY YMFY NX QNPJQ^ YT LJS JWFYJ MNLM JSLFLJRJSY KWTR ZXJWX YMNX TKYJS NSHQZIJX XJSXFYNTSFQ HTSYWT[JWXNFQ TW JRTYNTSFQQ^ charged content, including MDM (as the algorithms don’t discern between accurate and inaccu WFYJ NSKTWRFYNTS 78 9MJ FQLTWNYMRX YFNQTW HTSYJSY KJJIX YT NSIN[NIZFQ ZXJWX UWJKJWJSHJX FSI UFXY interactions, creating echo chambers where users are repeatedly exposed to similar viewpoints, NSHQZINSL UTYJSYNFQQ^ MFWRKZQ 2)2 9MJ FQLTWNYMRX FWJ IJXNLSJI YT UWTRTYJ HTSYJSY YMFY NX QNPJQ^ YT LJSJWFYJ XYWTSL WJFHYNTSX 9MNX HFS QJFI YT YMJ FRUQN ܪ HFYNTS TK J]YWJRJ [NJ\X NSHQZINSL RNXNSKTWRFYNTS FSI HTSXUNWFH^ YMJTWNJX FX YMJXJ TKYJS UWT[TPJ RTWJ JSLFLJRJSY YMFS RTIJWFYJ TW SZFSHJI HTSYJSY 9MJ FQLTWNYMRX do not differentiate between credible information and MDM, as long as the content keeps users JSLFLJI 79 Social media algorithms exploit human psychological ten dencies by amplifying types of information from which hu mans are biased to learn from, i.e. prestigious, in-group, moral, and emotional information (or “PRIME”). 8THNFQ RJINF FQLTWNYMRX J]UQTNY MZRFS UX^HMTQTLNHFQ YJSIJSHNJX G^ FRUQNK^NSL Y^UJX TK NSKTWRF YNTS KWTR \MNHM MZRFSX FWJ GNFXJI YT QJFWS KWTR N J UWJXYNLNTZX NS LWTZU RTWFQ FSI JRTYNTSFQ NSKTWRFYNTS TW Ѧ57.2*ѧ 9MNX J]UQTNYFYNTS HFS QJFI YT YMJ XUWJFI TK 2)2 FX ZXJWX FWJ RTWJ QNPJQ^ to engage with and share content that aligns with their views or sparks strong emotional reactions J L UTXYX YMFY UWT[TPJ RTWFQ TZYWFLJ WJLFWIQJXX TK NYX [JWFHNY^ 80 How Fake Followers, Sock Puppets, and Troll Farms Are Used Especially in social media, the proliferation of fake followers, sock puppets, and troll farms rep WJXJSYX F XNLSN ܪ HFSY YMWJFY YT NSKTWRFYNTS NSYJLWNY^ Fake Followers +FPJ KTQQT\JWX FWJ FWYN ܪ HNFQ XTHNFQ RJINF FHHTZSYX TKYJS HWJFYJI ZXNSL GTYX NS ܫ FYNSL YMJ SZRGJW TK FHHTZSY KTQQT\JWX 9MNX HFS HWJFYJ YMJ NQQZXNTS TK UTUZQFWNY^ TW NS ܫ ZJSHJ FSI XUWJFI INX RNX FSI RFQ NSKTWRFYNTS TW UWTUFLFSIF 9JFR /TWLJ IJRTSXYWFYJX YMJ [FQZJ TK KFPJ KTQQT\JWX YT HWJFYTWX TK 2)2 9JFR /TWLJ NX F UWN [FYJ .XWFJQN YFXP KTWHJ XUJHNFQN_NSL NS H^GJW FHYN[NYNJX NSHQZINSL MFHPNSL XFGTYFLJ FSI GTY KFWRJI XTHNFQ RJINF INXNSKTWRFYNTS HFRUFNLSX ZXJI YT RFSNUZQFYJ YMJ WJXZQYX TK JQJHYNTSX 9MJ LWTZUѣX FHYN[NYNJX \JWJ J]UTXJI NS +JGWZFW^ FKYJW OTZWSFQNXYX HTSIZHYJI F XJHWJY ZSIJWHT[JW TUJWF YNTS 9MJ LWTZU MFX GJJS FHYN[J XNSHJ FY QJFXY FSI HQFNRX YT MF[J RFSNUZQFYJI UWJXNIJS YNFQ JQJHYNTSX FWTZSI YMJ \TWQI NS HFXJX FQQJLJIQ^ XZHHJXXKZQQ^ 4SJ TK YMJ HTRUFS^ѣX RTXY JX XJSYNFQ YTTQX NX YMJ XTKY\FWJ UWTLWFR &I[FSHJI .RUFHY 2JINF 8TQZYNTSX &.28 81

78 https://neurosciencenews.com/social-media-behavior-misinformation-23752/. 79 https://www.fastcompany.com/90943919/the-science-behind-why-social-media-algorithms-warp-our-view-of-the-world. 80 https://www.fastcompany.com/90943919/the-science-behind-why-social-media-algorithms-warp-our-view-of-the-world. 81 MYYUX \\\ YMJLZFWINFS HTR \TWQI KJG FNRX XTKY\FWJ F[FYFWX YJFR OTWLJ INXNSKTWRFYNTS KFPJ UWT ܪ QJX .

14 Data Mining, Algorithms, and Deepfakes AI Roundtable Page 170