🛡️ Ukuphepha kwe-Codex: manje kusekuhlolweni kocwaningo ↗
I-OpenAI iqalise ukuhlola kuqala kocwaningo lwe-Codex Security kubasebenzisi be-ChatGPT Pro, Enterprise, Business, kanye ne-Edu ngokusebenzisa uhlelo lokusebenza lwewebhu lwe-Codex. Inkulumo iqondile impela - vumela i-ejenti ye-AI ukuthi ithole futhi ilungise izinkinga zokuphepha kohlelo lokusebenza ngaphambi kokuba zibe yikhanda elibuhlungu kakhulu. (OpenAI)
Iqala ngokusetshenziswa kwamahhala enyangeni yokuqala, okuzwakala sengathi kuwukucindezela ngamabomu ukwenza amaqembu azame ngokushesha. Abashayeli bezindiza abasizayo bezokuphepha sebeqala ukubukeka njengomjaho wezikhali ezincane wabo manje. (OpenAI)
🇬🇧 I-UK kufanele isekele indlela yokunikeza amalayisense kuqala yokuqeqeshwa kwe-AI, kusho ikomidi eliphezulu lendlu ↗
Ikomidi leNdlu yamaKhosi yase-UK lithe uhulumeni kufanele akhethe imodeli yokuqala yelayisensi yokuqeqeshwa kwe-AI, kunokuncika ezintweni ezibanzi ezingavumela izinkampani ukuthi zisebenzise izinto ezinelungelo lobunikazi ngokukhululeka. Lokho kuyisibonakaliso esikhulu empini yelungelo lobunikazi - mhlawumbe hhayi ekhangayo, kodwa ebaluleke kakhulu. (Reuters)
Lesi sincomo senezela ingcindezi kongqongqoshe ukuthi babonise ukuthi abagcini nje ngokuvula indlela yezinkampani ze-AI ngezindleko zabadali. Ngendlela ethule, lokhu kungathinta ukuziphatha kwe-AI ngaphezu kwengxenye yokwethulwa kwemodeli exakile. (Reuters)
🏗️ I-Oracle ne-OpenAI bayeka uhlelo lokukhulisa isikhungo sedatha saseTexas, ngokusho kweBloomberg News ↗
Kubikwa ukuthi i-Oracle ne-OpenAI bahoxise izinhlelo zokwandisa isayithi lesikhungo sedatha saseTexas. Lokho akusho ukuthi ukuchuma kwengqalasizinda sekuphelile - kude nakho - kodwa kusikisela ukuthi ngisho nokwakhiwa kwe-AI okukhulu kudlula ezinkingeni ezinzima, ukubalwa kabusha, noma kokubili. (Reuters)
Njengoba isidingo se-AI sisakhula, noma yikuphi ukuhoxa kuphrojekthi enkulu yokubala kuyagqama. Kuyisikhumbuzo sokuthi ikusasa le-AI akulona nje amamodeli nama-demos - umhlaba, amandla, ukupholisa, izinkontileka... izinto okuthiwa ziyisicefe ezizwakala ziyisicefe ngokuzumayo. (Reuters)
📈 Amasheya eMarvell Tech akhuphuka ngombono we-AI chip weminyaka eminingi okhuphukile ↗
UMarvell wagxuma ngemuva kokunikeza isibikezelo seminyaka eminingi esihambisana nesidingo sama-chip e-AI enziwe ngokwezifiso avela kumakhasimende amakhulu obuchwepheshe. Abatshalizimali bazwile ngokucacile lokho ababefuna ukukuzwa: Ukusetshenziswa kwezimali kwingqalasizinda ye-AI kusaphila kakhulu, mhlawumbe kungakhululekile kancane. (Reuters)
Lesi sinyathelo siphinde siqinise ukuthi indaba ye-AI ingaphansi kwengqimba ye-chatbot. Ama-chips, ukuxhumana, kanye ne-silicon eyenziwe ngokwezifiso kuyaqhubeka nokuphakamisa ukunaka - kanye nemali - ngoba yilapho ingxenye enkulu yenzuzo ikhona khona. (Reuters)
🇨🇳 Umnotho wesifundazwe omkhulu kunayo yonke eShayina uthembisa ukwakha kabusha imboni nge-AI ↗
IJiangsu, umnotho wesifundazwe omkhulu kunayo yonke eShayina, ithe ifuna ukusebenzisa i-AI ukwakha kabusha imboni. Ukugcizelelwa akugcini nje emikhiqizweni yabathengi ekhangayo kodwa nasekusetshenzisweni kwezimboni, okuyinto engathandeki kangako, ngokuqinisekile, kodwa mhlawumbe lapho ushintsho olukhulu lwezomnotho luzofika khona. (Reuters)
Lesi simemezelo sifanelana nokucindezela okubanzi kwaseShayina kokuhambisa i-AI ngokujulile emikhakheni yokukhiqiza kanye nemikhakha eyinhloko. Ngakho-ke yebo, enye indaba yenqubomgomo - kodwa enesisindo esikhulu, ngoba isikali eJiangsu sifana nokuqondisa umkhumbi wezimpahla ngenetha yemizwa eboshelwe kuwo. Mhlawumbe isithombe esingavamile, yize sibamba iphuzu. (Reuters)
💬 Ngemuva kweYurophu, i-WhatsApp izovumela izinkampani ze-AI ezincintisana nazo ukuthi zinikeze ama-chatbot eBrazil ↗
I-Meta ivula i-WhatsApp eBrazil ukuze incintisane nabahlinzeki be-chatbot ye-AI ngemali, kulandela isinyathelo esifanayo eYurophu. Lokho kuyaphawuleka ngoba kususa i-WhatsApp ekubeni nje indawo ye-Meta AI iye kokuthile okufana nesendlalelo sokusabalalisa sabathengisi abaningi be-AI. (TechCrunch)
Futhi kuyisibonakaliso esithule sokuthi ukulawulwa kwepulatifomu kanye nokufinyelela kwe-AI kuyaxakeka kakhulu. Izimpi ze-chatbot azisekho nje mayelana nokuthi ubani onemodeli ehlakaniphe kakhulu - zimayelana nokuthi ubani ongumnikazi womnyango. (TechCrunch)
🦊 Ukubambisana neMozilla ukuthuthukisa ukuphepha kweFirefox ↗
U-Anthropic uthe uClaude Opus 4.6 usebenze nabacwaningi beMozilla futhi wathola ubuthakathaka obungu-22 beFirefox emavikini amabili, kanti iMozilla ibeke abangu-14 babo njengobunzima obukhulu. Lokho kuwumphumela omkhulu, futhi kunikeza i-"AI yokuphepha" inkomba enkulu kunezithembiso ezivamile ezingacacile. (Anthropic)
Okubalulekile ukuthi amathuluzi e-AI aqala ukufakazela ukubaluleka kwawo emisebenzini emincane, enenani eliphezulu lapho isivinini sibalulekile futhi umphumela ungahlolwa ngabantu. Irobhothi elincane lomlingo, umzingeli wezinambuzane one-caffeine kakhulu. (Anthropic)
Imibuzo Evame Ukubuzwa
Kuyini i-Codex Security futhi kungani kubalulekile emaqenjini okuphepha kwezinhlelo zokusebenza?
I-Codex Security yethulwa njenge-ejenti ye-AI engathola futhi ilungise izinkinga zokuphepha kohlelo lokusebenza ngaphambi kokuba zibe izinkinga ezinkulu. Kulokhu kuqoqwa, yethule ekuhlolweni kocwaningo lwabasebenzisi be-ChatGPT Pro, Enterprise, Business, kanye ne-Edu ngohlelo lokusebenza lwewebhu lwe-Codex. Inyanga yokuqala yokusetshenziswa kwamahhala iphinde iphakamise ukuthi i-OpenAI ifuna amaqembu ayihlole ngokushesha emisebenzini yangempela.
Kungani inyanga yokuqala yamahhala ye-Codex Security iyindaba enkulu kangaka?
Inyanga yokuqala yamahhala yehlisa isithiyo samaqembu ezokuphepha ukuthi azame ngaphandle kwengcindezi yesabelomali esheshayo. Lolu hlobo lwesu lokuqalisa luvame ukusiza abathengisi ukuthi baqoqe ukusetshenziswa, impendulo, kanye nezifundo zamacala okuqala ngokushesha. Ngokwezinto ezingokoqobo, luguqula ilukuluku libe yizilingo futhi lungasheshisa ukwamukelwa uma amaqembu ebona imiphumela ewusizo kusenesikhathi.
Kusho ukuthini ukuphakanyiswa kokuqala kwelayisensi e-UK ngokuqeqeshwa kwe-AI kanye nelungelo lobunikazi?
Ikomidi leNdlu yamaKhosi libonisa ukuthi ukuqeqeshwa kwe-AI kufanele kuncike kakhulu ekutholeni amalayisense kunokungavunyelwe okubanzi kwe-copyright. Lokho kubalulekile ngoba kusunduza inqubomgomo ekufinyeleleni okuxoxiswana ngakho esikhundleni sokusetshenziswa ngokukhululekile kwezinto ezivikelwe. Kubadali nezinkampani ze-AI, isincomo singabumba indlela idatha yokuqeqesha etholakala ngayo, ebekwa ngayo intengo, futhi ephikiswana ngayo.
Kungani le ngxoxo yezindaba ze-AI igxile kakhulu engqalasizinda, hhayi kuma-chatbot kuphela?
Izinto eziningana zikhomba ngaphansi kwesendlalelo se-chatbot ezinhlelweni ezenza i-AI ibe nokwenzeka ngezinga. Ukuguqulwa kwesikhungo sedatha saseTexas kugqamisa umhlaba, amandla, ukupholisa, kanye nezinkontileka, kuyilapho isibikezelo sikaMarvell sikhomba kuma-chips kanye nokuxhumana njengezindawo ezinkulu zokusebenzisa amandla. Ndawonye, zikhombisa ukuthi ukukhula kwe-AI kuncike kakhulu engqalasizinda ebonakalayo kanye nakuma-demo amamodeli.
Kungani i-OpenAI ne-Oracle kungabalulekile ukulahla uhlelo lokukhulisa isikhungo sedatha saseTexas?
Ukwandiswa okukhanseliwe kuvelele ngoba isidingo se-AI sisachazwa njengesiqinile. Kusikisela ukuthi ngisho namaphrojekthi engqalasizinda ye-AI amakhulu kakhulu angafinyelela imingcele esebenzayo noma adinge ukubalwa kabusha kwesu. Okubaluleke kakhulu ukuthi ukukhula kwe-AI akukona okungenazingxabano, ikakhulukazi lapho ukwakheka kwezibalo kuncike ezinsizeni zangempela eziyinkimbinkimbi.
Ukwanda kukaMarvell kusho ukuthini ngokuthi imali ye-AI igeleza kuphi?
Ukuhlangana kukaMarvell kukhombisa ukuzethemba kwabatshalizimali ekufuneni okuqhubekayo kwama-chip e-AI enziwe ngokwezifiso avela kumakhasimende amakhulu obuchwepheshe. Lokho kuqinisa umqondo wokuthi ingxenye enkulu yezindleko ze-AI isagxile kwihadiwe, ukuxhumana, kanye nengqalasizinda. Ngisho noma i-AI ebhekene nabathengi iba yizindaba, indaba yemali engenayo ejulile ivame ukuhlala ezinhlelweni ezingaphansi.
Kungani ukukhuthazwa kwe-AI esifundazweni saseJiangsu eShayina kubalulekile emakethe ebanzi ye-AI?
IJiangsu ichazwa njengomnotho wesifundazwe omkhulu kunayo yonke eShayina, ngakho-ke ukushukuma kwe-AI yezimboni lapho kuthwala izinga langempela. Kugxilwe ekushintsheni imikhakha yokukhiqiza kanye nemikhakha eyinhloko kunokwakha imikhiqizo yabathengi ekhangayo kuphela. Lokho kubalulekile ngoba umthelela omkhulu wezomnotho we-AI ungavela ekusetshenzisweni okuthule ngaphakathi kwemboni, ezokuthutha, kanye nemisebenzi yokukhiqiza.
Ukuvulwa kwe-WhatsApp kuma-chatbot ancintisanayo eBrazil kuwushintsha kanjani umjaho weplatifomu ye-AI?
Kusikisela ukuthi i-WhatsApp isiba ngaphezu kwekhaya le-Meta AI futhi isifana nesiteshi sokusabalalisa sabahlinzeki be-AI abaningi. Lokho kushintsha kubalulekile ngoba ukulawulwa kokufinyelela kwabasebenzisi kungabaluleka njengekhwalithi yemodeli. Ezimakethe eziningi ze-AI, ukuba nomnyango wabasebenzisi kungabumba umncintiswano njengoba nje nokuba nomsizi ohlakaniphe kakhulu.