La ubuhlakani bokufakelwa (AI) ithuthuka ngokweqa nangemikhawulo. Baya ngokuhlakanipha futhi banezinhlelo zokusebenza eziningi. Ngokuhamba kwesikhathi bazogcina sebethathe inqwaba yemisebenzi manje ephethwe ngabantu. Ubungcweti obunjengokuhlela noma ukuthuthukiswa kwehardware kungenziwa ngenye yalezi zinhlelo njengoba uGithub Copilot ekhombisile.
Imishini ingahlela njengoba lolu hlelo lusebenzisa i-GPT-3, imodeli entsha yolimi lwe-OpenAI esetshenziselwa ukufunda okujulile, ikhombisile ukukhiqiza imibhalo elingisa ukubhala komuntu. Noma kunjalo, ku UGithub Copilot baqhubekele phambili kancane futhi bakwenza kwaba namandla okukhiqiza ikhodi yomthombo ngokuzimela ngokuphelele, ngaleyo ndlela basize abathuthukisi bonge isikhathi futhi basebenze kahle.
Abaningi bayazibuza ngingaya kuphi IGithub Copilot noma amaphrojekthi afanayo ngokuzayo. Uma sebevele bekwazi ukwenza lokho, i-AI eminyakeni embalwa ingenza imisebenzi eminingi ingasebenzi, yize kuyiqiniso futhi ukuthi izokhiqiza emisha engaziwa njengamanje.
Njengokwazi kwakho, I-GPT-3 ikhiqiza umbhalo kusuka ekuqeqeshweni kwangaphambilini, kufana nokuthi ngumbhalo obhalwe ngumuntu. Le phrojekthi iyakwazi ukukhuluma nokubhala ngendlela ekholisayo, okuyinika inqwaba yezicelo ezisebenzayo. Enye yazo yile, okwenza i-AI ibhale ikhodi kube sengathi ingunjiniyela.
Kubukeka kuthembisa, futhi Microsoft, umnikazi wamanje we-GitHub, uthuthukile ukuthi uzoyihlanganisa namathuluzi wayo wokuthuthuka. Okuthile okungazongena esikhundleni sohlelo okwamanje, kepha ungabhala izingxenye zekhodi ngokuzenzekelayo, wenze iziphakamiso, nokuthi intuthuko isebenza kahle futhi iyashesha.
Kuyakwazi guqula amazwana abe yikhodi. Umqambi uzochaza umqondo abafuna ukuwusebenzisa futhi iGitHub Copilot izokhiqiza ikhodi edingekayo. Ungaqedela ngokuzenzakalela, njengama-IDE amaningi namuhla, kepha ngendlela ethuthuke kakhulu, noma ubonise ezinye izindlela zekhodi.
Imininingwane engaphezulu ngeGitHub Copilot - Iwebhu Esemthethweni
Kuyathakazelisa kakhulu, kepha enye indlela yokusheshisa intuthuko futhi engaqhathaniswa nokwakha ama-API aphezulu kakhulu, ngokwesibonelo, i-copilot ingathatha ukuphawula okufana nokuthi "xhuma ku-database bese ukhombisa iminyaka engaphezu kwengu-50" futhi lokhu kuzokhiqiza okulungile ikhodi noma ekusebenzeni ngakunye bekuzokhiqiza ikhodi ehlukile njalo, kepha uma sinomsebenzi othile ojwayelekile odabuka olimini uqobo, njenge-show_database_result ("midatabase", SHOW_AGE_GREATER (50)); Sizoba nalokhu kusebenza okufanayo emshweni owodwa futhi bekuzolingana, uma sibeka imisebenzi emikhulu efana nalena ngezilimi zamanje, kungaba ngqo kakhulu ukuletha i-algorithm kwi-PC, ngokwesibonelo ngolimi olungamanga
I-var db yohlobo lwe-String ne-val "mibase"
xhuma_ku ("192.168.1.1")
extract_regs_ Where ("ID = 24") bese ugcina umphumela
uhlu (umphumela)
vala (db)
----------
var cube uhlobo Cube3D
show_cube (cube)
colorize_face (2 kuye ku-3) ngombala RED
susa_buso (1)
intambo (jikelezisa_ukuphila ku-axis x (cube) nge-INC 0.5)
bonisa kukhonsoli ("Ngizungezisa ikhekhe")
uma (uchofoze i-cube.face (6))
stop_rotation_of (cube)