Umsebenzi othuthukisiwe we-AI: Ingabe izinhlelo zokufunda zomshini zingaba umlingani wethu weqembu ongcono kakhulu?

• About the Author:
• Igama lomlobi

  I-Quantumrun Foresight
• November 10, 2023

Isifinyezo sokuqonda

Ukushintshashintshashintsha phakathi kwabantu nemishini kuyathuthuka, ngobuhlakani bokwenziwa (AI) bengena ezindimeni ezikhulisa amakhono omuntu futhi ziguqule ubudlelwano obujwayelekile bethuluzi lomsebenzisi bube ukusebenzisana okwengeziwe. Ukusuka ekunakekelweni kwezempilo kuye ekuthuthukisweni kwesoftware, indima ye-AI iguqukela ekubeni ngumsizi obalulekile, esiza emisebenzini efana nokuhlaziya idatha, ukuphatha amarekhodi esiguli, noma ngisho nokufunda ukubhala amakhodi. Lolu shintsho luphinde luveze imithelela ehlukahlukene, okuhlanganisa isidingo sezinhlaka ezintsha zokulawula, ukufunda okuqhubekayo kwabasebenzi, kanye namandla ezinqubo zokusebenza ezisebenza kahle neziphephile kuyo yonke imikhakha eyahlukene.

I-AI-augmented work content

Ukusebenzisana phakathi kwabantu nemishini bekulokhu kuyinkundla yezingxoxo, ikakhulukazi ngokufika kwe-AI kanye nobuchwepheshe bokufunda ngomshini (ML). Ukwesaba okuvamile ukuthi i-AI ingaba yindawo yokuzalanisa imininingwane engamanga noma izindaba ezingamanga, ebhebhezela ukungathembani phakathi kwabantu. Kodwa-ke, i-AI ikhombisa amandla amakhulu ekwandiseni amakhono abantu kanye nokuqhubekisela phambili ubuhlakani nokuqamba izinto ezintsha. Ochwepheshe abaningi baphikisa ngokuthi ukusetshenziswa kwamanje kwe-AI akukafinyeleli esicongweni; ivamise ukwehliselwa ebudlelwaneni bethuluzi lomsebenzisi nje kunokuba ubambiswano lokuhlanganyela.

I-AI manje ihlanganisa amandla okucabanga ayinkimbinkimbi kanye nezenzo ezizimele, iyenze ibe ibhizinisi elisebenzayo kunokuba ithuluzi elingenzi lutho elibhekelela izidingo zabantu kuphela. Ushintsho lubheke ekusebenzelaneni ngokubambisana lapho abantu ne-AI behlanganyela engxoxweni yezindlela ezimbili, okuvumela ukuthathwa kwezinqumo nokwenza imisebenzi okufanele kwabelwane ngayo. Ngokwenza kanjalo, abantu bangabuyekeza futhi balungise izimpendulo ze-AI, bacwenge izinjongo zabo ngokusekelwe emininingwaneni ehlinzekwa yi-AI. Le paradigm entsha ingase iholele ekuchazweni kabusha kokuhlukaniswa kwezabasebenzi phakathi kwabantu nemishini ehlakaniphile, ikhulise amandla akho kokubili. 

Phakathi kwentuthuko ephawulekayo kulesi sizinda kukhona izinhlobo zezilimi ezinkulu (LLMs). I-OpenAI's ChatGPT, isibonelo, ingacubungula futhi ikhiqize umbhalo ofana nomuntu ngokusekelwe olwazini olunikezwa yona, inikeze imininingwane ebalulekile, okusalungiswa, noma iziphakamiso ezingonga isikhathi futhi zigqugquzele ukucabanga okuhlakaniphile. Ngaleso sikhathi, ijeneretha yesithombe i-DALL-E 3 ingakha izithombe ezingokoqobo, amahlaya, ngisho nama-meme. Inkampani yezokubonisana i-Deloitte ihlanganisa lobu budlelwano obuguqukayo ngokuphakamisa ukuthi abantu manje sebengakwazi ukusebenza emishinini, ngemishini, kanye nasemishinini, ikhomba ngekusasa lapho ukusebenzisana kwethu ne-AI kuxhumene kakhulu futhi kucebisa ngokufanayo.

Umthelela ophazamisayo

U-Tom Smith, umnikazi wokuqalisa we-AI, waqala ukuhlola umklami we-software we-OpenAI, i-Codex, futhi wathola ukuthi ukusetshenziswa kwayo kudlula amandla okuxoxa nje. Njengoba ejula, wathola i-Codex inekhono lokuhumusha phakathi kwezilimi ezihlukene zokuhlela, ebonisa ukuthi kungenzeka kube ngcono ukusebenzisana kwekhodi nokwenza lula ukuthuthukiswa kwenkundla ehlukahlukene. Okuhlangenwe nakho kwakhe kwamholela esiphethweni sokuthi esikhundleni sokubeka usongo kubahleli bezinhlelo abangochwepheshe, ubuchwepheshe obufana ne-Codex bungasebenza njengezisusa zokukhiqiza kwabantu. 

Emkhakheni wezokunakekelwa kwempilo, ukusetshenziswa kwe-AI kuveza indlela ethembisayo yokwengeza ukunemba kokuxilonga nokusebenza kahle kodokotela. Nakuba i-AI ingase intule ukuthinta okunembile kodokotela abangabantu, imi njengenqolobane yedatha yamacala adlule nemilando yokwelashwa, elungele ukufinyelelwa ukuze yazise izinqumo zomtholampilo ezingcono. Usizo ludlulela ekulawuleni amarekhodi eziguli zezokwelapha kanye nomlando wemithi, umsebenzi obaluleke kakhulu kodwa odla isikhathi kodokotela abamatasa. Ngale kwalezi zinsiza eziqondene nomsebenzi othile, ukwethulwa kwamarobhothi asebenzisanayo anamandla e-AI noma amakhobhothi kuzindawo zokukhiqiza noma zokwakha kwazisa ukwehliswa okukhulu kwezingozi zokulimala.

Ngaleso sikhathi, ikhono le-AI lokwenza imephu, ukwenza kahle, kanye nokwengamela ukuhamba komsebenzi okuyinkimbinkimbi kuma njengobufakazi bendima engaba khona ekuthuthukiseni ukusebenza kahle kokusebenza. Izinhlelo zokusebenza ezixube izimboni, kusukela ekuthuthukisweni kwesofthiwe kuya ekunakekelweni kwezempilo kanye nokusebenza kwezimboni, zigcizelela ukushintshela ekusebenzisaneni okuthe xaxa kwemishini yabantu. Njengoba ama-LLM kanye nombono wekhompiyutha kukhula kangcono futhi kwanda, kungase kuholele ekucabangeni kabusha kwendima yomuntu ngamunye kodwa nasekuguquleni okubanzi kwenhlangano.

Imithelela yomsebenzi we-AI-augmented

Imithelela engaba khona yomsebenzi we-AI-augmented ingabandakanya: 

• Ukukhuphuka kwe-AI njengomsizi obalulekile ezizindeni ezihlukahlukene, okuhlanganisa abasizi ababonakalayo, ama-chatbots, nama-coding aid, okunikela ekuthuthukisweni ukusebenza kahle nokukhiqiza kuyo yonke imikhakha eminingi.
• Ukuqaliswa kwezinhlaka zokulawula ezizungezile ubudlelwano bokusebenza phakathi kwabantu ne-AI, okuchaza ububanzi nemikhawulo yemisebenzi, ekhuthaza indawo yokusebenza echazwe kahle kanye nokucaca ekuhlukaniseni indima.
• Ukusatshalaliswa kwe-AI ezindimeni zokuhlaziya idatha, ukuletha imininingwane ebalulekile kwezezimali nezimboni kanye nokusiza ekwakhiweni kwamasu aqhutshwa yidatha kanye nezinqubo zokuthatha izinqumo ezinolwazi.
• Ukuthuthukiswa kobuchwepheshe obusiza kakhulu kumalebhu e-AI, okuthuthukisa amandla e-AI njengabalingani beqembu ababalulekile, ikakhulukazi ekunakekelweni kwezempilo, okungaholela ekunakekelweni okungcono kweziguli nokusebenza kahle kwesibhedlela.
• Ukushintshela ekufundeni okuqhubekayo kanye nokuthuthukiswa kwamakhono phakathi kwabasebenzi ukuze bahambisane nentuthuko ye-AI, okugqugquzela isiko lokufunda impilo yonke nokuzivumelanisa nezimo.
• Ushintsho olungase lube khona kumamodeli ebhizinisi njengoba izinkampani zingase zisebenzise i-AI ukuze zehlise izindleko zokusebenza, zithuthukise ukusebenzelana kwamakhasimende, futhi zinikeze izinsizakalo noma imikhiqizo emisha, okudale ushintsho kumamodeli agxile kakhulu kudatha.
• Izinzuzo zezomnotho ezivela ekusebenzeni kahle okuthuthukisiwe kwe-AI zingaholela ekongeni izindleko kubathengi, okungenzeka kuhumushele emananini aphansi ezimpahla namasevisi kanye nezinga eliphezulu lokuphila.
• Ushintsho lwezepolitiki njengoba ohulumeni bebandakanya i-AI ukuze ihlaziye inqubomgomo engcono, ukulethwa kwezinsiza zomphakathi, nokwenza izinqumo ezinolwazi, nakuba kunezinselele eziphathelene nobumfihlo bedatha nokucatshangelwa kwezimiso zokuziphatha.
• Izinzuzo ezingaba khona zezemvelo njengoba i-AI ingasiza ekuthuthukiseni ukwabiwa kwezinsiza, ekwehliseni ukumosha nokufaka isandla emikhubeni yokusebenza esimeme kakhulu ezimbonini.

Imibuzo okufanele icatshangelwe

• I-AI ingayithuthukisa kanjani futhi imisebenzi yabantu?
• Yimiphi imikhawulo engaba khona yokusebenza ngamasistimu e-AI?