Ukugcwala kwe-AI Kuyinto Yangempela. Nakhu Okwenzeka Ebuchosheni Bakho.

Kukhona okwashintsha cishe ngo-2024. Inani labantu abachaza uhlobo oluthile lwe-ukukhathala kwengqondo — okungeyona ngempela i-burnout, okungeyona ngempela ukugcwala kolwazi, kodwa into ephakathi kwalokhu kokubili — lenyuka kakhulu. Izincazelo ziyefana: umuzwa wokugcwala kwengqondo, ubunzima bokucabanga ngokuzimela, ukwehla kokubekezelela ukungaqiniseki, nokuthembela okungakhululekile kumathuluzi e-AI okuthi ngandlela thile kwenze ukucabanga kuzwakale kunzima kakhulu, hhayi kube lula.

Lokhu akukhona ukwesaba ubuchwepheshe noma ukulangazelela izikhathi zakudala. Kubonisa izindlela zangempela zobuchopho ezibikezelwayo uma sibheka lokho esikwaziyo ngendlela ubuchopho bucubungula ngayo ulwazi, bulondoloza kanjani ukusebenza kwengqondo, futhi busabela kanjani osizweni oluhlale lutholakala. Nakhu okwenzekayo — nokuthi ubufakazi busikisela ukuthini mayelana nokubhekana nakho.

Indawo entsha yokucabanga

Kusukela ngo-2025, abasebenzi bolwazi ezimbonini eziningi basebenzisa amathuluzi e-AI — i-ChatGPT, i-Copilot, i-Gemini, i-Claude, namanye amathuluzi akhethekile amaningi — izikhathi eziningi ngosuku. Umsebenzisi ojwayelekile uthumela imibuzo emashumini ambalwa ngosuku. Okuqukethwe okukhiqizwe yi-AI manje sekuyingxenye enkulu yokusalungiswa kwama-imeyili, izifinyezo zemihlangano, ikhodi, imibhalo ebhaliwe, nemiphumela yokusesha.

Lena iyindawo entsha ngempela yokucabanga. Ngokokuqala ngqa emlandweni, ingxenye enkulu yomsebenzi wokucabanga owayekade udinga umzamo oqhubekayo — ukubhala okusalungiswa, ukufingqa, ukusesha, ukuhlela, nokulungisa amaphutha — manje ingadluliselwa cishe ngokushesha. Umbuzo uwukuthi lokhu kwenzani ebuchosheni ngokuhamba kwesikhathi.

Impendulo, ngokusekelwe kulokho esikwaziyo ngokusebenza kwengqondo nomthwalo wengqondo, iyinkimbinkimbi kakhulu kunenguqulo enethemba (i-AI yandisa ikhono lomuntu) noma le enethemba elincane (i-AI inciphisa ikhono lomuntu). Kokubili kuyenzeka, ngezindlela ezihlukene, ezinhlotsheni ezihlukene zokucabanga.

Kungani amathuluzi e-AI engase andise umthwalo wengqondo esikhundleni sokuwunciphisa

Umcabango ovamile uthi ukwelula umsebenzi ku-AI kunciphisa umthwalo wokuqonda. Ngandlela-thile kunjalo. Kodwa ucwaningo lwethiyori yomthwalo wokuqonda (Sweller, 1988) luhlukanisa phakathi komthwalo ongaphakathi (ubunzima bomsebenzi uqobo), umthwalo ongadingekile (izimfuno zokuqonda ezidalwa indlela umsebenzi owethulwa ngayo), kanye nomthwalo owakha ukuqonda (umzamo wokuqonda owakha ukuqonda namakhono).

Amathuluzi e-AI ikakhulukazi anciphisa umthwalo ongaphakathi nowakha ukuqonda — enza umsebenzi obezodinga umzamo futhi ulethe ukufunda. Kodwa evame ukwandisa umthwalo ongadingekile: ukuhlola ukuthi okukhiqizwe yi-AI kunembile yini, ukuhlanganisa izingxenye eziningi ezikhiqizwe yi-AI zibe umsebenzi ohlangene, ukuphatha umongo kuwo wonke amathuluzi ahlukahlukene, ukunquma ukuthi yini ongayiphathisa i-AI nokuthi yini okufanele uyenze wena mathupha, nokugcina iso ezinqubweni ze-AI. Lezi yizimfuno ezintsha zokuqonda ezazingekho ngaphambili.

Umphumela kubantu abaningi uyisiphithiphithi esingajwayelekile: okukhiqizwayo okuningi, kodwa ukukhathala okukhulu kwengqondo. Umsebenzi uzizwa ulula ezikhathini ezithile, kodwa ukhathaza kakhulu uma kubhekwa usuku lonke. Abaningi bakuchaza lokhu njengokuthi ukukhathala kokwenza izinqumo: uchungechunge olude lwezinqumo ezincane luzwakala lukhathaza njengalezo ezimbalwa ezinkulu. Ukuthi kungani lokho kwenzeka kuseyimpikiswano ocwaningweni, kodwa okuhlangenwe nakho kokukhathala kwengqondo ngemva kwamahora ezinqumo ezincane eziqhubekayo kuyinto evamile.

Umsebenzi osizwa yi-AI wandisa kakhulu inani lezinqumo ezincane: yamukela lesi siphakamiso, wenqabe leso, buyekeza le ndima, buza kabusha lowo myalelo. Isinqumo ngasinye sishesha, kodwa umthwalo ohlangene kuwo wonke amakhulu ezinqumo ezincane zansuku zonke mkhulu kakhulu.

Indida ye-automation nokuncipha kwamakhono

Kukhona into eyaziwa kahle ocwaningweni lwe-ergonomics nezici zabantu ebizwa nge-automation paradox noma i-ironies of automation (Bainbridge, 1983): lapho i-automation iphethe umsebenzi ojwayelekile, abasebenzisi abangabantu bagcina bengasazijwayezi kahle amakhono adingekayo ezimeni ezingajwayelekile. Abashayeli bezindiza abathembele ku-autopilot ezindizeni zabo eziningi babonisa amakhono aphansi okuphatha ngezandla. Odokotela be-radiology abasebenzisa amathuluzi okuxilonga e-AI babonisa amazinga aphansi okuthola lapho i-AI ingakutholi okuthile.

Umfutho ofanayo usebenza nasemsebenzini wokuqonda. Ukubhala, ukucabangisisa ngezinkinga, ukuhlela, nokuhlanganisa imibono ngobuciko ngamakhono akhula ngokuzijwayeza futhi ayaphela uma kungabikho lokho. Uma i-AI ithatha uhlaka lokuqala, ukuhlanganisa, noma ukucabanga okuhlelekile, lezo zindlela zengqondo azitholi umsebenzi owanele. Okukhathazayo akukukhulu — akusho ukuthi abantu bayakhohlwa ukucabanga — kodwa ukuncipha kancane kancane kwamakhono asetshenziswayo kuletha ushintsho lwangempela emandleni engqondo nasekuzethembeni ngokuhamba kwesikhathi.

Lokhu kuhlukile kunokumane ungabe usajwayele ukubhala ngesandla. Ukubhala — ukwakha umbhalo uqobo, hhayi ukutshela omunye noma ukuhlela — kusebenzisa inkumbulo esebenzayo, ukuhleleka kwemibono, nokucabanga ngokuzindla ngendlela ukubuyekeza nokuhlela umphumela we-AI okungakukopishi ngokugcwele. Umsebenzi wengqondo wokwakha umbhalo unika okusha; umsebenzi wengqondo wokuhlela uhlola. Kokubili kunenani; akushintshiseki.

Ungqimba lokuhlakazeka kokunaka

Ngaphandle komthwalo wengqondo nokuwohloka kwamakhono, amathuluzi e-AI engeza olunye ungqimba lokuhlakazeka kokunaka endaweni esivele igcwele ukuphazamiseka. Yonke i-interface yengxoxo, umsizi wokubhala ikhodi, kanye nethuluzi lokubhala kungaba umthombo wokuphazamiseka — umbuzo wokukhiqiza, umphumela wokuwuhlola, nokulandelayo ukuze kuwulungiswe.

I-prefrontal cortex, ephethe ukunaka okuqhubekayo nomsebenzi wokulawula nokwenza izinqumo, ayihlukanisi phakathi kokuphazamiseka okuvela kubantu nokuphazamiseka okuvela ku-AI. Ucwaningo ngezindleko zokushintsha phakathi kwemisebenzi (Rubinstein et al., 2001) lukhombisa ukuthi ukushintsha phakathi kwemisebenzi — ngisho okwesikhashana — kubeka umthwalo wengqondo ohlanganayo osuke ukhula usuku lonke. Ukwengeza amathuluzi amaningi e-AI endaweni yomsebenzi esivele ihlakazekile kwenza lokhu kube kubi nakakhulu, kanti abantu abaningi abazi ukuthi ukukhathala kwabo kuvela kuphi.

Okuxakayo wukuthi amathuluzi e-AI avame ukusetshenziswa ngenhloso yokunciphisa umthwalo wengqondo. Kodwa ukungezwa kwamathuluzi amasha, ama-interface amasha, namaphuzu amasha okwenza izinqumo kuvame ukukhulisa umthwalo ophelele wengqondo kunokuwunciphisa — okungenani kuze kube yilapho la mathuluzi esehlelwe kahle ngokwanele ukuze aphathwe ngokuzenzakalela kunokucatshangelwa ngokucophelela.

Ongakwenza ngempela ngakho

Hlukanisa phakathi kokunciphisa nokudlulisela omunye umsebenzi. Akusilo lonke usizo lwe-AI oluba nomphumela ofanayo engqondweni. Ukusebenzisa i-AI ukuphatha imisebenzi enenani eliphansi ngempela (ukufometha, umbhalo ojwayelekile, ukusesha okuphindaphindayo) kunciphisa umthwalo ngaphandle kwezindleko ezinkulu zamakhono. Ukusebenzisa i-AI emakhonweni okucabanga ofuna ukuwagcina — ukubhala okuyinkimbinkimbi, ukuhlaziya okunengqondo, ukuxazulula izinkinga ngokudala — kunciphisa umthwalo kodwa kubiza ikhono osulijwayeze. Ukucabangisisa kahle ngokuthi ukusetshenziswa okuthile kungena kusiphi isigaba kuyisiqalo.

Vikela isikhathi sokucabanga ungasekelwa lutho. Amakhono engqondo asengcupheni enkulu ngenxa yokudlulisela umsebenzi ku-AI yilawo kanye afaneleka kakhulu ukuwagcina: ukucabanga ngokuzimela, ukubhala isikhathi eside, nokuhlela izinto eziyinkimbinkimbi. Ukuzilolonga ngamabomu kulawa makhono ngaphandle kosizo lwe-AI — isikhathi esibekiwe, emisebenzini ebalulekile — akusona ukwesaba ubuchwepheshe; kuwumqondo ofanayo nomdwebi ohlinzayo ozilolonga amasu okusebenza ngesandla noma sekukhona amathuluzi erobhothi.

Yehlisa inani lamathuluzi e-AI asetshenziswa okwamanje. Izindleko zengqondo zokuphatha izixhumi ze-AI ezinhlanu ezihlukene ziphakeme kunokuphatha eyodwa ekhethwe kahle. Ukusabalala kwamathuluzi e-AI kudala inkinga efanayo yokuhlukana kokunaka njengokusabalala kwama-app nezaziso. Ukuwahlanganisa kunciphisa umthwalo wengqondo ongadingekile.

Sebenzisa izimiso ezifanayo nalezo zokuphila kahle kwedijithali okujwayelekile. I-AI overload iwuhlobo oluthile lwenkinga ebanzi i-Unwire eyibhekayo: izindawo zengqondo ezifuna okungaphezu kwalokho ezikunikezayo, ukuhlakazeka kokunaka okuqeda amandla ngaphandle kokuwabuyisela, kanye nesikhuthazo esihlala sitholakala esenza ukuzihlukanisa kubonakale kungenzeki. Izinyathelo ezisizayo ziyefana: izikhathi ezichaziwe zokungasebenzisi, ukuhlela indawo okunciphisa ukusebenzisana ne-AI okusemoyeni, nesikhathi esijwayelekile sokuphumula lapho ubuchopho bungaqondiswa khona.

Umbono wesikhathi eside

Ukuhlanganiswa kwe-AI emsebenzini osekelwe olwazini ngeke kubuyele emuva. Umbuzo uwukuthi abantu bazoyiphatha kanjani indawo eyakhelwe, ngenhloso, ukuba ikhange kakhulu futhi ibe nokuphazamiseka okuncane kakhulu — nokuthi lokho kwenzani empilweni yengqondo eminyakeni eminingi, hhayi ezinsukwini.

Ucwaningo ngenhlalakahle yedijithali ngokubanzi luphakamisa ukuthi abantu abakwazi kahle ukuzulazula ezindaweni zobuchwepheshe obuphezulu akubona labo abasebenzisa amathuluzi amaningi kakhulu noma ambalwa kakhulu, kodwa yilabo abagcina ukulawula ngenhloso ezimweni zokunaka kwabo: ukuthi bayatholakala nini, nini bengatholakali, yini abayiphathisa abanye, nokuthi yini abayivikela.

Lokho kwenza ngamabomu kuba nzima ukukugcina endaweni egcwele i-AI kunasesimeni esigcwele ama-smartphone kuphela. Kodwa umgomo oyisisekelo — vikela amandla akho okucabanga ngokuzimela, buyisela ubuchopho esimweni sazo esijwayelekile ngokuphumula kweqiniso, futhi wakhe indawo okuzungezile kunokuphendula kuyo — uhlala ufana.

Sources

1. Sweller, J. (1988). Cognitive load during problem solving: effects on learning. Cognitive Science, 12(2), 257–285.
2. Bainbridge, L. (1983). Ironies of automation. Automatica, 19(6), 775–779.
3. Rubinstein, J.S., et al. (2001). Executive control of cognitive processes in task switching. Journal of Experimental Psychology: Human Perception and Performance, 27(4), 763–797.
4. Wiehler, A., et al. (2022). A neuro-metabolic account of why daylong cognitive work alters the control of economic decisions. Current Biology, 32(16), 3564–3575.
5. Baumeister, R.F., et al. (1998). Ego depletion: is the active self a limited resource? Journal of Personality and Social Psychology, 74(5), 1252–1265.

Indlela esicwaninga futhi sibuyekeze ngayo lokhu