{"id":8785,"date":"2024-05-28T00:00:00","date_gmt":"2024-05-27T23:00:00","guid":{"rendered":"https:\/\/instrid.sk\/uncategorized\/europska-rada-schvalila-zakon-na-regulaciu-umelej-inteligencie\/"},"modified":"2024-05-28T00:00:00","modified_gmt":"2024-05-27T23:00:00","slug":"europska-rada-schvalila-zakon-na-regulaciu-umelej-inteligencie","status":"publish","type":"post","link":"https:\/\/instrid.sk\/en\/archiv\/europska-rada-schvalila-zakon-na-regulaciu-umelej-inteligencie\/","title":{"rendered":"Eur\u00f3pska rada schv\u00e1lila z\u00e1kon na regul\u00e1ciu umelej inteligencie. Modely rozdel\u00ed pod\u013ea rizika"},"content":{"rendered":"

Umel\u00e1 inteligencia doslova zachv\u00e1tila svet, pri\u010dom mnoho spolo\u010dnost\u00ed sa sna\u017e\u00ed zu\u017eitkova\u0165 novodob\u00fd fenom\u00e9n vo svoj prospech. V\u00fdvin nov\u00fdch a\u00a0efekt\u00edvnej\u0161\u00edch modelov umelej inteligencii je zrejme nezastavite\u013en\u00fd. Napriek tomu sa v\u0161ak Eur\u00f3pska \u00fania sna\u017e\u00ed o\u00a0regul\u00e1ciu<\/strong>.<\/p>\n

Rada Eur\u00f3pskej \u00fanie schv\u00e1lila v\u00a0utorok 21. m\u00e1ja z\u00e1kon zameran\u00fd na zos\u00faladenie pravidiel t\u00fdkaj\u00facich sa umelej inteligencie na \u00fazem\u00ed \u010dlensk\u00fdch kraj\u00edn \u00fanie. Akt o\u00a0umelej inteligencii<\/strong> sa m\u00e1 riadi\u0165 pr\u00edstupom zalo\u017een\u00fdm na pos\u00faden\u00ed rizika<\/strong>, informuje rada v\u00a0tla\u010dovej spr\u00e1ve.<\/p>\n

Nov\u00fd z\u00e1kon preto rozli\u0161uje nieko\u013eko druhov umelej inteligencie. Na syst\u00e9my s limitovan\u00fdm rizikom by sa vz\u0165ahovali ve\u013emi mierne povinnosti t\u00fdkaj\u00face sa transparentnosti. Vysoko rizikov\u00e9 syst\u00e9my by boli povolen\u00e9, ale podliehali pr\u00edsnej\u0161\u00edm po\u017eiadavk\u00e1m.<\/p>\n

Osobitnou kateg\u00f3riou s\u00fa syst\u00e9my umelej inteligencie s\u00a0vyu\u017eit\u00edm na kognit\u00edvne behavior\u00e1lne manipul\u00e1cie a\u00a0soci\u00e1lne sk\u00f3rovanie<\/strong>. Tieto modely bud\u00fa v\u00a0zak\u00e1zan\u00e9. To ist\u00e9 plat\u00ed o\u00a0pou\u017e\u00edvan\u00ed umelej inteligencie na predikt\u00edvny policajn\u00fd doh\u013ead zalo\u017een\u00fd alebo kategoriz\u00e1ciu \u013eud\u00ed.<\/p>\n

Prv\u00fd svojho druhu<\/h3>\n

\u201eTento prelomov\u00fd z\u00e1kon, prv\u00fd svojho druhu na svete, rie\u0161i glob\u00e1lnu technologick\u00fa v\u00fdzvu, ktor\u00e1 z\u00e1rove\u0148 vytv\u00e1ra pr\u00edle\u017eitosti pre na\u0161e spolo\u010dnosti a\u00a0hospod\u00e1rstva,\u201c <\/em>uviedol belgick\u00fd \u0161t\u00e1tny tajomn\u00edk pre digitaliz\u00e1ciu Mathieu Michel. Hovor\u00ed, \u017ee ide o\u00a0v\u00fdznamn\u00fd m\u00ed\u013enik pre \u00faniu.<\/p>\n

Kv\u00f4li aktu sa z\u00e1rove\u0148 zriadilo nieko\u013eko riadiacich org\u00e1nov. Prv\u00fdm z nich je \u00darad pre umel\u00fa inteligenciu<\/strong>, ktor\u00fd spad\u00e1 pod Eur\u00f3psku komisiu. \u010eal\u0161\u00edm z nich je Rada pre umel\u00fa inteligenciu<\/strong> so z\u00e1stupcami \u010dlensk\u00fdch \u0161t\u00e1tov, ktor\u00e1 m\u00e1 poskytova\u0165 poradenstvo pri jeho uplat\u0148ovan\u00ed.<\/p>\n Modely umelej inteligencie na v\u0161eobecn\u00e9 \u00fa\u010dely, ako napr\u00edklad ChatGPT, nemaj\u00fa predstavova\u0165 systematick\u00e9 riziko. Zdroj: unsplash.com\/@jupp\n

Okrem toho vznikla vedeck\u00e1 skupina tvoren\u00e1 z\u00a0nez\u00e1visl\u00fdch odborn\u00edkov<\/strong>, aby podporili \u010dinnosti s\u00favisiace s\u00a0presadzovan\u00edm pr\u00e1va a\u00a0poradn\u00e9 f\u00f3rum pre zainteresovan\u00e9 strany<\/strong>. To bude poskytova\u0165 technick\u00e9 a\u00a0odborn\u00e9 znalosti Rade pre umel\u00fa inteligenciu a\u00a0Eur\u00f3pskej komisii.<\/p>\n

V pr\u00edpade poru\u0161enia nov\u00e9ho z\u00e1kona hrozia spolo\u010dnostiam pokuty. Ich v\u00fd\u0161ku stanovili ako percento z celkov\u00e9ho ro\u010dn\u00e9ho obratu za minul\u00fd rok<\/strong>, pr\u00edpadne ako vopred ur\u010den\u00e1 suma pod\u013ea toho, ktor\u00e1 z nich je vy\u0161\u0161ia, uv\u00e1dza port\u00e1l \u017div\u00e9.sk na z\u00e1klade spr\u00e1vy z TASR<\/p>\n

Technologicky sektor sa ob\u00e1va<\/h3>\n

\u201eAktom o umelej inteligencii Eur\u00f3pa zd\u00f4raz\u0148uje v\u00fdznam d\u00f4very, transparentnosti a zodpovednosti pri pr\u00e1ci s nov\u00fdmi technol\u00f3giami a z\u00e1rove\u0148 zabezpe\u010duje, aby sa t\u00e1to r\u00fdchlo sa meniaca technol\u00f3gia mohla rozv\u00edja\u0165 a podporova\u0165 eur\u00f3pske inov\u00e1cie,\u201c <\/em>dod\u00e1va Michel.<\/p>\n

L\u00eddri technologick\u00e9ho sektoru s\u00fa v\u0161ak vo\u010di nov\u00e9mu skeptick\u00ed<\/strong>. \u201eV z\u00e1kone E\u00da o umelej inteligencii a na r\u00f4znych in\u00fdch miestach existuj\u00fa ustanovenia, ktor\u00e9 reguluj\u00fa v\u00fdskum a\u00a0v\u00fdvoj,\u201c povedal pre CNN \u0161\u00e9f spolo\u010dnosti Meta pre umel\u00fa inteligenciu Yann LeCun.<\/p>\n

Hovor\u00ed, \u017ee to nie je spr\u00e1vne. Podobn\u00fd n\u00e1zor zast\u00e1va aj technologick\u00fd riadite\u013e spolo\u010dnosti Amazaon Werner Vogels. \u201eExistuje cel\u00fd rad oblast\u00ed, v ktor\u00fdch s\u00fa pod\u013ea m\u0148a rizik\u00e1 minim\u00e1lne a mali by sme v nich necha\u0165 inov\u00e1cie be\u017ea\u0165,“<\/em> povedal CNN na VivaTech Conference.<\/p>\n Eur\u00f3pska komisia v s\u00fa\u010dasnosti pre\u0161etruje Amazon, kv\u00f4li mo\u017en\u00e9mu poru\u0161eniu Nariadeniu o digit\u00e1lnych trhoch. Zdroj: unsplash.com\/@christianw\n

Z\u00e1rove\u0148 v\u0161ak dodal, \u017ee v\u00a0in\u00fdch oblastiach, kde chybovos\u0165 modelov umelej inteligencie by mohla ma\u0165 v\u00e4\u010d\u0161ie dopady na \u013eudsk\u00e9 \u017eivoty, by sa hrozby mali riadi\u0165 \u201ejedine\u010dne pre dan\u00fa oblas\u0165\u201c<\/em>. \u00a0Zd\u00f4raznil, \u017ee Amazon regul\u00e1ciu v\u00edta, ale varoval E\u00da pred nadmernou regul\u00e1ciou<\/strong>.<\/p>\n

Pouk\u00e1zal pri tom na z\u00e1kon o\u00a0ochrane osobn\u00fdch \u00fadajov, zn\u00e1my ako GDPR, pri\u010dom ho ozna\u010dil za ve\u013emi \u201ehrub\u00fa\u201c knihu<\/strong>. Eur\u00f3pska \u00fania je zn\u00e1ma svojimi z\u00e1konmi, medzi nimi aj Nariadenie o\u00a0digit\u00e1lnych trhoch (DMA), o\u00a0ktorom sme v\u00e1s informovali v\u00a0predch\u00e1dzaj\u00facom \u010dl\u00e1nku.<\/p>\n

Potrebujeme glob\u00e1lnu regul\u00e1ciu?<\/h3>\n

Pou\u017e\u00edvanie, v\u00fdvoj a v\u00fdskum umelej inteligencie by mohli regulova\u0165 aj jednotliv\u00e9 \u0161t\u00e1ty<\/strong> pod\u013ea toho, ako uznaj\u00fa za vhodn\u00e9. Toto rie\u0161enie by v\u0161ak mohlo sp\u00f4sobi\u0165 v\u00e4\u010d\u0161\u00ed chaosu v pravidl\u00e1ch a nariadeniach. Spolo\u010dnosti by sa toti\u017e museli prisp\u00f4sobova\u0165 ka\u017edej krajine zvl\u00e1\u0161\u0165.<\/p>\n

Okrem toho sa v\u0161ak na\u010drt\u00e1 \u010fal\u0161\u00ed probl\u00e9m, ktor\u00e9ho podstata je ekonomick\u00e1. Port\u00e1l britskej telev\u00edzie SkyNews informuje o\u00a0tom, \u017ee regula\u010dn\u00e9 org\u00e1ny v\u00a0Spojenom kr\u00e1\u013eovstve maj\u00fa by\u0165 pofinancovan\u00e9<\/strong>. Upozornil na to \u0161\u00e9f vedeck\u00e9ho v\u00fdboru britsk\u00e9ho parlamentu Greg Clark.<\/p>\n

Vid\u00ed v\u0161ak aj \u010fal\u0161\u00ed probl\u00e9m. Novovzniknut\u00fd AI Safety Institute nemal prist\u00fap k modelom niektor\u00fdch v\u00fdvoj\u00e1rov pred ich uveden\u00edm na trh<\/strong>. To im znemo\u017enilo otestova\u0165 ich v\u00fdstupy na zaujatos\u0165, a \u201eaby sa zistilo, \u010di maj\u00fa neprijate\u013en\u00e9 d\u00f4sledky\u201d<\/em>.<\/p>\n

\u201eJe spr\u00e1vne pracova\u0165 prostredn\u00edctvom existuj\u00facich regula\u010dn\u00fdch org\u00e1nov, ale bud\u00faca vl\u00e1da by mala by\u0165 pripraven\u00e1 r\u00fdchlo prija\u0165 z\u00e1kon, ak sa uk\u00e1\u017ee, \u017ee niektor\u00fd z mnoh\u00fdch regula\u010dn\u00fdch org\u00e1nov nem\u00e1 z\u00e1konn\u00e9 pr\u00e1vomoci, aby bol \u00fa\u010dinn\u00fd,\u201c<\/em> skon\u0161tatoval \u0161\u00e9f v\u00fdboru Clarck.<\/p>","protected":false},"excerpt":{"rendered":"

Umel\u00e1 inteligencia doslova zachv\u00e1tila svet, pri\u010dom mnoho spolo\u010dnost\u00ed sa sna\u017e\u00ed zu\u017eitkova\u0165 novodob\u00fd fenom\u00e9n vo svoj prospech. V\u00fdvin nov\u00fdch a\u00a0efekt\u00edvnej\u0161\u00edch modelov umelej inteligencii je zrejme nezastavite\u013en\u00fd.<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[51],"tags":[],"class_list":["post-8785","post","type-post","status-publish","format-standard","hentry","category-archiv","two-columns"],"_links":{"self":[{"href":"https:\/\/instrid.sk\/en\/wp-json\/wp\/v2\/posts\/8785","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/instrid.sk\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/instrid.sk\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/instrid.sk\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/instrid.sk\/en\/wp-json\/wp\/v2\/comments?post=8785"}],"version-history":[{"count":0,"href":"https:\/\/instrid.sk\/en\/wp-json\/wp\/v2\/posts\/8785\/revisions"}],"wp:attachment":[{"href":"https:\/\/instrid.sk\/en\/wp-json\/wp\/v2\/media?parent=8785"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/instrid.sk\/en\/wp-json\/wp\/v2\/categories?post=8785"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/instrid.sk\/en\/wp-json\/wp\/v2\/tags?post=8785"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}