{"id":705497,"date":"2026-07-29T07:26:03","date_gmt":"2026-07-29T07:26:03","guid":{"rendered":"https:\/\/microscopemedia.com\/?p=705497"},"modified":"2026-07-29T07:26:03","modified_gmt":"2026-07-29T07:26:03","slug":"peste-1-000-de-angajati-ai-openai-google-si-meta-cer-incetinirea-dezvoltarii-inteligentei-artificiale-dupa-un-incident-de-securitate","status":"publish","type":"post","link":"https:\/\/microscopemedia.com\/?p=705497","title":{"rendered":"Peste 1.000 de angaja\u021bi ai OpenAI, Google \u0219i Meta cer \u00eencetinirea dezvolt\u0103rii inteligen\u021bei artificiale dup\u0103 un incident de securitate"},"content":{"rendered":"
\n
\n
<\/div>\n<\/p><\/div>\n

Printre semnatarii peti\u021biei se num\u0103r\u0103 reprezentan\u021bi ai OpenAI, Anthropic, Google DeepMind \u0219i Meta AI. Potrivit Euronews<\/a>, \u00een documentul adresat autorit\u0103\u021bilor americane, semnatarii solicit\u0103 sprijin pentru un efort interna\u021bional care s\u0103 dezvolte instrumente tehnice \u0219i mecanisme de guvernan\u021b\u0103 capabile s\u0103 controleze ritmul dezvolt\u0103rii inteligen\u021bei artificiale de ultim\u0103 genera\u021bie.<\/p>\n

Solicitarea \u2013 pentru o dezvoltare mai lent\u0103 \u0219i mai sigur\u0103<\/h2>\n

Potrivit peti\u021biei, marile companii din domeniu consider\u0103 c\u0103 se apropie momentul \u00een care inteligen\u021ba artificial\u0103 va putea automatiza chiar procesul de cercetare \u00een AI. Acest lucru ar putea accelera dezvoltarea tehnologiei \u00eentr-un ritm dificil de anticipat \u0219i de controlat. \u201eExist\u0103 un risc real ca dezvoltarea capacit\u0103\u021bilor AI s\u0103 avanseze mai rapid dec\u00e2t abilitatea noastr\u0103 de a \u00een\u021belege sau controla aceste sisteme\u201d, avertizeaz\u0103 semnatarii.<\/p>\n

Incidentul care a schimbat percep\u021bia<\/h2>\n

Apelul vine la scurt timp dup\u0103 ce OpenAI a dezv\u0103luit un incident produs \u00een timpul unor teste interne. Compania a anun\u021bat c\u0103 dou\u0103 dintre modelele sale de inteligen\u021b\u0103 artificial\u0103 au reu\u0219it s\u0103 desf\u0103\u0219oare un atac cibernetic autonom, dep\u0103\u0219ind m\u0103surile de securitate create special pentru mediul de testare. Potrivit OpenAI, modelele au ie\u0219it din mediul izolat \u00een care erau testate, s-au conectat la internet \u0219i au accesat platforma de partajare a codului Hugging Face. De\u0219i incidentul s-a produs \u00eentr-un cadru controlat, acesta a fost considerat unul dintre cele mai serioase semnale de alarm\u0103 privind comportamentul autonom al sistemelor AI.<\/p>\n

<\/p>\n

Aplic\u021bia ChatGPT, imagine ilustrativ\u0103. Foto: Algi Febri Sugita\/ZUMA Press Wire<\/em><\/p>\n<\/div>\n

Sam Altman: Societatea are nevoie de timp<\/h2>\n

Directorul general al OpenAI, Sam Altman, nu a semnat peti\u021bia, \u00eens\u0103 a transmis un mesaj similar. El a declarat c\u0103 dezvoltatorii de inteligen\u021b\u0103 artificial\u0103 ar putea fi nevoi\u021bi s\u0103 \u00eencetineasc\u0103 voluntar ritmul progresului tehnologic. Este important acest demers, pentru a permite societ\u0103\u021bii s\u0103 se adapteze. Altman a afirmat c\u0103 incidentul de securitate a fost primul care l-a f\u0103cut s\u0103 resimt\u0103 \u00een mod direct riscurile asociate dezvolt\u0103rii AI \u0219i s-a declarat surprins c\u0103 reac\u021biile publice nu au fost mai puternice.<\/p>\n

\u00cengrijor\u0103rile cresc \u00een industrie<\/h2>\n

Nu este pentru prima dat\u0103 c\u00e2nd un model de inteligen\u021b\u0103 artificial\u0103 manifest\u0103 comportamente care dep\u0103\u0219esc scenariile anticipate de dezvoltatori. Totu\u0219i, incidentul recent este considerat de mul\u021bi speciali\u0219ti unul dintre cele mai \u00eengrijor\u0103toare de p\u00e2n\u0103 acum. Problema implic\u0103 un comportament autonom \u00eentr-un domeniu extrem de sensibil: securitatea cibernetic\u0103. Semnatarii peti\u021biei avertizeaz\u0103 c\u0103, f\u0103r\u0103 mecanisme eficiente de supraveghere \u0219i cooperare interna\u021bional\u0103, evolu\u021bia rapid\u0103 a inteligen\u021bei artificiale<\/a> ar putea dep\u0103\u0219i capacitatea societ\u0103\u021bii de a gestiona riscurile asociate acestei tehnologii.<\/p>\n<\/p><\/div>\n","protected":false},"excerpt":{"rendered":"

Printre semnatarii peti\u021biei se num\u0103r\u0103 reprezentan\u021bi ai OpenAI, Anthropic, Google DeepMind \u0219i Meta AI. Potrivit Euronews, \u00een documentul adresat autorit\u0103\u021bilor americane, semnatarii solicit\u0103 sprijin pentru un efort interna\u021bional care s\u0103 … Read More<\/a><\/p>\n","protected":false},"author":1,"featured_media":705498,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"Default","format":"standard","meta":[],"categories":[1],"tags":[],"_links":{"self":[{"href":"https:\/\/microscopemedia.com\/index.php?rest_route=\/wp\/v2\/posts\/705497"}],"collection":[{"href":"https:\/\/microscopemedia.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/microscopemedia.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/microscopemedia.com\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/microscopemedia.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=705497"}],"version-history":[{"count":0,"href":"https:\/\/microscopemedia.com\/index.php?rest_route=\/wp\/v2\/posts\/705497\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/microscopemedia.com\/index.php?rest_route=\/wp\/v2\/media\/705498"}],"wp:attachment":[{"href":"https:\/\/microscopemedia.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=705497"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/microscopemedia.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=705497"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/microscopemedia.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=705497"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}