{"id":8964,"date":"2024-07-15T06:31:35","date_gmt":"2024-07-15T06:31:35","guid":{"rendered":"https:\/\/www.slovensko.ai\/?p=8964"},"modified":"2024-07-18T06:12:44","modified_gmt":"2024-07-18T06:12:44","slug":"nove-pokroky-v-ai-od-mit","status":"publish","type":"post","link":"https:\/\/www.slovensko.ai\/en\/nove-pokroky-v-ai-od-mit\/","title":{"rendered":"Nov\u00e9 pokroky v AI od MIT"},"content":{"rendered":"<div class='epvc-post-count'><span class='epvc-eye'><\/span>  <span class=\"epvc-count\"> 631<\/span><span class='epvc-label'> Viden\u00ed<\/span><\/div><div class=\"flex-shrink-0 flex flex-col relative items-end\">\n<div>\n<div class=\"pt-0.5 juice:pt-0\">\n<div class=\"gizmo-bot-avatar flex h-6 w-6 items-center justify-center overflow-hidden rounded-full juice:h-8 juice:w-8\">\n<h2 class=\"relative p-1 rounded-sm flex items-center justify-center bg-token-main-surface-primary text-token-text-primary h-8 w-8\">Pokroky v AI<\/h2>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"group\/conversation-turn relative flex w-full min-w-0 flex-col agent-turn\">\n<div class=\"flex-col gap-1 md:gap-3\">\n<div class=\"flex flex-grow flex-col max-w-full\">\n<div class=\"min-h-[20px] text-message flex flex-col items-start whitespace-pre-wrap break-words [.text-message+&amp;]:mt-5 juice:w-full juice:items-end overflow-x-auto gap-2\" dir=\"auto\" data-message-author-role=\"assistant\" data-message-id=\"3700cb61-d3e1-47f2-a97a-e44977e79976\">\n<div class=\"flex w-full flex-col gap-1 juice:empty:hidden juice:first:pt-[3px]\">\n<div class=\"markdown prose w-full break-words dark:prose-invert dark\">\n<p>MIT (Massachusetts Institute of Technology) sa v uplynulom t\u00fd\u017edni predstavil s nieko\u013ek\u00fdmi zauj\u00edmav\u00fdmi novinkami v oblasti umelej inteligencie (AI), ktor\u00e9 demon\u0161truj\u00fa ich pokrok a inovat\u00edvne pr\u00edstupy k rie\u0161eniu komplexn\u00fdch probl\u00e9mov. Po\u010fme sa na nich pozrie\u0165.<\/p>\n<h4>V\u00fdskum schopnost\u00ed ve\u013ek\u00fdch jazykov\u00fdch modelov (LLMs)<\/h4>\n<p>Jedn\u00fdm z hlavn\u00fdch objavov je \u0161t\u00fadia z MIT CSAIL (Computer Science and Artificial Intelligence Laboratory), ktor\u00e1 pouk\u00e1zala na to, \u017ee ve\u013ek\u00e9 jazykov\u00e9 modely, ako s\u00fa tie, ktor\u00e9 poh\u00e1\u0148aj\u00fa popul\u00e1rne chatboty a asistentov, \u010dasto prece\u0148ujeme v ich schopnostiach uva\u017eova\u0165. V\u00fdskum uk\u00e1zal, \u017ee tieto modely vynikaj\u00fa pri rie\u0161en\u00ed \u00faloh, ktor\u00e9 s\u00fa im zn\u00e1me, ale maj\u00fa probl\u00e9my s nov\u00fdmi a nezn\u00e1mymi scen\u00e1rmi, \u010do nazna\u010duje, \u017ee ich \u00faspech je viac zalo\u017een\u00fd na memorovan\u00ed ne\u017e na skuto\u010dnom uva\u017eovan\u00ed. Tento objav je d\u00f4le\u017eit\u00fd pre pochopenie limit\u00e1ci\u00ed s\u00fa\u010dasn\u00fdch AI syst\u00e9mov a nasmerovanie bud\u00faceho v\u00fdvoja\u200b.<\/p>\n<h4>Ceny za inov\u00e1ciu v AI pre vesm\u00edrne aplik\u00e1cie<\/h4>\n<p>MIT ARCLab (Aerospace Robotics and Control Laboratory) ozn\u00e1mil v\u00ed\u0165azov svojej prvej ceny za inov\u00e1ciu v AI pre vesm\u00edrne aplik\u00e1cie. T\u00e1to s\u00fa\u0165a\u017e vyz\u00fdvala t\u00edmy, aby vyvinuli AI algoritmy na sledovanie a predpovedanie pohybov\u00fdch vzorcov satelitov na obe\u017enej dr\u00e1he pomocou pas\u00edvne zhroma\u017e\u010fovan\u00fdch d\u00e1t. V\u00ed\u0165azn\u00e9 projekty predstavili inovat\u00edvne rie\u0161enia, ktor\u00e9 by mohli zlep\u0161i\u0165 presnos\u0165 a spo\u013eahlivos\u0165 sledovania satelitov, \u010do m\u00e1 ve\u013ek\u00fd v\u00fdznam pre bud\u00face vesm\u00edrne misie a komunika\u010dn\u00e9 syst\u00e9my\u200b\u200b.<\/p>\n<h4>Generat\u00edvna AI pre datab\u00e1zy<\/h4>\n<p>\u010eal\u0161ou v\u00fdznamnou inov\u00e1ciou od MIT je v\u00fdvoj nov\u00e9ho n\u00e1stroja generat\u00edvnej AI pre anal\u00fdzu komplexn\u00fdch tabu\u013ekov\u00fdch d\u00e1t. Tento n\u00e1stroj zjednodu\u0161uje pr\u00e1cu s rozsiahlymi d\u00e1tov\u00fdmi s\u00fabormi a umo\u017e\u0148uje u\u017e\u00edvate\u013eom r\u00fdchlej\u0161ie a efekt\u00edvnej\u0161ie z\u00edskava\u0165 d\u00f4le\u017eit\u00e9 poznatky. Tento pr\u00edstup by mohol n\u00e1js\u0165 \u0161irok\u00e9 uplatnenie v r\u00f4znych oblastiach od financi\u00ed po zdravotn\u00edctvo, kde je anal\u00fdza d\u00e1t kritick\u00e1 pre rozhodovanie\u200b.<\/p>\n<h4>Zlep\u0161enie neistoty v AI modeloch<\/h4>\n<p>MIT tie\u017e pracuje na zlep\u0161en\u00ed presnosti odhadov neistoty v modeloch strojov\u00e9ho u\u010denia, \u010do by mohlo pou\u017e\u00edvate\u013eom pom\u00f4c\u0165 lep\u0161ie rozhodova\u0165 o tom, kedy a ako pou\u017e\u00edva\u0165 tieto modely v re\u00e1lnom svete. Presn\u00e9 odhady neistoty m\u00f4\u017eu zn\u00ed\u017ei\u0165 riziko nespr\u00e1vnych rozhodnut\u00ed a zlep\u0161i\u0165 d\u00f4veru v AI syst\u00e9my\u200b\u200b.<\/p>\n<p>Tieto <strong>Pokroky v AI<\/strong> ukazuj\u00fa, \u017ee MIT na\u010falej hr\u00e1 k\u013e\u00fa\u010dov\u00fa rolu v pos\u00favan\u00ed hran\u00edc umelej inteligencie, prin\u00e1\u0161aj\u00fac rie\u0161enia, ktor\u00e9 maj\u00fa potenci\u00e1l transformova\u0165 r\u00f4zne priemyseln\u00e9 odvetvia a zlep\u0161i\u0165 ka\u017edodenn\u00fd \u017eivot.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n","protected":false},"excerpt":{"rendered":"<p>631 Viden\u00ed Pokroky v AI MIT (Massachusetts Institute of Technology) sa v uplynulom t\u00fd\u017edni predstavil s nieko\u013ek\u00fdmi zauj\u00edmav\u00fdmi novinkami v oblasti umelej inteligencie (AI), ktor\u00e9 demon\u0161truj\u00fa ich pokrok a inovat\u00edvne pr\u00edstupy k rie\u0161eniu komplexn\u00fdch probl\u00e9mov. Po\u010fme sa na nich pozrie\u0165. V\u00fdskum schopnost\u00ed ve\u013ek\u00fdch jazykov\u00fdch modelov (LLMs) Jedn\u00fdm z hlavn\u00fdch objavov je \u0161t\u00fadia z MIT CSAIL [&hellip;]<\/p>\n","protected":false},"author":2,"featured_media":9256,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[51],"tags":[],"class_list":["post-8964","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-blog"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.slovensko.ai\/en\/wp-json\/wp\/v2\/posts\/8964"}],"collection":[{"href":"https:\/\/www.slovensko.ai\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.slovensko.ai\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.slovensko.ai\/en\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.slovensko.ai\/en\/wp-json\/wp\/v2\/comments?post=8964"}],"version-history":[{"count":1,"href":"https:\/\/www.slovensko.ai\/en\/wp-json\/wp\/v2\/posts\/8964\/revisions"}],"predecessor-version":[{"id":9135,"href":"https:\/\/www.slovensko.ai\/en\/wp-json\/wp\/v2\/posts\/8964\/revisions\/9135"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.slovensko.ai\/en\/wp-json\/wp\/v2\/media\/9256"}],"wp:attachment":[{"href":"https:\/\/www.slovensko.ai\/en\/wp-json\/wp\/v2\/media?parent=8964"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.slovensko.ai\/en\/wp-json\/wp\/v2\/categories?post=8964"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.slovensko.ai\/en\/wp-json\/wp\/v2\/tags?post=8964"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}