Theubuhlakani bokufakelwa kuyinqubo yokwakha imishini ekhaliphile esuka kumthamo omkhulu wedatha. Amasistimu afunda ekufundeni nasezintweni ezedlule futhi enze imisebenzi efana neyomuntu. Ithuthukisa isivinini, ukunemba nokusebenza kwemizamo yabantu. I-Artificial intelligence isebenzisa ama-algorithms ayinkimbinkimbi nezindlela zokwakha imishini ekwazi ukwenza izinqumo ngokwayo. Ukufunda ngomshini kanye ne deep learning zakha umnyombo weubuhlakani bokufakelwa.
I-Artificial intelligence manje isetshenziswa cishe kuyo yonke imikhakha yebhizinisi:
Manje njengoba usuyazi ukuthi ubuhlakani bokwenziwa buyini ngempela, ake sibheke ukuthi yiziphi izinhlobo ezihlukene zobuhlakani bokwenziwa?
Ubuhlakani bokwenziwa bungahlukaniswa ngokusekelwe emandleni nokusebenza.
Kunezinhlobo ezintathu ze-AI ezisuselwe emandleni:
Ngaphansi kwezici, sinezinhlobo ezine ze-Artificial Intelligence:
Okokuqala, sizobheka izinhlobo ezahlukene ze-AI esekelwe kumakhono.
I-AI encane, ebizwa nangokuthi i-AI ebuthakathaka, igxile emsebenzini omncane futhi ayikwazi ukusebenza ngaphezu kwemingcele yayo. Iqondise isethi engaphansi eyodwa yamakhono okuqonda nokuthuthuka kuwo wonke lowo mkhakha. Izicelo ze-AI emincane ziya ngokuya zivame kakhulu ezimpilweni zethu zansuku zonke njengoba izindlela zithuthuka yokufunda ngomshini kanye deep learning qhubeka nokuthuthukisa.
Apple Siri iyisibonelo se-AI encane esebenza ngebanga elilinganiselwe lemisebenzi yangaphambilidefikusihlwa. USiri uvame ukuba nezinkinga ngemisebenzi engaphezu kwamandla akhe. 
SiriIBM Watson esinye isibonelo se-AI ewumngcingo. Sebenzisa ikhompuyutha yokuqonda, ukufunda ngomshini kanyeukucubungulwa kwezilimi zemvelo ukucubungula ulwazi nokuphendula imibuzo yakho. IBM Watson wake wamehlula imbangi yakhe engumuntu Ken Jennings ekubeni ngumpetha wohlelo lwe-TV oludumile Jeopardy!. 
Narrow AI IBM WatsonNarrow AI faka phakathi Google Translate, isofthiwe yokuqaphela izithombe, izinhlelo zokuncoma, izihlungi zogaxekile kanye ne-algorithm yezinga lekhasi le-Google.Narrow AI Google TranslateI-Artificial general intelligence, eyaziwa nangokuthi ubuhlakani bokwenziwa obuqinile, iyakwazi ukuqonda nokufunda noma yimuphi umsebenzi wobuhlakani umuntu angawenza. Ivumela umshini ukuthi usebenzise ulwazi namakhono ezimeni ezahlukene. Kuze kube manje, abacwaningi be-AI abakwazanga ukuzuza i-AI eqinile. Kuzodingeka bathole indlela yokwenza imishini iqaphele ngokuhlela isethi ephelele yamakhono okuqonda. I-General AI ithole ukutshalwa kwezimali okungu-$1 billion Microsoft inqubo OpenAI.
Fujitsu wakha i K computer, enye yamakhompiyutha amakhulu ashesha kakhulu emhlabeni. Kungomunye wemizamo ebalulekile yokuthola ubuhlakani bokwenziwa obuqinile. Kuthathe cishe imizuzu engama-40 ukulingisa isekhondi elilodwa nje lomsebenzi wemizwa. Ngakho-ke, kunzima ukunquma ukuthi i-AI eqinile izokwazi yini noma nini maduze.
Fujitsu K ComputerTianhe-2 iyi-supercomputer eyakhiwe yi-China National Defence Technology University. Iphethe irekhodi le-cps (izibalo ngomzuzwana) elinama-petaflops angu-33,86 (quadrillion cps). Nakuba kuzwakala kuthakazelisa, kulinganiselwa ukuthi ubuchopho bomuntu buyakwazi ukwenza i-exaflop eyodwa, okungukuthi, ama-cps ayizigidi eziyinkulungwane.
tianhe-2ISuper AI idlula ubuhlakani bomuntu futhi ingenza noma yimuphi umsebenzi kangcono kunomuntu. Umqondo wokuhlakanipha okwenziwayo ubona ubuhlakani bokwenziwa buguqukela ekubeni bufane nemizwa nokuhlangenwe nakho komuntu kangangokuthi kwenza okungaphezu nje kokukuqonda; iphinde ivuse imizwa yomuntu, izidingo, izinkolelo kanye nezifiso. Ukuba khona kwayo kusewukucatshangelwa. Ezinye zezici ezibucayi ze-super AI zifaka ukucabanga, ukuxazulula izindida, ukwenza izahlulelo, nokwenza izinqumo ezizimele.
Manje sizobheka izinhlobo ezahlukene ze-AI esekwe kwisici.
Ukuze uchaze izinhlobo ezihlukahlukene zezinhlelo ze-Artificial Intelligence kuyadingeka ukuze uzihlukanise ngokusekelwe emisebenzini yazo.
Umshini osebenzayo uwuhlobo oluyinhloko lobuhlakani bokwenziwa olungagcini izinkumbulo noma lusebenzise okuhlangenwe nakho kwesikhathi esidlule ukuze kunqunywe izenzo zesikhathi esizayo. Isebenza kuphela ngedatha ekhona. Babona umhlaba futhi basabela kuwo. Imishini esebenzayo inikezwa imisebenzi ethile futhi ayinawo amandla angaphezu kwaleyo misebenzi.
Deep Blue dell 'IBM owahlula ugogo wechess Garry Kasparov kungumshini osebenzayo obona izingcezu ze-chessboard bese usabela kuzo. Deep Blue akakwazi ukubhekisela kunoma yikuphi kokuhlangenwe nakho kwakhe kwangaphambili noma ukuthuthukisa ngokuprakthiza. Ingakwazi ukubona izingcezu ebhodini le-chess futhi yazi ukuthi zihamba kanjani. I-Deep Blue ingenza izibikezelo mayelana nokuthi umnyakazo olandelayo ungaba yini kuye nombangi wakhe. Ziba yonke into ngaphambi kwesikhashana samanje bese ubheka izingcezu ze-chessboard njengoba zikhona kulesi sikhathi bese ukhetha phakathi kokunyakaza okulandelayo okungenzeka.
I-Limited Memory AI iqeqesha kusukela kudatha edlule ukuze yenze izinqumo. Inkumbulo yezinhlelo ezinjalo ihlala isikhathi esifushane. Bangasebenzisa le datha yesikhathi esidlule isikhathi esithile, kodwa abakwazi ukuyengeza kulabhulali yokuhlangenwe nakho kwabo. Lolu hlobo lobuchwepheshe lusetshenziswa ezimotweni ezizishayelayo.
Mitsubishi Electric kade ezama ukuthola indlela yokuthuthukisa lobo buchwepheshe bezinhlelo zokusebenza ezifana nezimoto ezizishayelayo.
Ithiyori yengqondo yobuhlakani bokwenziwa imele isigaba sezobuchwepheshe esithuthukisiwe futhi ikhona kuphela njengomqondo. Lolu hlobo lwe-AI ludinga ukuqonda okujulile ukuthi abantu nezinto ezingaphakathi kwendawo kungashintsha imizwa nokuziphatha. Kufanele iqonde imizwa, imizwa nemicabango yabantu. Nakuba kuye kwenziwa ukuthuthukiswa okuningi kulo mkhakha, lolu hlobo lobuhlakani bokwenziwa alukapheleli ngokuphelele.
Kismet. Kismet iyinhloko yerobhothi eyenziwe ngasekupheleni kwawo-90 ngumcwaningi ovela Massachusetts Institute of Technology. Kismet angakwazi ukulingisa imizwa yabantu futhi ayibone. Womabili la makhono amele intuthuko eyinhloko ku-artificial intelligence theory, kodwa Kismet ayikwazi ukulandela amehlo noma ukudonsa ukunaka kwabantu.
Kismet MITSophia di Hanson Robotics ngesinye isibonelo lapho inkolelo-mbono yobuhlakani bokwenziwa kwengqondo iye yasetshenziswa. Amakhamera asemehlweni kaSophia, ahlanganiswe ne-algorithms yekhompiyutha, amvumela ukuthi abone. Ingagcina ibheke emehlweni, ibone abantu futhi ilandelele ubuso.I-self-awareness AI ikhona ngokucabangela kuphela. Izinhlelo ezinjalo ziqonda izici zazo zangaphakathi, izifunda kanye nezimo futhi zibona imizwa yomuntu. Le mishini izohlakanipha kakhulu kunomqondo womuntu. Lolu hlobo lwe-AI ngeke lukwazi nje kuphela ukuqonda nokuvusa imizwa kulabo oluxhumana nabo, kodwa luzoba nemizwa, izidingo nezinkolelo zalo.
Ucwaningo lobuhlakani bokwenziwa luthuthukise ngempumelelo amasu asebenzayo okuxazulula izinkinga eziningi, kusukela kwezemidlalo kuya ekuxilongweni kwezokwelapha.
Kunamagatsha amaningi obuhlakani bokwenziwa, ngalinye linokugxila kwalo kanye nesethi yamasu. Amanye amagatsha abalulekile wobuhlakani bokwenziwa afaka:
Machine learning: iphathelene nokuthuthukiswa kwama-algorithms akwazi ukufunda kudatha. Ama-algorithms e-ML asetshenziswa ezinhlelweni ezihlukahlukene, okuhlanganisa ukubonwa kwesithombe, ukuhlunga ogaxekile, nokucubungula ulimi lwemvelo.Deep learning: Kuyigatsha lokufunda komshini esebenzisa amanethiwekhi okwenziwa emizwa ukuze athole ulwazi kudatha. I-algorithms ye- deep learning baxazulula ngempumelelo izinkinga ezihlukahlukene, okuhlanganisa i-NLP, ukuqashelwa kwesithombe, nokuqashelwa kwenkulumo.Robotica: iwumkhakha wobunjiniyela obhekene nokuklama, ukwakhiwa nokusebenza kwamarobhothi. Amarobhothi angenza imisebenzi ngokuzenzakalelayo emikhakheni eyahlukene, ehlanganisa ezokukhiqiza, ezokunakekelwa kwezempilo nezokuthutha.I-Generative AI ihlukile kwezinye izinhlobo ze-AI ekhonweni layo lokukhiqiza okuqukethwe okusha nokoqobo, njengezithombe, umbhalo noma umculo, ngokusekelwe kumamodeli afundwe kudatha yokuqeqeshwa, ebonisa ukusungula nokusungula izinto ezintsha.
Amajeneretha obuciko be-AI aqoqa idatha ezithombeni, ezisetshenziswa ukuqeqesha i-AI ngemodeli ye deep learning.
Le phethini ikhomba amaphethini, njengesitayela esihlukile sezinhlobo ezahlukene zobuciko.
I-AI bese isebenzisa lezi zifanekiso ukuze idale izithombe ezihlukile ngokusekelwe ezicelweni zomsebenzisi.
Le nqubo iyaphindaphinda futhi ikhiqiza izithombe eziningi ukuze icwenge futhi izuze umphumela oyifunayo.
Iningi lamajeneretha e-AI linikeza izinguqulo zesilingo samahhala, kodwa futhi kukhona amajeneretha wobuciko we-AI amaningana amahhala atholakalayo.
Ezinye zazo zihlanganisa i-Bing Image Creator, i-Craiyon, i-StarryAI, i-Stablecog, nezinye.
Ijeneretha ye-AI ngayinye inemibandela yayo yokuthengisa umsebenzi wobuciko owenziwe nge-AI kuwebhusayithi yayo.
Nakuba amanye amajeneretha emisebenzi yobuciko engenamingcele ekuthengiseni isithombe njengesakho, njenge-Jasper AI, amanye awakuvumeli ukwenza imali ngomsebenzi wobuciko abawenzayo.
BlogInnovazione.it
I-Coveware ye-Veeam izoqhubeka nokuhlinzeka ngezinsizakalo zokuphendula izigameko zokuntshontshwa kwe-inthanethi. I-Coveware izohlinzeka ngama-forensics kanye nekhono lokulungisa…
Ukulungiswa okuqagelayo kuguqula umkhakha kawoyela negesi, ngendlela emisha nesebenzayo yokuphatha izitshalo.…
I-CMA yase-UK ikhiphe isexwayiso mayelana nokuziphatha kwe-Big Tech emakethe yezobunhloli bokwenziwa. Lapho…
Isinqumo esithi "Case Green", esakhiwe yi-European Union ukuze kuthuthukiswe ukusebenza kahle kwamandla ezakhiwo, siphothule inqubo yaso yomthetho ngokuthi...