{"id":39178,"date":"2025-11-10T13:36:16","date_gmt":"2025-11-10T12:36:16","guid":{"rendered":"https:\/\/kinit.sk\/stredoskolaci-z-sos-handlova-v-kinite\/"},"modified":"2025-11-18T09:01:33","modified_gmt":"2025-11-18T08:01:33","slug":"stredoskolaci-z-sos-handlova-v-kinite","status":"publish","type":"post","link":"https:\/\/kinit.sk\/sk\/stredoskolaci-z-sos-handlova-v-kinite\/","title":{"rendered":"Stredo\u0161kol\u00e1ci z SO\u0160 Handlov\u00e1 v KInITe"},"content":{"rendered":"
\n

Ned\u00e1vno n\u00e1s nav\u0161t\u00edvili stredo\u0161kol\u00e1ci z <\/strong>SO\u0160 Handlov\u00e1<\/strong><\/a>, ktor\u00ed mali mo\u017enos\u0165 stretn\u00fa\u0165 sa s na\u0161imi PhD \u0161tudentami a dozvedie\u0165 sa viac o modern\u00fdch technol\u00f3gi\u00e1ch, umelej inteligencii a najm\u00e4 o ich v\u00fdskume. <\/strong><\/p>\n<\/div>\n\n

\n

Na \u00favod \u0161tudentov priv\u00edtal n\u00e1\u0161 v\u00fdskumn\u00edk Jakub \u0160imko<\/a>, ved\u00faci WUDAP t\u00edmu, ktor\u00fd \u0161tudentom predstavil KInIT a uviedol ich do t\u00e9m, ktor\u00fdm sa v in\u0161tit\u00fate venujeme.<\/p>\n<\/div>\n\n

\n

N\u00e1sledne sa \u0161tudentom prihovoril n\u00e1\u0161 PhD \u0161tudent R\u00f3bert Belanec<\/a>, absolvent tejto \u0161koly. R\u00f3bert predstavil KInIT z poh\u013eadu mlad\u00e9ho v\u00fdskumn\u00edka a porozpr\u00e1val im viac o svojej pr\u00e1ci a v\u00fdskumn\u00fdch projektoch:<\/p>\n<\/div>\n\n

\n

 \u201cAk sa be\u017ene nevenujete AI, len \u0165a\u017eko si viete predstavi\u0165, \u010do sa skr\u00fdva pod v\u0161etk\u00fdmi \u00farov\u0148ami abstrakcie, ktor\u00e9 m\u00f4\u017eeme vidie\u0165 v n\u00e1strojoch, ako je ChatGPT. To m\u00f4\u017ee vytv\u00e1ra\u0165 dojem, \u017ee AI je nejak\u00e1 ve\u0161teck\u00e1 gu\u013ea alebo \u017ee je naozaj inteligentn\u00e1. V\u010faka tomu sa m\u00f4\u017ee spo\u010diatku \u0161tudentom zda\u0165 AI ako nie\u010do nedosiahnute\u013en\u00e9 a ve\u013emi abstraktn\u00e9, nie ako nie\u010do, \u010do si vedia ohmata\u0165 a spusti\u0165 na svojom notebooku,\u201d <\/em>vysvetlil R\u00f3bert.<\/p>\n<\/div>\n\n

\n

\u201cV r\u00e1mci mojej predn\u00e1\u0161ky som \u0161tudentom pribl\u00ed\u017eil, ako AI v skuto\u010dnosti vyzer\u00e1 ‚pod kapotou‘, pre\u010do s\u00fa AI modely neefekt\u00edvne (t\u00e9ma, ktorej sa venujem v KInITe), a odporu\u010dil som materi\u00e1ly, ktor\u00e9 im pom\u00f4\u017eu si AI rozbeha\u0165, pozrie\u0165 sa do jej vn\u00fatra a lep\u0161ie jej porozumie\u0165. Z\u00e1rove\u0148 som si na z\u00e1klade ot\u00e1zok \u0161tudentov v\u0161imol, \u017ee mnoh\u00fdm \u013eu\u010fom nemus\u00ed by\u0165 \u00faplne jasn\u00e9, ako m\u00f4\u017ee vyzera\u0165 be\u017en\u00fd de\u0148 v\u00fdskumn\u00edka v umelej inteligencii, ako prebiehaj\u00fa experimenty, alebo ak\u00e9 n\u00e1stroje pou\u017e\u00edvame pri v\u00fdskume. Toto a mnoho \u010fal\u0161ieho som \u0161tudentom pribl\u00ed\u017eil po\u010das ich n\u00e1v\u0161tevy v KInITe. Te\u0161ili ma reakcie \u0161tudentov a pr\u00edjemn\u00e1 atmosf\u00e9ra po\u010das cel\u00e9ho d\u0148a.\u201d<\/em><\/p>\n<\/div>\n\n

\n
\"\"<\/figure>\n<\/div>\n\n
\n

Na t\u00e9mu zodpovedn\u00e9ho pou\u017e\u00edvania umelej inteligencie nadviazal Branislav Pecher<\/a>, ktor\u00fd u n\u00e1s \u00faspe\u0161ne dokon\u010dil svoje PhD \u0161t\u00fadium. \u0160tudentom pribl\u00ed\u017eil, ako ve\u013ek\u00e9 jazykov\u00e9 modely  (ne)funguj\u00fa v praxi. \u201cV predn\u00e1\u0161ke som sa zameral najm\u00e4 na probl\u00e9my spojen\u00e9 s ve\u013ek\u00fdmi jazykov\u00fdmi modelmi a s t\u00fdm, ako sa o nich v spolo\u010dnosti rozpr\u00e1va. Uk\u00e1zal som napr\u00edklad, ako jednoducho dok\u00e1\u017eeme obch\u00e1dza\u0165 \u2018mantinely\u2019 (anglicky guardrails), teda z\u00e1brany, ktor\u00e9 maj\u00fa v sebe tieto modely zabudovan\u00e9, aby negenerovali nebezpe\u010dn\u00e9 odpovede – a ako ich mo\u017eno prin\u00fati\u0165 vytvori\u0165 napr\u00edklad  dezinforma\u010dn\u00fd \u010dl\u00e1nok,\u201d pribl\u00ed\u017eil Branislav.<\/em><\/p>\n<\/div>\n\n

\n

\u201cTie\u017e som sa venoval aj tomu, ako sa dnes \u010dasto prece\u0148uj\u00fa schopnosti t\u00fdchto modelov a \u017ee by sme ich nemali vn\u00edma\u0165 ako v\u0161emocn\u00e9 n\u00e1stroje. I\u0161lo teda sk\u00f4r o tak\u00e9 filozofickej\u0161ie zhrnutie toho, \u010do dobr\u00e9 aj zl\u00e9 modely prin\u00e1\u0161aj\u00fa, a \u010do n\u00e1s s nimi \u010dak\u00e1, na \u010do som dostal aj ve\u013emi pozit\u00edvnu sp\u00e4tn\u00fa v\u00e4zbu. Ver\u00edm, \u017ee to in\u0161pirovalo \u0161tudentov aby slepo neverili modelom ani ich tvorcom, ktor\u00ed ich prezentuj\u00fa ako najv\u00e4\u010d\u0161iu z\u00e1chranu sveta, ale aby sa na ne pozerali kriticky. To je aj jedna z d\u00f4le\u017eit\u00fdch t\u00e9m, ktorej som sa v r\u00e1mci KInITu zaoberal \u2013 lep\u0161ie pochopi\u0165, \u010do tieto ve\u013ek\u00e9 jazykov\u00e9 modely dok\u00e1\u017eu, kde robia chyby, a ako ich mo\u017eno napravi\u0165.\u201d<\/em><\/p>\n<\/div>\n\n

\n

PhD \u0161tudentka Ivana Be\u0148ov\u00e1<\/a> vo svojej predn\u00e1\u0161ke upozornila na to, \u017ee umel\u00e1 inteligencia nie je dokonal\u00e1 a m\u00e1 st\u00e1le svoje limity. <\/p>\n<\/div>\n\n

\n

\u201cMoja predn\u00e1\u0161ka sa zamerala na to, ako AI st\u00e1le m\u00f4\u017ee robi\u0165 chyby. Bolo skvel\u00e9 vidie\u0165, ako sa menil n\u00e1zor \u0161tudentov na umel\u00fa inteligenciu v\u010faka mojej predn\u00e1\u0161ke. Napriek tomu, \u017ee to boli \u0161tudenti zo strednej \u0161koly so zameran\u00edm na informatiku, tak si mysleli, \u017ee umel\u00e1 inteligencia u\u017e dnes dok\u00e1\u017ee takmer v\u0161etko. S\u00fa\u010das\u0165ou mojej prezent\u00e1cie bola uk\u00e1\u017eka, kde v\u0161ade e\u0161te umel\u00e1 inteligencia m\u00f4\u017ee robi\u0165 chyby. Mysl\u00edm si, \u017ee pr\u00e1ve to mohlo \u0161tudentov ve\u013emi zauja\u0165, a hlavne vidie\u0165, \u017ee je tu priestor na zlep\u0161ovanie umelej inteligencie. To je napr\u00edklad aj jedna z t\u00e9m, ktorej sa u n\u00e1s v KInITe m\u00f4\u017eu v bud\u00facnosti podrobnej\u0161ie venova\u0165,\u201c povedala Ivana.<\/em><\/p>\n<\/div>\n\n

\n

\u201cMysl\u00edm si, \u017ee vyu\u010dovanie o AI na stredn\u00fdch \u0161kol\u00e1ch e\u0161te st\u00e1le ch\u00fdba. A aj po rozhovore so \u0161tudentmi som mala pocit, \u017ee nemaj\u00fa o tom e\u0161te nejak\u00fd ve\u013ek\u00fd preh\u013ead, a o to viac ma te\u0161\u00ed, \u017ee k n\u00e1m pri\u0161li.  Je skvel\u00e9, \u017ee \u0161koly vyu\u017e\u00edvaj\u00fa tak\u00fato mo\u017enos\u0165 – pr\u00eds\u0165 k n\u00e1m na cel\u00fd de\u0148, porozpr\u00e1va\u0165 sa o tom, ako funguje umel\u00e1 inteligencia, a odnies\u0165 si nie\u010do navy\u0161e oproti be\u017en\u00e9mu vyu\u010dovaniu.\u201d<\/em><\/p>\n<\/div>\n\n

\n
\"\"<\/figure>\n<\/div>\n\n
\n

V\u010faka OZ BrAIn<\/a> m\u00f4\u017eeme podporova\u0165 mlad\u00e9 talenty a motivova\u0165 ich k vzdel\u00e1vaniu v oblasti AI. Sme hrd\u00ed, \u017ee aj na\u0161i mlad\u00ed PhD \u0161tudenti a v\u00fdskumn\u00edci m\u00f4\u017eu odovzda\u0165 svoje poznatky \u010fal\u0161\u00edm gener\u00e1ci\u00e1m.<\/p>\n<\/div>\n\n

\n

Svoje dojmy z n\u00e1v\u0161tevy KInITu pridal aj Peter Fabian, u\u010dite\u013e odborn\u00fdch informatick\u00fdch predmetov na SO\u0160 Handlov\u00e1: <\/p>\n<\/div>\n\n

\n

\u201cJe to nevyhnutn\u00e9 pre mlad\u00fdch \u013eud\u00ed, najm\u00e4 aby nepodliehali r\u00f4znym dezinform\u00e1ci\u00e1m a aby sa nau\u010dili overova\u0165 si v\u0161etky inform\u00e1cie a fakty. Musia sa nau\u010di\u0165 nekonzumova\u0165 v\u0161etko, \u010do im umel\u00e1 inteligencia resp. Chatboty pon\u00fakaj\u00fa ako spr\u00e1vnu odpove\u010f na ich ot\u00e1zky. Informatika ide ka\u017ed\u00fdm d\u0148om dopredu a tie\u017e aj u\u010dite\u013e sa mus\u00ed neust\u00e1le u\u010di\u0165. Nov\u00e9 technol\u00f3gie n\u00e1m v tom vedia pom\u00f4c\u0165 – umo\u017e\u0148uj\u00fa n\u00e1m vyh\u013eada\u0165 si nov\u00e9 inform\u00e1cie, ktor\u00e9 potom m\u00f4\u017eeme posun\u00fa\u0165 \u0161tudentom \u010falej,\u201d<\/em> zhodnotil Peter.<\/p>\n<\/div>\n\n

\n
\"\"<\/figure>\n<\/div>\n\n
\n

Nesmierne n\u00e1s te\u0161\u00ed, \u017ee sp\u00e4tn\u00e1 v\u00e4zba od \u0161tudentov bola ve\u013emi pozit\u00edvna. Niektor\u00e9 t\u00e9my z oblasti AI ich zaujali nato\u013eko, \u017ee by najrad\u0161ej hne\u010f u n\u00e1s zostali a za\u010dali pracova\u0165 na novom v\u00fdskume. Ver\u00edme, \u017ee sa k n\u00e1m \u0161tudenti op\u00e4\u0165 vr\u00e1tia, aby mohli pokra\u010dova\u0165 vo vzdel\u00e1van\u00ed a objavovan\u00ed mo\u017enost\u00ed umelej inteligencie. <\/p>\n<\/div>\n\n

\n

\u010eakujeme za n\u00e1v\u0161tevu a te\u0161\u00edme sa na \u010fal\u0161ie stretnutia!<\/p>\n<\/div>\n\n

\n
\"\"<\/a><\/figure>\n<\/div>\n\n
\n

The visit was organised in collaboration with the lorAI project, one of the goals of which is to attend, contribute to, or organise events aimed at promoting science among young people and the general public.<\/em><\/p>\n<\/div>\n\n

\n
\n\t
\n\t\t
\n\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t
\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t
\n\t\t\t\t\t\t\t\t\t
\n\t\t\t\t\t\t\t\t\t\t\"\"\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t
\n\t\t\t\t\t\t\t\t\t
\n\t\t\t\t\t\t\t\t\t\t\"\"\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t