\nRecent, pe 30 noiembrie, ChatGPT a \u00eemplinit trei ani de la lansare. Debutul s\u0103u \u00een 2022 a constituit una dintre cele mai importante transform\u0103ri tehnologice ale secolului: pentru prima dat\u0103, un sistem automatizat putea conversa, elabora texte, imagini sau prezent\u0103ri la fel de natural ca un om. \u00cens\u0103, dincolo de entuziasm, r\u0103m\u00e2ne o \u00eentrebare fundamental\u0103: c\u00e2t de \u201einteligent\u0103\u201d este, \u00een realitate, inteligen\u0163a artificial\u0103?<\/p>\n
\nCercet\u0103ri recente subliniaz\u0103 un aspect crucial: modelele lingvistice nu au o \u00een\u0163elegere real\u0103 a lumii. Ele doar se limiteaz\u0103 la a genera propozi\u0163ii care par coerente. Acestea identific\u0103 \u015fabloane statistice, f\u0103r\u0103 a procesa semnifica\u0163ii. De exemplu, un studiu condus de echipe de la universit\u0103\u0163ile Ca\u2019 Foscari \u015fi Cardiff, prezentat \u00een La Stampa, a ar\u0103tat c\u0103 sistemele AI nu reu\u015fesc s\u0103 detecteze ironia sau jocurile de cuvinte \u2013 nuan\u0163e care depind de context, inten\u0163ionalitate \u015fi interpretare, elemente ce dep\u0103\u015fesc simpla analiz\u0103 lingvistic\u0103.<\/p>\n
\nDe ce este etichetat\u0103 aceast\u0103 tehnologie drept \u201einteligent\u0103\u201d, dac\u0103 nu posed\u0103 inteligen\u0163\u0103 \u00een accep\u0163iunea uman\u0103? Explica\u0163ia este direct\u0103: gigan\u0163ii din tehnologie au un interes major \u00een a sus\u0163ine o poveste grandioas\u0103. Afirma\u0163ii precum \u201ene apropiem de superinteligen\u0163\u0103\u201d, \u201evom face descoperiri extraordinare\u201d sau \u201evom extinde durata vie\u0163ii cu ajutorul IA\u201d servesc la captarea aten\u0163iei, atragerea investi\u0163iilor \u015fi men\u0163inerea interesului public. Totu\u015fi, astfel de promisiuni, de\u015fi atractive, nu au la baz\u0103 dovezi \u015ftiin\u0163ifice concrete.<\/p>\n
\nAceasta nu este prima situa\u0163ie \u00een care o tehnologie este promovat\u0103 drept o solu\u0163ie miraculoas\u0103. \u00cen anii 2000, internetul era prezentat ca un instrument universal pentru consolidarea democra\u0163iei \u015fi accesul la educa\u0163ie. Consecin\u0163ele reale au fost \u00eens\u0103 diferite: inegalit\u0103\u0163ile s-au ad\u00e2ncit, dezinformarea a proliferat, iar corpora\u0163iile din Silicon Valley au acumulat o putere f\u0103r\u0103 precedent. Ast\u0103zi, un discurs similar este construit \u00een jurul inteligen\u0163ei artificiale.<\/p>\n
\nInteligen\u0163a este mult mai complex\u0103 dec\u00e2t simpla st\u0103p\u00e2nire a limbajului. Neuro\u015ftiin\u0163a \u015fi psihologia arat\u0103 c\u0103 procese precum g\u00e2ndirea, creativitatea \u015fi \u00een\u0163elegerea mediului se dezvolt\u0103 cu mult \u00eenainte ca un copil s\u0103 poat\u0103 vorbi. Dup\u0103 cum explic\u0103 psiholoaga Alison Gopnik, cei mici investigheaz\u0103, formuleaz\u0103 ipoteze \u015fi analizeaz\u0103 realitatea asemenea unor cercet\u0103tori \u00een miniatur\u0103, totul \u00eenainte de a st\u0103p\u00e2ni limbajul. \u00cen opozi\u0163ie, IA generativ\u0103 se limiteaz\u0103 la a imita vorbirea, lipsindu-i mecanismele fundamentale ale cogni\u0163iei umane.<\/p>\n
\nA\u015fa cum scria Galileo Galilei, \u201enumele trebuie s\u0103 corespund\u0103 lucrurilor, nu lucrurile numelor\u201d. Aceast\u0103 observa\u0163ie este extrem de relevant\u0103 \u00een contextul actual. De\u015fi inteligen\u0163a artificial\u0103 este o tehnologie remarcabil\u0103, ea nu posed\u0103 inteligen\u0163\u0103 \u00een accep\u0163iunea uman\u0103. Pentru a evalua corect impactul s\u0103u social, este esen\u0163ial s\u0103 \u00eencet\u0103m s\u0103 ne l\u0103s\u0103m indu\u015fi \u00een eroare de o denumire atr\u0103g\u0103toare, dar \u00een\u015fel\u0103toare.<\/p>\n","protected":false},"excerpt":{"rendered":"
De trei ani, Inteligen\u0163a artificial\u0103 generativ\u0103 s-a integrat \u00een peisajul nostru digital. Totu\u015fi, succesul ei fulminant se sprijin\u0103 pe o ne\u00een\u0163elegere fundamental\u0103: confuzia dintre capacitatea de a simula limbajul uman \u015fi inteligen\u0163a autentic\u0103. Datele \u015ftiin\u0163ifice contrazic aceast\u0103 percep\u0163ie.<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[510],"tags":[25316,7853],"class_list":["post-242823","post","type-post","status-publish","format-standard","hentry","category-actualitate","tag-inteligenta-artificiala","tag-studiu"],"_links":{"self":[{"href":"https:\/\/bm.dev.synology.me\/index.php?rest_route=\/wp\/v2\/posts\/242823","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/bm.dev.synology.me\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/bm.dev.synology.me\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/bm.dev.synology.me\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/bm.dev.synology.me\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=242823"}],"version-history":[{"count":0,"href":"https:\/\/bm.dev.synology.me\/index.php?rest_route=\/wp\/v2\/posts\/242823\/revisions"}],"wp:attachment":[{"href":"https:\/\/bm.dev.synology.me\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=242823"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/bm.dev.synology.me\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=242823"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/bm.dev.synology.me\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=242823"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}