- Ukufinyelela okuvulekile kumamodeli nedatha kuvumela ama-SME kanye namaphrofayili angewona awezobuchwepheshe ukuthuthukisa izixazululo ze-AI ezincintisanayo.
- Ukwehlukanisa abadali bobuchwepheshe kubalulekile ekunciphiseni ukucwasa kwe-algorithmic kanye nokuthuthukisa ukulingana komphakathi.
- Ukufakwa kwe-AI ekhiqizayo kusebenza njengesikhuthazo sokwandisa umkhiqizo kanye nobuciko emsebenzini.
Uma sikhuluma ngokushintsha ubuhlakani bokwenziwa bube yintando yeningi, asibhekiseli ohlelweni lokuvota, kodwa sibhekisela ekuvuleni iminyango yentuthuko ukuze kungabi yiklabhu ekhethekile yabambalwa abakhethiwe. Empeleni, kumayelana nokuqinisekisa ukuthi ikhono lokudala nokusebenzisa la mathuluzi aligcini kumaseva ezinkampani ezinkulu zobuchwepheshe, kodwa kunalokho lifinyelela noma ubani onomqondo omuhle, kungakhathaliseki amakhono abo okuhlela, ngisho nokuvumela ukusetshenziswa kobuhlakani bokwenziwa ngaphandle kokudala i-akhawunti ukuze kuqedwe izithiyo zokufinyelela.
Lokhu kuthutha kubalulekile ngoba uma sivumela abambalwa kuphela ukuthi baqondise umkhumbi, sizibeka engcupheni yokudala ukwahlukana kwedijithali okungenakuqhathaniswa . Ngokwenza ubuchwepheshe butholakale kalula, senza ukusungula izinto ezintsha kuchume kusuka ezinhlangothini ezingalindelekile, okwenza i-AI ibe yinjini yentuthuko ebandakanya wonke umuntu hhayi ithuluzi lokulawula labambalwa abanelungelo.
Kusho ukuthini ngempela ukwenza i-AI ibe yentando yeningi?
Lo mqondo unezingqimba eziningana. Ngakolunye uhlangothi, kukhona isici sobuchwepheshe: ukunikeza abacwaningi kanye nososayensi bedatha izinsiza zokubala ezingabizi noma zamahhala, kususa isithiyo sezomnotho esinganqotshwa yizinkampani ze-Big Tech kuphela njengamanje kalula. Kodwa kukhona ungqimba olujulile: ukufakwa kwabasebenzisi abangebona ochwepheshe . Lokhu kuhilela ukudala izixhumi namathuluzi avumela umuntu ongenalo ulwazi ngama-algorithms ayinkimbinkimbi ukuthi aklame izinhlelo zakhe zokusebenza ze-AI.
Lesi simo sifana kakhulu nalokho okwenzeka ngokusetshenziswa kwemininingwane ngentando yeningi ezinkampanini, lapho noma yisiphi isisebenzi, kungakhathaliseki ukuthi sivelaphi, manje singasebenzisa isayensi yedatha ukwenza izinqumo ezisekelwe ebufakazini, hhayi nje ukuthembela ekuqondeni. Empeleni, ukushintsha kusuka kumodeli "yochwepheshe abavalelwe embhoshongweni wezinyo lendlovu" kuya kumodeli yokudala ngokubambisana okuvulekile nokuhlanganyela.
Indima ebalulekile yedatha evulekile kanye nomthombo ovulekile
Ukuze i-AI ifinyeleleke kalula, idinga uphethiloli, nokuthi uphethiloli uyidatha. Yilapho izinhlelo ze -Open Data zingena khona , njengalezo ezikhuthazwa yi-European Union ngama-portals afana ne-data.europa.eu. Uma idatha isesidlangalaleni, izinhlangano zingakha izixazululo ezintsha ngaphandle kokutshala izigidi ekuqoqeni ulwazi kusukela ekuqaleni, okuthuthukisa umnotho wedijithali futhi kuqinisekiswe ukuthi izinzuzo zobuchwepheshe zisatshalaliswa ngokulinganayo.
Isibonelo esiyinhloko yiphrojekthi ye-AI4EU, edala uhlelo lokusebenza ngokubambisana ukuze kuthuthukiswe konke kusukela ekuxilongweni kwezokwelapha kuya ekukhiqizweni okuhlakaniphile. Okungeziwe kulokhu ukwanda kwamamodeli omthombo ovulekile , njenge-Llama kaMeta noma i-Gemma kaGoogle. Lawa mathuluzi avumela noma yimuphi unjiniyela ukuthi enze ngezifiso i-AI ngokwezidingo zakhe, okukhuthaza isiko lokuhlola okusheshayo okungaxhomekile kumalayisense obunikazi abizayo.
Umthelela emhlabeni womsebenzi kanye ne-AI yokukhiqiza
I-AI ekhiqizayo ibe yimbangela yangempela yalolu shintsho. Ngokuvumela ukusebenzisana okulula ngolimi lwemvelo, kwenze ubuchwepheshe bube lula kakhulu ukuyisebenzisa . Akusekho sidingo sokubhala ikhodi eyinkimbinkimbi; udinga nje ukwazi ukuthi ucela kanjani izinto. Lokhu kuguqula yonke imikhakha: emkhakheni wezomthetho, kucutshungulwa amanani amakhulu emibhalo ngemizuzwana, kanti kwezempilo, ukuxilongwa kuyathuthukiswa, uma nje i-AI isebenza njengokusekela udokotela hhayi njengokufaka esikhundleni.
- Usizo lomuntu siqu: Ukuphathwa kwe-ajenda kanye nokubhala umbhalo ngokushesha.
- Imfundo yomuntu siqu: Ukwenza lula imiqondo enzima kumfundi.
- Ukusekelwa okuzenzakalelayo: Izimpendulo ezisheshayo zemibuzo evame ukubuzwa ezinkampanini.
- Ukukhula kokukhiqiza: Ukuqeda imisebenzi ephindaphindwayo ukuze kugxilwe ekudaleni.
Abaningi besaba ukuthi lokhu kuzosho ukuphela kwemisebenzi yabo, kodwa umbono onethemba usikisela ukuthi, njengaseNguquko Yezimboni, simane siphakamisa izinga lokucabanga okucashile . Ngokunikeza imishini imisebenzi ejwayelekile, abantu bangagxila emizamweni enamasu amaningi negxile kubantu, bafinyelele ulwazi olwandisiwe oluthuthukisa amakhono ochwepheshe.
Ukuziphatha, ukucwasa kanye nokulwa nohulumeni abazimele
Enye yezinkinga ezibaluleke kakhulu ukulwa nobandlululo lwe-AI. Lapho ubuchwepheshe buthuthukiswa njalo ngamaqembu afanayo abantu, ubandlululo lomphakathi luyangena kuma-algorithms, kunciphise amaqembu angamelwe kahle. Ngokuthuthukisa intuthuko ngentando yeningi, sifaka imibono eyahlukahlukene esiza ekudaleni izinhlelo ezifanele nezilinganayo. Kuyindaba yokusinda komphakathi ukuvimbela ubuchwepheshe ekuqhubekiseleni phambili ubandlululo.
Ngaphezu kwalokho, ukuvula imakethe kuyivimbela ekubeni yi-oligopoly. Uma ama-SME ekwazi ukufinyelela amathuluzi anamandla, angancintisana ngokulinganayo nezinkampani ezinkulu. Lokhu akugcini nje ngokuthuthukisa umnotho kodwa futhi kwenza ukuziphatha kube yinkinga ebalulekile, njengoba ukusetshenziswa kabanzi kwe-AI kuhilela ukuqapha okuqinile kakhulu kweqembu.
Izinselele kanye nezingozi endleleni
Akuyona yonke into enhle. Ukuvulwa okuphelele kwedatha kungqubuzana ngqo nelungelo lobumfihlo. Ngenkathi iYurophu ine-GDPR, amanye amazwe anemithetho ehlukene, okwenza kube nzima ukudalwa kwendinganiso yokulawula yomhlaba wonke . Ukuze i-Open Data isebenze, kumele kube nesiqinisekiso sangempela sokuthi ubumfihlo abuthengiswa kumuntu onikela kakhulu; idatha iyona ebaluleke kakhulu, kodwa ayikwazi ukunyathela inkululeko yomuntu ngamunye futhi idinga ukuphathwa kahle kwesimo sokuphepha kwedatha.
Ngakolunye uhlangothi, ukuthandwa kwala mathuluzi kuthwala izingozi ezibonakalayo, njengokwanda kokuhlaselwa kwe-cyber okuyinkimbinkimbi noma ukusetshenziswa kwe-AI ukudala ulwazi olungesilo iqiniso kanye nezenzo ezingekho emthethweni. Kuyibhalansi ebucayi phakathi kokufuna wonke umuntu abe nokufinyelela olwazini nokuvimbela lokho kufinyelela ukuthi kusetshenziselwe ukuvula iminyango okufanele ivalwe , ikakhulukazi lapho kubhekene nezingozi zokuphepha kwe-cyber esikhathini sedijithali.
Ukushintshela ekuhlakanipheni kokwenziwa okufinyelelekayo kuguqula isisekelo somphakathi, kuqinisekisa ukuthi ubuhlakani kanye nokusebenza kahle akusaxhomekile kusayizi wesabelomali senkampani. Ngokuhlanganisa idatha evulekile, amamodeli okubambisana, kanye nombono wokuziphatha olwa nokubandlulula, kwakhiwa ikusasa lapho ikhono lokuqamba izinto ezintsha liyilungelo, hhayi ilungelo, okwenza ubuchwepheshe busebenze ngempela ukuthuthukisa ikhwalithi yokuphila kwawo wonke umuntu.