Wake wazibuza ukuthi kwathatha isikhathi esingakanani ukuthi lelo philisi oliphuze ekuseni lifinyelele ezandleni zakho? Okokuqala, benza ucwaningo olunzulu ezindaweni zokucwaninga. Bese belihlola ezilwaneni. Kodwa ngaphambi kokuba livunyelwe ukusetshenziswa kubantu, kufanele balihlole kubantu. Lokhu yilokho esikubiza ngokuthi ukuhlolwa kwezokwelapha. Kuyinqubo eyinkimbinkimbi kakhulu futhi ebiza kakhulu.
Kalula nje, kuyini ukuhlolwa kwezokwelapha?
Kalula nje, isivivinyo somtholampilo sisebenza kanje: Abacwaningi baqasha iziguli ezinesifo esiqondiswe yilo muthi osanda kuthuthukiswa. Laba bavolontiya bahlukaniswa ngokungahleliwe babe amaqembu amabili.
1. Iqembu lokuqala linikezwa umuthi omusha ukuze uhlolwe.
2. Iqembu lesibili , i-"Control Group", linikezwa i-placebo engenaso isithako esisebenzayo futhi ebukeka ifana ncamashi nomuthi. Lokhu sikubiza ngokuthi "i-Placebo".
Uma, ekupheleni kwalolu cwaningo, izimpawu zalabo abathathe umuthi omusha sezithuthuke kakhulu kunalabo abathathe i-placebo, khona-ke lokhu kufakazela ukuthi umuthi omusha uyasebenza.
Enye yezinselele ezinkulu esivivinyweni esinjalo ukuthola iziguli ezanele ezifanele kahle isivivinyo. Ngezinye izikhathi ngisho nodokotela abazi ukuthi yiziphi izivivinyo ezifanele iziguli zabo. Futhi, izici zeziguli ezifuna ukuhlanganyela zingase zingahambisani nezidingo zesivivinyo. Kodwa-ke, ubuhlakani bokwenziwa (AI) bungenza lo msebenzi ube lula kakhulu.
Hlangana "noDigital Twin" wakho
Cabanga ukuthi unemodeli yakho yekhompyutha, okungukuthi, ikhophi yakho yedijithali. Yilokho okubizwa ngokuthi "iDigital Twin". Lawa "amaModeli eKhompyutha" nje alingisa into noma uhlelo lwangempela kukhompyutha. Kodwa uma ebhekwa ngezinombolo, aziphatha ngendlela efanayo nento yangempela. Isibonelo esihle kakhulu yilapho ithangi lomoya-mpilo emkhathini we-Apollo 13 liqhuma, i-NASA yasiza ekulungiseni nge-"Digital Twin" yesikhala esifanayo esakhiwe eMhlabeni.
Manje, ososayensi bangakha "iDijithali Twin" yomuntu, uma benikezwe "idatha" eyanele. Lokhu kusebenzisa ubuchwepheshe bobuhlakani bokwenziwa obubizwa ngokuthi "Ukufunda Komshini". Kalula nje, lokhu kuhilela ukunikeza uhlelo lwekhompyutha idatha eningi bese luyivumela ukuthi ifunde kuyo.
Amawele edijithali eziguli adalwa ngokuqeqesha lawa mamodeli e-Machine Learning esebenzisa idatha evela ezivivinyweni zomtholampilo zangaphambilini kanye namarekhodi ezokwelapha esiguli ngasinye. Imodeli ingabe isibikezela ukuthi impilo yesiguli izoshintsha kanjani phakathi nokuhlolwa uma inikezwa i-placebo.Okusho ukuthi, iqembu lokulawula elenziwe ngekhompyutha (Iqembu Lokulawula Elilingiswayo) lenzelwe leso siguli esifanayo.
Manje ake sibone ukuthi lokhu kusebenza kanjani. Ake sithi othile ogama lakhe lingu-"Samadhi" ukhethwe ukuba abe seqenjini elithola umuthi omusha. "Iwele ledijithali" likaSamadhi, imodeli yekhompyutha, liseqenjini lokulawula. Imodeli ibikezela ukuthi kuzokwenzekani uma uSamadhi enikwa "i-Placebo" esikhundleni somuthi wangempela. Ekugcineni, umehluko phakathi kwempendulo yangempela kaSamadhi kumuthi nalokho imodeli eyabikezela ukuthi kuzokwenzeka uma ethatha "i-Placebo" iyisilinganiso sokuthi umuthi usebenza kangakanani kuSamadhi.
Enye yezinzuzo ezinkulu zale ndlela ukuthi cishe wonke umuntu ohlanganyela esivivinyweni somtholampilo unethuba lokuthola umuthi omusha, ngoba "i-placebo" inikezwa iwele lakhe ledijithali.
Ake sibheke leli thebula ukuze sibone ngokucacile izinzuzo zokusebenzisa ``Amawele Edijithali'' esikhundleni samaqembu okulawula avamile.
| Isici | Indlela yendabuko | Indlela Yedijithali Yamawele |
|---|---|---|
| Iqembu Lokulawula | Kunabantu bangempela. Bathola (i-Placebo). | Iqukethe amamodeli ekhompyutha (amawele edijithali). |
| Inzuzo yeziguli | Kunethuba elingu-50% lokuthola i-placebo. Lokho kusho ukuthi ungawuyeka umuthi omusha. | Abantu abaningi banethuba lokuzama umuthi wangempela. |
| Inani labavolontiya elidingekayo | Kudingeka abantu abengeziwe (kuwo womabili amaqembu). | Ukuhlolwa kungenziwa nenani elincane labantu. |
| Isikhathi kanye nezindleko | Kuphezulu kakhulu. | Kungancishiswa kakhulu. |
Ngisho noma lobu buchwepheshe buhle, azikho yini izinkinga?
Nakuba lobu buchwepheshe buthembisa kakhulu esikhathini esizayo, abukakasetshenziswa kabanzi. Kungase kube nezizathu ezinhle zalokhu. Njengoba uDkt. Daniel Neal, uchwepheshe we-"(Machine Learning)," esho, la mamodeli athembele emininingwaneni eminingi. Futhi, kunzima kakhulu ukuthola idatha esezingeni eliphezulu kubantu ngabanye.
"Abantu bavame ukuzibika ngokwabo ulwazi mayelana nezinto ezifana nokudla kwabo kanye nemikhuba yabo yokuzivocavoca. Abantu abakhulumi iqiniso ngaso sonke isikhathi. Bavame ukweqisa inani lokuzivocavoca abakwenzayo futhi banciphise inani lokudla okungenamsoco abakudlayo."
Enye inkinga imiphumela emibi engavamile nengalindelekile engabangwa umuthi. Lokhu akuvamile ukumbozwa yimodeli eyenzelwe iqembu lokulawula.
Kodwa ukwesaba okukhulu kukaDkt. Neal ukuthi le modeli yokubikezela isekelwe "ebhizinisini njengenjwayelo." Cabanga ngesenzakalo esikhulu, esingalindelekile njengobhubhane lwe-COVID-19 esishintsha ukuziphatha kwawo wonke umuntu futhi sigulise abantu. "Amamodeli okulawula awakunaki lokhu," usho kanje. Ngemuva kwalokho, ngenxa yalezi zenzakalo iqembu lokulawula elingazange lizinake, imiphumela yokugcina yokuhlolwa ingaba ehluke ngokuphelele.
Ngakho-ke lokhu kungasetshenziswa ngempela? Ikusasa lizosiphathelani?
U-Eric Topol, uchwepheshe ekusetshenzisweni kobuchwepheshe bedijithali kwezempilo, ucabanga ukuthi lo mbono muhle, kodwa isikhathi asikafiki. "Angicabangi ukuthi izivivinyo zemitholampilo zizoshintsha maduze. Ngoba kudinga izingqimba eziningi zedatha, ngale kwamarekhodi ezempilo nje kuphela, njengokulandelana kwe-genome, i-gut microbiome, idatha yezemvelo."
Ngokusho kukaCharles Fisher, umsunguli we-Unlearn.AI, iphayona kulobu buchwepheshe, sekuvele kunezixazululo zezinkinga zobumfihlo bedatha kanye nokubandlulula. "Ubumfihlo bulula. Sisebenza kuphela ngedatha esivele ingaziwa."
Uma ekhuluma ngokucwasa, uthi yize le nkinga ingakaxazululwa ngokuphelele, ayibalulekile emiphumeleni yocwaningo. Ngenxa yokuthi izifundo zenziwa ngokungahleliwe, noma ngabe kukhona ukucwasa okuthile kudatha, akuthinti imiphumela. Kodwa-ke, abanye ochwepheshe abavumelani nalo mbono. Bathi,Ngenxa yokuthi uqhathanisa umuntu wangempela nemodeli yekhompyutha, uma kukhona iphutha kuleyo modeli, lingathinta ngqo imiphumela.
Kodwa-ke, i-Unlearn.AI isivele isebenza nezinkampani zemithi ukuklama izivivinyo zemitholampilo zezifo zezinzwa ezifana ne-Alzheimer's, i-Parkinson's, kanye ne-multiple sclerosis. Bakhethe lezi zifo njengendawo enhle yokuqala ngoba banedatha eningi kunezinye. UFisher uthi uma le ndlela ingasetshenziswa ngelinye ilanga kuzo zonke izifo, inganciphisa kakhulu isikhathi esithathwa yimithi emisha ukuthi ize emakethe.
Umlayezo Wokuya Nawe Ekhaya
- Iwele ledijithali liyimodeli yekhompyutha yesiguli sangempela esisetshenziswa ezivivinyweni zemitholampilo ezihlola imithi emisha.
- Lokhu kuvumela ukuthi lawa mamodeli edijithali asetshenziswe esikhundleni sokusebenzisa abantu bangempela eqenjini lokulawula.
- Inzuzo enkulu yale ndlela ukuthi abahlanganyeli abaningi ocwaningweni banethuba lokuthola umuthi wangempela ohlolwayo.
- Kodwa-ke, kusenezinselele, njengobunzima bokuthola idatha esezingeni eliphezulu kanye nokungakwazi ukwenza amamodeli wezehlakalo ezingalindelekile.
- Nakuba lobu buchwepheshe buthembisa kakhulu ikusasa lezokwelapha, busesebuncane. Uma unemibuzo noma ukungabaza, kubalulekile ukukhuluma nodokotela wakho .
💬 Comments (0)
Awekho amazwana athunyelwe okwamanje. Faka amazwana akho lapha okokuqala ngqa.
Engeza amazwana akho