🌍 I-OpenAI ifuna ukwandisa ukusetshenziswa kwe-AI emhlabeni wonke empilweni yansuku zonke ↗
I-OpenAI iqhubela phambili umkhankaso wokwamukelwa “okugxile emazweni” - inkulumo ethi: yeka ukuphatha i-AI njengethoyizi lelebhu bese uqala ukuyihlanganisa nezinsizakalo zomphakathi abantu abazithinta nsuku zonke.
Izibonelo ezimbalwa ezifakiwe zifaka phakathi ukuthunyelwa kwezemfundo (cabanga: amathuluzi esitayela se-ChatGPT asetshenziswa ngezinga elikhulu), kanye nomsebenzi ezinhlelweni ezisebenzayo njengezixwayiso zakuqala zezinhlekelele ezihlobene namanzi. Kungumyalezo othi “lokhu kufanele kube wusizo buthule”, engingenakukugwema ukuwuthanda… noma ngabe futhi, uyazi, kuyisu lebhizinisi.
⚡ I-OpenAI ithembisa ukukhokha izindleko zamandla ze-Stargate ↗
I-OpenAI ithi izokhokhela izindleko zamandla ezihlobene nokwakhiwa kwesikhungo sayo sedatha se-Stargate, isithi ayifuni ukwenyusa amanani kagesi endawo kuwo wonke umuntu.
Lokho kungasho ukuxhasa amandla amasha kagesi, isitoreji, ukuthuthukiswa kokudlulisela, noma enye ingqalasizinda evimbela igridi ukuthi ingakhali ngaphansi komthwalo. Ingxenye iwukuma kwezakhamuzi ezinhle, ingxenye ipolitiki yokuzikhethela - zombili lezi zinto zingaba yiqiniso ngesikhathi esisodwa, ngendlela ecasulayo.
🎧 I-OpenAI ihlose ukuthumela idivayisi yayo yokuqala ngo-2026, futhi ingaba ama-earbud ↗
Amahemuhemu ehadiwe ye-OpenAI ayaqhubeka nokukhuphuka, futhi lona uthi idivayisi yokuqala ingaba ama-earbud - igajethi "yokuqala ye-AI" oyigqokayo, hhayi uhlelo lokusebenza oluvulayo kuphela.
Kukhona izingxoxo mayelana negama lekhodi (“Sweet Pea”) kanye nezifiso ze-silicon ezinolaka, lapho ukucubungula okuningi kwenzeka kudivayisi kunokuba kuhlale kugxuma efwini. Ukushintsha imikhuba yabantu yokusebenzisa ama-earbud kufana nokuzama ukushintsha indlela yomfula ngesipuni… kodwa, ukubheja kwemikhiqizo yasendle kusebenze.
📜 'Umthethosisekelo' omusha ka-Anthropic kaClaude: yiba usizo futhi uthembeke, futhi ungabhubhisi isintu ↗
I-Anthropic yashicilela “uMthethosisekelo kaClaude” obuyekeziwe - ngokuyisisekelo uhlelo lwezindinganiso nokuziphatha oluhloselwe ukwakha indlela uClaude aphendula ngayo, acabange ngayo futhi enqabe ngayo.
Ibeka imingcele ebukhali ezungeze izinto eziyingozi ngempela (izikhali, ukulimala okukhulu, ukubamba amandla, ukushaya okukhulu okubi), kuyilapho futhi ithambekele “ezimisweni” ezicacile kunemithetho eqinile. Ingxenye eyodwa ebabayo ngokumangazayo: iveza obala ukungaqiniseki ngokuthi ama-AI athuthukile angadinga ukucatshangelwa ngokuziphatha - hhayi ukuthi ayakufanelekela, imane iphathe lo mbuzo njengophilayo.
🛡️ Izici ezintsha zokuphepha kanye nezici zokuthola i-AI ze-Google Workspace for Education ↗
I-Google imemezele izici ezintsha zokuphepha ze-Workspace for Education, ngokugxila okuphawulekayo ekutholeni nasekuqinisekiseni okuqukethwe okukhiqizwe yi-AI - ngoba ulwazi olungelona iqiniso lwesikole lusakazeka njenge-glitter futhi aluphumi ngokuphelele kukhaphethi.
Esinye isici esivelele ukuqinisekiswa okuhlobene ne-SynthID yezithombe namavidiyo akhiqizwe yi-AI kuhlelo lokusebenza lwe-Gemini, kanye nokuvikelwa okuqinile njengokutholwa kwe-ransomware kanye nokutholwa kwamafayela okulula kwe-Drive kudeskithophu. Kuyinhlanganisela yomsebenzi wokuphepha ongathi sína kanye nokuthi “sicela usithembe emakilasini,” okuyinto… elungile, uma ngikhuluma iqiniso.
🚫 I-CEO ye-Google DeepMind iphinda ithi 'akukho zinhlelo' zezikhangiso ze-Gemini, imangazwe ukuthi i-ChatGPT izifake 'kusenesikhathi' ↗
UDemis Hassabis we-DeepMind uphinde wathi iGemini ayinazo izinhlelo zezikhangiso, futhi uzwakale emangele kancane ukuthi i-ChatGPT ishintshele ezikhangisweni “kusenesikhathi.”
Kuyisinyathelo esihlanzekile sokubeka izinto - umsizi njengomkhiqizo, hhayi umsizi njengebhodi lokukhangisa. Lokho kusho ukuthi, “akukho zinhlelo” kuyigama lenkampani elithi “ngibuze futhi kamuva,” ngakho… sizobona ukuthi leso sithembiso siqine kangakanani.
🏛️ Isiqondiso esisha sizosiza i-UK ukulawula i-AI ngempumelelo nangendlela efanele ↗
I-Alan Turing Institute ishicilele uhlaka lwamandla okulawula kanye nethuluzi lokuzihlola elihlose ukusiza abalawuli base-UK ukuthi bahlole ukuthi bakulungele yini ukuqondisa i-AI.
Akukuhle neze - uhlu lokuhlola, izici zamandla, izitatimende "zokuthi kubukeka kanjani okuhle" - kodwa yilokho kanye okushiwoyo. Ukubusa kwangempela kuvame ukubukeka njengephepha elinemiphetho ebukhali, hhayi amagama asemqoka acacile... noma kubonakala kanjalo.
Imibuzo Evame Ukubuzwa
Iyini inhloso ye-OpenAI “egxile emazweni” yokwandisa ukusetshenziswa kwe-AI emhlabeni wonke empilweni yansuku zonke?
Kuyindlela yokususa ukuphatha i-AI njengomkhiqizo ozimele bese uyihlanganisa nezinsizakalo zomphakathi abantu abathembele kuzo nsuku zonke. Inhloso iwukuhlanganisa i-AI nezinhlelo ezisebenzayo nezithinta kakhulu, kunokuyishiya ivaleleke kuma-demos “elabhorethri”. Izibonelo ezishiwo zifaka phakathi ukuthunyelwa kwemfundo enkulu kanye namathuluzi okuphepha abhekene nomphakathi. Ihlelwe njengewusizo buthule, noma ngabe isekela isu elibanzi lokwamukelwa kwe-OpenAI.
Uhlelo lwe-OpenAI lungabonakala kanjani ezinkonzweni zomphakathi njengezemfundo noma ekuphenduleni ezinhlekeleleni?
Kwezemfundo, umqondo ngamathuluzi esitayela se-ChatGPT asetshenziswa ngezinga elikhulu, ukuze izikole zikwazi ukugoqa i-AI zibe yimisebenzi yokufunda evamile. Ekuphenduleni inhlekelele, i-OpenAI igcizelele izinhlelo zokuxwayisa zakuqala ezihlobene namanzi njengesibonelo sengqalasizinda "esebenzayo ngokuthula". Ingxoxo efanayo usizo olulethwa lapho kudingeka khona, ngamathuluzi afakwe ezinsizakalweni ezikhona. Impumelelo cishe izoncika ekuthembekeni, ekuqondiseni, nasekuhlanganisweni okusekelwe.
Kungani i-OpenAI ithi izokhokhela izindleko zamandla zokwakhiwa kwesikhungo sedatha se-Stargate?
I-OpenAI ikubeke njengomzamo wokungakhuphuli amanani kagesi endawo ngokufaka isidingo esikhulu esisha kugridi. Ukumboza izindleko kungafaka phakathi ukuxhasa amandla amasha kagesi, isitoreji, ukuthuthukiswa kokudlulisela, noma enye ingqalasizinda enciphisa ukucindezeleka. Kubekwe njengesinyathelo somakhelwane omuhle kanye nendlela yokuvimbela ukucindezelwa kwezepolitiki kanye nomphakathi. Empeleni, kubonisa ukuthi imiphumela yamandla manje iyisisekelo sezinhlelo zokukhulisa i-AI.
Yini yakamuva ngamahemuhemu e-hardware e-OpenAI - ingabe ama-earbud “okuqala nge-AI” anengqondo ngempela?
Umbiko usikisela ukuthi i-OpenAI ihlose ukuthumela idivayisi yayo yokuqala ngo-2026, ngama-earbud njengesici esingaba khona. Lo mqondo uyinto egqokwayo "yokuqala nge-AI" oyisebenzisa njalo, hhayi uhlelo lokusebenza oluvulayo kuphela. Kukhona izingxoxo mayelana negama lekhodi ("i-Sweet Pea") kanye nezifiso zokucubungula okwengeziwe kudivayisi kunokuthembela ngokuphelele efwini. Leyo ndlela ingenza ukuxhumana kuzwakale kushesha futhi kuyimfihlo, kuye ngokusetshenziswa.
Uyini “uMthethosisekelo kaClaude” ka-Anthropic obuyekeziwe, futhi yini eshintshayo?
Kuyipulani yezindinganiso nokuziphatha eshicilelwe eyenzelwe ukwakha indlela uClaude aphendula ngayo, acabange futhi enqabe ngayo izicelo. Lesi sibuyekezo sigcizelela imingcele ebukhali ezungeze izindawo eziyingozi ngempela njengezikhali, ukulimala okukhulu, kanye nokuziphatha kokufuna amandla. Siphinde sithambekele "ezimisweni" ezicacile kakhulu esikhundleni semithetho eqinile. Okuphawulekayo ukuthi siphatha udaba lokucatshangelwa kokuziphatha kwama-AI athuthukile njengendaba evulekile kunesimangalo esixazululiwe.
Yiziphi izici ezintsha zokuphepha kanye nokutholwa kwe-AI i-Google ezifake ku-Workspace for Education?
I-Google imemezele ukuvikelwa okuqinile kwe-Workspace for Education, okuhlanganisa nokugxila ekutholakaleni nasekuqinisekisweni kokuqukethwe okukhiqizwe yi-AI. Okunye okuvelele ukuqinisekiswa okuhlobene ne-SynthID yezithombe namavidiyo akhiqizwe yi-AI kuhlelo lokusebenza lwe-Gemini. Lesi sibuyekezo siphinde sikhulume ngokutholwa kwe-ransomware kanye nokutholwa kwamafayela okulula kwe-Drive kudeskithophu. Sekukonke, ibekwe njengomsebenzi wokuphepha ogxile ekilasini ohlose ukusiza izikole ukuphatha ulwazi olungelona iqiniso kanye nezigameko ngempumelelo enkulu.
Ingabe i-Gemini kaGoogle izobonisa izikhangiso, futhi kungani i-DeepMind yaphawula ngokuzifaka kwe-ChatGPT?
I-CEO ye-DeepMind uDemis Hassabis uphinde wathi i-Gemini “ayinazo izinhlelo” zezikhangiso futhi waveza ukumangala ukuthi i-ChatGPT ishintshele ezikhangisweni “kusenesikhathi.” Lokho kuthumela imiyalezo kubeka i-Gemini njengomsizi njengomkhiqizo kunokuba kube umsizi njengebhodi lokukhangisa. Ngesikhathi esifanayo, “akukho zinhlelo” kushiya isikhala sezinguquko zesikhathi esizayo ngaphandle kokwenza isithembiso esibophayo. Kubasebenzisi, kugqamisa ukuhlukana kwemodeli yebhizinisi okuvelayo kubasizi be-AI yabathengi.
Yini okufanele yenziwe uhlaka lwamandla okulawula i-AI lwe-Alan Turing Institute?
Luhlaka lwamandla okulawula kanye nethuluzi lokuzihlola elihlose ukusiza abalawuli base-UK ukuthi bahlole ukuthi bahlonyiswe yini ukuphatha i-AI ngempumelelo. Le ndlela ayidumisi ngamabomu: uhlu lokuhlola, izici zamandla, kanye nezitatimende "zokuthi kubukeka kanjani okuhle". Umgomo ukwenza ukubusa kusebenze, hhayi nje okufiswayo. Kuwusizo kakhulu kubalawuli kanye namaqembu enqubomgomo adinga ukuhlonza izikhala ngaphambi kokuba kukhuphuke ukuphoqelelwa noma ukuqapha.