Zininzi izinto zokusebenza zobukrelekrele ezenza ukuba ubomi bethu bemihla ngemihla bube lula kakhulu kwaye bunemveliso ngakumbi. Kwisantya sokwenza izinto ezintsha, yonke into inokwenziwa ngomyalelo omnye.
I-AI ifikeleleka ngakumbi kwinani elikhulayo labantu kwihlabathi liphela, kodwa njengoko obu buchwephesha budala amathuba amatsha okuphucula kunye noncedo lwemihla ngemihla, ukuqhubela phambili kwayo kuphakamisa imibuzo malunga nokuba isebenza njani - umzekelo, zeziphi iingxaki ezinokubangela ukuba ayiphuhliswanga ngokufanelekileyo.
Njengoko igama lisitsho, bubukrelekrele obenziwe ngumntu, kodwa obenziwa ngoomatshini kwaye, ukuya kuthi ga kwinqanaba elithile, ubuchule obufanayo nobabantu: iyafunda, iyaphucula kwaye iyakwazi ukusebenza kwiindawo ezithile.
Xa sithetha ngobukrelekrele bokuzenzela, izikolo ezimbini ezinkulu zokucinga zingqubana: abo bacinga ukuba sisixhobo, akusekho, kwaye abo bakholelwa ukuba yinto nje yexesha ngaphambi kokuba ibe sisoyikiso kuluntu.
Njengoko amandla ethu e-AI kunye namathuba okwanda, tSiza kubona ukuba iyakusetyenziselwa iinjongo eziyingozi okanye ezinobungozi. Yiyo loo nto abo babona obu buchwephesha njengesoyikiso bayijonga ngokukrokra kwaye boyika ifuthe elinokuba nalo kubomi babo. Abanye abantu abadumileyo njengo-Elon Musk baphakathi kwabo.
Umphathi weTesla kunye ne-SpaceX sele elumkise ngaphezulu kwesihlandlo esinye: I-AI iya kuphumelela ubuchule bokuqonda kwabantu. UMusk ukholelwa ukuba obu buchwephesha buza kusongela abantu kwixa elizayo, ngakumbi emsebenzini.
Kungenxa yesi sizathu le nto inkampani yakhe iNeuralink isebenza kubuchwephesha boomatshini bobuchwephesha obuza kufakwa kukhakhayi ukulungiselela uluntu ikamva "elibulalayo" apho iirobhothi ziya kulilawula khona. Inyani yile, kukho ezinye iimuvi ze-sci-fi ezoyikisayo abantu nazo, zibonisa ikamva le-dystopi apho i-AI ifumana ukulawula abantu.
Abaphandi bathi i-AI ayinakulindeleka ukuba ibonise iimvakalelo zabantu njengothando okanye inzondo kwaye akukho sizathu sokulindela ukuba i-AI izakuba intle okanye ngenjongo.
Ngale ndlela, UGoogle ukhathazekile malunga nengozi enokubangelwa yi-AI xa ungakhuli ngenkathalo kunye nendlela onxibelelana ngayo nabantu. I-AI kufuneka ifunde njengomntu, kodwa ihlale isebenza kakuhle kwaye ungabi ngumatshini oyingozi. UGoogle ube ngumdlali ophambili kuphuhliso lwe-AI.
Ngenkqubo yayo yophando yePentagon, "Iprojekthi Maven," inkampani iye "yaqeqesha" i-AI ekuhleleni izinto kwimifanekiso yedrone. Ngamanye amagama, ifundise iidrones ukuba ziqonde ukuba zijonge ntoni.
UGoogle ngoku uthi ubuGcisa bobuGcisa buya kuhlala noBias kwaye inkampani ifuna ukwenza okuthile ngayo. Ukwenza oku, uGoogle ubeke iinkqubo ezifanelekileyo kumxholo othi "Uxanduva lwe-AI".
Isiseko sesibini se-AI yeGoogle "kukuba noxanduva ebantwini" kwaye "kuthintelwe ukudala okanye ukomeleza ucalucalulo olungenabulungisa". Oku kubandakanya ukuphuhliswa kweenkqubo zobuntlola obutolikekayo obubeka abantu phambili kuwo onke amabakala enkqubo yophuhliso, ngelixa kuqinisekiswa ukuba ucalucalulo olungenabulungisa umntu anokuba nalo alubonakali kwiziphumo zemodeli.
Ngokwesi sikhokelo, uGoogle uzabalazela ukukhulisa ubukrelekrele bokuzenzela ngokuthe gabalala kwaye useka iindawo ezithile zesicelo ezingayi kuzilandela, njengokungasebenzisi ubukrelekrele bokufakelwa kubuchwephesha obunokubangela ukwenzakala okanye ukwenzakala kwabantu.
Uphando izakuzama ukuqinisekisa ukuba ulwazi olufumanekayo ngeemodeli ze-AI zichaneke kwaye zikumgangatho ophezulu. Ngapha koko, itekhnoloji "kufuneka inoxanduva lokuphendula kubantu, kuxhomekeke kulwalathiso lomntu nolawulo."
Ubuchwephesha bobuchwephesha bobugcisa kunye neeseti zedatha zinokubonakalisa, zomeleze, kwaye zinciphise ucalucalulo olungenabulungisa. Ngale ndlela, uGoogle uya kuzama ukunqanda iimpembelelo ezingafanelekanga ebantwini, ngakumbi ezo zinxulumene neempawu ezibuthathaka ezinje ngohlanga, ubuhlanga, isini, umvuzo, ubuzwe okanye iinkolelo zopolitiko okanye ezenkolo, phakathi kwabanye.
Umthombo: https://ai.google