🤖 Umholi we-hardware ye-OpenAI wesula esikhundleni ngemuva kwesivumelwano nePentagon ↗
Umholi we-hardware kanye ne-robotics ye-OpenAI, uCaitlin Kalinowski, wesula esikhundleni ngemuva kwesivumelwano senkampani iPentagon, ethi lesi sinyathelo asizange sithole izingxoxo ezanele zangaphakathi ngaphambi kokuba amamodeli avunyelwe amanethiwekhi okuvikela ayimfihlo. Ukukhathazeka kwakhe okukhulu kwakuqondile impela - ukubhekwa kwabantu, imingcele yokuqapha, kanye nengozi yokuzimela okubulalayo okuqhubeka ngokushesha okukhulu. (Reuters)
I-OpenAI ithi lesi sivumelwano sihlanganisa ezinye izindlela zokuvikela futhi sisavimbela ukuqapha kwasekhaya kanye nokusetshenziswa kwezikhali ezizimele. Noma kunjalo, ukwesula kwenza konke kungabonakali njengokubuyekezwa kwenqubomgomo okubushelelezi kodwa kufane nomnyango womlilo ovalwa phakathi nomhlangano. (Reuters)
🏗️ I-Oracle ne-OpenAI bayeka uhlelo lokukhulisa isikhungo sedatha saseTexas, ngokusho kweBloomberg News ↗
Kubikwa ukuthi i-Oracle ne-OpenAI bahoxise izinhlelo zokwandisa isikhungo sabo sedatha se-AI esikhulu sase-Abilene, eTexas ngemuva kokuba izingxoxo zokuxhasa ngezimali ziqhubekile futhi izidingo zokubala ze-OpenAI zishintshile. Indawo yayiyingxenye yomkhankaso omkhulu wengqalasizinda ye-Stargate, ngakho lokhu akukuncane neze. (Reuters)
Ikhampasi ekhona isasebenza, futhi izinkampani zithi ukwakhiwa okubanzi kwama-gigawatt amaningi kuyaqhubeka nakwezinye izindawo. Ngesikhathi esifanayo, ukunwetshwa okushiywe kungavulwa manje indlela ye-Meta, okubikwa ukuthi ixoxe ngokuqasha indawo esikhundleni sayo. Ingqalasizinda ye-AI iyaqhubeka nokuhamba njengosimende omanzi - osindayo, obizayo, futhi ongakaze uxazululeke kahle. (Reuters)
⚖️ IPentagon iqoka i-Anthropic njengengozi yochungechunge lokuphakelwa ↗
IPentagon yabiza ngokusemthethweni i-Anthropic ngokuthi ingozi yokuthengwa kwempahla, yavimbela uClaude ekusebenziseni izinkontileka zoMnyango Wezokuvikela ngemuva kokuba inkampani yenqabe ukukhulula izindlela zokuvikela izikhali ezizimele kanye nokuqapha abantu abaningi. Yilokho okuyinhloko - impi yokuthengwa kwempahla eyaba yimpi enkulu kakhulu yezindinganiso. (Reuters)
U-Anthropic uthi lesi sinyathelo asilungile ngokomthetho futhi uhlela ukusiphonsela inselelo enkantolo. Lesi sikhundla sinzima ngendlela engavamile kwilebhu ye-AI yase-US, futhi sifika phakathi kokuhlukana okukhulu ngokuthi izinkampani ezisebenzisa imodeli yemingcele kufanele zibhale imithetho eqinile ezinkontilekeni zokuzivikela, noma zithembele emthethweni okhona ukuthi wenze umsebenzi... okuyinto akuwona wonke umuntu ozimisele ukuyenza. (Reuters)
💬 I-Meta izovumela izimbangi ze-AI ku-WhatsApp ukuze kuvinjelwe isenzo se-EU ↗
I-Meta ithi izovumela ama-chatbot e-AI ancintisanayo ukuthi angene ku-WhatsApp eYurophu unyaka wonke, esebenzisa i-Business API, ngemuva kwengcindezi evela kubalawuli be-EU abaphikisana nokungathembani. Abancintisana nabo bakhononde ngokuthi i-Meta yayibavimba ngempumelelo ngenkathi ikhetha umsizi wayo - imininingwane ecacile, uma kubhekwa ukuthi ukusatshalaliswa okuphakathi kuye kwaba kanjani ku-AI. (Reuters)
Okubalulekile ukuthi ukufinyelela kuza nezimali, futhi abagxeki bathi lokho kusashiya izimbangi ezincane zisesimweni esingesihle. Abalawuli bayanquma ukuthi leso sivumelwano sanele yini, ngakho-ke lokhu akufani nesivumelwano sokuthula kodwa kufana nenkinobho yokumisa ehlanganiswe emaphepheni. (Reuters)
📜 I-xAI ilahlekelwa yisicelo sokumisa umthetho wokudalulwa kwedatha ye-AI yaseCalifornia ↗
I-xAI ilahlekelwe umzamo wayo wokuvimba umthetho waseCalifornia odinga izinkampani ze-AI ukuthi zidalule okwengeziwe ngedatha esetshenziselwa ukuqeqesha amamodeli azo. Ukwenqaba kwenkantolo ukumisa umthetho kusho ukuthi i-xAI manje ibhekene nengcindezi efanayo yokungafihli lutho abanye abakhi bemodeli abaye bazama, ngezinye izikhathi ngendlela exakile, ukuyigwema. (Reuters)
Kubalulekile ngoba izimpi zokuqeqeshwa kwedatha aziseyona nje eyezemfundo noma eseduze namalungelo okushicilela - seziba yinkinga yokuthobela imithetho ebukhoma. Ku-xAI, esivele ibhekene nezingxabano zokuphepha eziphindaphindwayo eziphathelene neGrok, lokhu kunezela esinye isisindo esifana nomlawuli esikhwameni. (Reuters)
Imibuzo Evame Ukubuzwa
Kungani umholi we-OpenAI we-hardware kanye ne-robotics esula esikhundleni ngemuva kwesivumelwano sePentagon?
UCaitlin Kalinowski wesula esikhundleni ngoba wayekholelwa ukuthi isinqumo sokususa amamodeli amanethiwekhi okuzivikela ayimfihlo asizange sithole ukucatshangelwa okwanele kwangaphakathi. Ukukhathazeka kwakhe kugxile ekuqaphelweni kwabantu, imikhawulo ekuqapheni, kanye nengozi yokuzimela okubulalayo okuqhubekela phambili ngokushesha okukhulu. I-OpenAI ithe isivumelwano sisavimbela ukuqapha kwasekhaya kanye nokusetshenziswa kwezikhali ezizimele, kodwa ukwesula kusho ukuthi impikiswano yangaphakathi ayizange ixazululwe ngokugcwele.
Kusho ukuthini ngempela ukuthi iPentagon ibiza i-Anthropic ngokuthi ingozi yochungechunge lokuhlinzeka?
Kusho ukuthi uClaude uzovinjelwa ukusetshenziswa ezinkontilekeni zoMnyango Wezokuvikela, okuyisithiyo esikhulu sokuthengwa kwempahla. Ngokusho kwalesi sihloko, impikiswano igxile ekwenqabeni kuka-Anthropic ukubuthakathaza izindlela zokuvikela ezihlobene nezikhali ezizimele kanye nokuqapha okukhulu. U-Anthropic uthi ukuqokwa akuvumelekile ngokomthetho futhi uhlela ukukuphikisa, ngakho-ke lokhu kungaba yimpi yenkantolo kanye nokuhlolwa kwenqubomgomo.
Kungani ukukhanselwa kwesikhungo sedatha saseTexas kubalulekile ezindabeni zomkhakha we-AI?
Ukwandiswa kwe-Abilene okushiywe ngaphandle kubalulekile ngoba kwakuyingxenye yokucindezela ingqalasizinda ebanzi ye-Stargate, okuyibopha ekuhleleni kokubala kwe-AI yesikhathi eside. I-Reuters ibika ukuthi ikhampasi ekhona isasebenza, kodwa ukubambezeleka kwezimali kanye nokushintsha kwezidingo zokubala kuholele i-Oracle ne-OpenAI ekuyekeni ukwandiswa. Kubonisa nokuthi ukubheja okukhulu kwengqalasizinda ye-AI kungashintsha ngokushesha kangakanani, ngisho noma ukwakhiwa okubanzi kuqhubeka kwenye indawo.
Kungani iMeta ivumela ama-chatbot e-AI ayizimbangi zayo ku-WhatsApp eYurophu?
I-Meta yenze lesi sinyathelo ngemuva kokucindezelwa ngabalawuli be-EU abalwa nokungathembani kanye nezikhalazo zokuthi ithanda umsizi wayo ngenkathi ikhawulela izimbangi. Ukufinyelela kungokwesikhashana futhi kugijima nge-Business API, ngakho-ke akusho ukuthi inqubomgomo evulekile ngokuphelele. Imali isasebenza, okusho ukuthi abancintisana abancane bangaqhubeka nokubhekana nezinkinga ezingokoqobo ngisho nangemva kwesivumelwano.
Usho ukuthini umthetho waseCalifornia wokudalulwa kwedatha yokuqeqeshwa kwe-AI nge-xAI?
Ukwenqaba kwenkantolo ukuvimba umthetho kusho ukuthi i-xAI manje kufanele ibhekane nezidingo ezengeziwe zokucaca mayelana nedatha yokuqeqesha. Lokho kubalulekile ngoba abathuthukisi bemodeli bavame ukuzama ukugwema ukudalulwa okuningiliziwe, ikakhulukazi lapho ukuthola idatha kubucayi noma kuphikiswana ngakho. Ngokwemibandela engokoqobo, idatha yokuqeqesha isiba yinkinga esebenzayo yokuthobela imithetho kunokuba ihlale iyimpikiswano mayelana nemigomo yocwaningo noma ingozi ye-copyright.
Yini ebaluleke kakhulu oyifundayo ezindabeni zemboni ye-AI yaleli sonto?
Ingxoxo evamile ukuthi i-AI ayibunjwa kakhulu ukwethulwa komkhiqizo kuphela kodwa ibunjwa kakhulu ukubusa, ingqalasizinda, kanye nemithethonqubo. Ubambiswano lokuvikela ludala ukungqubuzana kwangaphakathi, ukwanda kwekhompyutha kuyabiza futhi kuyashintshashintsha, kanti abalawuli bacindezela kakhulu ukufinyelela nokudalulwa. Kuyo yonke i-OpenAI, i-Anthropic, i-Meta, kanye ne-xAI, lesi sihloko sikhomba embonini ephoqeleka ukwenza ukuhwebelana kwayo kubonakale kakhudlwana.