🤖 I-OpenAI ithumela i-GPT-5.1 ngokucabanga okubukhali nokwenza kube ngokwakho
I-OpenAI yethule i-GPT-5.1, iletha ukucabanga okuguquguqukayo okuzizwa kungaqinile ezingxoxweni ezinde futhi okulawulwa kalula amaqembu. Kuyanda, yebo-kodwa kucwebezelisiwe ngempela.
Ukwenza ngokwezifiso kube lula - kuxhunywe emisebenzini yokusebenza ngama-hoops ambalwa, izivikelo eziqinile, kanye nomuzwa wokuthi ilalela ngempela. Ayiphelele... kodwa ayiqinile kangako.
🔗 Funda kabanzi
🏗️ I-Anthropic yenza u-$50B ezikhungweni zedatha ze-US AI
I-Anthropic ithi ithela u-$50B ezikhungweni zedatha ezenziwe ngokwezifiso eTexas naseNew York nomlingani wayo uFluidstack - ukubheja okukhulu kakhulu ekukhuleni kukaClaude. Inani elikhulu; isifiso esikhulu.
Uhlelo lugxile ekusebenzeni kahle kokuqeqeshwa kwemingcele kanye, yebo, imisebenzi eminingi emisha. Uhlobo lwesikali esenza izikweletu zikagesi zishise.
🔗 Funda kabanzi
🎶 Abalaleli abakwazi ukusho umculo we-AI ovela kubantu... kakhulu
Ucwaningo lwe-Deezer-Ipsos luthole ukuthi abantu abangu-97% abakwazi ukubona izingoma ezenziwe nge-AI ngendlela ethembekile. Lokho kuyaxaka - futhi kuyaphazamisa kancane abaculi namapulatifomu ngokufanayo.
Amalungelo, ukwahlukaniswa kwemali engenayo, izinjini zezincomo - konke lokhu kufanele kucatshangelwe kabusha uma izindlebe zingakwazi ukubona noma kunjalo. Noma kubonakala kanjalo.
🔗 Funda kabanzi
🧠 I-SPICE ye-Meta igudluzela ama-LLM ekuzifundeleni
Abacwaningi be-Meta baveze i-SPICE - uhlaka lokuzidlalela lapho imodeli eyodwa idala izinkinga ezinzima, ezisekelwe kumadokhumenti kanti enye izama ukuzixazulula. Udube oluhle; izinzuzo ezinkulu.
Izinombolo zokuqala zibonisa cishe ukwehla okungu-10% ezilinganisweni zokucabanga ngokugwema izihibe zedatha yokwenziwa. Igumbi elincane le-echo, umbhalo wezwe langempela.
🔗 Funda kabanzi
💼 I-'Cameyo by Google' ihlala ne-Chrome Enterprise + Gemini
I-Google ifake i-Cameyo yajula kwi-Chrome Enterprise, yengeza izici ze-Gemini ukuze isakaze izinhlelo zokusebenza ze-Windows kuma-Chromebook ngaphandle kwe-VDI faff ephelele. Kuhle - futhi mhlawumbe kushibhile.
Kumayelana nokunikeza i-IT indlela elula yokuletha izinhlelo zokusebenza ezindala ngenkathi ithembele ku-AI yokuphatha kanye nosizo lwabasebenzisi. Ama-ejenti ambalwa, izinhlungu ziyancipha.
🔗 Funda kabanzi
⚖️ I-OpenAI ibuyisela emuva ukuze inikeze izingxoxo ze-ChatGPT
Empikiswaneni yezomthetho ne-New York Times, i-OpenAI ilwa nomyalo wokuguqula izigidi zezingxoxo zabasebenzisi. Ama-alamu obumfihlo yonke indawo, ngokwemvelo.
Izingqinamba zizwakala zinkulu kunethemba elilodwa lomsebenzisi, ukuphepha kwedatha, kanye nalapho izinkantolo zidweba khona umugqa ebufakazini bokuqeqeshwa kwe-AI. Kubalulekile, kungcolile.
🔗 Funda kabanzi