Pi i-chatbot ehlose ukugxila kakhulu ezingxoxweni "zabantu".
Ezinsukwini ezimbalwa ezedlule kwavela lokho I-Inflection AI, inkampani yaseCalifornia eyasungulwa umsunguli we-LinkedIn u-Reid Hoffman kanye nomsunguli we-DeepMind u-Mustafa Suleyman, yethule i-Pi, i-chatbot ehlose ukubeka phambili izingxoxo zabantu ngezinga eliphezulu lobuhlakani bemizwa, kusho inkampani. I-Pi inokusetshenziswa okulinganiselwe kune-ChatGPT, i-Google Bard, noma i-Bing chatbot ye-Microsoft, kodwa ingcono kunalezo zinhlelo zokusebenza zokuqhuba izingxoxo ngendlela yomuntu.
Nge-Pi, i-Inflection AI ijoyina umjaho nezinkampani zobuchwepheshe ezinkulu nezincane ukuthuthukisa imikhiqizo ye-chatbot ngokusekelwe kubuhlakani bokwenziwa okukhiqizayo.
Mayelana nePi
I-Pi (personal intelligence) iyi-chatbot entsha eqale ukubonakala emakethe yezingxoxo ye-AI eqhudelana kakhulu. Pngihloselwe ukwenza njengomlaleli okhuthele, isiza abasebenzisi ukuthi babuze imibuzo noma baxazulule izinkinga ngengxoxo eya phambili naphambili abazoyikhumbula kamuva. Umgomo uwukwazi abasebenzisi.
I-Inflection AI ibeke phambili kakhulu ekuthuthukiseni ubuhlakani obungokomzwelo be-chatbot kanye nokuqinisekisa ukuthi uziphatha ngendlela efanele kuzo zonke izimo. Inkampani uqeqeshe uPi ukuthi aqaphele ukuziphatha kwakhe futhi ugweme noma yikuphi ukwephulwa kwezinqubomgomo zabo, okuhlanganisa nokungazibandakanyi ezingxoxweni zothando. Umgomo uwukuqinisekisa ukuthi i-chatbot ayilokothi izame ukulingisa umuntu futhi ikhumbuza njalo abasebenzisi ukuthi i-AI.
“UPi uzimisele ngokukusiza ngayo yonke indlela angakwazi ngayo. Khuluma noPi mayelana nesihloko osithakaselayo, kusukela ku-quantum physics kuya ku-trivia ye-movie. Cela iseluleko ngezinqumo ezinkulu nezincane empilweni yakho. Noma thola amathiphu awusizo angakusiza ufunde izinto ezintsha zokuzilibazisa namakhono, ”kusho inkampani ku-tweet.
mustafa suleman, i-CEO ye-Inflection AI, uchaze i-chatbot njengebhodi elizwakalayo elinozwela, esikhundleni somthombo wolwazi. Uthe i-chatbot yenkampani inamacala okusebenzisa alinganiselwe, ayenza iphephe futhi kube lula ukuyilawula.
Pi yakhelwe ukubeka phambili izingxoxo ezifana nezomuntu ngezinga eliphezulu lobuhlakani bemizwa, kuhlanganise umusa nokusekela.
“U-Pi unobhalansi futhi akachemi ezindabeni zezepolitiki noma ezihlokweni ezibucayi, kodwa futhi angaba ahlekisayo, ayisiphukuphuku futhi asungule ngezinye izikhathi. Ilungele imisebenzi yomuntu yansuku zonke,” kusho uSuleyman. Uthe inkampani inemikhawulo ebekiwe yokuqinisekisa ukuthi i-AI ayiphuli imigomo yayo yokuziphatha, okuhlanganisa nokungena ezingxoxweni zothando.
“Inhloso wukuqinisekisa ukuthi i-AI ihlala iyazi ukuthi iyi-AI futhi ayilokothi izame ukuzenza umuntu. Ngakho-ke kuvame ukukhumbuza umsebenzisi ongumuntu ukuthi bayi-AI, ”kusho yena. Ngokusho kwakhe, abasebenzisi bangaxhumana ne-Pi ezinkundleni ezahlukene, okuhlanganisa iwebhusayithi yayo, uhlelo lokusebenza, kanye nezinkundla zokuxhumana ezifana ne-Instagram. Isevisi imahhala, futhi ukuqalisa kungase kuqalise okubhaliselwe kwe-premium esikhathini esizayo. Le nkampani ithi isebenzisa idatha yomsebenzisi, okuhlanganisa nokuqukethwe kwengxoxo, ukuqeqesha izinhlelo zayo ze-AI. I-Pi okwamanje ayixhunyiwe ku-inthanethi.
Ngesikhathi sombukiso obukhoma we I-Pi, yazehlukanisa kwamanye ama-chatbots afana ne-ChatGPT noma i-Bard ngenxa yokuthi wayevame ukuphetha izimpendulo zakhe ngombuzo womsebenzisi, ngaleyo ndlela ekhuthaza inkhulumomphendvulwano.
"Yilokho uPi akwenza kahle kakhulu, kuyasiza ukwenza ucwaningo lwabo," kusho uSuleyman. Kodwa-ke, akanikezi izingcaphuno noma izinkomba, nakuba u-Suleyman ethi lokho kuzoshintsha. Ngaphezu kwalokho, uPi naye kwesinye isikhathi wenza amaqiniso. Kalula nje, kuyamangaza njengoba kwenza kuzo zonke izinhlobo zezilimi ezinkulu (ama-LLM), ubuchwepheshe obusekela isizukulwane esilandelayo sama-chatbots anamandla e-AI.
Ama-Chatbots asebenzisa ubuchwepheshe be-AI okukhiqizayo abe yisihloko esishisayo kusukela i-OpenAI's ChatGPT yakhululwa ngoNovemba odlule. Ngokusebenzisa amamodeli ezilimi ezinkulu, ukusebenzisa amanani amakhulu ombhalo ukuze kufinyezwe ulwazi futhi kukhiqizwe okuqukethwe, ama-chatbots enza abantu bakwazi ukuba nezingxoxo ezijulile zebhizinisi nezidingo zabo siqu.
Okokugcina Uma unesifiso sokwazi okwengeziwe ngakho, ungabheka imininingwane Kulesi sixhumanisi esilandelayo.