Friends

MarcLee

Mayelana nalolu hlelo lokusebenza

NgoDisemba 2018, i-NVIDIA yanyakazisa umhlaba ngokubonisa ukuthi i-Artificial Intelligence ingadala kanjani izithombe ezingokoqobo zabantu abangekho.

Abangane baphakamisa imiphumela yalolu cwaningo futhi banikeze ulwazi olujulile lokuhlola inani elikhulu lokuqukethwe okukhiqizwa yi-AI. Ngokusebenzisa umakhalekhukhwini, ubuso obungenakubalwa bukhiqizwa futhi bugqolozele umsebenzisi kunoma iyiphi indlela. Zonke izithombe ezijwayelekile ezibukekayo zabantu zingamanga: zenziwa ngokungahleliwe yi-AI.
Izithombe zihlaziywa zibe endaweni ye-3D ekwazi ukuzulazula futhi ziyazungeza ukuze zihlale zibuka umsebenzisi, ukuze zibhekise izithombe zephrofayela yenkundla yezokuxhumana.

Abasebenzisi benkundla yezokuxhumana banamuhla bahlale bebhekene nezinkundla ezifuna ukuveza izinto abazithandayo (okuthandayo, ukubalwa kwabalandeli nabalandelayo...) ukuze benze inzuzo engapheli. Lezi zinkundla seziphenduke amathuluzi omhlaba ezokuxhumana kanye nokufinyelela olwazini, izindlela esixhumana ngazo sodwa nokufunda ngomhlaba. Izingxenyekazi eziklanyelwe ngokuyinhloko ukukhiqiza ukubandakanya nokukhula okuningi ngangokunokwenzeka. Singaqonda kanjani kangcono - futhi ngaleyo ndlela sikwazi ukumelana - ukuthi lezi zinhlelo zinethonya kanjani ukuthi singobani nesikwenzayo? Yiziphi izindlela zokumelana? Ingabe kufanele sifake ugaxekile kumaphrofayela ethu ngokuqhubeka sishintsha okuqukethwe okungelona iqiniso ukuze sisebenzisane nama-algorithms awo?

Ngasikhathi sinye, Abangane bahlose ukukhombisa amandla okuphazamisa e-AI kanye nokusebenza kwayo okubanzi. Ngokubeka ukugcizelela ezimisweni zokuziphatha ze-AI, Abangane basikhumbuza ngemithelela yokuziphatha engaphansi kwezinye zezinto eziyimpikiswano zokusetshenziswa kobuchwepheshe obusha: kusukela ekuziphatheni kwedatha kuya ekwesabeni "imishini ebusa umhlaba". Njengoba singasondeli nakancane ekubusweni ngomshini, impela kube khona izehlakalo zokucekelwa phansi komphakathi yidatha emibi. Futhi uma i-AI elawulwa ngokuziphatha nangokuziphatha iyilokho esikudingayo, ingabe i-AI kwezobuciko kufanele ibe nesimilo nesimilo? Noma ingabe ubuciko kufanele buhlale bulwela ukweqa imingcele yokuziphatha neyokuziphatha yomphakathi?

Ngokusebenzisa iselula noma ithebulethi, izithombe eziningi zokuma ngobude zihlanganiswa ngokungahleliwe kuhlelo lokusebenza ngezicelo ze-HTTP. Bonke bangibhekile njalo. Isithombe ngasinye sithola igama nesibongo esingahleliwe, izinhlamvu ezintathu ngasinye. Ukugqwayiza nemisindo kulandela ukunyakaza komsebenzisi: indawo ebonakalayo iyazungeza lapho umsebenzisi ezungezisa idivayisi. Isibhakabhaka sivela uma idivayisi iyiswa phezulu. Ngokutshekisa idivayisi phansi, kuvela phansi. Imvelo ebonakalayo ayipheli futhi ingazulazula yonke indawo.
Umsindo wakhelwe uhlelo lokusebenza futhi usabela kukho konke lokhu kunyakaza kanye nezivinini zokuzulazula.
Isibonisi se-Mobile App singavezwa odongeni olulodwa noma ngaphezulu endaweni yombukiso.

IZIKWELETU
UMarc Lee ngokubambisana noShervin Saremi (Umsindo)

IWEBHUSI
https://marclee.io/en/friends/