<\/p>\n <\/a><\/p><\/div>\n Mandati i tij p\u00ebrfshin mbik\u00ebqyrjen e Google Research dhe Google Labs.<\/p>\n Pra, ai \u00ebsht\u00eb njeriu pas inovacioneve m\u00eb ambicioze t\u00eb Google n\u00eb AI, informatik\u00eb dhe shkenc\u00eb me p\u00ebrgjegj\u00ebsi.<\/p>\n Po ashtu ai bashk\u00eb-kryeson Bordin K\u00ebshillimor t\u00eb AI t\u00eb Kombeve t\u00eb Bashkuara , i krijuar p\u00ebr t\u00eb ndihmuar komunitetin nd\u00ebrkomb\u00ebtar t\u00eb qeveris\u00eb inteligjenc\u00ebn artificiale.<\/p>\n N\u00eb nj\u00eb intervist\u00eb t\u00eb dh\u00ebn\u00eb p\u00ebr Corriere della Sera, e para intervist\u00eb me nj\u00eb gazet\u00eb italiane, Manyika flet p\u00ebr zhvillimet e frikshme q\u00eb AI do t\u2019i sjell\u00eb njer\u00ebzimit, nga sh\u00ebndeti te lufta kund\u00ebr ndryshimeve klimatike, duke na lejuar t\u00eb krijojm\u00eb materiale t\u00eb reja, p\u00ebrfshir\u00eb ato p\u00ebr nd\u00ebrtimin e baterive elektrike, ose p\u00ebr t\u00eb parashikuar p\u00ebrmbytjet shtat\u00eb dit\u00eb p\u00ebrpara.<\/p>\n Pa fshehur rreziqet e k\u00ebsaj teknologjie t\u00eb re, e cila \u201cduhet t\u00eb rregullohet dhe t\u00eb qeveriset n\u00eb nivel global\u201d.<\/p>\n A mund t\u00eb na jepni nj\u00eb ide se \u00e7far\u00eb na pret fal\u00eb AI-s\u00eb n\u00eb t\u00eb ardhmen e af\u00ebrt?<\/em><\/p>\n \u201cNj\u00eb nga gj\u00ebrat mahnit\u00ebse \u00ebsht\u00eb se kjo teknologji l\u00ebviz kaq shpejt. Por do t\u00eb doja t\u00eb filloja me disa gj\u00ebra q\u00eb kemi b\u00ebr\u00eb s\u00eb fundmi. N\u00eb dhjetor t\u00eb vitit t\u00eb kaluar ne prezantuam Gemini Ultra, i cili ishte modeli i par\u00eb n\u00eb bot\u00eb q\u00eb ia kalonte njer\u00ebzve n\u00eb t\u00eb ashtuquajtur\u00ebn MMLU (Massive Multitask Language Understanding)\u201d.<\/p>\n \u00c7far\u00eb do t\u00eb thot\u00eb?<\/em><\/p>\n \u201cMMLU \u00ebsht\u00eb nj\u00eb pik\u00eb referimi e r\u00ebnd\u00ebsishme, sepse analizon t\u00eb kuptuarit e arsyetimit, nga ligji n\u00eb matematik\u00eb, n\u00eb 57 fusha. N\u00ebse dikush do t\u00eb kishte par\u00eb k\u00ebt\u00eb pik\u00eb referimi nj\u00eb vit m\u00eb par\u00eb, sistemet m\u00eb t\u00eb mira t\u00eb hapura t\u00eb AI do t\u00eb kishin arritur nj\u00eb rezultat prej rreth 35-36%, krahasuar me nivelin e ekspertiz\u00ebs njer\u00ebzore, i cili arrin nj\u00eb rezultat prej rreth 88%. Gemini, me 90%, \u00ebsht\u00eb modeli i par\u00eb q\u00eb ia kalon. Do t\u00eb p\u00ebrmend tre gj\u00ebra t\u00eb tjera q\u00eb kan\u00eb ndodhur n\u00eb k\u00ebt\u00eb fush\u00eb n\u00eb tre jav\u00ebt e fundit, t\u00eb cilat mund t\u00eb japin nj\u00eb ndjenj\u00eb shpejt\u00ebsie. Por s\u00eb pari duhet t\u00eb b\u00ebj nj\u00eb premis\u00eb\u201d.<\/p>\n Po. Thuaje. <\/em><\/p>\n \u201cGj\u00ebja e par\u00eb \u00ebsht\u00eb se modeli n\u00eb vetvete \u00ebsht\u00eb b\u00ebr\u00eb multimodal. Do t\u00eb thot\u00eb q\u00eb i nj\u00ebjti model tani nuk prodhon m\u00eb vet\u00ebm tekst, por video, imazhe, audio dhe n\u00eb t\u00eb dy drejtimet. Ju shkruani tekst dhe merrni nj\u00eb imazh ose anasjelltas; ju vendosni nj\u00eb imazh dhe merrni nj\u00eb tekst. Dhe at\u00ebher\u00eb tani kemi t\u00eb ashtuquajturin \u2018kontekst i gjat\u00eb\u2019: p\u00ebrpara se t\u00eb dh\u00ebnat t\u00eb ishin t\u00eb kufizuara, tani p\u00ebr her\u00eb t\u00eb par\u00eb kemi demonstruar se mund t\u00eb ofrojm\u00eb nj\u00eb kontekst p\u00ebr deri n\u00eb 2 milion argumente. \u00cbsht\u00eb befasuese. Mund t\u00eb vendos nj\u00eb video 3-or\u00ebshe si k\u00ebrkes\u00eb dhe shablloni shpjegon se p\u00ebr \u00e7far\u00eb b\u00ebhet fjal\u00eb. Kjo jo vet\u00ebm q\u00eb ju lejon t\u00eb merrni rezultate shum\u00eb m\u00eb t\u00eb sakta, por mund t\u00eb arsyetoni me modelin dhe t\u2019i kuptoni m\u00eb mir\u00eb problemet. Nj\u00eb vit m\u00eb par\u00eb kjo ishte teknikisht e pamundur\u201d.<\/p>\n Cilat jan\u00eb tre pikat e kthes\u00ebs q\u00eb p\u00ebrmend\u00ebt?<\/em><\/p>\n \u201cTre jav\u00eb m\u00eb par\u00eb ne prezantuam nj\u00eb model t\u00eb quajtur Med Gemini, i krijuar posa\u00e7\u00ebrisht p\u00ebr fush\u00ebn e mjek\u00ebsis\u00eb. P\u00ebrve\u00e7 gj\u00ebrave baz\u00eb, ajo tani kupton imazhet e radiologjis\u00eb, grupet e t\u00eb dh\u00ebnave t\u00eb ADN-s\u00eb dhe at\u00eb lloj gj\u00ebje dhe \u00ebsht\u00eb shum\u00eb m\u00eb mir\u00eb se shumica e ekspert\u00ebve mjek\u00ebsor\u00eb njer\u00ebzor\u00eb n\u00eb 11 nga 15 standardet\u201d.<\/p>\n Po i dyti?<\/em><\/p>\n \u201cDy jav\u00eb m\u00eb par\u00eb prezantuam Alpha Fold 3. Ne prezantuam Alpha Fold rreth nj\u00eb vit m\u00eb par\u00eb: ishte programi q\u00eb zgjidh 58 nga problemet e m\u00ebdha t\u00eb biologjis\u00eb. P\u00ebr shembull: t\u00eb kuptuarit e struktur\u00ebs s\u00eb proteinave. T\u00eb kuptuarit e procesit t\u00eb nj\u00eb proteine \u200b\u200bt\u00eb vetme zakonisht k\u00ebrkon 3 ose 4 vjet pun\u00eb n\u00eb laborator dhe p\u00ebrpara se biolog\u00ebt e Alpha Fold t\u00eb kishin kuptuar struktur\u00ebn e rreth 110 mij\u00eb proteinave n\u00eb mbar\u00eb bot\u00ebn. Nj\u00eb vit m\u00eb par\u00eb Alpha Fold parashikoi me sakt\u00ebsi struktur\u00ebn e t\u00eb gjitha proteinave t\u00eb njohura p\u00ebr shkenc\u00ebn, af\u00ebrsisht 200 milion\u00eb. Por dy jav\u00eb m\u00eb par\u00eb njoftuam se Alpha Fold 3 jo vet\u00ebm q\u00eb kupton struktur\u00ebn e proteinave, por mund t\u00eb kuptoj\u00eb edhe t\u00eb gjitha molekulat, ADN-n\u00eb, ARN-n\u00eb, etj\u201d.<\/p>\n \u201cN\u00eb biologji, Alpha Fold 3 \u00ebsht\u00eb k\u00ebrkimi n\u00eb Google p\u00ebr proteinat. Dhe \u00ebsht\u00eb falas dhe i disponuesh\u00ebm p\u00ebr t\u00eb gjith\u00eb. Sot, 1.8 milion\u00eb biolog\u00eb n\u00eb 190 vende p\u00ebrdorin Alpha Fold. Disa po punojn\u00eb p\u00ebr vaksinat e malaries, por edhe p\u00ebr ila\u00e7e t\u00eb tjera\u201d.<\/p>\n Dhe gj\u00ebja e tret\u00eb?<\/em><\/p>\n \u201c\u00cbsht\u00eb harta 3D e sinapseve t\u00eb trurit t\u00eb njeriut Ka ende shum\u00eb pun\u00eb p\u00ebr t\u00eb b\u00ebr\u00eb, por t\u00eb kuptuarit e struktur\u00ebs s\u00eb trurit hap mund\u00ebsi t\u00eb paprecedent\u00eb n\u00eb neuroshkenc\u00eb. Ndodhi vet\u00ebm nj\u00eb jav\u00eb m\u00eb par\u00eb. Lista e progresit \u00ebsht\u00eb shum\u00eb e gjat\u00eb. Gjasht\u00eb muaj m\u00eb par\u00eb, programi yn\u00eb i AI n\u00eb shkencat materiale zbuloi 2.2 milion\u00eb kristale t\u00eb reja q\u00eb nuk kan\u00eb ekzistuar kurr\u00eb m\u00eb par\u00eb. Nga k\u00ebto 380 mij\u00eb jan\u00eb mjaftuesh\u00ebm t\u00eb q\u00ebndrueshme p\u00ebr t\u2019u sintetizuar\u201d.<\/p>\n A thua se mund t\u00eb sintetizohen edhe materiale t\u00eb reja p\u00ebr t\u00eb z\u00ebvend\u00ebsuar tokat e rralla?<\/em><\/p>\n \u201cKa disa kompani q\u00eb tashm\u00eb po p\u00ebrpiqen t\u00eb sintetizojn\u00eb disa kristale p\u00ebr t\u00eb prodhuar materiale t\u00eb reja p\u00ebr teknologjin\u00eb e baterive p\u00ebr panelet diellore. Por un\u00eb mund t\u00eb jap shembuj t\u00eb modelimit t\u00eb klim\u00ebs dhe shum\u00eb gj\u00ebra t\u00eb tjera: aplikimet e AI n\u00eb shkenc\u00eb jan\u00eb thjesht fantastike. Le t\u00eb marrim parashikimin e p\u00ebrmbytjeve. Deri m\u00eb tani, meteorolog\u00ebt kan\u00eb dh\u00ebn\u00eb vet\u00ebm 2 ose 3 dit\u00eb paralajm\u00ebrim. \u00cbsht\u00eb nj\u00eb moment historik sepse \u00ebsht\u00eb koha e mjaftueshme p\u00ebr t\u00eb shp\u00ebtuar jet\u00eb njer\u00ebzore, aq m\u00eb e r\u00ebnd\u00ebsishme duke pasur parasysh shpesht\u00ebsin\u00eb me t\u00eb cil\u00ebn ato ndodhin p\u00ebr shkak t\u00eb ndryshimeve klimatike\u201d.<\/p>\n \u201cNe filluam me nj\u00eb projekt pilot nj\u00eb vit e gjysm\u00eb m\u00eb par\u00eb n\u00eb Bangladesh dhe m\u00eb pas e zgjeruam at\u00eb n\u00eb t\u00eb gjith\u00eb Indin\u00eb jugore. Sot ne e p\u00ebrdorim at\u00eb n\u00eb 83 vende t\u00eb bot\u00ebs, 28 prej t\u00eb cilave jan\u00eb n\u00eb Afrik\u00eb, por ne gjithashtu punojm\u00eb n\u00eb disa vende t\u00eb Evrop\u00ebs dhe n\u00eb disa zona t\u00eb SHBA-ve. Parashikimi i nj\u00eb p\u00ebrmbytjeje nj\u00eb jav\u00eb p\u00ebrpara ka nj\u00eb ndikim t\u00eb madh n\u00eb shoq\u00ebri, por p\u00ebrfitimi p\u00ebr vendet me t\u00eb ardhura t\u00eb ul\u00ebta \u00ebsht\u00eb shum\u00eb m\u00eb i r\u00ebnd\u00ebsish\u00ebm\u201d.<\/p>\n Inteligjenca artificiale mund t\u00eb ndryshoj\u00eb ngadal\u00ebsimin e produktivitetit dhe t\u00eb nxis\u00eb nj\u00eb revolucion t\u00eb ri ekonomik. Duke pasur parasysh parashikimet e zymta t\u00eb rritjes globale, \u00e7far\u00eb do t\u00eb duhet p\u00ebr t\u00eb par\u00eb nj\u00eb ndikim n\u00eb ekonomi dhe sa koh\u00eb do t\u00eb zgjas\u00eb ?<\/em><\/p>\n \u201cPotenciali p\u00ebr rritjen e produktivitetit \u00ebsht\u00eb i konsideruesh\u00ebm. Por q\u00eb ndikimi i rritjes t\u00eb ndodh\u00eb shpejt, AI duhet t\u00eb adoptohet nga sektor\u00eb mjaft t\u00eb m\u00ebdhenj t\u00eb ekonomis\u00eb, si kujdesi sh\u00ebndet\u00ebsor, transporti, etj. Megjithat\u00eb, deri m\u00eb tani, AI \u00ebsht\u00eb p\u00ebrqafuar kryesisht nga sektori i teknologjis\u00eb, i cili p\u00ebr shembull n\u00eb SHBA p\u00ebrfaq\u00ebson vet\u00ebm 4% t\u00eb fuqis\u00eb pun\u00ebtore. Kjo industri \u00ebsht\u00eb shum\u00eb produktive, por nuk ka r\u00ebnd\u00ebsi sepse \u00ebsht\u00eb shum\u00eb e vog\u00ebl\u201d.<\/p>\n \u201cP\u00ebr m\u00eb tep\u00ebr, ne nuk duhet t\u00eb harrojm\u00eb m\u00ebsimet e s\u00eb kaluar\u00ebs, ku ne shohim AI kudo, p\u00ebrve\u00e7 n\u00eb statistikat e produktivitetit. Do t\u00eb duhet koh\u00eb dhe shum\u00eb ndryshime n\u00eb sektor\u00ebt e ekonomis\u00eb, kompanit\u00eb do t\u00eb duhet t\u00eb modifikojn\u00eb proceset dhe strukturat p\u00ebr t\u00eb p\u00ebrfituar nga teknologjia e re, do t\u00eb duhet t\u00eb riorganizojn\u00eb m\u00ebnyr\u00ebn e pun\u00ebs\u201d.<\/p>\n \u201cEdhe fuqia pun\u00ebtore do t\u00eb duhet t\u00eb ndryshoj\u00eb. Pra, nevojitet nj\u00eb adoptim i gjer\u00eb dhe nj\u00eb riorganizim i thell\u00eb. E gjith\u00eb kjo k\u00ebrkon koh\u00eb. Por varet gjithashtu nga lloji i produktivitetit q\u00eb duam\u201d.<\/p>\n \u00c7far\u00eb do t\u00eb thot\u00eb?<\/em><\/p>\n \u201cDuhet t\u00eb pyesim veten se p\u00ebr \u00e7far\u00eb e p\u00ebrdorim inteligjenc\u00ebn artificiale. N\u00ebse nj\u00eb kompani miraton AI vet\u00ebm p\u00ebr t\u00eb reduktuar fuqin\u00eb e saj pun\u00ebtore, por jo p\u00ebr t\u00eb rritur prodhimin, kjo nuk ka asnj\u00eb ndikim n\u00eb rritje. Thjesht po b\u00ebhet m\u00eb efikase, jo domosdoshm\u00ebrisht m\u00eb produktive. Por nj\u00eb gj\u00eb e kat\u00ebrt \u00ebsht\u00eb gjithashtu e nevojshme.<\/p>\n Kush?<\/em><\/p>\n \u201cPolitikan\u00ebt duhet ta b\u00ebjn\u00eb t\u00eb gjith\u00eb k\u00ebt\u00eb t\u00eb ndodh\u00eb, politik\u00ebb\u00ebr\u00ebsit duhet t\u00eb krijojn\u00eb stimujt e duhur p\u00ebr p\u00ebrdorimin e teknologjis\u00eb. Ata gjithashtu duhet t\u00eb inkurajojn\u00eb investimet plot\u00ebsuese, duke p\u00ebrfshir\u00eb infrastruktur\u00ebn. Sepse pa mund\u00ebsuar infrastruktur\u00ebn, vet\u00ebm teknologjia nuk ka r\u00ebnd\u00ebsi. N\u00ebse do t\u00eb arriheshin k\u00ebto kat\u00ebr gj\u00ebra, mendoj se ndikimi n\u00eb ekonomi do t\u00eb ishte v\u00ebrtet shum\u00eb i shpejt\u00eb\u201d.<\/p>\n Shum\u00eb kan\u00eb frik\u00eb se shum\u00eb vende pune do t\u00eb plot\u00ebsohen nga AI. Ne tashm\u00eb po shohim q\u00eb kompanit\u00eb e teknologjis\u00eb, p\u00ebrfshir\u00eb Google, t\u00eb pushojn\u00eb nga puna shum\u00eb njer\u00ebz. A parashikoni val\u00eb t\u00eb m\u00ebdha pushimesh nga puna t\u00eb pakt\u00ebn n\u00eb faz\u00ebn e tranzicionit? Cil\u00ebt jan\u00eb sektor\u00ebt m\u00eb t\u00eb rrezikuar? Cila \u00ebsht\u00eb e ardhmja e pun\u00ebs, pasi ju bashk\u00ebkryesuat Komisionin p\u00ebr t\u00eb Ardhmen e Pun\u00ebs n\u00eb Kaliforni n\u00eb vitin 2019?<\/em><\/p>\n \u201cShkarkimet nga puna n\u00eb teknologji nuk kan\u00eb t\u00eb b\u00ebjn\u00eb fare me t\u00eb. Ato jan\u00eb kryesisht nj\u00eb korrigjim pas bumit t\u00eb p\u00ebrjetuar gjat\u00eb pandemis\u00eb, kur sektori u rrit jasht\u00ebzakonisht shum\u00eb. Por \u00ebsht\u00eb e v\u00ebrtet\u00eb q\u00eb duhet t\u2019i b\u00ebjm\u00eb vetes pyetjen se cila \u00ebsht\u00eb e ardhmja e pun\u00ebs. Shumica e hulumtimeve tregojn\u00eb se disa vende pune do t\u00eb ulen, t\u00eb tjera do t\u00eb rriten, shumica do t\u00eb ndryshojn\u00eb. M\u00eb shum\u00eb vende pune ka t\u00eb ngjar\u00eb t\u00eb krijohen sesa do t\u00eb humbasim. Nuk do t\u00eb thot\u00eb q\u00eb disa vende pune nuk do t\u00eb bien, por rezultati neto do t\u00eb jet\u00eb pozitiv. Megjithat\u00eb, rikualifikimi i pun\u00ebtor\u00ebve do t\u00eb jet\u00eb vendimtar. Hulumtimi i ILO-s (Organizata Bot\u00ebrore e Pun\u00ebs), i kryer me t\u00eb dh\u00ebna nga 150 vende, arrin n\u00eb p\u00ebrfundimin se inteligjenca artificiale gjeneruese nuk do t\u00eb shkat\u00ebrroj\u00eb vendet e pun\u00ebs, por do t\u2019i rris\u00eb ato gjasht\u00ebfish. Nj\u00eb artikull nga Universiteti Stanford argumenton se ajo q\u00eb ka m\u00eb shum\u00eb r\u00ebnd\u00ebsi sesa dizajni i teknologjis\u00eb \u00ebsht\u00eb m\u00ebnyra se si p\u00ebrdoret. E kam fjal\u00ebn p\u00ebr pun\u00ebdh\u00ebn\u00ebsit. Prandaj duhet t\u00eb mendojm\u00eb edhe se cilat duhet t\u00eb jen\u00eb stimujt p\u00ebr pun\u00ebdh\u00ebn\u00ebsit. Kjo do t\u00eb b\u00ebj\u00eb nj\u00eb ndryshim\u201d.<\/p>\n \u201cSot, stimujt favorizojn\u00eb kapitalin mbi pun\u00ebn, duke favorizuar automatizimin me shkat\u00ebrrimin e vendeve t\u00eb pun\u00ebs n\u00eb vend t\u00eb rritjes s\u00eb aft\u00ebsive njer\u00ebzore. Lajmi i mir\u00eb \u00ebsht\u00eb se studimet e fundit tregojn\u00eb se AI gjeneruese p\u00ebrfiton m\u00eb shum\u00eb pun\u00ebtor\u00ebt m\u00eb pak t\u00eb aft\u00eb, ndryshe nga e kaluara. Edhe ata q\u00eb nuk din\u00eb t\u00eb shkruajn\u00eb kode, mund t\u00eb p\u00ebrdorin AI dhe t\u2019i k\u00ebrkojn\u00eb modelit q\u00eb t\u00eb shkruaj\u00eb kodin p\u00ebr ta\u201d.<\/p>\n \u00c7far\u00eb mund t\u00eb shkoj\u00eb keq? Cilat jan\u00eb rreziqet m\u00eb t\u00eb m\u00ebdha t\u00eb AI-s\u00eb?<\/em><\/p>\n \u201cKa rreziqe dhe kompleksitete. Nd\u00ebr t\u00eb par\u00ebt un\u00eb p\u00ebrfshij rreziqet e performanc\u00ebs. Kjo ndodh kur rezultatet e AI nuk jan\u00eb t\u00eb besueshme. Ne e dim\u00eb se LLM-t\u00eb ndonj\u00ebher\u00eb halucinojn\u00eb dhe mund t\u00eb japin p\u00ebrgjigje t\u00eb pav\u00ebrteta. Ka shum\u00eb nga k\u00ebto probleme t\u00eb performanc\u00ebs, ne jemi ende n\u00eb dit\u00ebt e para t\u00eb teknologjis\u00eb, duhet t\u2019i rregullojm\u00eb sepse mund t\u00eb shkaktojn\u00eb d\u00ebme. P\u00ebr m\u00eb tep\u00ebr, ne kemi rezultate t\u00eb shtremb\u00ebruara nga paragjykimi. Nj\u00eb kategori tjet\u00ebr rreziqesh jan\u00eb ato q\u00eb un\u00eb i quaj keqaplikime dhe keqp\u00ebrdorime, q\u00eb ndodhin kur aktor\u00ebt e k\u00ebqij p\u00ebrdorin teknologjin\u00eb p\u00ebr q\u00ebllime t\u00eb ndryshme nga ato p\u00ebr t\u00eb cilat \u00ebsht\u00eb krijuar p\u00ebr aktivitet kriminal dhe dezinformim, gj\u00eb q\u00eb \u00ebsht\u00eb ve\u00e7an\u00ebrisht e r\u00ebnd\u00ebsishme n\u00eb nj\u00eb vit zgjedhor si ky, miliona njer\u00ebz do t\u00eb votojn\u00eb\u201d.<\/p>\n \u201cP\u00ebr t\u00eb luftuar k\u00ebto, si Google, ne kemi prezantuar nj\u00eb lloj mjeti q\u00eb sh\u00ebnon gjith\u00e7ka q\u00eb \u00ebsht\u00eb krijuar me AI, qoft\u00eb tekst, imazh, video apo audio. Megjithat\u00eb, nd\u00ebrlikimet kan\u00eb t\u00eb b\u00ebjn\u00eb me ndikimin dhe ndryshimet q\u00eb AI shkakton n\u00eb sisteme, nga arsimi te tregu i pun\u00ebs e deri te siguria komb\u00ebtare\u201d.<\/p>\n Nj\u00eb nga problemet me AI \u00ebsht\u00eb edhe se konsumon shum\u00eb energji.<\/em><\/p>\n \u201c\u00cbsht\u00eb e v\u00ebrtet\u00eb. Ka shum\u00eb pun\u00eb p\u00ebr t\u00eb b\u00ebr\u00eb dhe nj\u00eb pjes\u00eb tashm\u00eb \u00ebsht\u00eb duke u b\u00ebr\u00eb. P\u00ebr t\u00eb p\u00ebrmir\u00ebsuar efikasitetin e k\u00ebtyre modeleve dhe gjithashtu p\u00ebr t\u00eb shpikur arkitektura t\u00eb reja q\u00eb nuk jan\u00eb aq intensive nga ana llogarit\u00ebse. \u00cbsht\u00eb nj\u00eb pun\u00eb n\u00eb progres, por mendoj se duhet t\u00eb kujtojm\u00eb gjithashtu se, n\u00eb t\u00eb nj\u00ebjt\u00ebn koh\u00eb, k\u00ebto teknologji do t\u00eb na ndihmojn\u00eb t\u00eb trajtojm\u00eb ndryshimin e klim\u00ebs dhe tranzicionin e energjis\u00eb. Ne kemi par\u00eb se p\u00ebrdorimi i teknologjive t\u00eb AI n\u00eb qendrat e t\u00eb dh\u00ebnave na ka lejuar t\u00eb p\u00ebrmir\u00ebsojm\u00eb efikasitetin e t\u00eb dh\u00ebnave me m\u00eb shum\u00eb se 40%, ne nuk do t\u00eb kishim arritur kurr\u00eb atje pa t\u00eb\u201d.<\/p>\n \u201cPor un\u00eb jam gati t\u00eb pranoj kompromisin p\u00ebr t\u00eb shpenzuar 1% m\u00eb shum\u00eb energji, por n\u00eb t\u00eb nj\u00ebjt\u00ebn koh\u00eb duke kursyer 10 her\u00eb ose 100 her\u00eb m\u00eb shum\u00eb gjurm\u00ebn ton\u00eb t\u00eb karbonit\u201d.<\/p>\n Meq\u00eb bashk\u00eb-kryeson Bordin K\u00ebshillimor t\u00eb AI t\u00eb Kombeve t\u00eb Bashkuara, si po shkon puna? Cilat jan\u00eb aspektet mbi t\u00eb cilat ka konvergjenc\u00eb dhe ato q\u00eb diskutohen n\u00eb komision?<\/em><\/p>\n \u201cDuke qen\u00eb i p\u00ebrb\u00ebr\u00eb nga 39 an\u00ebtar\u00eb nga 33 vende, me personalitete nga akademia, qeverit\u00eb, industria, shoq\u00ebria civile, perspektivat jan\u00eb shum\u00eb t\u00eb ndryshme. Si p\u00ebrmbledhje, mund t\u00eb them se i ashtuquajturi Jug Global, i cili p\u00ebrfshin Amerik\u00ebn Latine, Afrik\u00ebn, Azin\u00eb, n\u00eb p\u00ebrgjith\u00ebsi \u00ebsht\u00eb shum\u00eb m\u00eb optimist p\u00ebr AI dhe mund\u00ebsin\u00eb q\u00eb ajo t\u00eb mund t\u00eb transformoj\u00eb ekonomit\u00eb e tyre, duke p\u00ebrmir\u00ebsuar aksesin n\u00eb njohuri dhe sh\u00ebndet, n\u00eb krahasim me Evrop\u00eb dhe Amerik\u00ebn e Veriut. Kemi arritur pik\u00ebn e par\u00eb\u201d.<\/p>\n \u201cAsambleja e P\u00ebrgjithshme e OKB-s\u00eb pes\u00eb jav\u00eb m\u00eb par\u00eb votoi nj\u00ebz\u00ebri mbi nj\u00eb rezolut\u00eb mbi AI t\u00eb mb\u00ebshtetur kryesisht n\u00eb pun\u00ebn ton\u00eb, e cila prodhoi nj\u00eb Raport t\u00eb P\u00ebrkohsh\u00ebm t\u00eb paraqitur dhjetorin e kaluar, me projekt-rekomandime, duke marr\u00eb reagime t\u00eb gjera pozitive nga shtetet an\u00ebtare. Tani po japim nj\u00eb kontribut n\u00eb procesin e OKB-s\u00eb p\u00ebr p\u00ebrgatitjen e Paktit Digjital global, p\u00ebr t\u00eb cilin shpresohet se do t\u00eb miratohet nga Shtetet An\u00ebtare n\u00eb samitin e shtatorit. Bordi yn\u00eb \u00ebsht\u00eb duke p\u00ebrgatitur Raportin P\u00ebrfundimtar t\u00eb cilin do ta prezantojm\u00eb n\u00eb gusht. Shum\u00eb prisnin q\u00eb ne do t\u00eb rekomandonim nj\u00eb agjenci t\u00eb re p\u00ebr AI. Por ne ram\u00eb dakord q\u00eb nuk na nevojitet\u201d.<\/p>\n Si e vler\u00ebsoni Aktin e AI t\u00eb miratuar nga Bashkimi Evropian?<\/em><\/p>\n \u201cKur idet\u00eb ishin ende duke u formuluar, isha mjaft i shqet\u00ebsuar. Por rezultati p\u00ebrfundimtar \u00ebsht\u00eb m\u00eb i mir\u00eb se pik\u00ebnisja, fal\u00eb pun\u00ebs s\u00eb Italis\u00eb. M\u00eb p\u00eblqeu q\u00eb mor\u00ebn nj\u00eb qasje t\u00eb bazuar n\u00eb rrezik. Megjithat\u00eb, do t\u00eb kisha dashur t\u00eb inkurajoja pak m\u00eb shum\u00eb inovacionin. Pik\u00ebpamja ime, t\u00eb pakt\u00ebn p\u00ebr sa i p\u00ebrket rregulloreve, \u00ebsht\u00eb se ato duhet t\u00eb b\u00ebjn\u00eb gjithmon\u00eb dy gj\u00ebra. Duhet t\u00eb adresojn\u00eb rreziqet dhe sfidat q\u00eb ne nuk duam, por edhe t\u00eb mund\u00ebsojn\u00eb at\u00eb q\u00eb duam, q\u00eb \u00ebsht\u00eb inovacioni dhe avancimi i shkenc\u00ebs dhe ekonomis\u00eb\u201d.<\/p>\n Si mund t\u00eb triumfoj\u00eb Evropa me AI, duke pasur parasysh se lojtar\u00ebt e m\u00ebdhenj t\u00eb teknologjis\u00eb jan\u00eb n\u00eb SHBA dhe Kin\u00eb?<\/em><\/p>\n \u201cSigurisht q\u00eb nuk \u00ebsht\u00eb p\u00ebr shkak t\u00eb munges\u00ebs s\u00eb talentit. Por investimet jan\u00eb nj\u00eb pjes\u00eb e inovacionit. Ashtu si rregullorja q\u00eb, jo vet\u00ebm q\u00eb duhet t\u00eb merret me at\u00eb q\u00eb nuk duam, por edhe t\u00eb lejoj\u00eb at\u00eb q\u00eb duam. N\u00eb k\u00ebt\u00eb rast, shpresoj q\u00eb Evropa d\u00ebshiron t\u00eb b\u00ebhet m\u00eb inovative, m\u00eb transformuese, t\u00eb luaj\u00eb n\u00eb sken\u00ebn globale me kompani q\u00eb i sh\u00ebrbejn\u00eb bot\u00ebs dhe jan\u00eb ekuivalente me Google dhe shum\u00eb t\u00eb tjera\u201d.<\/p>\n N\u00ebse do t\u00eb kishit nj\u00eb djal\u00eb adoleshent, \u00e7far\u00eb do t\u2019i rekomandonit t\u00eb studionte?<\/em><\/p>\n \u201c\u00cbsht\u00eb thelb\u00ebsore t\u00eb studiohen l\u00ebnd\u00ebt STEM, dometh\u00ebn\u00eb shkenca, teknologjia, inxhinieria dhe matematika. Por edhe t\u00eb kombinohet me nj\u00eb njohuri t\u00eb thell\u00eb t\u00eb shkencave humane, sepse nj\u00eb nga gj\u00ebrat q\u00eb po b\u00ebhet e r\u00ebnd\u00ebsishme \u00ebsht\u00eb se shum\u00eb nga pyetjet mbi ndikimin dhe p\u00ebrdorimin e k\u00ebtyre teknologjive jan\u00eb tashm\u00eb nd\u00ebrdisiplinore: \u00ebsht\u00eb e v\u00ebshtir\u00eb t\u00eb ndash, p.sh. k\u00ebrkimin dhe zhvillimin n\u00eb AI nga filozofia dhe etika. Djali im \u00ebsht\u00eb 22 vje\u00e7 dhe jam shum\u00eb i lumtur q\u00eb ka nj\u00eb diplom\u00eb t\u00eb dyfisht\u00eb p\u00ebr informatik\u00eb dhe filozofi\u201d.<\/p>\n \u201cPor mesazhi im i fundit \u00ebsht\u00eb ky: ne duhet t\u00eb m\u00ebsojm\u00eb t\u00eb m\u00ebsojm\u00eb\u201d.<\/p>\n The post Manyika p\u00ebr inteligjenc\u00ebn artificiale: Do t\u00eb ndryshoj\u00eb jet\u00ebn e t\u00eb gjith\u00ebve<\/a> appeared first on Albanian Post<\/a>.<\/p>\n","protected":false},"excerpt":{"rendered":" Ka vite q\u00eb flitet p\u00ebr inteligjenc\u00ebn artificiale. Por tani po d\u00ebshmohet nj\u00eb \u201cp\u00ebrparim spektakolar me nj\u00eb shpejt\u00ebsi t\u00eb jasht\u00ebzakonshme\u201d, n\u00eb mjek\u00ebsi, n\u00eb shkencat e jet\u00ebs, n\u00eb fizik\u00eb. P\u00ebrparime k\u00ebto, q\u00eb kan\u00eb habitur edhe ekspert\u00eb si James Manyika, 58 vje\u00e7, me origjin\u00eb nga Zimbabve, PhD n\u00eb shkenca kompjuterike, AI dhe robotik\u00eb e marr\u00eb 25 vjet m\u00eb […]<\/p>\n","protected":false},"author":1,"featured_media":3605,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[],"class_list":["post-3604","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-uncategorized"],"_links":{"self":[{"href":"https:\/\/radiogjilani.com\/index.php\/wp-json\/wp\/v2\/posts\/3604","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/radiogjilani.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/radiogjilani.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/radiogjilani.com\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/radiogjilani.com\/index.php\/wp-json\/wp\/v2\/comments?post=3604"}],"version-history":[{"count":0,"href":"https:\/\/radiogjilani.com\/index.php\/wp-json\/wp\/v2\/posts\/3604\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/radiogjilani.com\/index.php\/wp-json\/wp\/v2\/media\/3605"}],"wp:attachment":[{"href":"https:\/\/radiogjilani.com\/index.php\/wp-json\/wp\/v2\/media?parent=3604"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/radiogjilani.com\/index.php\/wp-json\/wp\/v2\/categories?post=3604"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/radiogjilani.com\/index.php\/wp-json\/wp\/v2\/tags?post=3604"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}<\/p>\n