Ngaba ubulumko bokwenziwa buya kusetyenziselwa ububi?

Ngaba ubulumko bokwenziwa buya kusetyenziselwa ububi?

Nangona izibonelelo zayo ezininzi, ngaphezu kweengcali ze-350 kwintsimi yobukrelekrele bokwenziwa emhlabeni wonke, kubandakanywa nomdali we "ChatGPT", baye balumkisa ukuba le teknoloji inokukhokelela ekuphelisweni koluntu.

Kwisiteyitimenti esidibeneyo, exhaswa zii-CEO zeenkampani eziphambili ze-AI, bathi ukunciphisa lo mngcipheko "kufuneka kube yinto ephambili kwihlabathi liphela kunye neminye imingcipheko yoluntu, efana nobhubhane kunye nemfazwe yenyukliya."

Iingcali ezininzi zivakalise inkxalabo malunga nobungozi bokusebenzisa imifuziselo efana ne-ChatGBT ukusasaza ulwazi olungeyonyani kunye nolwaphulo-mthetho lwe-cyber, kunye nokudangala ukuphazamiseka kwemisebenzi kwinqanaba loluntu.

Impucuko iyawohloka

Ingxelo, eququzelelwa yi-US Centre for Artificial Intelligence Safety (CAIS), nayo yavuma ezi zinto zixhalabisayo, kodwa yathi izisongelo ezinzulu kodwa ezincinci kufuneka zixoxwe, kubandakanywa nokuba kungenzeka ukuba ukukhawuleza kwe-AI kunokukhokelela ekuweni kwempucuko.

Ezinye izazinzulu ngeekhompyutha zoyikela ukuba isazi esingesosemvelo esinezinto esinomdla kuzo ezingavisisaniyo nezo zabantu sinokusibuyisela endaweni yethu singaqondanga.

Abanye banexhala lokuba ukuthembela kakhulu kwiinkqubo esingaziqondiyo kusishiya sisengozini enkulu ukuba kukho into engahambi kakuhle.

UDan Hendricks, uMlawuli we-CAIS, wachaza ukuba le ngxelo yayiyindlela yokubonakalisa abaphandi abaninzi. “Abantu bebesoyika ukuthetha ngaphambili,” uqhube watsho.

Abanye baye bawathandabuza la mabango

Ngokwahlukileyo koko, ezinye izifundiswa zayikhaba ngawo omane le ngxelo njengengenancedo. UDkt. Mahiri Aitken, i-Ethics Ethics Fellow kwi-Alan Turing Institute, yabiza ngokuthi "ukuphazamiseka" kwiisongelo ezingakumbi ezivela kwi-AI.

Wongeze wathi, "Ibali le-super-AI licebo eliqhelekileyo elivela kwi-blockbusters yaseHollywood engenakubalwa."

UGq. Carissa Velez, weZiko leeNqobo ezisesikweni kwiYunivesithi yaseOxford kwiArtificial Intelligence, wayethandabuza iinjongo zabanye babatyikityi.

“Ndinexhala lokuba ukugxila kwisoyikiso esikhoyo kuphazamisa kwimiba engxamisekileyo, efana nokukhukuliseka okanye ukuphela kwedemokhrasi, ii-CEO zeenkampani ezithile abafuni kujongana nazo,” utshilo. "Ubukrelekrele bokwenziwa bunokubangela intshabalalo enkulu ngaphandle komngcipheko okhoyo," wongeze watsho.

Izixhobo zebhayoloji

Iinkxalabo malunga nezona zoyikiso zinzima ukusuka kwi-AI ukusuka ekunokwenzeka ukuba isetyenziswe ngabantu ukuyila izixhobo zebhayoloji, ukuya kwi-AI ngokwayo iceba ukuwa kwempucuko.

Phakathi kwabatyikitye kwisiteyitimenti nguSam Altman, i-CEO ye-OpenAI, eyila i-ChatGPT, kunye no-Demis Hassabis, i-CEO ye-Google Deepmind, eyila isoftware ukoyisa abona badlali babalaseleyo kwi-Go and chess. Kwasayinwa ziintloko ze-Anthropic AI, i-Inflection AI kunye ne-Stability AI iinkqubo kunye neeprofesa zesayensi yekhompyutha ezivela eCambridge, Oxford, Harvard, Yale naseStanford.

Uqikelelo lukaMaguy Farah lwehoroscope yonyaka ka-2023