🚀 I-Meta yembula imodeli yokuqala ye-AI evela eqenjini elibizayo le-superintelligence ↗
Ekugcineni iMeta ibonise isandla sayo neMuse Spark, imodeli yokuqala evela eqenjini elibizayo le-superintelligence eyalihlanganisa ukuze ibuyise amandla emjahweni wemingcele. Iqala ukusebenza kuhlelo lokusebenza lwe-Meta AI kanye newebhusayithi, bese idlulela ku-WhatsApp, Instagram, Facebook, kanye ne-smart glass stack - okungenzeka ukuthi yiyona ngxenye ebaluleke kakhulu.
Okuzwakala kuthakazelisa kakhulu ukuthi i-Meta igcine usayizi wemodeli ufihliwe futhi yanikeza ukubuka kuqala okuyimfihlo kuphela, ukuphambuka okubonakalayo ekumeni kwayo okuvulekile okukhulu. Ukuhlolwa okuzimele kusikisela ukuthi iyakwazi ukwenza imisebenzi yolimi neyokubona, kodwa isalandela ekubhaleni ikhodi nasekucabangeni okungaqondakali, ngakho-ke lokhu kuzwakala sengathi akubuyi ngokuphelele kodwa kufana ne-trailer yokubuya. (Reuters)
🛡️ I-OpenAI ikhipha uhlelo olusha lokuphepha ukubhekana nokwanda kokuxhashazwa kwezingane ngokocansi ↗
I-OpenAI ikhiphe i-Child Safety Blueprint ehlose ukutholakala okusheshayo, ukubika, kanye nokuphenya ngokuxhashazwa kwezingane okunikwe amandla yi-AI. Lesi sinyathelo sifika njengoba ingcindezi iqhubeka ikhula mayelana nendlela izinhlelo zokukhiqiza ezingasetshenziswa kabi ngayo ngezithombe ezingamanga, ukuzilungisa, kanye nokushushumbisa ngokocansi - indawo embi, ngokusobala, kodwa manje eyinhloko yengxoxo ye-AI.
Ingxenye ephawulekayo ukuthi lokhu akufakiwe njengesibuyekezo somkhiqizo kodwa njengomdlalo wenqubomgomo. I-OpenAI ibonakala ihlose ukukhombisa ukuthi ifuna imithetho, hhayi nje ukufinyelela, efundeka njengenomthwalo wemfanelo - futhi ehlelekile kahle. (TechCrunch)
🧩 I-Atlassian yethula amathuluzi e-AI abonakalayo kanye nama-ejenti eqembu lesithathu e-Confluence ↗
I-Atlassian isunduza i-AI ijule kakhulu endaweni amaqembu asevele ehlala kuyo - i-Confluence - ngethuluzi elisha elibonakalayo elibizwa ngokuthi i-Remix kanye nama-ejenti eqembu lesithathu axhunywe nge-MCPs. Iphimbo lilula: guqula amadokhumenti abe amashadi, ihluzo, ama-prototypes, izinhlelo zokusebenza zokuqala, kanye nama-slide decks ngaphandle kokushintsha amathebhu amahlanu njenge-squirrel ene-caffeine.
Uhlu lwabalingani lusho lukhulu futhi. I-Lovable, i-Replit, ne-Gamma zixhunyiwe ukuze ikhasi libe yisibonelo, i-app stub, noma isethulo esinobunzima obuncane, okuzwakala kakhulu njenge-"AI njengeglue yomsebenzi" kunokuba kube "AI njengethoyizi elicwebezelayo eliseceleni." (TechCrunch)
🏦 I-Citigroup ithi i-AI isiza ukusheshisa ukuvulwa kwama-akhawunti kanye nokuthuthukiswa kwezinhlelo ↗
I-Citi ithi i-AI isivele inciphisa isikhathi sokusebenza esibalulekile, ikakhulukazi ekufakweni nasekuhlanzweni kwesistimu endala. Inqubo eyodwa yokubuyekezwa kwemibhalo yokuvulwa kwama-akhawunti ase-US yehle isuka esikhathini esingaphezu kwehora yaya cishe emizuzwini eyi-15, okungeyona i-AI yabathengi ekhangayo - kodwa uhlobo lokuwina buthule amabhodi avame ukukuthanda.
Kukhona ungqimba lwesibili lapha: ibhange liphinde linciphise ukuthembela kosonkontileka be-IT bangaphandle ngenkathi lakha okwengeziwe ngamathimba angaphakathi. Ngakho-ke yebo, kuyindaba yokusebenza kahle, kodwa futhi kuyindaba yokulawula - i-AI njenge-wrench, umshanelo, futhi mhlawumbe nomhleli omusha wenkampani ngesikhathi esisodwa. (Reuters)
💼 I-Anthropic kungenzeka ukuthi ivale igebe lemali engenayo ku-OpenAI. Nakhu okushiwo yi-IPO yabo ↗
Esinye sezizathu ezinkulu zokwethuka kwebhizinisi: I-Anthropic manje ifuna imali engenayo engaphezu kwezigidigidi ezingama-30 zamaRandi ngonyaka, ngaphambi kwe-OpenAI ebikwe ngamaRandi ayizigidigidi ezingama-24. Lokho kuwukushintsha okukhulu kusukela esikhaleni sangaphambilini, futhi kusikisela ukuthi isidingo esikhulu sebhizinisi, esigxile kumakhodi singadlula udumo lwabathengi - noma okungenani sifinyelele isivinini esimangazayo.
Konke kubalulekile ngoba zombili izinkampani zibheke ekufinyeleleni amafasitela e-IPO angaba khona, futhi abatshalizimali manje baqhathanisa amabhizinisi amabili ahlukene kakhulu e-AI esikhundleni sokuphatha i-OpenAI njengomnqobi ozenzakalelayo. Abasebenzisi abavamile bakha ukunaka, ngokuqinisekile, kodwa imisebenzi yamabhizinisi afuna amathokheni ibonakala ikhiqiza imali. Lokho kubonakala kuyisibonakaliso, noma kunjalo. (Reuters)
📱 I-Poke yenza ukusebenzisa ama-ejenti e-AI kube lula njengokuthumela umbhalo ↗
I-Poke izama ukwenza ama-ejenti e-AI azizwe sengathi awafani nesofthiwe kodwa azizwe sengathi athumela imiyalezo kumngane onekhono kakhulu. Isebenza nge-iMessage, i-SMS, i-Telegram, nakwezinye izindawo i-WhatsApp, iphatha imisebenzi yansuku zonke efana nokuhlela, izikhumbuzi, izexwayiso ze-imeyili, ukulandelela impilo, kanye nezilawuli zekhaya elihlakaniphile ngolimi olulula.
Lokho kuzwakala kucocekile kakhulu, futhi mhlawumbe kunjalo - kodwa umqondo womkhiqizo kulula ukuwuqonda emakethe esacwila kuma-demo. Esikhundleni sokucela abantu ukuthi basebenzise isikhombikubona esisha ngokuphelele, i-Poke ifaka ama-ejenti emkhubeni omdala kakhulu efonini yakho: ukuthumela umlayezo. Kuhlakaniphile, kuhlakaniphile, mhlawumbe kuyingozi kancane, mhlawumbe kuphefumulelwe kancane. (TechCrunch)
Imibuzo Evame Ukubuzwa
Yini eyangithinta kakhulu kulezi zindaba ze-AI?
Iphethini ecacile kakhulu ukuthi intuthuko ye-AI yayibukeka kakhulu kwezohwebo nokusebenza kunokuba ibe sezingeni eliphezulu. I-Meta ifake imodeli entsha eyinhloko esixukwini sayo sabathengi, i-OpenAI igcizelele ukuphathwa kokuphepha, kanti i-Citi igcizelele ukonga isikhathi ezinqubweni zangaphakathi. Uma zihlanganiswa, lezi zindaba ziphakamisa ukuthi umjaho wamanje awusiwo nje amalungelo okuqhosha amamodeli, kodwa umayelana nokusabalalisa, ukwethembana, kanye nenani lebhizinisi elilinganiswayo.
Kungani ukwethulwa kwe-Muse Spark kaMeta kubaluleke kakhulu kunokukhishwa kwemodeli eyodwa nje?
I-Muse Spark ibalulekile ngoba i-Meta iyifaka ngqo emikhiqizweni abantu abayisebenzisa kakade, okuhlanganisa uhlelo lwayo lokusebenza, iwebhusayithi, amapulatifomu omphakathi, kanye nezibuko ezihlakaniphile. Lokho kunikeza imodeli indlela eyakhelwe ngaphakathi yokusetshenziswa kwangempela kunokuyishiya emjikelezweni wedemo ozimele. Ukwethulwa kuphinde kubonise ukuthi i-Meta ifuna ukubuyela engxoxweni yasemngceleni, noma ngabe lokhu kubukeka njengokusetha kabusha kunokushintsha ngokuphelele ubuholi.
Kungani iMeta ithule kakhulu ngeMuse Spark kunangesikhathi ikhipha amamodeli avulekile ngaphambilini?
Ukukhishwa okuthule kuvelele ngoba i-Meta ayizange idalule usayizi wemodeli futhi yagcina ukufinyelela kunqunyelwe ngokubuka kuqala kwangasese. Lokho kuphawula ithoni ehlukile ekubekweni kwayo kwemodeli evulekile yomphakathi esikhathini esidlule. Ngokusekelwe kulesi sihloko, umyalezo ongaba khona kakhulu uyisixwayiso: bonisa inqubekela phambili, hlola ukusebenza kwangempela, futhi ugweme ukubiza ngokweqile ngenkathi imodeli isabonakala ibuthakathaka ekubhaleni ikhodi nasekucabangeni okungaqondakali.
Yini ngempela eboniswa yi-OpenAI yokuphepha kwezingane?
Ayifani nokwethulwa kwesici kodwa ifana nesimo senqubomgomo. Uhlelo lugxile ekuthuthukiseni ukutholakala, ukubika, kanye nophenyo mayelana nokuxhashazwa kwezingane okunikwe amandla yi-AI, okuhlanganisa izithombe ezingamanga ezicacile, ukulungiswa, kanye nokushushumbiswa ngokocansi. Isibonakaliso esibanzi ukuthi i-OpenAI ifuna ukubonakala ihlanganyela ekubekeni imithetho, hhayi nje kuphela ekuthumeleni izinhlelo ezinamandla emakethe esheshayo nezwela kwezepolitiki.
Amathuluzi amasha e-Atlassian e-Confluence AI kufanele ashintshe kanjani umsebenzi wansuku zonke?
Isifiso sisekunciphiseni ukushintshashintsha komongo ngaphakathi kwemisebenzi yeqembu. Esikhundleni sokuhamba phakathi kwamathuluzi ahlukene, abasebenzisi bangaguqula amadokhumenti abe amashadi, ihluzo, ama-prototype, izinhlelo zokusebenza zokuqala, kanye nezethulo ezivela ngaphakathi kwe-Confluence. Njengoba ozakwethu abanjengoLovable, Replit, kanye ne-Gamma bexhunywe nge-MCPs, umqondo uwukuthi i-AI iba yi-workflow glue yamaqembu kunokuba umsizi ongaxhunyiwe ohlala kwenye ithebhu.
Lezi zindaba ze-AI zebhizinisi zisikiselani mayelana nokuthi i-AI iwusizo kuphi njengamanje?
Izibonelo eziqine kakhulu lapha ziyasebenza futhi ziyasebenza. I-Citi ichaze umsebenzi wokubuyekezwa kwemibhalo wokuvulwa kwama-akhawunti ase-US okwehla kusuka esikhathini esingaphezu kwehora kuya cishe emizuzwini eyi-15, futhi ixhumanise i-AI nokuthuthukiswa kwesistimu kanye nokunciphisa ukuthembela kosonkontileka. Lokho kukhomba ukuthi i-AI yebhizinisi iletha inani kuqala ngesivinini senqubo, ukulawulwa kwangaphakathi, kanye nokuthuthukiswa, hhayi kuphela kokuhlangenwe nakho kwe-chatbot okubhekene nomphakathi.
Kungani ukwenyuka kwemali engenayo ye-Anthropic kushintsha indlela abantu abayiqhathanisa ngayo ne-OpenAI?
Lesi sihloko sichaza lokhu njengoshintsho olukhulu ngoba i-Anthropic manje ithi inani lemali engenayo elitholakala minyaka yonke liphezulu kune-OpenAI. Lokho kubalulekile ngoba kuphonsela inselelo ukucabanga ukuthi ukubonakala kwabathengi kuholela ngokuzenzakalelayo emiphumeleni yebhizinisi enamandla kakhulu. Kuphinde kuqinise umehluko phakathi kwezinkampani ezimbili ze-AI ezinamandla ahlukene, ikakhulukazi njengoba abatshalizimali bebheke phambili emafasiteleni e-IPO angaba khona futhi baqhathanise isidingo sebhizinisi nokufinyelela kwabathengi.
Kungani ama-ejenti e-AI asekelwe kumbhalo afana ne-Poke ezizwa ebalulekile noma ngabe umqondo uzwakala ulula?
Inzuzo yabo ukujwayelana. Esikhundleni sokucela abantu ukuthi bafunde isikhombikubona esisha, iPoke ibeka ukuziphatha kwe-ejenti ngaphakathi kwemikhuba yokuthumela imiyalezo abasebenzisi asebeyazi kakade nge-iMessage, i-SMS, i-Telegram, kanye ne-WhatsApp. Lokho kwenza umkhiqizo ube lula ukuwuqonda, kuyilapho kuphakamisa izinkinga: lapho ama-ejenti ezizwa ekhululekile njengokuthumela imiyalezo, ukungezwani kokwamukela kuyehla, kodwa futhi nokuphumula abantu abavame ukukwenza ngaphambi kokunikeza imisebenzi kungaba njalo.