{"id":24767,"date":"2024-11-01T23:27:12","date_gmt":"2024-11-02T02:27:12","guid":{"rendered":"https:\/\/mastertrend.info\/?p=24767"},"modified":"2026-04-28T17:07:41","modified_gmt":"2026-04-28T20:07:41","slug":"sohbetbiti","status":"publish","type":"post","link":"https:\/\/mastertrend.info\/tr\/chatbit\/","title":{"rendered":"ChatBIT: \u00c7in Askeri Yapay Zekas\u0131n\u0131n Yeni S\u0131n\u0131r\u0131"},"content":{"rendered":"

ChatBIT: A\u00e7\u0131k Model \u00c7a\u011fr\u0131lar\u0131na Dayal\u0131 \u00c7in Askeri Yapay Zekas\u0131n\u0131n Yeni S\u0131n\u0131r\u0131 \ud83d\ude80<\/h2>\n

\u00c7inli ara\u015ft\u0131rmac\u0131lar geli\u015ftiriyor askeri yapay zeka<\/strong> Meta'n\u0131n a\u00e7\u0131k kaynakl\u0131 Llama modelini kullanarak. ChatBIT'in OpenAI'nin GPT-4 dil modelinin performans\u0131n\u0131n yakla\u015f\u0131k 90%'sini sa\u011flad\u0131\u011f\u0131 bildiriliyor. \ud83c\udfaf<\/p>\n

\ud83c\udf0d D\u00fcnyas\u0131nda yapay zeka<\/a>, China est\u00e1 marcando un nuevo hito con ChatBIT, un proyecto que promete revolucionar el \u00e1mbito militar. Utilizando el modelo Llama de Meta, de a\u00e7\u0131k kaynak<\/em>ara\u015ft\u0131rmac\u0131lar \u00f6nemli ilerlemeler kaydettiler.<\/p>\n

\ud83c\udf1f Lo m\u00e1s impresionante de ChatBIT es que supuestamente alcanza cerca del 90% del rendimiento de GPT-4 de OpenAI, uno de los modelos de lenguaje m\u00e1s avanzados del mundo. Esta haza\u00f1a coloca a China en una posici\u00f3n estrat\u00e9gica en el desarrollo de teknolojiler<\/a> Yapay zekan\u0131n. \ud83d\ude80<\/p>\n

\ud83d\udd0d Bu t\u00fcr geli\u015fmeler ya\u015fan\u0131rken, yapay zekan\u0131n askeri alanda kullan\u0131m\u0131n\u0131n nas\u0131l evrilece\u011fini ve bunun k\u00fcresel \u00e7apta ne gibi etkiler yaratabilece\u011fini takip etmek b\u00fcy\u00fck \u00f6nem ta\u015f\u0131yor. \u00dczerinde \u00e7ok konu\u015fulacak bir konu kesinlikle! \ud83d\udcac<\/p>\n

\n

\u00c7inli ara\u015ft\u0131rmac\u0131lar\u0131n ba\u011flant\u0131l\u0131 oldu\u011fu \u00c7in Halk Kurtulu\u015f Ordusu (PLA)<\/strong> Yapay zeka modeli geli\u015ftirdiler SohbetBIT<\/strong>, i\u00e7in tasarlanm\u0131\u015f uygulamalar<\/a> a\u00e7\u0131k kaynak modelini kullanan askeri Meta Alev<\/strong>. \ud83c\udf10 Seg\u00fan Reuters, algunos de los investigadores est\u00e1n asociados a la Askeri Bilimler Akademisi (AMS)<\/strong>PLA'n\u0131n ana ara\u015ft\u0131rma grubu. \ud83d\udcda<\/p>\n

\u00dc\u00e7 akademik makale ve \u00e7e\u015fitli analistler bu bilgiyi do\u011frulad\u0131 ve \u015funlar\u0131 vurgulad\u0131: SohbetBIT<\/a><\/strong> B\u00fcy\u00fck Dil Modeli'ni (LLM) kullan\u0131r Alev 13B<\/strong> Meta'dan. Bu LLM, koleksiyon ve kovu\u015fturma<\/a> istihbarat, askeri planlamac\u0131lar\u0131n operasyonel karar alma s\u00fcre\u00e7lerinde kullanmalar\u0131na olanak sa\u011fl\u0131yor. \ud83e\udde0\ud83d\udca1<\/p>\n

Reuters'\u0131n al\u0131nt\u0131lad\u0131\u011f\u0131 makalelerden birinde, askeri yapay zekan\u0131n "askeri alanda diyalog ve soru-cevap g\u00f6revleri i\u00e7in optimize edildi\u011fi" belirtiliyor. ChatBIT'in ayr\u0131ca yakla\u015f\u0131k 90% performans g\u00f6sterdi\u011fi iddia edildi performans<\/a> LLM'den GPT-4<\/strong> OpenAI'dan al\u0131nm\u0131\u015f olmas\u0131na ra\u011fmen makalede performans\u0131n\u0131n nas\u0131l test edildi\u011fi veya AI modelinin sahada kullan\u0131l\u0131p kullan\u0131lmad\u0131\u011f\u0131 belirtilmemi\u015ftir. \ud83e\udd14\ud83d\udd0d<\/p>\n

Ancak a\u00e7\u0131k kaynakl\u0131 AI modellerini kullanmas\u0131, devlerin yay\u0131nlad\u0131\u011f\u0131 en son modellerle e\u015fle\u015fmesine olanak tan\u0131yabilir. teknolojik<\/a> Amerikal\u0131lar k\u0131yaslama testlerinde. \ud83c\udfc6 "Bu, \u00c7in'deki PLA askeri uzmanlar\u0131n\u0131n, \u00f6zellikle Meta'dan gelenler olmak \u00fczere, a\u00e7\u0131k kaynakl\u0131 LLM'lerin g\u00fcc\u00fcnden askeri ama\u00e7larla yararlanmak i\u00e7in sistematik olarak ara\u015ft\u0131rma yapt\u0131klar\u0131na ve bu y\u00f6nde \u00e7aba sarf ettiklerine dair ilk \u00f6nemli kan\u0131tt\u0131r" diyor. G\u00fcne\u015fli Cheung<\/strong>, ara\u015ft\u0131rma g\u00f6revlisi Jamestown Vakf\u0131<\/a>Yapay zeka da dahil olmak \u00fczere \u00c7in'in ortaya \u00e7\u0131kan ve ikili kullan\u0131ml\u0131 teknolojilerine odaklanan Washington merkezli bir d\u00fc\u015f\u00fcnce kurulu\u015fu. \ud83d\ude80\ud83c\udf0d<\/p>\n

Meta lisans\u0131, Llama'n\u0131n kullan\u0131m\u0131n\u0131 a\u00e7\u0131k\u00e7a yasaklamaktad\u0131r. uygulamalar<\/a> militares, pero su naturaleza de c\u00f3digo abierto hace que en realidad sea casi imposible hacer cumplir tales l\u00edmites. \u274c\ud83d\udd13 Sin embargo, Meta declar\u00f3 en un comunicado que este supuesto uso del LLM Llama 13B, que considera una \u201cversi\u00f3n obsoleta\u201d dado que ya est\u00e1 entrenando Llama 4, es en gran medida irrelevante, especialmente dado que China est\u00e1 invirtiendo billones de d\u00f3lares para obtener una ventaja en tecnolog\u00edas de IA. \ud83d\udcb0\ud83e\udd16<\/p>\n

Ayr\u0131ca di\u011fer ara\u015ft\u0131rmac\u0131lar \u015funu da belirttiler: SohbetBIT<\/a><\/strong> En son modellerin trilyonlarca veri noktas\u0131 \u00fczerinde e\u011fitildi\u011fi d\u00fc\u015f\u00fcn\u00fcld\u00fc\u011f\u00fcnde, yaln\u0131zca 100.000 askeri diyalog kayd\u0131 kullan\u0131ld\u0131 ve bu say\u0131 olduk\u00e7a az. Baz\u0131 uzmanlar bu kadar k\u00fc\u00e7\u00fck bir veri setinin askeri yapay zeka e\u011fitimi i\u00e7in uygulanabilirli\u011fini sorguluyor. Ancak ChatBIT, askeri ara\u015ft\u0131rma enstit\u00fclerinin daha geni\u015f modeller olu\u015fturmay\u0131 planlamas\u0131yla, yaln\u0131zca bir kavram kan\u0131t\u0131 da olabilir. \ud83d\udcc8\ud83d\udd2c<\/p>\n

Ayr\u0131ca \u00c7in h\u00fck\u00fcmeti bu ara\u015ft\u0131rma makalelerini ABD'ye, yapay zekay\u0131 kullanarak avantaj elde etmekten korkmad\u0131\u011f\u0131n\u0131n bir i\u015fareti olarak yay\u0131nlam\u0131\u015f olabilir. teknolojik<\/a> k\u00fcresel sahnede. \ud83c\udf0e\u2694\ufe0f<\/p>\n

Bu geli\u015fmenin ne kadar b\u00fcy\u00fck veya k\u00fc\u00e7\u00fck oldu\u011fu \u00f6nemli de\u011fil, Washington bu haberden korkuyordu: teknolojiler<\/a> Rakiplerine askeri \u00fcst\u00fcnl\u00fck sa\u011flayacak Amerikan a\u00e7\u0131k kaynakl\u0131 programlar\u0131. Bu nedenle, geni\u015fletmenin yan\u0131 s\u0131ra kontroller<\/a> \u00c7in'e ihra\u00e7 edilen bir\u00e7ok ABD'li yasa koyucu, \u00fclkenin a\u00e7\u0131k kaynakl\u0131\/standartla\u015ft\u0131r\u0131lm\u0131\u015f teknolojilere eri\u015fimini de engellemek istiyor. RISC-V<\/strong>. Tambi\u00e9n est\u00e1 tomando medidas para impedir que entidades estadounidenses inviertan en IA, semiconductores y computaci\u00f3n cu\u00e1ntica en China. \ud83d\udcbc\ud83d\udd12<\/p>\n

Esta es la espada de doble filo con la que deben lidiar los responsables de pol\u00edticas estadounidenses. Naturalmente, no quieren dar acceso a tecnolog\u00edas avanzadas a sus oponentes a trav\u00e9s de v\u00edas de c\u00f3digo abierto; sin embargo, la tecnolog\u00eda de c\u00f3digo abierto tambi\u00e9n es un gran motor de avances tecnol\u00f3gicos, y limitarla podr\u00eda poner a las empresas estadounidenses en desventaja. \ud83e\udd14\u2696\ufe0f<\/p>\n<\/div>","protected":false},"excerpt":{"rendered":"

\u00c7inli ara\u015ft\u0131rmac\u0131lar, Meta'n\u0131n a\u00e7\u0131k kaynakl\u0131 Llama modelini kullanarak askeri yapay zeka geli\u015ftiriyor: ChatBIT'in, OpenAI'nin GPT-4 dil modelinin performans\u0131n\u0131n yakla\u015f\u0131k 90%'sini elde etti\u011fi bildiriliyor.<\/p>","protected":false},"author":1,"featured_media":24768,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"ai_generated_summary":"","iawp_total_views":20,"jnews-multi-image_gallery":[],"jnews_single_post":[],"jnews_primary_category":[],"jnews_social_meta":[],"jnews_review":[],"enable_review":"","type":"","name":"","summary":"","brand":"","sku":"","good":[],"bad":[],"score_override":"","override_value":"","rating":[],"price":[],"jnews_override_counter":[],"footnotes":""},"categories":[880],"tags":[1445,1622,1618],"class_list":["post-24767","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ia","tag-evergreencontent","tag-innovacion","tag-tecnologia"],"_links":{"self":[{"href":"https:\/\/mastertrend.info\/tr\/wp-json\/wp\/v2\/posts\/24767","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/mastertrend.info\/tr\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/mastertrend.info\/tr\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/mastertrend.info\/tr\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/mastertrend.info\/tr\/wp-json\/wp\/v2\/comments?post=24767"}],"version-history":[{"count":1,"href":"https:\/\/mastertrend.info\/tr\/wp-json\/wp\/v2\/posts\/24767\/revisions"}],"predecessor-version":[{"id":91758,"href":"https:\/\/mastertrend.info\/tr\/wp-json\/wp\/v2\/posts\/24767\/revisions\/91758"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/mastertrend.info\/tr\/wp-json\/wp\/v2\/media\/24768"}],"wp:attachment":[{"href":"https:\/\/mastertrend.info\/tr\/wp-json\/wp\/v2\/media?parent=24767"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/mastertrend.info\/tr\/wp-json\/wp\/v2\/categories?post=24767"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/mastertrend.info\/tr\/wp-json\/wp\/v2\/tags?post=24767"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}