![\"\"](\"https:\/\/www.hashtag.al\/wp-content\/uploads\/2026\/02\/mustafa.jpg\")
<\/p>\n<\/p>\n
Mustafa Suleyman, CEO i Microsoft AI, ka parashikuar se Inteligjenca Artificiale (IA) do t\u00eb jet\u00eb e aft\u00eb t\u00eb automatizoj\u00eb shumic\u00ebn d\u00ebrrmuese t\u00eb detyrave profesionale t\u00eb \u201cjakave t\u00eb bardha\u201d (punonj\u00ebsve t\u00eb zyrave) brenda 12 deri n\u00eb 18 muajve t\u00eb ardhsh\u00ebm.<\/span><\/p>\n The Financial Times raporton se Mustafa Suleyman, i cili drejton divizionin e IA n\u00eb Microsoft, ka b\u00ebr\u00eb nj\u00eb parashikim t\u00eb guximsh\u00ebm mbi ndikimin afatshkurt\u00ebr t\u00eb IA n\u00eb profesionet intelektuale.<\/span><\/p>\n N\u00eb nj\u00eb intervist\u00eb p\u00ebr Times t\u00eb publikuar k\u00ebt\u00eb jav\u00eb, Suleyman deklaroi se pret q\u00eb shumica, n\u00ebse jo t\u00eb gjitha detyrat e kryera nga punonj\u00ebsit e zyrave, do t\u00eb automatizohen plot\u00ebsisht nga IA brenda 12 deri n\u00eb 18 muajve.<\/span><\/p>\n Sipas Suleyman, sistemet e IA do t\u00eb arrijn\u00eb performanc\u00eb n\u00eb nivel njer\u00ebzor n\u00eb nj\u00eb gam\u00eb t\u00eb gjer\u00eb detyrash profesionale.<\/span><\/p>\n \u201cMendoj se do t\u00eb kemi performanc\u00eb t\u00eb nivelit njer\u00ebzor n\u00eb shumic\u00ebn, n\u00ebse jo n\u00eb t\u00eb gjitha detyrat profesionale,\u201d tha Suleyman n\u00eb intervist\u00eb. \u201cPra, puna e \u2018jakave t\u00eb bardha\u2019, ku ulesh para nj\u00eb kompjuteri, qoft\u00eb si jurist, kontabilist, menaxher projekti apo person n\u00eb marketing \u2014 shumica e k\u00ebtyre detyrave do t\u00eb automatizohen plot\u00ebsisht nga nj\u00eb IA brenda 12 deri n\u00eb 18 muajve t\u00eb ardhsh\u00ebm.\u201d<\/em><\/span><\/p>\n Inxhinieria e softuerit si tregues i par\u00eb<\/strong><\/span><\/p>\n Shefi i Microsoft AI vuri n\u00eb dukje inxhinierin\u00eb e softuerit si nj\u00eb tregues t\u00eb hersh\u00ebm t\u00eb k\u00ebtij trendi. Ai vuri re se programuesit po p\u00ebrdorin tashm\u00eb kodimin e asistuar nga IA p\u00ebr shumic\u00ebn e prodhimit t\u00eb kodit t\u00eb tyre, gj\u00eb q\u00eb p\u00ebrfaq\u00ebson nj\u00eb ndryshim themelor n\u00eb m\u00ebnyr\u00ebn se si kryhet puna. \u201c\u00cbsht\u00eb nj\u00eb marr\u00ebdh\u00ebnie krejt\u00ebsisht ndryshe me teknologjin\u00eb, dhe kjo ka ndodhur n\u00eb gjasht\u00eb muajt e fundit,\u201d tha ai.<\/span><\/p>\n P\u00ebrparimi i shpejt\u00eb i inteligjenc\u00ebs artificiale gjat\u00eb viteve t\u00eb fundit ka filluar tashm\u00eb t\u00eb transformoj\u00eb pun\u00ebn e zyrave n\u00eb m\u00ebnyra t\u00eb dukshme. Raportimet e fundit kan\u00eb theksuar shfaqjen e asaj q\u00eb disa e quajn\u00eb \u201clodhja nga IA\u201d (AI fatigue) mes inxhinier\u00ebve t\u00eb softuerit, ku teknologjia ka sjell\u00eb rritje t\u00eb produktivitetit, por edhe rritje t\u00eb rraskapitjes, pasi punonj\u00ebsit p\u00ebrballen me presionin p\u00ebr t\u00eb p\u00ebrballuar ngarkesa m\u00eb t\u00eb m\u00ebdha pune n\u00eb m\u00ebnyr\u00eb simultane.<\/span><\/p>\n Rreziqet e siguris\u00eb dhe \u201chalucinacionet\u201d<\/strong><\/span><\/p>\n Microsoft \u00ebsht\u00eb pozicionuar n\u00eb ball\u00eb t\u00eb integrimit t\u00eb IA n\u00eb vendin e pun\u00ebs. Kompania ka zhvilluar produkte si Copilot dhe ka b\u00ebr\u00eb investime t\u00eb r\u00ebnd\u00ebsishme n\u00eb kompani si OpenAI dhe Anthropic.<\/span><\/p>\n Megjithat\u00eb, Breitbart Ne\u00ebs ka raportuar m\u00eb par\u00eb se Microsoft e ka integruar IA n\u00eb \u00cbindo\u00ebs pa i kuptuar plot\u00ebsisht rreziqet e siguris\u00eb q\u00eb kjo krijon. Shqet\u00ebsimet e siguris\u00eb burojn\u00eb nga defektet e njohura n\u00eb shumic\u00ebn e modeleve t\u00eb m\u00ebdha gjuh\u00ebsore (LLM). K\u00ebrkuesit kan\u00eb demonstruar vazhdimisht se LLM-t\u00eb mund t\u00eb japin p\u00ebrgjigje faktikisht t\u00eb gabuara dhe ilogjike, nj\u00eb sjellje e njohur si \u201chalucinacione.\u201d Kjo do t\u00eb thot\u00eb se p\u00ebrdoruesit nuk mund t\u2019u besojn\u00eb plot\u00ebsisht rezultateve t\u00eb asistent\u00ebve si Copilot, Gemini ose Claude, dhe duhet t\u2019i verifikojn\u00eb informacionet n\u00eb m\u00ebnyr\u00eb t\u00eb pavarur.<\/span><\/p>\n Nj\u00eb tjet\u00ebr problem \u00ebsht\u00eb vulnerabiliteti ndaj \u201cinjeksioneve t\u00eb komandave\u201d (prompt injections), ku hakerat mund t\u00eb shfryt\u00ebzojn\u00eb k\u00ebt\u00eb dob\u00ebsi p\u00ebr t\u00eb mbjell\u00eb udh\u00ebzime dashakeqe n\u00eb faqe interneti apo email-e, t\u00eb cilat IA i ndjek pa b\u00ebr\u00eb dallimin mes komandave t\u00eb vlefshme dhe p\u00ebrmbajtjes s\u00eb pasigurt.<\/span><\/p>\n Paralajm\u00ebrimet p\u00ebr papun\u00ebsi masive<\/strong><\/span><\/p>\n Parashikimi i Suleyman bashkohet me z\u00ebrat e lider\u00ebve t\u00eb tjer\u00eb t\u00eb industris\u00eb. Stuart Russell, nj\u00eb shkenc\u00ebtar kompjuteri i shquar, deklaroi vitin e kaluar se udh\u00ebheq\u00ebsit politik\u00eb po p\u00ebrballen me mund\u00ebsin\u00eb e nj\u00eb papun\u00ebsie prej 80 p\u00ebr qind t\u00eb shkaktuar nga IA. Russell sugjeroi se pozicionet q\u00eb variojn\u00eb nga kirurg\u00ebt deri te drejtor\u00ebt ekzekutiv\u00eb mund t\u00eb jen\u00eb n\u00eb rrezik z\u00ebvend\u00ebsimi, shkruan Breitbart<\/a>.<\/span><\/p>\n N\u00eb m\u00ebnyr\u00eb t\u00eb ngjashme, Dario Amodei, CEO i Anthropic, ka paralajm\u00ebruar se IA mund t\u00eb eliminoj\u00eb p\u00ebraf\u00ebrsisht gjysm\u00ebn e t\u00eb gjitha pozicioneve t\u00eb nivelit fillestar (entry-level) p\u00ebr \u201cjakat e bardha\u201d. \u201cNe, si prodhues t\u00eb k\u00ebsaj teknologjie, kemi detyr\u00ebn dhe obligimin t\u00eb jemi t\u00eb sinqert\u00eb p\u00ebr at\u00eb q\u00eb po vjen,\u201d tha Amodei. \u201cNuk mendoj se kjo \u00ebsht\u00eb ende n\u00eb radarin e njer\u00ebzve.\u201d<\/span><\/p>\n","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"excerpt":{"rendered":" Mustafa Suleyman, CEO i Microsoft AI, ka parashikuar se Inteligjenca Artificiale (IA) do t\u00eb jet\u00eb e aft\u00eb t\u00eb automatizoj\u00eb shumic\u00ebn d\u00ebrrmuese t\u00eb detyrave profesionale t\u00eb \u201cjakave t\u00eb bardha\u201d (punonj\u00ebsve t\u00eb zyrave) brenda 12 deri n\u00eb 18 muajve t\u00eb ardhsh\u00ebm. The Financial Times raporton se Mustafa Suleyman, i cili drejton divizionin e IA n\u00eb Microsoft, ka […]<\/p>\n","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"author":51,"featured_media":786208,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[53],"tags":[],"class_list":["post-786207","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-fast-foodi"],"gt_translate_keys":[{"key":"link","format":"url"}],"_links":{"self":[{"href":"https:\/\/www.hashtag.al\/index.php\/wp-json\/wp\/v2\/posts\/786207","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.hashtag.al\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.hashtag.al\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.hashtag.al\/index.php\/wp-json\/wp\/v2\/users\/51"}],"replies":[{"embeddable":true,"href":"https:\/\/www.hashtag.al\/index.php\/wp-json\/wp\/v2\/comments?post=786207"}],"version-history":[{"count":99999,"href":"https:\/\/www.hashtag.al\/index.php\/wp-json\/wp\/v2\/posts\/786207\/revisions"}],"predecessor-version":[{"id":786209,"href":"https:\/\/www.hashtag.al\/index.php\/wp-json\/wp\/v2\/posts\/786207\/revisions\/786209"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.hashtag.al\/index.php\/wp-json\/wp\/v2\/media\/786208"}],"wp:attachment":[{"href":"https:\/\/www.hashtag.al\/index.php\/wp-json\/wp\/v2\/media?parent=786207"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.hashtag.al\/index.php\/wp-json\/wp\/v2\/categories?post=786207"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.hashtag.al\/index.php\/wp-json\/wp\/v2\/tags?post=786207"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}