🛡️ Abalawuli base-UK bashesha ukuhlola izingozi zemodeli yakamuva ye-Anthropic AI, ngokusho kwemibiko ye-FT ↗
Kubikwa ukuthi abaqaphi bezezimali baseBrithani baqhubeka ngokushesha nohlelo olusha lwe-Anthropic lwe-Mythos ngemuva kokukhathazeka ngokuthi lungadalula ubuthakathaka obukhulu ezinhlelweni ezibalulekile. Izingxoxo zihilela iBhange LaseNgilandi, i-FCA, i-Treasury, kanye ne-National Cyber Security Centre - okuwuhlu olukhulu lokubuka kuqala uhlobo olulodwa.
Okukhathazayo ukuthi iMythos inekhono elingavamile ekutholeni ubuthakathaka besofthiwe, yize i-Anthropic ithi ukuqaliswa kwayo kulinganiselwe kakhulu futhi kuhloswe emsebenzini wokuvikela. Amabhange, izinkampani zomshuwalense, kanye nezindawo zokushintshana kulindeleke ukuthi zinikezwe ulwazi, ngakho-ke lokhu manje kuzwakala sengathi akusafani nenhlebo yelabhorethri kodwa kufana nenqubomgomo yengqalasizinda.
🤖 Engqungqutheleni ye-HumanX, wonke umuntu wayekhuluma ngoClaude ↗
Kubonakala sengathi i-anthropic ibeka isimo sengqondo ku-HumanX - leyo kwakuyindaba. Kuwo wonke amaphaneli, izingxoxo zabathengisi, kanye nezingxoxo zasephasishi, uClaude wayelokhu ebonakala njengemodeli abantu abayifunayo kakhulu emisebenzini yebhizinisi kanye neyokubhala amakhodi.
Okwagqama kakhulu kwaba ukuthula okuhlobene ne-ChatGPT. Lo mbhalo usikisela ukuthi i-OpenAI isenayo imali kanye nesilinganiso, kodwa isimo sengqungquthela sathambekela kuClaude njengethuluzi elinamandla okusebenza kwebhizinisi okwamanje - noma kubonakala kanjalo.
📄 I-Claude for Word ka-Anthropic ingenye inselele embusweni wesofthiwe ye-Microsoft ↗
I-Anthropic yethule inguqulo ye-beta ye-Claude ye-Word, iqhubekela phambili kusofthiwe yansuku zonke yehhovisi esikhundleni sokuhlala ivaliwe njengethuluzi lonjiniyela. Ingaphendula imibuzo mayelana nedokhumenti enezingcaphuno zesigaba ezingachofozeka, ihlele umbhalo okhethiwe ngaphandle kokufometha okuphukile, futhi iphathe izinguquko ezilandelelwe.
Izilaleli eziqondiwe zicacile: amaqembu agcwele amadokhumenti, ikakhulukazi ezomthetho kanye nezezimali. Lokho kwenza lokhu kungaphezu nje kwesengezo sezici - kuyisinyathelo esiqondile endaweni yasekhaya yeMicrosoft, futhi kuyisinyathelo esihlekisayo impela.
🌐 Ama-AI Bots Asebenzisa I-Strip-Mineraving Web, Futhi I-Anthropic Ihola Iqembu ↗
Ukubukwa okusha kwedatha ye-Cloudflare kuphikisana ngokuthi izinkampani ze-AI zikhipha okuningi kakhulu kuwebhu kunalokho ezikubuyisayo. Isibalo esiyinhloko yisilinganiso se-crawl-to-refer - ukuthi ama-bots aklwebha kangaki amakhasi uma kuqhathaniswa nokuthi athumela kangaki abasebenzisi emuva - futhi i-Anthropic iphumelela kakhulu.
Isilinganiso sika-Anthropic sibikwa njengo-8,800 ku-1, kanti i-OpenAI nayo iphakeme kodwa iphansi kakhulu. Iphuzu elibanzi lifika kanzima: isivumelwano esidala sewebhu sasisithi skrola, bese sithumela ithrafikhi. Izinjini zempendulo ze-AI zigcina ingxenye yokuqala bese zivumela eyesibili ibe yinkungu.
⚔️ Ukubuka umncintiswano wezikhali ze-AI okhula ngokushesha emhlabeni wonke, njengoba i-US, i-China, i-Russia, kanye nabanye bephuthuma ukwakha izikhali ezizimele kanye nezinhlelo zokuzivikela ezisekelwa yi-AI ↗
Umbiko wakamuva wezempi ne-AI awudingi kakhulu ama-demo akhangayo kodwa umayelana nokusheshisa amasu. Amazwe amakhulu asebenzisa izinhlelo zokuzivikela ezizimele kanye nezinhlelo zokuzivikela ezisizwa yi-AI ngesikhathi esisodwa, okwenza yonke into izwakale njengomjaho wokugijima lapho wonke umuntu ebamba khona induku futhi ekhohlwa amabhuleki.
Lokho kuphakamisa imibuzo evamile - ukulawula, ukukhuphula izinga, ukuziphendulela - kodwa ngokuphuthuma okukhulu manje. Uma i-AI isingene ngokujulile ekuqondisweni, ekuqapheni, nasekuthathweni kwezinqumo enkundleni yempi, "ukuhlola" kuqala ukuzwakala njengelebula lephepha kokuthile okuvele kuhamba.
🏭 I-SoftBank yaseJapan yethula iyunithi yokuthuthukisa i-AI ekhulele ekhaya ↗
I-SoftBank ihlanganyele ne-Sony, i-Honda, nezinye izinkampani zaseJapan ukwakha ibhizinisi elisha le-AI eligxile "ku-AI ebonakalayo" - amamodeli ahloselwe ukulawula imishini namarobhothi, hhayi ukuphendula ngombhalo kuphela. Lowo umzila ohlukile kumncintiswano ojwayelekile we-chatbot, futhi mhlawumbe othakazelisa kakhulu.
Isifiso sikhulu, njengoba umzamo kuthiwa uhlose imodeli yesisekelo enkulu kakhulu kanye nomzila omude wokusetshenziswa kwezimboni. Ingxenye yayo isu lezobuchwepheshe bezwe, ingxenye ukuthwebula inyanga, ingxenye i-anime ye-mech nemihlangano yebhodi.
Imibuzo Evame Ukubuzwa
Kungani abalawuli base-UK bebheka eduze imodeli ye-Anthropic's Mythos?
Imibiko iphakamisa ukuthi iziphathimandla zase-UK zibheka i-Mythos njengengozi engaba khona ngoba ingase ibe nekhono elingavamile ekutholeni ubuthakathaka besofthiwe. Kubikwa ukuthi izingxoxo zihilela iBhange LaseNgilandi, i-FCA, i-Treasury, kanye ne-National Cyber Security Centre. Lokho kuhlanganiswa kwezikhungo kubonisa ukukhathazeka ngemiphumela engaba khona ezinhlelweni ezibalulekile, ikakhulukazi kwezezimali kanye nengqalasizinda.
Yini eyenza imodeli ye-Anthropic Mythos ihluke ekukhishweni kwe-AI okuvamile?
Ukukhathazeka akukhona nje ukuthi kuyimodeli entsha, kodwa ukuthi ibonakala inekhono kakhulu emsebenzini wokuvikela ohlobene nokuthola ubuthakathaka besofthiwe. Kubikwa ukuthi i-Anthropic ithi ukufinyelela kuvinjelwe kakhulu, okubonisa ukuthi ukuqaliswa kuphathwa ngokucophelela kakhulu kunokuqaliswa kwabathengi abaningi. Empeleni, lokho kwenza kuzizwe kuseduze nekhono lokuphepha elilawulwayo kunesibuyekezo esijwayelekile se-chatbot.
Kungani uClaude ayethola ukunakwa okungaka engqungqutheleni ye-HumanX?
Ngokusho kwalesi sihloko, uClaude uvele kaningi kumaphaneli, ezingxoxweni zabathengisi, nasezingxoxweni ezingakahleleki njengenketho enamandla yemisebenzi yebhizinisi kanye neyokubhala amakhodi. Lo mbhalo wethula i-Anthropic njengenomfutho ophawulekayo kubasebenzisi bebhizinisi kulowo mcimbi. Uphinde uphakamise ukuthi isimo sengqondo esizungeze i-HumanX sasinciphile kakhulu ku-ChatGPT, yize i-OpenAI isenesilinganiso esikhulu kanye nezinsizakusebenza.
Yini uClaude for Word ayenzela amaqembu adinga amadokhumenti amaningi?
Inguqulo ye-beta echazwe lapha yenzelwe ukusebenza ngaphakathi kwamadokhumenti kunokumane ikhiqize umbhalo ozimele. Ingaphendula imibuzo mayelana nedokhumenti enezingcaphuno zesigaba, ihlele izindima ezikhethiwe ngaphandle kokuphazamisa ukufometha, futhi iphathe izinguquko ezilandelelwe. Lokho kuyenza ifaneleke kakhulu kumaqembu ezomthetho, ezezimali, namanye adinga ukubuyekezwa ngokucophelela kunokubhala phansi.
Ithini isilinganiso se-crawl-to-refer mayelana nama-AI bots kanye newebhu?
Lesi sihloko sinikeza lesi silinganiso njengendlela yokuqhathanisa ukuthi izinkampani ze-AI zikhasa kangakanani amawebhusayithi uma kuqhathaniswa nokuthi zithumela kangaki abasebenzisi emuva. Kulolo hlaka, isilinganiso esiphezulu kakhulu sibonisa ukukhipha ngethrafikhi encane yokubuyisa. I-Anthropic ichazwa njengengasebenzi kahle ikakhulukazi ngalesi silinganiso, esiqinisa ukugxekwa okubanzi kokuthi izinjini zezimpendulo ze-AI zingase zenze buthaka ukushintshana kwewebhu okudala kokufaka izinkomba ngokushintshana nokudluliselwa.
Kungani abantu bexhumanisa intuthuko ye-AI nezikhali ezizimele kanye ne-AI ebonakalayo?
Lesi sihloko sikhomba ezindleleni ezimbili ezixhumene: ohulumeni abasheshisa izinhlelo zokuzivikela ezisekelwa yi-AI, kanye nezinkampani ezifana namamodeli okusekela eSoftBank aklanyelwe ukulawula imishini namarobhothi. Ndawonye, zibonisa i-AI idlulela ngale kwengxoxo futhi isesha ezinhlelweni ezibonakalayo ezingazwa, zinqume, futhi zenze. Lokho kuphakamisa imibuzo emikhulu mayelana nokulawula, ukuziphendulela, ukusetshenziswa kwezimboni, nokuthi lezi zinhlelo zishintsha ngokushesha kangakanani kusukela ekuhlolweni kuya ekusebenzeni.