🏫 Kwezinye izikole, ama-chatbot abuza abafundi ngomsebenzi wabo. Kodwa inguquko ye-AI ikhathaze othisha ↗
Ezinye izikole zase-Australia zenza ushintsho olubukhali ku-"bonisa umsebenzi wakho" - abafundi baqeda umsebenzi, bese i-chatbot ye-AI ibabuza imibuzo ngokuzumayo ukuze bahlole ukuthi bayaqonda yini abakuthumele. Empeleni kuyi-oral viva, kodwa ngomshini obuza ukuthi "usho ukuthini ngalokho?" uze uqale ukuthuthumela.
Okukhathaza kakhulu ukukhishwa okungalingani: iphepha lemboni lixwayisa ngokuthi abantu abamukela ngokushesha bangagijima phambili kuyilapho abanye bebambezeleka, okudala uhlelo lwemfundo olunesivinini esikabili. Othisha baqhubeka nokumaka ukukhathazeka okujwayelekile - nokusabalulekile -: ubumfihlo, inhlalakahle, kanye nendlela yokumisa "usizo lwe-AI" ukungena ku-"AI did it".
👮 Uhlangane namaphoyisa esebenzisa amathuluzi e-AI anikezwe yi-Palantir ukuze abonise ukuziphatha kabi kwamaphoyisa ↗
I-Met yaseLondon isebenzisa i-AI ehlinzekwa yi-Palantir ukuhlaziya amaphethini abasebenzi bangaphakathi - ukugula, ukungabikho emsebenzini, isikhathi esengeziwe - ifuna izimpawu ezingase zihlobane nezinkinga zezindinganiso zobungcweti. Ibekwe njengokuhlola okulinganiselwe isikhathi futhi, okubaluleke kakhulu, i-Met ithi abantu basazenza izinqumo zangempela ... kodwa lokho "kusasebenza" kwenza umsebenzi omningi.
I-Police Federation ayijabule, ikubiza ngokuthi “ukusola okuzenzakalelayo” futhi ixwayisa ngamathuluzi angacacile achaza kabi ukucindezeleka kanye nomsebenzi njengokungalungile. Kungenye yalezo zikhathi lapho ubuchwepheshe buzwakala bucocekile ku-slide, bese ubona ngeso lengqondo ukuthi buzizwa kanjani phansi bese buphenduka bube bukhali.
🛡️ Inhlangano eqapha ubumfihlo e-UK ixwayisa ngezithombe ezikhiqizwe yi-AI esitatimendeni esihlanganyelwe ↗
I-ICO yase-UK ijoyine abanye abalawuli esixwayisweni esihlangene mayelana nezithombe ezikhiqizwe yi-AI ezibonisa abantu abaziwayo ngaphandle kwemvume - ngokuyisisekelo: ungaphathi "okwenziwayo" njengezaba zomlingo. Umyalezo uwukwakha izivikelo kusenesikhathi futhi ubandakanye abalawuli esikhundleni sokwenza isinyathelo sobuchwepheshe esijwayelekile sokuthumela kuqala bese uxolisa kamuva.
Ziphinde zigqamise izingozi ezinkulu ezinganeni, okunengqondo, ngoba uma lokhu kusakazeka kufana nokuzama ukuqoqa i-glitter ebhodweni. Inhloso ukuthi ubumfihlo kanye nesithunzi kuhloselwe ukusinda emjikelezweni wemfashini, noma kubonakala kanjalo.
🪖 UNobhala Wezokuvikela wase-US uHegseth ubize i-CEO ye-Anthropic ngezingxoxo ezinzima mayelana nokusetshenziswa kukaClaude kwezempi, kubika i-Axios ↗
I-Reuters ibika ukuthi uNobhala Wezokuvikela wase-US uhlelelwe ukuhlangana ne-CEO ye-Anthropic ngalokho okuchazwa njengengxoxo ephikisanayo mayelana nokusetshenziswa kwezempi kukaClaude. Kubikwa ukuthi iPentagon ifuna abahlinzeki abakhulu be-AI benze amamodeli atholakale kumanethiwekhi ayimfihlo anemikhawulo embalwa - futhi i-Anthropic ibilokhu ilwa nanoma yikuphi ukuxegiswa kwezivikelo.
Umbhalo ongezansi ufundeka njengenkinga evamile ethi “hamba ngokushesha kodwa… ngemicibisholo ehilelekile”. Uma izinhlangothi zingavumelani ngezivikelo, ubudlelwano bungaba nzima noma buqhekeke - nokho i-Anthropic nayo ikhombisa ukuthi izingxoxo nohulumeni zisalokhu zakha, ngakho-ke kushubile, kodwa akuzona imibhalo yokuhlukana ngokuphelele (okwamanje).
🧠 I-ChatGPT - Amanothi Okukhishwa ↗
Amanothi okukhishwa kwe-OpenAI akhomba iwindi lokuqukethwe elandisiwe lemodi ethi “Thinking” - okusho ukuthi ungondla imodeli ngombhalo omningi futhi ugcine izingxoxo ezinde zihambisana ngaphandle kokuthi ilahle imininingwane yangaphambilini. Uma uke wabuka imodeli “ikhohlwa” into osanda kuyisho (ubuhlungu), lokhu kuhloswe ngakho ngqo.
Kuyintuthuko eyinkimbinkimbi kakhulu eshintsha buthule lokho okusebenzayo: amadokhumenti amade, amaphrojekthi axakile, izikhathi zocwaningo ezibanzi. Akuyona into ekhangayo, kodwa uhlobo lokushintsha amapayipi okwenza indlu yonke izwakale inkulu ngokuzumayo.
Imibuzo Evame Ukubuzwa
Kusho ukuthini uma izikole zisebenzisa ama-chatbot “ukubuza” abafundi ngemisebenzi?
Kwezinye izikole zase-Australia, abafundi baqeda umsebenzi othile bese bedlula emuva naphambili nge-chatbot ye-AI ehlola ukucabanga kwabo. Isebenza njenge-oral viva, icindezela abafundi ukuthi bachaze lokho abakubhalile nokuthi kungani. Inhloso ukuqinisekisa ukuqonda, hhayi nje umphumela. Ingaveza nezikhala lapho umfundi engakwazi ukuvikela izimangalo ezibalulekile.
Ama-chatbot e-AI ezikoleni angaludala kanjani uhlelo lwemfundo “olunesivinini esikabili”?
Iphepha lemboni lixwayisa ngokuthi abantu abasebenzisa ngokushesha bangaqhubeka ngenkathi izikole ezihamba kancane zihlehla kakhulu. Uma ezinye izikole zisebenzisa amathuluzi e-AI ukuhlola ukuqonda, ukusekela ukufunda, noma ukuvimbela ukusetshenziswa kabi, zingase ziphakamise imiphumela ngokushesha. Ezinye zingase zithole isabelomali, amandla abasebenzi, noma ukucaca kwenqubomgomo ukuze kuqaliswe izinhlelo ezifanayo. Lokho kusetshenziswa okungalingani kungandisa izikhala ezikhona zezinsizakusebenza kanye nokufinyelela.
Othisha bazama kanjani ukuvimba ukuthi “i-AI iyasiza” iphenduke “i-AI ikwenzile”?
Indlela evamile ukushintsha ukuhlola kuye enkambisweni nasencazelweni, hhayi nje umkhiqizo oqediwe. Imibuzo yesitayela se-Chatbot, ama-draft, ukuzindla, noma ukuhlolwa ngomlomo kungenza kube nzima ukuthumela umsebenzi ongawuqondi. Othisha bangase babeke nemingcele eqinile ekusizeni okwamukelekayo, njengokucabanga ngokuqhathanisa nokubhala kabusha okugcwele. Ukugcizelela kuhlala ekuboniseni ukuqonda, hhayi ukuqondisa wonke amathuluzi.
Yiziphi izinkinga zobumfihlo kanye nenhlalakahle eziphakamayo lapho kusetshenziswa i-AI emakilasini?
Othisha baqhubeka nokuphakamisa ukukhathazeka ngokuthi yiluphi ulwazi lwabafundi oluqoqwayo, ukuthi lugcinwa kuphi, nokuthi ubani ongalufinyelela. Kukhona futhi nokukhathazeka ngenhlalakahle uma abafundi bezizwa beqashwe njalo noma becindezelwa imibuzo eholwa ngumshini. Izingane ziyiqembu elikhethekile elisengozini, ngoba ulwazi olubucayi lungasakazeka noma luqhubeke ezinhlelweni ngezindlela ezingalindelekile. Izikole eziningi zihlose ukubeka izindlela zokuvikela kusenesikhathi kunokuzama ukuzilungisa kamuva.
I-Met isebenzisa kanjani i-AI ehlinzekwa yi-Palantir ukukhombisa ukuziphatha kabi kwamaphoyisa okungenzeka?
I-Met yaseLondon isebenzisa i-AI ehlinzekwa yi-Palantir ukuhlaziya amaphethini abasebenzi njengokugula, ukungabikho emsebenzini, kanye nesikhathi esengeziwe. Inhloso ukubona izimpawu ezingase zihlobane nezinkinga zezindinganiso zobungcweti. Ichazwa njengokuhlola okulinganiselwe isikhathi, futhi i-Met ithi abantu basazenza izinqumo. Abagxeki bathi leli thuluzi lingazwakala “njengokusola okuzenzakalelayo” uma libheka ukucindezeleka njengokwenza okubi.
Kungani izinyunyana zamaphoyisa zikhathazekile “ngokusola okuzenzakalelayo” okuvela kumathuluzi okuqapha e-AI?
Ukukhathazeka kwe-Police Federation ukuthi amamodeli angacaci angase aphethe ngokuthi ukuziphatha kabi kubangelwa amaphethini anezincazelo ezingembi. Ukwanda komthwalo womsebenzi, izinkinga zempilo, kanye nokucindezeleka kungashintsha ukuba khona noma isikhathi esengeziwe ngaphandle kokusho ukuthi kukhona okungalungile. Uma ukucabanga kwaleli thuluzi kungacacile, kuba nzima ngamaphoyisa ukuphikisa amafulegi noma kubaphathi ukulinganisa imingcele efanele. Ngisho nezinqumo zabantu, izimpawu ezizenzakalelayo zisengakha imiphumela.
Isexwayiso se-ICO sase-UK simayelana nani ngezithombe ezikhiqizwe yi-AI zabantu abaziwayo?
Inhlangano eqapha ubumfihlo e-UK ijoyine abanye abalawuli ekuxwayiseni ngokuthi "ukwenziwa" akusho ukuthi akukho miphumela. Uma isithombe esenziwe nge-AI sibonisa umuntu obonakalayo ngaphandle kwemvume, izingozi zobumfihlo nesithunzi zisasebenza. Abalawuli bagcizelela ukwakhiwa kwezivikelo kusenesikhathi kanye nokubandakanya ukuqapha kunokuthumela kuqala. Baphinde baveze izingozi ezinkulu ezinganeni, lapho ukulimala kunganda khona ngokushesha uma okuqukethwe kusakazeka.
Yini ebangela ingcindezi ebikiwe yePentagon ku-Anthropic mayelana nokusetshenziswa kukaClaude kwezempi?
I-Reuters ibika ukuthi uNobhala Wezokuvikela wase-US uhlelelwe ukuhlangana ne-CEO ye-Anthropic ukuze kuxoxwe kanzulu ngokusetshenziswa kukaClaude kwezempi. Kubikwa ukuthi iPentagon ifuna abahlinzeki abakhulu be-AI benze amamodeli atholakale kumanethiwekhi ayimfihlo anemikhawulo embalwa. I-Anthropic ibilokhu imelana nezivikelo ezikhululekile, ikhomba ukungqubuzana ngezivikelo. Ubudlelwano buzwakala bunzima, yize izinhlangothi zombili zibonisa ukuthi izingxoxo zisakha.
Yini eshintshile kumanothi okukhishwa kwe-ChatGPT mayelana nobude bomongo wemodi ethi “Thinking”?
Amanothi okukhishwa kwe-OpenAI akhomba efasiteleni lomongo elandisiwe lemodi ethi “Thinking”, okuvumela umbhalo omningi ukuthi uhlale ngaphakathi kwentambo eyodwa. Empeleni, lokho kungasiza ngamadokhumenti amade, amaphrojekthi ayinkimbinkimbi, kanye nocwaningo lwezinyathelo eziningi lapho imininingwane yangaphambilini ibalulekile. Kuyintuthuko yesitayela samapayipi kunokuba kube yisici esikhazimulayo. Inzuzo eyinhloko imininingwane embalwa elahlekile ngezikhathi ezinde.