- I-CEO ye-OpenAI uSam Altman uthi umbuzo ngamnye we-ChatGPT usebenzisa malunga neelitha ze-0,00032 zamanzi, uthelekisa lo mqulu kunye "nesinye kwishumi elinesihlanu itispuni."
- Ukusetyenziswa kwamandla okusebenzisana ne-ChatGPT kujikeleze i-0,34 watt-iiyure, kufana nokusebenzisa ibhalbhu yokukhanya kwe-LED imizuzu embalwa.
- Iingcali kunye namalungu enzululwazi achaza ukuba akukho bungqina bucacileyo obuye banikwa ukuxhasa la manani, kwaye indlela yawo yokusebenza ayizange ichazwe.
- Ingxoxo-mpikiswano malunga nempembelelo yendalo ye-AI ihlala iqhubeka, ngakumbi malunga nokupholisa iziko ledatha kunye noqeqesho lweemodeli ezinkulu.
Ukuqhubela phambili ngokukhawuleza kobukrelekrele bokwenziwa kuzise etafileni inkxalabo ngempembelelo yawo kwimekobumeingqalelo ekhethekileyo Ukusetyenziswa kwamandla kunye namanzi kubandakanyeka ekusebenzeni iimodeli ezidumileyo njengeChatGPT, iphuhliswe yi-OpenAI. Kwiinyanga zamva nje, i-CEO yenkampani, uSam Altman, uye wazama ukucacisa ubungakanani betekhnoloji yokusetyenziswa kwemithombo yendalo, nangona kungekho mpikiswano okanye ukungabikho kwemibuzo.
Iingxelo zika-Altman kwibhlog yakhe yobuqu ziye zabangela impikiswano enkulu kwicandelo letekhnoloji nesayensi.Njengoko udumo lwe-ChatGPT luqhubeka lukhula kwihlabathi, uluvo loluntu kunye namajelo eendaba agxile kwi-ecological footprint yombuzo ngamnye, kunye nokuba ingaba idatha enikeziweyo ibonisa ngokwenene ifuthe lemo engqongileyo obunokuba nalo ubukrelekrele bokwenziwa kubomi bemihla ngemihla.
Angakanani amanzi esetyenziswa yiChatGPT ngombuzo ngamnye?
Kutshanje, uSam Altman uthe Ngalo lonke ixesha umsebenzisi enxibelelana ne-ChatGPT, ukusetyenziswa kwamanzi okuhambelanayo kuncinci.Njengoko wachazayo, Ukubonisana okukodwa kudla malunga ne-0,00032 ilitha zamanzi, ngokumalunga nokulingana "nesinye kwishumi elinesihlanu letisipuni." Le mali isetyenziswe ngokuyinhloko kwiinkqubo zokupholisa zamaziko edatha apho iiseva ziqhuba kwaye zivelise iimpendulo ze-AI.
Ukupholisa kubalulekile ukuthintela ukufudumeza kwezinto ze-elektroniki, ngakumbi xa sithetha ngeziseko ezingundoqo ezinkulu eziqhuba ngokuqhubekayo nangokupheleleyo. Le mfuneko yokupholisa oomatshini ngamanzi ayiphelelanga kwi-ChatGPT kuphela, kodwa iqhelekile kubo bonke yonke i-computing yefu kunye necandelo le-AI. Nangona kunjalo, ubukhulu bemibuzo yemihla ngemihla-izigidi, ngokutsho kwe-OpenAI-buthetha ukuba nokuba ukusetyenziswa kancinci kuqokelela impembelelo encomekayo.
Nangona u-Altman wayefuna ukugxininisa ukuba ixabiso lomsebenzisi ngamnye liphantse lingabinamsebenzi, Iingcali kunye nezifundo zangaphambili zipapashe amanani aphezulu kuphando oluzimeleyoNgokomzekelo, uhlalutyo lwakutsha nje lweeyunivesithi zaseMelika lucebisa ukuba Ukuqeqesha iimodeli ezinkulu ezifana ne-GPT-3 okanye i-GPT-4 kunokufuna amakhulu amawaka eelitha zamanzi., nangona ukusetyenziswa okuthe ngqo kokubonisana kwansuku zonke kuphantsi kakhulu.
Impikiswano yamanani: amathandabuzo malunga nokungafihli kunye neendlela
Iingxelo zika-Altman zifunyenwe ngononophelo luluntu lwezenzululwazi kunye neendaba ezikhethekileyo, ngenxa yoku ukunqongophala kwengcaciso ecacileyo yokuba la maxabiso afunyenwe njaniAmanqaku amaninzi abonisa ukuba i-OpenAI ayizange ipapashe indlela echanekileyo yokubala ukusetyenziswa kwamanzi kunye nombane, nto leyo ekhokelele ukuba amanye amajelo eendaba kunye nemibutho ifune ukucaca ngakumbi kule ndawo.
Ushicilelo lwemidiya olunje ngeWashington Post, iVerge kunye neeyunivesithi ezinjengeMIT okanye iCalifornia zikhombe kuqikelelo oluphezulu, olufikelela phakathi Iilitha ze-0,5 kuzo zonke ii-20-50 zokubonisana (kwimeko yeemodeli zangaphambili ezifana ne-GPT-3) kunye namakhulu amawaka eelitha kwisigaba soqeqesho se-AI.
Ingxoxo yamandla: ukusebenza kakuhle, umxholo kunye nothelekiso
Enye yeengongoma ezibhekiswe nguSam Altman yile Ukusetyenziswa kwamandla okuhambelana nentsebenziswano nganye kunye neChatGPT. Ngokoqikelelo lwabo, Ukubonisana okuqhelekileyo kubandakanya malunga ne-0,34 watt-yure, kufana namandla asetyenziswe yi-bulb ye-LED kwimizuzu emibini okanye i-oven yasekhaya eshiywe isekhondi enye. Ukuqonda ngcono iimpembelelo ze-AI, unokubonisana impembelelo yobukrelekrele bokwenziwa ekuzinzeni.
Nangona kunjalo, Ukusebenza kweemodeli kuye kwanda kwiminyaka yamuva Kwaye i-hardware yanamhlanje iyakwazi ukucubungula izicelo ngamandla amancinci kuneminyaka embalwa edlulileyo. Oku kuthetha ukuba, nangona ukusetyenziswa komntu kuphantsi, umceli mngeni ulele kumthamo omkhulu wokunxibelelana okwenzeka ngaxeshanye okwenzeka kumaqonga afana neChatGPT, Gemini, okanye Claude.
Uphononongo lwakutsha nje luxhasa ukucutha okuthile kumndilili wokusetyenziswa ngokubonisana, nangona begxininisa ukuba Isikhangeli ngasinye, isixhobo ngasinye, kunye nommandla ngamnye unokuba namanani ahlukeneyo. kuxhomekeke kuhlobo lweziko ledatha kunye nenkqubo yokupholisa esetyenzisiweyo.
Inyathelo elongezelelekileyo kunye nomngeni wozinzo lwexesha elide
Eyona ngxaki ivela xa kusenziwa la manani mancinane ngokubonisana kwinani lilonke lonxibelelwano lwemihla ngemihla kwihlabathi liphela. Isixa sezigidi zamathontsi amancinci anokuba sisixa esikhulu samanzi., ngakumbi njengoko i-AI isetyenziselwa imisebenzi entsonkothileyo eyandayo kwaye inabela kumacandelo afana nezemfundo, ulonwabo, kunye nokhathalelo lwempilo.
Ngaphezu koko, Inkqubo yoqeqesho lweemodeli ze-AI zanamhlanje ezifana ne-GPT-4 okanye i-GPT-5 iyaqhubeka ifuna kakhulu izibonelelo., zombini ngokombane kunye namanzi, ukunyanzelisa iinkampani zeteknoloji ukuba zikhangele imithombo yamandla emitsha-njengamandla enyukliya-kunye nokuqwalasela iindawo zamaziko abo edatha apho iziseko zamanzi ziqinisekisiwe.
La Ukunqongophala kwemigangatho ecacileyo, amanani asemthethweni kunye nokungafihli kwizibalo kuyaqhubeka ukuphembelela iingxabanoImibutho efana ne-EpochAI kunye neefemu zokubonisana ziye zazama ukuqikelela impembelelo, kodwa akukho mvumelwano malunga neendleko zendalo ezisingqongileyo zokusebenzisana ne-AI yokuvelisa kwizinga elikhulu. Okwangoku, impikiswano ivula ifestile yokucamngca ngekamva leteknoloji kunye noxanduva lokusingqongileyo lwabasekeli bayo abaphambili.
Ingxoxo malunga USam Altman kunye ne-AI ngokubanzi iqaqambisa ukungavisisani phakathi kokuyila kobugcisa kunye nozinzo. Ngelixa amanani anikezelwe nguSam Altman efuna ukuqinisekisa uluntu malunga nefuthe eliphantsi lokubonisana komntu ngamnye, ukunqongophala kokungafihli kunye nomgangatho wenkonzo yehlabathi kugcina imbonakalo yemfuno yokubeka iliso kunye nobungqongqo besayensi xa kuvavanywa unyawo lwenkqubo yendalo esele iyinxalenye yobomi bethu bemihla ngemihla.
Ndingumntu othanda itekhnoloji ojike umdla wakhe we "geek" waba ngumsebenzi. Ndichithe ngaphezulu kweminyaka eli-10 yobomi bam ndisebenzisa itekhnoloji yokusika kwaye ndikhenkceza ngazo zonke iintlobo zeenkqubo ngenxa yokufuna ukwazi okumsulwa. Ngoku ndiqeqeshelwe ubugcisa bekhompyutha nakwimidlalo yevidiyo. Oku kungenxa yokuba ngaphezu kweminyaka emi-5 ndibhalela iiwebhusayithi ezahlukeneyo kwitekhnoloji kunye nemidlalo yevidiyo, ndisenza amanqaku afuna ukukunika ulwazi oludingayo ngolwimi oluqondakalayo kuye wonke umntu.
Ukuba unayo nayiphi na imibuzo, ulwazi lwam lusuka kuyo yonke into enxulumene nenkqubo yokusebenza yeWindows kunye ne-Android yeefowuni eziphathwayo. Kwaye ukuzinikela kwam kukuwe, ndihlala ndikulungele ukuchitha imizuzu embalwa kwaye ndikuncede usombulule nayiphi na imibuzo onokuba nayo kweli lizwe le-intanethi.