- Ukwethulwa kwezikhangiso ku-ChatGPT kuphendula isidingo sokuxhasa izindleko ezikhulayo, ngamamodeli e-CPM kanye ne-CPC asezingeni eliphezulu asekelwe ekuhlosweni okuphezulu kwabasebenzisi.
- I-OpenAI ithembisa izikhangiso ezihlukene nezinemalebula, kodwa izingozi zokubandlulula kwezentengiselwano, ukulahlekelwa ukungathathi hlangothi, kanye nemiphumela engaba khona ekuhlobaneni kwezimpendulo ziyaqhubeka.
- Abancintisana nabo njengo-Anthropic basebenzisa ukwenqatshwa kokukhangisa njengebhanela labo, kuyilapho i-Apple isebenza njengesihlungi ngokuhlanganisa i-ChatGPT ne-Gemini ngaphandle kokubonisa izikhangiso ku-interface yayo.
- Ukuthembela ekunembileni kwe-ChatGPT kuzoncika ekubonakaleni obala, ukuhlolwa kwangaphandle, kanye nekhono labasebenzisi lokuqinisekisa ulwazi olubalulekile.
Ukufika kokukhangisa ku-ChatGPT kubangele impikiswano enkulu mayelana nokuthembeka, ukungathathi hlangothi, kanye nemodeli yebhizinisi yama-chatbot e-AI. Lokho okwakubhekwa njengethuluzi "elingathathi hlangothi" ngaphambilini sekuqala ukufana, kwabaningi, nenjini yokusesha yakudala enezisusa zokukhangisa ezicacile ngemuva kwezincomo zonke.
Ngesikhathi esifanayo, kwi-OpenAI nakwezinye izinkampani kulo mkhakha, ukubonisa izikhangiso akuyona nje into eyenzeka ngenkani: kuyindlela eqondile yokuxhasa izindleko ezinkulu zekhompyutha nokwenza isevisi esetshenziswa yizigidi zabantu nsuku zonke. Phakathi kwabasebenzisi abaqaphile, izinhlobo ezilindelekile, kanye nabancintisana abamele noma "abamelene" nezikhangiso, umbuzo omkhulu usobala: konke lokhu kuthinta kanjani ukunemba kwe-ChatGPT lapho kuhilelekile ukukhangisa?
Okushintshile: kusukela ekubeni umsizi "ohlanzekile" kuya engxoxweni enezikhangiso ezihlanganisiwe
Okwesikhathi esithile, isimo esisemthethweni se-OpenAI kwakuwukugwema ukukhangisa okuqondile ngaphakathi kwe-ChatGPT. USam Altman waze wakuveza njengomugqa obomvu ukuze agweme ukungcolisa ulwazi lomsebenzisi . Kodwa isimo sezomnotho siye saqhubekela phambili ngenye indlela: ukugcina i-AI esetshenziswa kakhulu emhlabeni kuhilela ukutshalwa kwezimali okuyizigidi zamaRandi kumaseva, ama-chip, kanye namandla, futhi inkampani ekugcineni isiweqe lowo mkhawulo.
Namuhla, izikhangiso sezivele zisebenza kubasebenzisi abadala ababhalisiwe e-United States , kokubili kunguqulo yamahhala kanye nakumapulani aphakathi njenge-ChatGPT Go. Isihluthulelo ukuthi zihlanganiswa kanjani: azizona izibhengezo zendabuko noma amabhlogo esitayela sokusesha akudala, kodwa imiyalezo exhasiwe efakwe ekugelezeni kwengxoxo futhi ihambisana nalokho umsebenzisi akubuzayo ngaleso sikhathi.
I-OpenAI igcizelela ukuthi lezi zikhangiso zibhalwe ngokucacile njengokuqukethwe "Okuxhasiwe" futhi zibekwe ngaphansi kwempendulo ekhiqizwe yimodeli, hhayi ngaphakathi kombhalo wempendulo uqobo. Ngamanye amazwi, isithembiso esisemthethweni ukuthi imodeli iphendula ngokwemvelo kuqala, bese kuphela lapho isikhangiso siboniswa khona, ngefomethi ebonakalayo ehlukaniswe ngokucacile.
Ngaphezu kwalokho, inkampani igcizelela ukuthi izikhangiso ngeke ziphazanyiswe ezingxoxweni eziphathelene nezihloko ezibucayi njengempilo ejwayelekile, impilo yengqondo, noma ezombusazwe, nokuthi ngeke zinikezwe noma ubani ongaphansi kweminyaka engu-18. Ngokwethiyori, izikhangiso zisekelwe kumongo (ezisekelwe engxoxweni) futhi zingasebenzisa umlando wengxoxo kanye nokuxhumana kwangaphambilini nokunye okuqukethwe okuxhasiwe, kodwa kunezinketho zokunciphisa ukwenza kube ngokwakho noma ukususa imemori yezikhangiso.
Amamodeli ebhizinisi okukhangisa: kusukela ku-CPM ye-premium kuya ku-CPC kanye nokuqondiswa kwengxoxo
Isigaba sokuqala sebhizinisi lokukhangisa le-OpenAI sithembele kumodeli yakudala yezindleko-nge-nkulungwane-yemibono (CPM) enendlela ephezulu kakhulu: ukutshalwa kwezimali okuphansi kakhulu kanye namanani angaphezulu kwemakethe ngombono ngamunye . I-CPM yokuqala yayicishe ibe ngu-$60, kanti imali yokungena yayicishe ibe ngu-$200.000-$250.000, ihlose kuphela imikhiqizo emikhulu esigabeni sokuhlola esilawulwayo.
Ngokuhamba kwesikhathi, leso sithiyo sokungenela sehlisiwe. Utshalomali oluncane selwehle lwaya cishe ku-$50.000 kanti i-CPM yehle yaya cishe ku-$25. Lesi sinyathelo sibonisa okuthile okucacile: I-OpenAI idinga ukwandisa ngokushesha isisekelo sayo sabakhangisi futhi yandise inani lemikhankaso esebenzayo, ngenkathi igcina ukulandisa kwayo kwempahla ephezulu.
Nokho, ushintsho olubaluleke kakhulu ushintsho oluvela kumodeli esekelwe ekucabangeni kuphela kuya kumodeli yezindleko ngokuchofoza (CPC) engaba ngu-$3-5 ngokuchofoza ngakunye . Lolu hlu lubeka i-ChatGPT ngqo eqenjini elifanayo lezifiso njenge-Google Search yezigaba ezithile ezinezinhloso eziphezulu: umshwalense, abameli abakhethekile, izinsizakalo zezimali eziphuthumayo, njalo njalo.
Uma sibala i-CPM esebenzayo evela kulawo ma-CPC, sibona ukuthi amanani angacacile aphezulu kakhulu uma kuqhathaniswa nezinye iziteshi , okuqinisa umqondo wokuthi i-OpenAI ihlose ukuthengisa i-ChatGPT njengendawo yenhloso efanelekile kakhulu: abasebenzisi baxoxa ngenkuthalo ngezinqumo eziqondile, hhayi nje ukuvakasha okungaqondile kwiwebhusayithi.
Indlela okunqunywa ngayo ukuthi yisiphi isikhangiso umsebenzisi ngamunye asibonayo: ukuqondiswa kwengxoxo kanye nokulawula okusobala
Isisekelo sokuhlukaniswa kwe-ChatGPT yilokho i-OpenAI ekubiza ngokuthi "ukuqondiswa kwengxoxo ." Esikhundleni sokuthembela kuphela kuma-cookies noma kumaphrofayili endabuko yabantu, uhlelo lugxila kulokho umsebenzisi akubuzayo nakwenzayo ngesikhathi sangempela engxoxweni.
Le ndlela ihlanganisa izimpawu zomongo ezivela engxoxweni (amagama angukhiye, inhloso ebonakalayo, uhlobo lomsebenzi) nedatha yokusebenzisana yangaphambilini, njengezikhangiso ezibukwe ngaphambilini, ukuchofoza, noma umlando wengxoxo yomsebenzisi , lapho umsebenzisi enikeze imvume yokufinyelela imemori. Ngale ndlela, imodeli ingaletha izikhangiso, ngokombono, ezifanela kangcono isidingo esisheshayo.
I-OpenAI iqinisekisa ukuthi umsebenzisi ulawula imvelo: kungenzeka ukukhubaza ingxenye yokwenza ngezifiso izikhangiso, ukususa idatha ehlotshaniswa nemikhankaso , ukulahla izikhangiso ezithile ezinempendulo, noma ngisho nokukhubaza inkumbulo ukuze izingxoxo ezedlule zingasetshenziswa njengesisekelo sokuhlukanisa.
Kodwa-ke, kukhona okubambekayo: abasebenzisi bamahhala abakhawulela kakhulu ukukhangisa noma ukwenza kube ngokwakho bangathola ukufinyelela kwabo ezicini ezithile kuvinjelwe noma ivolumu yemibuzo yabo incishisiwe , ngaphandle kokuthi bathuthukele ohlelweni olukhokhelwayo olusezingeni eliphezulu. Ekugcineni, umyalezo oyimfihlo uwukuthi ukungabikho kwezikhangiso kuza nentengo, kungaba yimali noma izici.
Umthelela ongaba khona ekunembeni: lapho umugqa ophakathi kokusebenziseka kanye nokubandlulula weqa khona
Ukwesaba okukhulu phakathi kwabasebenzisi abaningi nochwepheshe ukuthi, njengoba ukukhangisa kuyanda, izimpendulo ze-ChatGPT zizoshintsha kusukela ekubekeni phambili ikhwalithi yolwazi ziye ekuthandeni izintshisekelo zezentengiselwano . Kubalulekile ukuhlukanisa phakathi kwezingqimba eziningana zethonya elingaba khona.
Esikhathini esifushane, ingozi ebonakala kakhulu yileyo yobandlululo olucashile endleleni ezinye izindlela ezi-oda noma ezethulwa ngayo . Nakuba i-OpenAI ithi izikhangiso azifakwanga embhalweni wempendulo, akukho okuvimbela amafomethi ahlanganisiwe njenge-"impendulo enconyiwe" noma iziphakamiso eziqhubekela phambili ezinkethweni ezinezinzuzo zomnotho ezingcono ekubeni yinto evamile ngokuhamba kwesikhathi.
Okukhathaza kakhulu ukuthi kungenzeka ukuthi idatha etholakala emikhankasweni (yiziphi izikhangiso ezisebenza kahle kakhulu, ngaphansi kwaziphi izimo, ngezinhlobo zemibuzo) ingagcina ithonya izimpawu zokufaneleka noma ngisho nedatha yokuqeqeshwa yesikhathi esizayo . Uma imodeli ijwayela ukubona imikhiqizo noma izinsizakalo ezithile ezihlotshaniswa nezinhloso ezithile, ingafaka lezo zixhumanisi ngaphakathi futhi izikhiqize ngale kwamabhulokhi "amakwe" njengaxhasiwe.
Lokhu kungqubuzana akukona okwe-OpenAI kuphela. Noma iyiphi ipulatifomu ehlanganisa ulwazi, izincomo, kanye nokukhangisa okukhokhelwayo ibhekene nenkinga efanayo eyisisekelo: inketho ekhulisa imali engenayo ayihambisani njalo naleyo ekhulisa inzuzo yomsebenzisi. I-Google ibilokhu izulazula kule bhalansi amashumi eminyaka, ngemiphumela engabazisayo kakhulu lapho iqhathanisa ama-SERP eminyaka engamashumi amabili edlule nemiphumela yanamuhla, egcwele amabhlogo okukhangisa.
Ngaphezu kwalokho, kunengozi yokuthi izikhangiso zizosusa imithombo ezimele noma izimo ezibucayi . Uma ungqimba lokuqala olubonakalayo ngemva kwempendulo kuyizipho ezikhokhelwayo, ingxenye yezilaleli ingase ihlale kulolo daba futhi ingafinyeleli ulwazi oluyinkimbinkimbi kakhulu, okonakalisa ikhwalithi iyonke yesistimu yolwazi.
Izingozi ezithile kumsebenzisi nokuthi ungazithola kanjani
Izinkinga ezivela kulolu xhumano lolwazi kanye nokukhangisa zingahlukaniswa ngezigaba eziningana, zonke ezithinta ngqo ulwazi lomsebenzisi kanye nokunemba okubonakalayo kwe-ChatGPT :
- Ukubandlulula kwezentengiselwano: izincomo ezikhetha ngokuhlelekile izinhlobo ezimbalwa zemikhiqizo, amapulatifomu, noma izinsizakalo ngaphandle kokuchaza ukuthi kungani zingcono kunezinye izindlela ezingathathi hlangothi noma ezivulekile.
- Ulwazi olungesilo iqiniso noma ukuntuleka komongo: Izimpendulo ezingenazo izinkomba ezicacile, ngaphandle kokucaphuna imithombo eqinisekisiwe, noma ezishiya izincazelo ezibucayi ngokuzama ukwenza kube lula kakhulu.
- Izingozi zobumfihlo: ukusetshenziswa kwedatha yengxoxo kanye nezimpawu zokuziphatha ukuze kuhlukaniswe izikhangiso ngokunemba okwandayo, noma ngabe umsebenzisi akakwazi ngokugcwele.
- Okuhlangenwe nakho okubi: ukuphazamiseka okuphindaphindiwe, ukuvimba izikhangiso ezingafanele, noma umuzwa wokuthi umsizi "ucindezela" okuqukethwe okungangezi inani.
Ukuze uzivikele, kunezindlela eziningana eziwusizo. Eyokuqala ukuphatha i-ChatGPT njengendawo yokuqala, hhayi umlamuli wokugcina : ukubhekisela kunoma iyiphi idatha efanele nemithombo eyinhloko, amadokhumenti asemthethweni, abezindaba abakhethekile, noma izizindalwazi zezemfundo lapho isinqumo sibalulekile.
Kunconywa futhi ukucela imodeli ukuthi ibhale imithombo abayisebenzisayo futhi inikeze izinkomba ezithile (isixhumanisi, umbhali, usuku) lapho yenza izincomo ezibucayi: kusukela kwezempilo kanye nezezimali kuya ezinkontilekeni noma iseluleko sezomthetho. Uma bengenakusho lutho oluqinile, kuwuphawu olucacile lokuthi akufanele sithembele impendulo yabo ngokungacabangi.
Ngokuphathelene nobumfihlo, kuyalulekwa ukuthi ubuyekeze ngokucophelela izilungiselelo zokwenza kube ngokwakho, imemori kanye nokusetshenziswa kwedatha yezikhangiso , ukhethe izinguqulo ezikhokhelwayo ezithembisa ukusetshenziswa okuncane kolwazi kwezentengiselwano, futhi ugweme ukwabelana ngemininingwane yomuntu siqu ebucayi kuzixhumi ezibonakalayo zamahhala ezisekelwa yizikhangiso.
Okokugcina, lapho isincomo sihilela izinqumo ezithinta kakhulu (ukwelashwa kwezempilo, ukutshalwa kwezimali, izinyathelo zomthetho eziyinkimbinkimbi), kuwukuhlakanipha ukubonisana nochwepheshe babantu futhi usebenzise i-chatbot njengokusekela kuphela , hhayi njengokufaka esikhundleni sombono wochwepheshe.
Izikhangiso ze-ChatGPT ngokombono wemikhiqizo nama-ejensi
Kumabhrendi, i-ChatGPT Ads iyitshe legolide elibalulekile. Sisesimeni lapho abasebenzisi bebuza imibuzo ecacile, enezinhloso eziphakeme futhi bafuna izixazululo eziqondile — into ukumaketha obekuzama ukuyibamba iminyaka eminingi ezinjinini zokusesha nakuma-social media. Ukuvela ngaleso sikhathi esiqondile ngesipho esifanele kungaba namandla amakhulu.
Kodwa-ke, iqiniso liwukuthi sisesigabeni sokuqala, ngesilinganiso esilinganiselwe, amanani aphezulu, kanye namakhono okulinganisa asathuthuka . Abakhangisi bokuqala basebenza kumodeli yokuhlola, enemibiko egxile kakhulu emibonweni nasekuchofozweni, kodwa intula ukujula kwedatha yokusebenza amaqembu okusebenza ajwayelene nayo.
Ama-ejensi afana ne-DAC ahumusha lesi siteshi esisha njengelabhorethri yokuqonda ukutholwa kwengxoxo : ukuthi i-ChatGPT ingena kanjani ku-funnel yokuguqulwa, ukuthi iyidlala yiphi indima maqondana ne-Google noma amanethiwekhi omphakathi, nokuthi amasu okudala nawokubhida kufanele azivumelanise kanjani nalesi simo esihluke kakhulu.
Isincomo esijwayelekile ukuthi uqhubeke ngokuqapha: klama izivivinyo ezihlelwe kahle ezinemibono ecacile , futhi uqaphele ukuthi, okwamanje, okubaluleke kakhulu akukhona ukukhulisa ivolumu kodwa ukufunda ukuthi hlobo luni lwemiyalezo, izingxenye, kanye nezikhathi engxoxweni ezikhiqiza umthelela wekhwalithi ovumela lawo ma-CPM nama-CPC e-premium.
Ngesikhathi esifanayo, izixazululo seziqala ukuvela ezisebenzisa amamodeli afana ne-GPT ukuze kuthuthukiswe ukuhlukaniswa kwezilaleli kanye nokwakhiwa komkhankaso kwezinye iziteshi. Ngokusebenzisa amakhasi okufikela, ikhophi yesikhangiso, amabhanela, kanye nedatha yokusebenza komlando, la mathuluzi angakhipha izimfanelo, achaze ubuntu bomthengi, aphakamise ukuhlolwa kwe-A/B, futhi aphakamise imingcele yokubhida kanye nezikhathi zokuguqulwa okuphezulu, ahlanganiswe namapulatifomu okukhangisa kanye nezinhlelo ze-BI ezifana ne-Power BI ukuze kuqashwe umthelela ngesigaba.
UClaude, uGemini, iMeta AI kanye nempi yokukhuthaza
Isinqumo se-OpenAI sokwamukela ukukhangisa siholele ekutheni izimbangi zayo zithathe isinyathelo esicacile . Isibonelo, i-Anthropic imemezele ukuthi umsizi wayo uClaude ngeke afake okuqukethwe okuxhasiwe ezingxoxweni, eziveza njengenye indlela "engenazikhangiso".
Inkampani kaDario Amodei ithi ukuxuba ukukhangisa nezimpendulo kungaba okungafaneleki noma kube yingozi ezimweni ezithile, nokuthi kungaqeda ukwethenjwa kwabasebenzisi, okubenza bahlale bezibuza ukuthi isincomo siqotho noma sinobandlululo yini esivumelwaneni sezentengiselwano.
I-Anthropic ize yaqala imikhankaso yomphakathi egxeka isinqumo se-OpenAI, ngemiyalezo efana nokuthi " Izikhangiso zifinyelela ku-AI, kodwa hhayi ku-Claude " kanye nezibonelo zendlela okuqukethwe okuxhaswe ngayo okungaphambukisa ngayo izingxoxo mayelana nokuzihlonipha, imfundo, noma inhlalakahle yemizwa. Umyalezo wabo oyisisekelo ukuthi bafuna ukugcina izisusa zabo zihambisana nenhlalakahle yabasebenzisi, hhayi ukuzibandakanya kwezikhangiso.
USam Altman uphendule ngokukhulu ukucasuka, esola u-Anthropic ngokungathembeki futhi ethatha isimo esicishe sibe "sobushiqela" ngokuzama ukuyalela abasebenzisi ukuthi i-AI kufanele isetshenziswe kanjani. Uphinde waveza intengo kanye nokufinyeleleka , esikisela ukuthi uClaude ungumkhiqizo obizayo ohloselwe izilaleli ezicebile, kuyilapho i-ChatGPT ihlose ukufinyelela izilaleli eziningi kakhulu, noma ngabe kusho ukuxhasa ingxenye yensizakalo ngokukhangisa.
Okwamanje, iziqhwaga ezifana ne-Google ne-Meta zinomlando wazo wokukhangisa, futhi kuyindaba yesikhathi nje ngaphambi kokuba i-Gemini noma i-Meta AI ihlanganise izikhangiso ngandlela thile. Lezi zinkampani zakhe imibuso yazo ngokuqondile ekwenzeni imali ngokunaka, ngakho akekho olindele ukuthi zishiye lowo mfula wemali engenayo endaweni yengxoxo engaba yindlela entsha yokungena kuwebhu, njengoba nje uNick Srnicek abikezela eminyakeni edlule.
I-Apple njengesihlungi: sebenzisa i-ChatGPT ngaphandle kwezikhangiso (futhi cishe ngendlela engabonakali)
Kuvela umdlali ohlukile kule nkundla yokudlala: i-Apple. Inkampani ye-Cupertino ikhethe, okwamanje, hhayi ukwakha i-AI enkulu ekhiqizayo kusukela ekuqaleni, kodwa kunalokho ukuhlanganisa amamodeli angaphandle njenge-ChatGPT ne-Gemini ohlelweni lwayo lwe-ecosystem , ngezivumelwano zezigidi zamaRandi kodwa ngaphandle kokuthwala zonke izindleko zezikhungo zedatha kanye nezisheshisi ze-AI.
Ngokuhlanganiswa kwe-ChatGPT ku-Siri kanye ne-Apple Intelligence, i-iPhone isebenza njengomlamuleli phakathi komsebenzisi kanye namaseva e-OpenAI . Umbuzo usuka kudivayisi uye kwingqalasizinda ye-OpenAI, kodwa wenza kanjalo ngaphansi kwezinqubomgomo zobumfihlo ze-Apple: Ikheli le-IP lifihliwe, imikhawulo yokusebenzisa ingxoxo ukuqeqesha amamodeli, kanye ne-interface yokukhangisa ye-OpenAI ayinakho ukuboniswa.
Lokhu kwenza i-Siri ibe uhlobo lokuvimba izikhangiso ezingeni lesistimu ye-ChatGPT. Uma ucela umsizi we-iPhone ukuthi abhale umbhalo esebenzisa i-ChatGPT, akhiqize isithombe, noma akusize ngomsebenzi oyinkimbinkimbi, imodeli ye-OpenAI isebenza ngemuva, kodwa izikhangiso ezizovela kuwebhu noma kuhlelo lokusebenza olusemthethweni aziboniswa kungqimba yesixhumi esibonakalayo se-Apple.
Lesi sinyathelo esibalulekile sikabili. Ngakolunye uhlangothi, i-Apple igwema ukuqala umjaho wokusebenzisa imali ngokungalawuleki njengowe-Google noma i-OpenAI, othwala izindleko ezinkulu zengqalasizinda yonyaka. Ngakolunye uhlangothi, inikeza abasebenzisi bayo ukufinyelela kumamodeli athuthukile "angenazo" izikhangiso eziqondile , okuqinisa indaba yobumfihlo kanye nolwazi oluphezulu.
Uma sibheka phambili, ukufika kwe- Siri evuselelwe ene-Gemini ehlanganisiwe kunganciphisa ukubaluleka kokuhlanganiswa kwe-ChatGPT ngaphakathi kwe-ecosystem ye-Apple. Uma umsebenzisi ekwazi ukufinyelela kumsizi onamandla, oyimfihlo, futhi ongenazikhangiso ngqo ngaphakathi kwesistimu, isisusa sokuvula i-ChatGPT kusiphequluli noma kuhlelo lokusebenza—lapho ukukhangisa kuzoba khona—sincipha kakhulu.
Imiphumela yocwaningo, umsebenzi wobungcweti, kanye nokuphila kwansuku zonke
Kulabo abasebenzisa i-ChatGPT njengethuluzi lokucwaninga, lokubhala, noma lokusekela lochwepheshe, ukwethulwa kokukhangisa kuphakamisa ukukhathazeka okunengqondo kakhulu. Kukhona ukwesaba ukuthi ipulatifomu izogcina iphinda indlela ye-Google : ukusuka esigabeni esibhekwa njengesingathathi hlangothi futhi esigxile ekusetshenzisweni kuya kwesilawulwa yisidingo sokukhipha imali engenayo yokukhangisa kukho konke ukuxhumana.
Emikhakheni efana nezemfundo, ezomthetho, noma ezempilo, lokhu kusola kunganciphisa ukuzimisela ukwethemba umsizi ngemisebenzi ebucayi. Abacwaningi nochwepheshe bangakhetha izinguqulo ezikhokhelwayo, ezingenazikhangiso, amathuluzi azisingathayo, noma abancintisana nabo abathembisa ukungathathi hlangothi , ikakhulukazi uma bedinga ukulandelelwa okucacile kwemithombo kanye neziqinisekiso zokuthi azikho izisusa ezifihliwe ngemuva kwesincomo.
Kumphakathi jikelele, umphumela ungaba sobala kakhulu: ukwenza kube khona izikhangiso ezingxoxweni ngendlela efanayo naleyo eziye zajwayeleka ngayo ezinkundleni zokuxhumana, izinsizakalo zokusakaza, noma izinjini zokusesha. Ingozi iwukuthi, ngokuhamba kwesikhathi, umugqa phakathi kweseluleko sochwepheshe neziphakamiso ezixhasiwe uzofiphala , futhi kuzoba nzima ngokwengeziwe ukubona ukuthi iyiphi ingxenye yesipiliyoni elungiselelwe ukusiza umsebenzisi nokuthi iyiphi ingxenye yokwenza imali.
Ngesikhathi esifanayo, kubalulekile futhi ukuqaphela ukuthi ngaphandle kwemithombo yemali engenayo eqinile, eziningi zalezi zinsizakalo bezingeke zihlale isikhathi eside. Isihluthulelo kuzoba ukuthi i-OpenAI kanye nabanye abadlali bangakwazi yini ukugcina izilawuli zekhwalithi ezanele, ukuhlolwa okuzimele, kanye nokucaca ukuze kuvinjelwe ukwethembana ohlelweni ukuthi lungawi njengoba ukuthembela ekukhangiseni kukhula.
Ekugcineni, ukunemba kwe-ChatGPT endaweni esekelwa izikhangiso kuzoncika ebhalansini elibucayi: ukuthi izikhuthazo zomnotho ziphumelela kangakanani ekugwemeni ukuhlanekezela ukwakheka komkhiqizo . Umlando wezinye izinkundla usikisela ukuthi ukungezwani phakathi kokusetshenziswa kanye nokwenza imali akupheli; kumane kuphathwa kangcono noma kubi. Kubasebenzisi, ukuzivikela okungcono kakhulu kuzoba ukuqhubeka nokusebenzisa ukucabanga okubucayi, ukuqinisekisa ulwazi olubucayi, kanye nokusebenzisa, uma kutholakala, ukufinyelela ezinguqulweni noma ezinye izindlela ezingenazikhangiso ezigcina ingxoxo ingenazo izixhaso.
