Uhlelo lwe-Artificial Intelligence: abadlali, ingqalasizinda kanye namathuba

Ubuhlakani bokwenziwa sebube yimbangela yoguquko lwezobuchwepheshe olungasagcini nje emalabhorethri noma ezinkampanini ezinkulu: lungena emnothweni, ezinkonzweni zomphakathi, ngisho nasezimpilweni zethu zansuku zonke. Ukuze uqonde ngempela umthelela walo, akwanele ukubheka i-algorithm eyodwa noma uhlelo lokusebenza; umuntu kumele aqaphele lonke inethiwekhi yabadlali, ingqalasizinda, imithethonqubo, idatha, kanye nezixazululo ezenza kube nokwenzeka: uhlelo lwe-ecosystem yobuhlakani bokwenziwa.

Lolu hlelo lokusebenza aluyona ipulatifomu eyodwa noma ubuchwepheshe obubodwa, kodwa luyinethiwekhi enamandla yezikhungo, izinkampani, abalawuli, ithalente, kanye nemiphakathi ebambisana futhi encintisana ngasikhathi sinye. Kusukela esu laseYurophu ukwakha ubuholi obunomthwalo wemfanelo ku-AI, kuya kuma-giants amakhulu e-cloud njenge-Google, i-Microsoft, ne-Amazon, kuya kubakhiqizi behadiwe njenge-NVIDIA ne-TSMC, kanye nezinkampani ezintsha eziqalayo ezithuthuke kakhulu, zonke zihlangana ndawonye njengezingcezu zephazili enkulu okudingeka siyiqonde uma sifuna ukusungula izinto ezintsha, ukulawula, ukutshala imali, noma ukugwema ukusala ngemuva.

Isu laseYurophu: uhlelo lwe-AI oluqinile, oluphephile noluhlangene

Ku-European Union, iKhomishini izibekele umgomo wokwakha uhlelo oluqinile noluxhumene lwezobunhloli bokwenziwa oluhambisana nezindinganiso zaseYurophu . Lokhu akugcini nje ngokuthuthukisa ukusungula izinto ezintsha, kodwa kumayelana nokwenza lokho ngokuzibophezela, ngezinhlaka zokuziphatha ezicacile nezomthetho kanye nesisekelo sezimboni nesayensi esincintisanayo emhlabeni wonke.

Ukuze kufezwe lokhu, iBrussels isebenzisa ingqalasizinda ebanzi exhumanisa ithalente, amabhizinisi, idatha, kanye namakhono obuchwepheshe . Umqondo uwukuthi ama-SME, izinhlangano ezinkulu, izazi zemfundo ephakeme, izinhlangano zomphakathi, kanye nezinkampani ezintsha ngokufanayo zizokwazi ukufinyelela izinsiza ze-AI ezithuthukisiwe ngaphandle kokuqala kusukela ekuqaleni isikhathi ngasinye.

Ngaphakathi kwalombono kukhona isu lokuguqula iYurophu ibe “izwekazi le-AI” langempela elinenethiwekhi yezinhlelo ezihlanganisiwe . Le nqubomgomo idlula imibhalo nje: ihumushela ezikhungweni, ezinhlelweni zokuqeqesha, ezindaweni zokuhlola ezilawulayo, izimboni ze-AI, kanye namapulatifomu ahlose ukunciphisa izithiyo zokungena nokusheshisa ukuguquka kusuka ocwaningweni kuya emakethe.

Le ndlela ehlelekile ihlose futhi ukuqinisa ukuzimela kweqhinga laseYurophu, ukuze izwekazi linganciki ngokuphelele kubalingisi bangaphandle ukuze bathole ingqalasizinda ebalulekile, idatha ebucayi, kanye namakhono abalulekile . Ngesikhathi esifanayo, i-EU izibophezele ekubambisaneni kwamazwe ngamazwe kanye nokuvumelanisa amazinga ayo nabanye abaholi bomhlaba, ihlale isekela ukuvikelwa kwamalungelo ayisisekelo kanye nokuphepha.

Izingxenye ezibalulekile ze-ecosystem ye-AI yaseYurophu

Enye yezinsika zesu laseYurophu yi- European Digital Innovation Hubs (EDIH) , eklanyelwe njengendawo eyodwa yamabhizinisi, ikakhulukazi ama-SME, ukufinyelela izinsizakalo zokuguqulwa kwedijithali ngokugxila kakhulu ku-AI. Lezi zikhungo zisiza izinkampani ukuthi zihlole ubuchwepheshe, zithole uxhaso, zixhumane nochwepheshe, futhi zinciphise ukwesaba ukuthatha isinyathelo.

Kanye nama-EDIH, iZindawo Zokuhlola Nokuhlola (ama-TEF) ziyagqama . Lezi yizindawo ezilawulwayo, ezihlonyiswe ngengqalasizinda esezingeni eliphezulu, lapho izixazululo ze-AI ezithuthukisiwe zingahlolwa ngaphansi kwezimo ezicishe zibe ngokoqobo: isibonelo, izinhlelo zezimoto, ezempilo, imboni, noma amadolobha ahlakaniphile, ukunciphisa izingozi kanye nokusheshisa ukuqinisekiswa kobuchwepheshe.

Esinye isici esibalulekile yilokho okubizwa ngokuthi “amafektri e-AI ,” izinhlelo ezisebenzisa amandla amakhulu e-supercomputing aseYurophu ukuletha i-AI kusuka elabhorethri iye ezimweni zokusetshenziswa zangempela. Umgomo wazo ukwenza kube lula ukuqeqeshwa kwamamodeli ayinkimbinkimbi, ukuthuthukiswa kwama-prototypes, kanye nokuhlolwa ngezinhlelo zokusebenza ezidinga amandla okusebenzisa ikhompyutha izinhlangano ezimbalwa ezingazikhokhela zodwa.

I-EU iphinde ikhuthaze “ama-sandbox e-AI,” noma izindawo zokuhlola ezilawulwayo , lapho izinkampani nezikhungo zingazama khona ubuchwepheshe obusha ngaphansi kohlaka lokulawula oluqondiswayo noluguquguqukayo. Lokhu kuvumela izixazululo zokuhlola emikhakheni elawulwayo (njengokunakekelwa kwempilo, ukuhamba, noma ezezimali) ngaphandle kokuthatha zonke izingozi zomthetho zokuthunyelwa kwabantu abaningi kusukela ekuqaleni.

Ukuze kuhlelwe lolu hlaka lonke, iKhomishini iqalise ipulatifomu ye-AI yaseYurophu "efunwayo" , esebenza njengendawo yokufinyelela ehlanganisiwe ezinsizeni, amathuluzi, idatha, imitapo yolwazi, izimo zokusetshenziswa, kanye nemikhuba emihle kakhulu. Umqondo uwukuthi abacwaningi, abathuthukisi, kanye nezinkampani babe nekhathalogi efanayo lapho bengathola khona izinsiza ezikhona, basebenzise kabusha izingxenye, futhi basebenzisane kalula.

Okwamanje, iYurophu ilungiselela i -AI Skills Academy , isikhungo samakhono se-AI esihlose ukuvumelanisa abasebenzi neqiniso elisha lobuchwepheshe. Umgomo wayo ukuthuthukisa izinhlelo zokuqeqeshwa ezithuthukisiwe kanye nokuthuthukiswa kobungcweti ukuze ochwepheshe bezobuchwepheshe kanye nabangebona abezobuchwepheshe bakwazi ukusebenzisana ne-AI ngempumelelo nangendlela yokuziphatha.

Isu elithi "Sebenzisa i-AI" lihambisana nakho konke okungenhla ngezinyathelo zokunciphisa isikhathi esithathwayo ukuze into entsha ifinyelele emakethe ngokuxhumanisa ingqalasizinda, idatha, kanye nezindawo zokuhlola. Lihlanganisa nezenzo zokuqinisa ichibi lamathalente elitholakala e-EU, ukukhuthaza isinyathelo "se-AI esemngceleni" esihlanganisa abadlali ababalulekile baseYurophu, kanye nokuguqula ama-EDIH abe yizindawo zobuchwepheshe be-AI zangempela ezisebenza njengezindawo zokubhekisela ezinkampanini nasezinhlanganweni.

Ngaphakathi kwalokho okubizwa ngokuthi i-Digital Omnibus, iKhomishini iphakamise ukwandisa izinyathelo zokuthobela imithetho ukuze abasunguli abaningi bakwazi ukusebenzisa izindawo zokuhlola ezineziqinisekiso , okuhlanganisa ne-sandbox ye-AI ebanzi ye-EU kusukela ngo-2028 kanye nokusetshenziswa kokuhlolwa okwengeziwe komhlaba wangempela ezimbonini ezibalulekile, njengezimoto.

Lonke lolu hlelo lwezinhlelo, izikhungo, amapulatifomu kanye nezinhlaka zomthetho ludala uhlelo lwe-AI lwaseYurophu oluguquguqukayo, oluxhumene nolugxile ebuholini olufuna ukuncintisana ngenkathi luvikela amalungelo futhi ludala ukwethembana komphakathi kulezi zobuchwepheshe.

Amapulatifomu kanye nezinhlelo zemvelo: izinhlangothi ezimbili zesu ledijithali

Eminyakeni yamuva nje, ukugxila kokusungula izinto ezintsha kuye kwashintshela kumapulatifomu edijithali kanye nezinhlelo zebhizinisi , lapho izinkampani ezinkulu zobuchwepheshe zitshala imali eningi kwingqalasizinda yamafu, ukuhweba nge-inthanethi, ukukhangisa kwedijithali, izitolo zezinhlelo zokusebenza, kanye namadivayisi axhunyiwe.

Amapulatifomu, aqondwa ngokuthi "amanethiwekhi emithetho kanye nengqalasizinda elekelela ukusebenzisana phakathi kwabasebenzisi benethiwekhi ," yindawo lapho idatha ikhiqizwa futhi ihlelwe khona, izinsizakalo zinikezwa, futhi ababambiqhaza abaningi bavunyelwe ukusebenzisana ngaphansi kwemithetho ehlanganyelwe. Empeleni, ayisakhiwo sobuchwepheshe kanye nesokulawula esisekela ukushintshaniswa kwenani.

Izinhlelo zemvelo, ngokwazo, zimele umphakathi ophilayo wabadlali abasebenza kulezi zinkundla : izinkampani, abaphakeli, abathuthukisi, izikhungo, izinkampani ezintsha, imiphakathi yabasebenzisi, njll. Ezingeni eliphakeme, zingachazwa njengemiphakathi eshintshashintshayo edala futhi ibambe inani ngokubambisana okuqhubekayo kanye nokuncintisana.

Esinye sezihluthulelo zempumelelo yamapulatifomu amakhulu kube yikhono lawo lokuhlela, ukukhulisa, nokuxhumanisa imvelo ebanzi . I-Alibaba, isibonelo, ikhuluma “ngobuhlakani bedatha” obuhlanganiswe “nokuxhumanisa inethiwekhi” ukuchaza ukuthi ixhumanisa kanjani ulwazi, izinsizakalo, kanye nabahlanganyeli, okukhuthaza umjikelezo omuhle lapho umsebenzi omningi ukhiqiza khona idatha eyengeziwe, okubuye kuthuthukise i-AI kanye nesiphakamiso senani.

Kulesi simo, i-AI isibe yingxenye ebalulekile: amapulatifomu awagcini nje ngokuhlanganisa ama-algorithms, kodwa futhi aziklama kabusha ukuze ahlangabezane nezidingo ze-AI , ikakhulukazi isidingo sedatha esezingeni eliphezulu nehlelekile evela emithonjeni eminingi.

Izinhlobo zezinhlelo zemvelo emnothweni wedijithali oqhutshwa yi-AI

Ngaphakathi kwendawo yamanje, izindlela eziningana zemvelo zingatholakala okufanele zihlukaniswe ukuze kuqondwe kangcono ukuthi amaphrojekthi kanye nezivumelwano ezizungeze i-AI zakhiwe kanjani.

Ngakolunye uhlangothi, kunezinhlelo zemvelo "ezinkulu" , ezivela ekuhlanganeni kwemikhakha ehlukene yendabuko ehlangana ndawonye ukudala iziphakamiso ezintsha zenani. Isibonelo esicacile ukubambisana phakathi kwezikhungo zikahulumeni, abaqhubi bezokuthutha, izinkampani zobuchwepheshe, kanye nabahlinzeki bamandla ukuthuthukisa amaphrojekthi edolobha ahlakaniphile noma ukwenza lula ukuhamba phakathi kwezindawo ngaphandle kokuphazamiseka.

Ngakolunye uhlangothi lwe-spectrum kukhona "ama-microecosystems " agxile kakhulu, adalwe ngezixazululo ezithile kakhulu zezinkinga ezithile. Isibonelo salokhu kungaba iphrojekthi yokushayela ezimele eyakhiwe ngumkhiqizi othile ngokubambisana nenqwaba yezinzwa ezikhethekile, isofthiwe, kanye nabahlinzeki bezinsizakalo.

Uhlobo lwesithathu yizinhlelo zemvelo ezintsha nokufunda , ezixhumanisa abadlali abagxile esikhathini esizayo (amabhizinisi amasha, izikhungo zocwaningo, izimali zokutshalwa kwezimali, imiphakathi evulekile) ukuze kusheshiswe ukuhlola. Lokhu kubonakala njengezisheshisi, izifutho, imiphakathi yezobuchwepheshe, noma izikhungo ezifana nalezo ezinikezelwe ku-AI kanye ne-blockchain, noma izinhlelo ezifana ne-StartMeUp, i-TED, i-MIT Media Lab, noma i-Singularity University.

Okokugcina, sithola izindawo zemvelo "ezinomphakathi" , ezihlelwe ukubhekana nezinselele ezihlangene ezidinga ukubambisana okuvela emikhakheni eminingi: uhulumeni, amabhizinisi, ama-NGO, izazi zemfundo ephakeme, kanye nezakhamuzi. Isibonelo esihle yi-Partnership on AI, ehlanganisa izinhlangano eziningi ukuze zichaze izindlela ezinhle kakhulu ezinhlelweni ze-AI futhi zikhuthaze imfundo yomphakathi mayelana nomthelela wazo.

Ubuhlakani bokwenziwa busetshenziswa kumapulatifomu abalulekile kanye nemikhakha

Inani lamapulatifomu edijithali lincike kakhulu ekhonweni lawo lokuhlanganisa, ukubhekisela, kanye nokuhlaziya idatha ekhiqizwe ukusebenzisana akusizayo: ukuthengwa kwe-e-commerce, ukuxhumana kwenethiwekhi yokuxhumana nabantu, uhambo kumapulatifomu okuhamba, ukushintshana kumapulatifomu ezemfundo, njll.

Lokhu kuqongelela kanye nokusetshenziswa kwedatha ngobuhlakani kube yimbangela yokwelula i-AI ezindaweni ezintsha kanye nokuthuthukisa ikhwalithi yamamodeli . Uma idatha efanele igeleza ngepulatifomu, kuba lula ukuqeqesha ama-algorithms anembile futhi aqinile, kuthuthukiswe ukuzenzekela kanye nokwenza izinqumo.

Ngokwesibonelo, kwezentengiso, iziqhwaga ezifana ne-Alibaba ne-Amazon seziqala ukuhlanganisa ukubonwa kobuso nokubonwayo nedatha yokuthenga eku-inthanethi nangaphakathi esitolo ukuze ziqonde kangcono ukuziphatha komthengi ngamunye. Nakuba eziningi zalezi zobuchwepheshe ezibonwayo zisakhula, umgomo uwukwenza kube ngokwakho okunikezwayo, ukwenza ngcono ulwazi olutholakala esitolo, kanye nokuthuthukisa isitokwe kanye nemikhankaso.

Enye indawo echumayo abasizi ababonakalayo abagxile kubathengi . Amaphrojekthi e-Amazon (Alexa), i-Google (Umsizi), i-Microsoft (Cortana), i-Apple (Siri), kanye ne-Meta atshale imali eningi kubasizi abaqonda ulimi lwemvelo, abenza imisebenzi, futhi abasebenza njengesixhumi esibonakalayo sengxoxo namasevisi namadivayisi. Ngemuva kwawo kunezinhlelo zemvelo zabathuthukisi, abakhiqizi bamadivayisi, kanye nabahlinzeki bokuqukethwe.

Emkhakheni wedolobha elihlakaniphile, ohulumeni basekhaya abaningi bakhetha amapulatifomu ajwayelekile, avulekile asekelwe endleleni "yohlelo lwezinhlelo" . Esikhundleni sokuthembela epulatifomu eyodwa eyimfihlo, bafuna ukuhlanganisa izingxenye ezahlukahlukene ezingasebenzisana, ezizokwakha uhlelo oluguquguqukayo lokuphatha ukuhamba, amandla, ukuphepha, imvelo, kanye nezinsizakalo zasemadolobheni.

I-Cloud computing ingenye insika eyisisekelo. Ngaphandle kwebhizinisi lokubamba nokugcina idatha, ifu lidlulela ekufundeni komshini njengesevisi , i-Intanethi Yezinto, ukuphepha kwe-inthanethi okuthuthukisiwe, kanye nezixazululo ze-blockchain . Zonke lezi zingxenye zikhuthaza ukuthuthukiswa kwama-algorithms e-AI futhi zandisa amakhono atholakala ezinkampanini ezingenazo ingqalasizinda yazo.

Ezinkonzweni zezezimali, ukwenza izinqumo zesikweletu kanye nezingozi kuya ngokuya kuzenzakalelayo ngamasethi edatha ahlukahlukene avela emithonjeni eminingi . I-Ant Financial (inkampani engaphansi kwe-Alibaba), isibonelo, ingacubungula izimali mboleko ezincane zama-SME ngemizuzu embalwa isebenzisa idatha yayo yesikhulumi kanye nolwazi olugunyazwe abathengisi ngokwabo.

Ama-giants obuchwepheshe: abaklami be-ecosystem ye-AI yomhlaba wonke

Uma sibheka isithombe esikhulu, iziqhwaga ezifana ne-Google, i-Microsoft, ne-Amazon zisebenza njengabakhi beqiniso be-"dolobha ledijithali" elisekelwe ku-AI . Ngayinye yakha ingqalasizinda ehlukene, amathuluzi, nezinsizakalo lapho abanye bangakha khona izinhlelo zokusebenza namamodeli ebhizinisi.

Isibonelo, i-Google ithuthukise i-TensorFlow, uhlaka oluvulekile lokwakha nokuqeqesha amamodeli e-AI manje asetshenziswa yiwo wonke umuntu kusukela ezinkampanini ezintsha kuya ezinkampanini ezinkulu. Ngokusekelwe kuloluhlaka, kuye kwadalwa abasizi bezwi abanjengo-Google Assistant, izinhlelo zokusekela izithombe zezokwelapha, izinjini zokuncoma, kanye nezinye izixazululo eziningi eziqondene nomkhakha.

I-Microsoft, nge-Azure, inikeza uhlelo lwe-cloud lapho amabhizinisi angathola khona izinsizakalo zedatha, ukuhlaziya, i-AI, ukuphepha, kanye nentuthuko . Kufana ne-workshop enkulu enamathuluzi akhethekile emisebenzi eminingi: ukucubungula ulwazi oluningi, ukuthola amaphethini, ukusebenzisa amamodeli e-AI, nokuwaxhumanisa nezinhlelo zokusebenza zebhizinisi.

I-Amazon, ngakolunye uhlangothi, iguqule i-AWS yaba yipulatifomu evumela ukufinyelela kumakhono e-AI athuthukile kuzo zonke izinhlobo zezinhlangano . Kusukela ekuboneni kwamakhompyutha kanye nezinsizakalo zokucubungula ulimi lwemvelo kuya kumathuluzi okufunda omshini aphethwe, umgomo wayo ukuvumela ngisho namabhizinisi amancane naphakathi nendawo (ama-SME) ukuthi ahlanganise i-AI emisebenzini yawo ngaphandle kokwakha konke kusukela ekuqaleni.

Lezi ziqhwaga azigcini nje ngokwethula imikhiqizo; zichaza amazinga obuchwepheshe, imikhuba emihle, kanye nezinhlaka ezithinta yonke imboni. Izinqumo zazo mayelana nezindawo zokusebenzelana, ama-API, amamodeli okuphatha idatha, kanye nezinqubomgomo zokuphepha ekugcineni zibeka ijubane kanye nesiqondiso sengxenye enkulu yesimiso somhlaba.

Ihadiwe, inkumbulo kanye ne-supercomputing: isisekelo esingokoqobo se-ecosystem

Ngemuva kwesofthiwe kanye nezinsizakalo zamafu kulele ungqimba olubalulekile: abahlinzeki behadiwe kanye nememori abenza i-AI enkulu ibe nokwenzeka . Ngaphandle kwabo, lonke isakhiwo besingawa.

I-NVIDIA izibonakalise njengomdlali obalulekile ngenxa yama-GPU ayo (amayunithi okucubungula ihluzo) alungiselelwe imisebenzi ye-AI . Singawacabanga njengenjini yemoto yezemidlalo uma kuqhathaniswa nenjini evamile: womabili akuhambisa lapho udinga ukuya khona, kodwa imoto yezemidlalo ifika lapho ngokushesha okukhulu. Kumamodeli okuqeqesha adinga ukucubungula izigidi zezithombe noma amasethi amakhulu ombhalo, lawa ma-GPU cishe awabalulekile.

I-TSMC, enye yabakhiqizi be-semiconductor abakhulu emhlabeni, ikhiqiza ama-chip asebenza "njengobuchopho" bamadivayisi namaseva . Kusukela kumaselula kuya kuma-supercomputer, ama-processor amaningi nama-accelerator asetshenziswa ku-AI adlula ezimbonini zayo, okwenza inkampani ibe yisixhumanisi esibalulekile kulolu chungechunge.

Izinkampani ezifana neMicron neHynix zigxile ekugcinweni nasekugcinweni kwememori okusebenzayo kakhulu , okubalulekile ukuze izinhlelo ze-AI ziphathe futhi zikhumbule ulwazi oluningi. Ngaphandle kwememori esheshayo nethembekile, ukucubungula kuyehla futhi amamodeli alahlekelwa ukusebenza kahle.

Inkumbulo ohlelweni lwe-AI isebenza njengesiphambano phakathi kwetafula lomsebenzi nomtapo wolwazi omkhulu . Ngakolunye uhlangothi, kunenkumbulo esheshayo kakhulu kodwa elinganiselwe, lapho idatha okudingeka imodeli ibe nayo itholakale kalula nganoma yisiphi isikhathi igcinwa khona; ngakolunye uhlangothi, kunesitoreji sesikhathi eside esinamandla amakhulu okugcina idatha yomlando, amamodeli, namarekhodi.

Ukusebenza kwesisombululo se-AI kuncike kakhulu endleleni iprosesa, i-GPU, inkumbulo esheshayo, kanye nesitoreji esikhulu esisebenzisana ngayo . Ukwakhiwa okuhle kwehadiwe kuvumela ukucutshungulwa kwedatha ngesikhathi sangempela kwemisebenzi efana nokubikezela imakethe, ukuhlaziywa kokukhwabanisa, noma ukuxilongwa kwezokwelapha ngokushesha, kuyilapho kugcinwa ukubambezeleka kanye nezindleko ngaphansi kokulawula.

Ukutshala imali ohlelweni lwe-AI: kusukela engqalasizinda kuya kuzicelo

Iketanga lenani lobuhlakani bokwenziwa linikeza izindawo eziningi zokungena zokutshalwa kwezimali , ezihlukahluka kuye ngesu lomtshali-zimali ngamunye, isifiso sokufaka engcupheni, kanye nesikhathi esinqunyiwe. Ukuqonda leli "ketanga lokukhiqiza" kusiza ekuboneni lapho kukhona khona amathuba amaningi nokuthi imakethe ingase igcwale khona kakade.

Esigabeni samanje, ingxenye enkulu yemali iqondiswa engqalasizinda eyinhloko: izikhungo zedatha, ama-chip, amanethiwekhi asheshayo, kanye namapulatifomu amafu . Izinkampani ezifana ne-Google, i-Microsoft, i-Amazon, i-Meta, kanye ne-Tesla zithembele kakhulu kuma-GPU asebenza kahle kanye nezinye izisheshisi kokubili ukuqeqesha amamodeli amakhulu kanye nokuwasebenzisa ekukhiqizeni.

Okulindelekile ukuthi, ngokuhamba kwesikhathi, ukutshalwa kwezimali kuzohlukahluka kuye ekusetshenzisweni nasezingqimbeni eziphakeme ze-stack . Njengoba ingqalasizinda iba yinto ejwayelekile futhi efinyeleleka kalula, ukugxila kuzoshintshela ezixazululweni eziqondile zezempilo, imfundo, ezezimali, imboni, ezokuthutha, ubuciko, kanye nokuzijabulisa, phakathi kweminye imikhakha.

I-AI isesigabeni sokukhula okukhulu kakhulu ngenani elikhulu lezicelo . Kulindeleke ukuthi iguqule ngokuphelele izimboni zonke: kusukela ekutholakaleni kwezidakamizwa nasekuxilongweni okusizwayo kuya emfundweni yomuntu siqu, ukuzenzekela kwezezimali, kanye nokudalwa kwezithombe nezethulo ezisebenzisa i-AI . Uhlelo ngalunye olusha, nalo, lukhulisa isidingo sesofthiwe eyengeziwe, ihadiwe eyengeziwe, kanye nezinsizakalo ezengeziwe.

Kulabo abahlaziya izitayela zomhlaba wonke ngombono ophelele (wezomnotho, wezenhlalo, nowezobuchwepheshe), uhlelo lwe-AI luvela njengomshayeli wesikhathi eside woguquko , hhayi imfashini edlulayo. Lokhu akusho ukuthi noma yini iyahamba, kodwa kunalokho kufanele kwenziwe umehluko oqaphile phakathi kwezendlalelo, amamodeli ebhizinisi, kanye nezingozi ezihlobene.

I-ecosystems ye-ejenti ye-AI: isibonelo se-Google Cloud

Ngaphakathi kwalesi simo esibanzi, kuvela izimiso ezithile zemvelo ezizungeze izinhlangano ezintsha ezifana nama -AI agents . I-Google Cloud, isibonelo, iqalise uhlelo lwezimiso zemvelo zama-AI agents oluklanyelwe ukusheshisa ukuthuthukiswa kwawo, ukuthunyelwa kwawo, kanye nokwamukelwa ngamakhasimende emhlabeni jikelele.

Ama-ejenti e-AI ayizinhlelo zokusebenza eziqhutshwa amamodeli obuhlakani bokwenziwa akwazi ukuqonda izicelo ngolimi lwemvelo futhi aphendule ngokuhambisanayo nangokusiza. Angenza imisebenzi ibe ngokuzenzakalela, enze ukusebenzisana kwabasebenzisi kube ngokwakho, futhi athuthukise ukusebenza kahle kwenqubo emikhakheni eyahlukahlukene njengenkonzo yamakhasimende, izinsiza zabantu, ukuthengisa, ukwesekwa kobuchwepheshe, kanye ne-logistics.

I-Google Cloud idale isigaba esizinikele "sama-AI Agents" ngaphakathi kweMakethe yayo , lapho ozakwethu bangashicilela khona izixazululo ezilungele ukusetshenziswa, futhi amakhasimende angazithola futhi azisebenzise kalula. Lokhu kunciphisa ukungqubuzana kwebhizinisi nobuchwepheshe futhi kukhuthaza uhlelo olusebenzayo lwabahlinzeki abakhethekile.

Ozakwethu abajoyina uhlelo bathola izinzuzo ezenzelwe ukunciphisa isikhathi sokumaketha nokwandisa ukubonakala . Lokhu kufaka phakathi ukufinyelela ngqo kumaqembu emikhiqizo nobunjiniyela be-Google, isiqondiso sobuchwepheshe, ukuqeqeshwa okuthuthukisiwe, ukufinyelela kusenesikhathi kubuchwepheshe obusha, izinhlelo zokuthengisa ngokubambisana, kanye nokubamba iqhaza ezinhlelweni zokumaketha eziqondiwe kanye nemicimbi ekhethekile.

Ukuqinisekisa izinga eliphansi lekhwalithi kanye nokuhambisana kwezobuchwepheshe, uhlelo lubeka izidingo eziqondile zama-ejenti e-AI : kumele axazulule umgomo othize (ngokuvamile ngamathuluzi, ukucabanga, noma amakhono okuhlela), asebenzise amamodeli e-Gemini noma amanye amamodeli atholakala ku-Model Garden futhi athembele kuphrothokholi yomongo wemodeli ye-MCP , asetshenziswe engqalasizinda ye-Google Cloud futhi athembele kumasevisi e-Vertex AI, noma okungenani abe nohlelo lokuthuthela kuwo ngale kokusetshenziswa okuyisisekelo kwe-LLM.

Ukujoyina uhlelo kungenziwa nge- Google Cloud Marketplace noma ngokuxhumana namaqembu abalingani , kanye nokuthumela ifomu lesicelo elithile. Lokhu kukhulisa umphakathi wama-ISV (abathengisi besofthiwe abazimele), abahlanganisi bezinhlelo, kanye nabathuthukisi abaqhubeka nokwandisa inani labalingani abatholakalayo.

Ukuziphatha, ukuphepha, kanye nezinhlaka zomthetho ohlelweni lwe-AI

Enye insika yesimiso sobuhlakani bokwenziwa umsebenzi wezindinganiso zokuziphatha, ukuphepha, kanye nokuhambisana nemithetho . Akwanele ukuthi ubuchwepheshe busebenze nje; kumele bube nokwethenjelwa, bube nobulungiswa, buchazeke, futhi buhloniphe amalungelo ayisisekelo.

Izinhlangano ezikhethekile kanye nezinhlangano zezimboni zisebenzisana nezinhlangano ezifana ne -Cloud Security Alliance, i-CoSAI, i-MITRE ATLAS noma i-OWASP ukuthuthukisa iziqondiso, izinhlaka zosongo, imikhuba emihle yokuphepha kanye nezindlela zokuhlola ubungozi ezinhlelweni ze-AI.

Lokhu kubambisana kuvumela izixazululo ezisezingeni eliphezulu ukuthi zakhiwe kusukela ekuqaleni ngokuvumelana nezinhlaka zomthetho kanye nezimiso zokuziphatha ezisafufusa , kunokuzama "ukulungisa" izinkinga ngemva kweqiniso. Ngakho-ke ukuphepha kulo lonke uchungechunge (idatha, amamodeli, ingqalasizinda, kanye nezinhlelo zokusebenza) kuba yimfuneko ebalulekile, hhayi isengezo esingazikhethela.

Ngesikhathi esifanayo, abadlali abaningi basebenza ezindaweni zemvelo zasendaweni ezigxile endaweni , njenge-AI Lab Granada noma i-aiMPULSA ecosystem, exhumanisa amanyuvesi, izinhlangano zikahulumeni, kanye namabhizinisi azungeze i-AI. La maphrojekthi asiza ukuhumusha izimiso zokuziphatha ezibanzi kanye nemithetho zibe yimikhuba eqondile evumelaniswe nezimo zasendaweni.

Umphumela uba yindawo lapho inqubomgomo yomphakathi, amazinga obuchwepheshe, kanye nezixazululo zezentengiselwano zithintana, zakha uhlelo lwe-AI olungafuni nje kuphela ukusebenza kahle nokusebenza kahle, kodwa futhi nokwethembana, ukucaca, kanye nokuzinza komphakathi.

Isu kanye nokuncintisana: indlela izimiso zemvelo ezibumba ngayo imakethe

Ngokombono webhizinisi, kuyasiza ukubuka amapulatifomu kanye nezinhlelo zemvelo njengezindlela ezimbili ezihlukile kodwa ezihlobene eduze zesu . Amapulatifomu ahlinzeka ngesisekelo sobuchwepheshe kanye nemithetho yomdlalo; izinhlelo zemvelo zinikela ngokuhlukahluka kwabadlali, ukusungula izinto ezintsha, kanye nekhono lokukhulisa.

Uhlelo oluphethwe kahle lungathuthukisa kakhulu inani leplatifomu ngokungeza amakhono amasha, idatha eyengeziwe, kanye nesivinini sokuthuthuka okuqhubekayo okuqhutshwa abahlanganyeli abaningi. Ngokulandelayo, ubuholi beplatifomu busebenza njengesisusa, buheha abalingani abaningi futhi buqinise isikhundla okunzima ukusiphonsela inselelo.

Ezimweni eziningi, izimiso zemvelo kanye nemifelandawonye iba yizindlela eziyisisekelo zokuzivikela ngokumelene nezikhulu ezinkulu . Njengoba bebhekene nokusunduza kwe-Amazon ekuthengiseni, isibonelo, abancintisana nabo abanjengoWalmart baphoqelekile ukuthuthukisa izimiso zemvelo ezincintisanayo ezihlanganisa abahlinzeki, izinkampani ezintsha zobuchwepheshe, kanye nabalingani bezokuthutha ukuze baphonsele inselelo ubukhulu bayo kanye nempahla yedatha.

Abadlali abakhulu bezobuchwepheshe nabo abamile: bavame ukunwebeka ngaphandle kokungabaza emigqeni yebhizinisi eseduze ukuze bavikele ukufinyelela kwedatha eyengeziwe futhi baqinise amakhono abo e-AI. Ukuthengwa kwe-Amazon kwe-Whole Foods kanye nesithakazelo se-Google kumadivayisi ahlakaniphile asekhaya kubonisa leli qhinga lokwehlukahluka eliqhutshwa ukusesha imithombo emisha yolwazi.

Konke lokhu kuqinisekisa ukuthi i-AI akuyona nje enye ubuchwepheshe, kodwa isici esibalulekile esithonya ukunyakaza kwezinkampani, ukuhlanganiswa, ukuthengwa, kanye nemifelandawonye . Noma ngubani olawula idatha, ingqalasizinda, kanye nohlelo lonjiniyela unesikhundla esikhethekile sokuhola imakethe.

Uhlelo olusafufusa lobuhlakani bokwenziwa, olusuka eBrussels luye eSilicon Valley kanye nezindawo eziyinhloko, seluvele luyithonya elikhulu ekwakheni umnotho wedijithali. Ukuqonda ukuthi izingxenye zalo—izinqubomgomo zomphakathi, amapulatifomu amafu, ihadiwe, amazinga okuziphatha, amanxusa e-AI, kanye nemiphakathi yokusungula izinto ezintsha—zihlangana kanjani kuyisihluthulelo sokwenza izinqumo ezinengqondo, kungakhathaliseki ukuthi zemithethonqubo, ukutshalwa kwezimali, ukuba ngusomabhizinisi, noma nje ukuguqula inhlangano ibe nekusasa lapho i-AI izoyeka ukuba yinto entsha futhi ibe yingqalasizinda eyisisekelo, engabonakali futhi ebalulekile njengogesi noma i-inthanethi.