[analyse_image type=”featured” src=”https:\/\/agerpres.ro\/data\/202609\/20260916101955_273949.jpg”]<\/p>\n
Directorul general al companiei Meta, Mark Zuckerberg, a respins mar\u021bi apelurile tot mai numeroase de suspendare sau m\u0103car de \u00eencetinire a dezvolt\u0103rii Inteligen\u021bei Artificiale (IA), el consider\u00e2nd c\u0103 mecanismele pie\u021bei \u0219i riscul unor ac\u021biuni judiciare sunt cele mai bune garan\u021bii \u00een fa\u021ba acestei tehnologii, relateaz\u0103 AFP.<\/p>\n
‘Oamenii nu vor dori s\u0103 foloseasc\u0103 agen\u021bi care nu sunt alinia\u021bi propriilor interese \u0219i care nu fac ceea ce le cer. Prin urmare, laboratoarele au un puternic stimulent natural s\u0103 fac\u0103 \u00een a\u0219a fel \u00eenc\u00e2t s\u0103-\u0219i alinieze mai mult modelele’ \u00een fa\u021ba cerin\u021belor, a scris Zuckerberg pe re\u021beaua de socializare X.<\/p>\n
Conceptul de ‘aliniere’ se refer\u0103 la eforturile de a face modelele de IA s\u0103 se comporte a\u0219a cum doresc oamenii. ‘Orice laborator care nu se concentreaz\u0103 pe aliniere va r\u0103m\u00e2ne \u00een urm\u0103’, anticipeaz\u0103 \u0219eful companiei Meta.<\/p>\n
\u00cengrijor\u0103rile din jurul IA au crescut \u00een ultimele s\u0103pt\u0103m\u00e2ni, dup\u0103 mai multe incidente de securitate \u0219i a unei noi capacit\u0103\u021bi a modelelor de IA de a se perfec\u021biona singure, f\u0103r\u0103 interven\u021bie uman\u0103.<\/p>\n
Dezbaterea are la ora actual\u0103 \u00een vedere \u0219i capacitatea sectorului de se auto-reglementa. Societ\u0103\u021bi precum OpenAI, Anthropic \u0219i Google studiaz\u0103 crearea unui organism comun de elaborare a unor standarde.<\/p>\n
Mai mul\u021bi reprezentan\u021bi influen\u021bi din sector, condu\u0219i de directorul general al Anthropic, Dario Amodei, pledeaz\u0103 pentru o \u00eencetinire a dezvolt\u0103rii IA.
\n\u00cen schimb, Zuckerberg consider\u0103 c\u0103 amenin\u021barea cu ac\u021biuni legale oblig\u0103 deja laboratoarele s\u0103 ac\u021bioneze \u00een mod responsabil. El d\u0103 ca exemplu \u00een acest sens decizia Meta de a am\u00e2na cu mai multe luni lansarea sistemului s\u0103u Muse AI, prev\u0103zut cu agen\u021bi personaliza\u021bi, pentru a-\u0219i consolida mai \u00eent\u00e2i m\u0103surile de securitate \u0219i siguran\u021b\u0103.<\/p>\n
De asemenea, Mark Zuckerberg s-a pronun\u021bat \u00een favoarea evalu\u0103rilor independente ale sistemelor de IA, subliniind c\u0103 Meta Superintelligence Labs, divizia de Inteligen\u021b\u0103 Artificial\u0103 a companiei Meta, folose\u0219te deja exper\u021bi externi.<\/p>\n
El a cerut crearea unui ‘ecosistem mai vast \u0219i mai diversificat de evaluatori’ pe ansamblul sectorului, o critic\u0103 voalat\u0103 la adresa Anthropic, companie c\u0103reia unii \u00eei repro\u0219eaz\u0103 c\u0103 favorizeaz\u0103 evaluatorii cu care men\u021bine leg\u0103turi str\u00e2nse.<\/p>\n
Compania Meta, care anul acesta va investi p\u00e2n\u0103 la 145 de miliarde de dolari \u00een infrastructura sa de IA, se pronun\u021b\u0103 \u00eempotriva oric\u0103rei reglement\u0103ri susceptibile s\u0103 \u00eencetineasc\u0103 dezvoltarea domeniului. Nicio regul\u0103 nu ar trebui s\u0103 \u00eent\u00e2rzie lansarea modelelor americane, ‘nici m\u0103car cu o lun\u0103’, a indicat deja fondatorul s\u0103u la \u00eenceputul lunii august.<\/p>\n
Opinia lui Mark Zuckerberg a fost salutat\u0103 de mai multe personalit\u0103\u021bi din Silicon Valley apropiate Casei Albe, care contest\u0103 la r\u00e2ndul lor cre\u0219terea aten\u021bion\u0103rilor privind pericolele poten\u021biale pe care le implic\u0103 Inteligen\u021ba Artificial\u0103 \u0219i apelurile de \u00eencetinire a dezvolt\u0103rii acesteia.AGERPRES\/(AS – redactor: Sorin Popescu, editor online: Ady Iva\u015fcu)<\/p>\n<\/div>\n
[analyse_source url=”https:\/\/agerpres.ro\/stiinta\/2026\/09\/16\/mark-zuckerberg-respinge-apelurile-de-incetinire-a-dezvoltarii-inteligentei-artificiale–1593932″]<\/p>\n","protected":false},"excerpt":{"rendered":"
[analyse_image type=”featured” src=”https:\/\/agerpres.ro\/data\/202609\/20260916101955_273949.jpg”] Directorul general al companiei Meta, Mark Zuckerberg, a respins mar\u021bi apelurile tot mai numeroase de suspendare sau m\u0103car de \u00eencetinire a dezvolt\u0103rii Inteligen\u021bei Artificiale (IA), el consider\u00e2nd c\u0103 mecanismele pie\u021bei \u0219i riscul unor ac\u021biuni judiciare sunt cele mai bune garan\u021bii \u00een fa\u021ba acestei tehnologii, relateaz\u0103 AFP. ‘Oamenii nu vor dori s\u0103 foloseasc\u0103 […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2],"tags":[37,226],"class_list":["post-2146943","post","type-post","status-publish","format-standard","hentry","category-politics","tag-agerpres-ro","tag-crawlmanager"],"_links":{"self":[{"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=\/wp\/v2\/posts\/2146943","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=2146943"}],"version-history":[{"count":0,"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=\/wp\/v2\/posts\/2146943\/revisions"}],"wp:attachment":[{"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=2146943"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=2146943"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=2146943"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}