Imbangi ye-OpenAI i-Anthropic yethula uClaude, umsizi we-AI ozothatha i-ChatGPT

I-Anthropic, isiqalo uxhaso nge-Google kanye yasungulwa ngabasebenzi bakudala be-OpenAI, namuhla yethule umsizi wayo wengxoxo we-AI olindelwe kakhulu, Claudeochwepheshe abaningi abayibheka njengembangi eyinhloko ye-OpenAI’s ChatGPT.

Ngokufanayo ne-ChatGPT, u-Claude angafinyelelwa ngokusebenzisa isixhumi esibonakalayo sengxoxo futhi uyakwazi ukwenza inhlobonhlobo yemisebenzi yokuxoxa neyokucubungula umbhalo. Isofthiwe yengxoxo yakhelwe ukusiza abasebenzisi ngokufingqa, ukusesha, ukubhala ngokuhlanganyela, i-Q&A, ukubhala amakhodi nokunye okuningi.

Elinye lamaphuzu abalulekile kaClaude okuhlukanisa ukuthi yakhelwe ukukhiqiza imiphumela eyingozi kakhulu kunamanye ama-chatbots amaningi e-AI afike ngaphambi kwayo. Le nkampani ichaza uClaude “njengohlelo lwe-AI oluwusizo, oluthembekile, nolungenabungozi.”

I-Anthropic ithi isebenze izinyanga ezimbalwa ezedlule nozakwethu abanjengoNotion, Quora, kanye ne-DuckDuckGo ku-alpha evaliwe ukuze kwandiswe amandla ayo. “Abasebenzisi bachaza izimpendulo zikaClaude njengezinemininingwane futhi ziqondakala kalula, futhi bathanda ukuthi ukuxoxisana kuzizwe njengengxoxo yemvelo,” kusho inhloko yabantu kanye namakhomikhali e-Quora, Autumn Besselman, esitatimendeni.

>>Landela okuqhubekayo kwe-VentureBeat i-AI ekhiqizayo ukufakwa <<

I-Anthropic inikeza uClaude kumabhizinisi nge-API

Enye yezinto ezibalulekile zesimemezelo sanamuhla ukuthi i-Anthropic manje isinikeza u-Claude nge-API ukuze isekele amabhizinisi nezingenzi-nzuzo. (Ungabhalisela ukufinyelela kwangaphambi kwesikhathi lapha.) Intengo ayikavezwa ukufinyelela i-API.

U-Anthropic uthe esimemezelweni sakhe sokuthi uClaude “mancane kakhulu amathuba okuthi akhiqize imiphumela eyingozi, kulula ukukhuluma nayo, futhi uyalawuleka – ukuze uthole okukhiphayo okufunayo ngomzamo omncane.” Le nkampani ithe ngaphezu kokufingqa, ukusesha, ukubhala ngobuciko kanye nokubhala amakhodi, ingakwazi futhi ukuthatha isiqondiso kubuntu, iphimbo nokuziphatha, iyenze ibe yikhandidethi eliyinhloko lenkonzo yamakhasimende nezinye izixazululo zebhizinisi ezibandakanya amakhasimende.

Inkampani okwamanje iyanikezela izinguqulo ezimbili zikaClaude: UClaude noClaude Instant. I-Claude iyimodeli esebenza kahle kakhulu, kuyilapho i-Claude Instant ilula, ingabizi kakhulu futhi ishesha kakhulu.

Izibopho ze-Anthropic no-Sam Bankman-Fried

Kwakuyi-Anthropic yasungulwa ngo-2021 ngabacwaningi abavula i-OpenAI. It yathola ukunakwa ngo-April odlule lapho, ngemva kwesikhathi esingaphansi konyaka ikhona, yamemezela kungazelelwe $580 million ngoxhaso – okuyinto, kuvela, ikakhulukazi bevela Sam Bankman-Othosiwe kanye abantu FTX, inkundla cryptocurrency manje-bankrupt osolwa ngokukhwabanisa. Kuke kwaba nemibuzo yokuthi leyo mali ingatholwa yini yinkantolo ebhedayo.

I-Anthropic, ne-FTX, nayo iboshelwe ku- I-Altruism Ephumelelayo ukunyakaza, okwabizwa umcwaningi wangaphambili we-Google u-Timnit Gebru muva nje ku Inezintambo umbono “njengophawu oluyingozi lokuphepha kwe-AI.”

I-Anthropic, ezichaza ngokuthi “isebenzela ukwakha amasistimu e-AI athembekile, achazekayo, futhi alawulekayo,” idale u-Claude esebenzisa inqubo ebizwa ngokuthi “I-AI Yomthethosisekelo,” ethi isekelwe emibonweni efana nokuhlomula, ukungabi nabesilisa kanye nokuzimela.

Ngokusho kwephepha le-Anthropic elinemininingwane I-AI yomthethosisekeloinqubo ihilela ukufunda okugadiwe kanye nesigaba sokufunda okuqiniswayo: “Ngenxa yalokhu siyakwazi ukuqeqesha umsizi we-AI ongenangozi kodwa ongabalekeleli ozibandakanya nemibuzo eyingozi ngokuchaza ukuphikisa kwawo kubo.”