- I-Apple PARS iyindlela yokuzifundela efunda ukwakheka kwesikhashana kwamasiginali we-EEG ngaphandle kwedatha echasisiwe.
- Le ndlela ihlanganisa i-ear-EEG kanye nama-earphone anelungelo lobunikazi nama-electrode ukukala umsebenzi wobuchopho kusuka endlebeni.
- Amamodeli aqeqeshwe nge-PARS afanayo noma enza kahle kakhulu ezindleleni zangaphambilini emisebenzini efana nokulala, isithuthwane, noma i-EEG engavamile.
- Lobu buchwepheshe bungaholela kuma-AirPods wesikhathi esizayo akwazi ukuqapha impilo yobuchopho nokuphila kahle nsuku zonke.
Umqondo wokuthi ama-headphone angakwazi Lalela umculo wakho futhi, ngesikhathi esifanayo, "lalela" ingqondo yakho Kubonakala sengathi kukhona okuphuma enganekwaneni yesayensi, kodwa i-Apple isivele ivula indlela ngenhlanganisela enamandla kakhulu: izinzwa ezintsha ezingaphakathi ezindlebeni namamodeli athuthukile we ukuhlakanipha okufakelwayoNgemuva kwakho konke lokhu kunendlela ebizwa ngokuthi I-PARS (I-Pairwise Relative Shift), indlela yokuzifundela evumela i-algorithm ukuthi iqonde umsebenzi kagesi wobuchopho ngaphandle kokuncika kochwepheshe abarekhoda idatha mathupha.
Esikhundleni sokugxila kugajethi ethile, ucwaningo lwe-Apple lugxile ekutheni imodeli ye IA kungaba funda ukwakheka kwesikhashana kwezimpawu ze-EEG (i-electroencephalography) bese usebenzisa lolo lwazi emisebenzini efana nokuhlukanisa izigaba zokulala noma ukuthola ukungahambi kahle kwemizwa. Futhi, nakuba ucwaningo lungasho ngokuqondile ama-AirPods, lwengeza kumalungelo obunikazi kanye nezibonelo ezikhomba ikusasa lapho ama-headphone alula angaba uhlobo "lwelabhorethri encane" yokuqapha ubuchopho endlebeni.
Iyini i-Apple PARS (Pairwise Relative Shift) futhi kungani ibalulekile?
Indlela ye-PARS iqhamuke ocwaningweni olwethulwe yithimba le-Apple nabahlanganyeli bezemfundo, ephepheni elamukelwe I-NeurIPS 2025 Foundation Models ye-workshop yobuchopho nomzimbaUcwaningo, olunesihloko esithi "Ukufunda ukwakheka okuhlobene kwamasiginali we-EEG usebenzisa ukuqeqeshwa kwangaphambili kokushintsha kwesihlobo ngokubili," luphakamisa indlela ehlukile yokuqeqesha amamodeli anamasignali e-electroencephalography ngaphandle kokusebenzisa amalebula abantu.
Empeleni, i-PARS iwuhlelo lokusebenzisa ukufunda ukuzigada isetshenziswe ku-EEG. Esikhundleni sokucela izazi zemizwa ukuthi zibonise mathupha ukuthi iyiphi ingxenye yesignali ehambisana nesigaba ngasinye sokulala noma ukuqala kokubanjwa kwesithuthwane, imodeli iqeqeshwa ngedatha engachazisiwe futhi iphoqeleka ukuxazulula inkinga yokwenziwa kodwa ewusizo kakhulu: bikezela ukuthi yiliphi ibanga lesikhathi elihlukanisa izingcezu zesignali ezimbili.
Umqondo oyisisekelo ukuthi uma imodeli ifunda ukulinganisa ukuthi singakanani isikhathi phakathi kwamafasitela amabili e-EEG, ekugcineni iyaqonda isakhiwo somhlaba kanye nokuncika ebangeni elide umsebenzi wobuchopho. Lokhu kamuva kuyivumela ukuthi yenze kangcono emisebenzini yangempela yomtholampilo, njengokuthola amaphethini okulala, ukuhlonza isithuthwane, noma ukubona izimpawu zemoto.
Ababhali bagcizelela ukuthi, ngokungafani nezindlela zakudala zokuzifundela ze-EEG, ezigxile kakhulu kuzo yakha kabusha izingxenye ezifihliwe zesignali (njengama-autoencoder afihliwe, MAE, do), i-PARS igxile ekwakhiweni kwesikhashana okuhlobene. Okusho ukuthi, ayigcini nje ngokugcwalisa "izikhala" zendawo, kodwa ithwebula ukuthi izingcezu ezihlukene zesiginali zihlangana kanjani ngokuhamba kwesikhathi.
Ezivivinyweni ezenziwe, amamodeli asuselwa ku-PARS akhombisa ukuthi ayakwazi ukufanisa noma ukweqa amasu adlule kumabhentshimakhi e-EEG ahlukahlukene, ikakhulukazi uma kutholakala amalebula ambalwa (isimo esivame kakhulu kwezokwelapha). Lokhu kwenza i-PARS ibe inketho ekhangayo kakhulu kunoma iyiphi isistimu efuna ukusizakala ngamavolumu amakhulu amasignali obuchopho ngaphandle kokuthembela ezichasiselweni eziphelele.
Isebenza kanjani i-PARS: ukusuka kumathokheni kuya esilinganisweni sokufuduka okuhlobene
Ukusebenzisa indlela ye-PARS ku-EEG, abacwaningi baklama i-architecture esekelwe Ama-Transformer anezinyathelo ezimbalwa ezibalulekileKonke kuqala ngokucubungula isiginali bese kuyiguqula ibe isethulo imodeli engakwazi ukusiphatha kalula.
Okokuqala, isignali ye-EEG ihlukaniswe yaba amafasitela esikhashana noma “amathokheni”Le nqubo yokwenza amathokheni ivumela ucezu ngalunye lwesignali ukuba lumelwe njengeyunithi lapho i-transformer ingasebenza khona, kufana nendlela umbhalo noma izithombe ezisetshenziswa ngayo. Lawa mathokheni abe esengezwa... ukushumeka kwesikhundla, nakuba ngendlela ethile, ngenxa yokuthi i-PARS idlala ngokunembile imaski kanye nokusetshenziswa kwalezi zikhundla zesikhashana.
Enye yezingxenye ezihlukile zendlela ukusetshenziswa kwe ukushumeka kwendawo okufihliweEsikhundleni sokunikeza imodeli imininingwane yendawo ngqo el tiempo Kuthokheni ngayinye, idatha ethile yendawo ifihliwe noma ishintshiwe. Lokhu kuphoqa isifaki khodi ukuthi cabangela ukwakheka kwesikhashana kokuqukethwengaphandle kokuthembela kuphela kunkomba noma isitembu sesikhathi esicacile.
Inhliziyo ye-PARS pre-Workout iwumsebenzi woku ukulinganisa kokushintsha kwesihlobo ngokubiliImodeli ithola amawindi e-EEG amabili akhishwe ngokungahleliwe ekurekhodeni okufanayo futhi kufanele abikezele ibanga lesikhashana phakathi kwawo. Akukhona nje ngokuqagela ukuthi asondelene noma aqhelelene kangakanani, kodwa mayelana nokufunda imephu eqhubekayo noma ehluzekile ebonisa isikhawu sesikhathi esihlobene.
Kulokhu, a idikhoda enezindlela zokunaka okuphambeneLe ngxenye ibhekisela kulwazi olufakwe ikhodi kusuka kuwo womabili amawindi futhi ifunda ukuhlobanisa izici zawo zangaphakathi ukuze kutholwe isikhathi eziwahlukanisayo. Ngenxa yale nqubo, i-transformer igcina yenza imodeli yokuncika kwesikhathi eside namaphethini wokuziphendukela kwemvelo komsebenzi wobuchopho okunwebeka kakhulu kunemvelo yendawo yama-millisecond ambalwa.
Ezigabeni zakamuva, imodeli ingashintshwa ukuze ibe nemisebenzi ehlukene ngokusebenzisa ukuhlela kahle iziteshi eziningi nokuhlola okuthileLokhu kusho ukuthi, uma isiqeqeshwe kusengaphambili ne-PARS ekurekhodweni kwe-EEG ehlukahlukene (okuhlanganisa nokusethwa kwe-multi-electrode), ilungiselelwa imisebenzi ethile efana ukuhlukaniswa kokulala, ukutholwa kwe-EEG engavamile noma isithuthwane.
I-athikili yezobuchwepheshe iphinde ichaze izici ezingokoqobo ezifana amasethi edatha asetshenzisiwe, ukwakheka okuqondile kwesishumeki sekhodi, uhlobo lwedekhoda ekhethiwe, imaski nezikimu zesampula zamasampula, kanye ne izinsiza zekhompyutha ezisetshenzisiweNgaphezu kwalokho, izinhlobonhlobo zezakhiwo ezihlukile ziqhathaniswa nezifundo zokukhishwa ukuze kutholwe ukuthi yiziphi izinqumo zedizayini ezikhiqiza ukusebenza okungcono kakhulu.
Ukuqhathanisa nezinye izindlela: MAE, MP3, DropPos nezinye
Ucwaningo aluchazi nje i-PARS, kodwa luyayiqhathanisa nayo Izindlela eziyisethenjwa ekuzifundiseni kwakho kwe-EEGPhakathi kwezindlela eziqhathaniswe yilezi ama-autoencoders (MAE), MP3 kanye ne-DropPos, ngayinye inefilosofi ehlukile uma kuziwa ekufundeni kudatha engenamalebula.
Ama-MAE agxile ku yakha kabusha izingxenye ezifihliwe zesignaliNgesikhathi sokuqeqeshwa kwangaphambili, izingxenye zokufakwayo ziyafihlwa, futhi imodeli izama ukuzibuyisela kumongo. Lokhu kuphoqa isifaki khodi ukuthi sifunde izethulo eziphusile, kodwa sithambekele kakhulu kumaphethini endawo—okungukuthi, ukugcwalisa "izikhala" eziseduze kunokuqonda ubudlelwano bebanga elide.
I-MP3 nezinye izindlela ezifanayo nazo ziyahlola amasu ezesabisa wokuthwebula ulwazi lwesakhiwoKodwa-ke, ngokuya ngemiphumela yephepha, zisasebenza kancane kune-PARS ekufanekiseni izikhawu zesikhathi esihlobene phakathi kwamasegimenti akude esiginali.
I-DropPos, ngakolunye uhlangothi, iguqula noma isuse imininingwane esobala yesimo ku-transformer ngomqondo wokuthi yenza imodeli iqine ngokwengeziwe endaweni eqondileNakuba lolu hlobo lwesu lusiza amanethiwekhi ukuthi anganciki ngokweqile ekushumekeni kwezikhundla, kubonisiwe ukuthi, ngokwalo, akwanele ukuxhaphaza ngokufanele ukwakheka kwesikhashana kwamasignali e-EEG.
Ukuhlolwa kokuhlola kubonisa ukuthi amamodeli eqeqeshwe kusengaphambili nge-PARS ilingana noma idlule lezi ezinye izindlela kumabhentshimakhi amathathu kwamane e-EEG asetshenzisiwe. Lapho kukhanya khona kakhulu kusezimo ze ukusebenza kahle kwelebulaOkusho ukuthi, lapho kuphela ingxenye encane yokurekhodwa etholakalayo. Lokhu kubalulekile esimweni somtholampilo, lapho ukulebula ngokunembile yonke iminithi ye-EEG kudla isikhathi futhi kudinga ochwepheshe abakhethekile.
Isithasiselo somsebenzi sichaza ngokuningiliziwe ukucushwa kwesisekelo ngasinye, ama-hyperparameter ahloliwe, umthelela wamazinga e-mask ahlukene noma i-decoder architectures ehlukene kanye nemiphumela yokugcina yobuningi. Umlayezo ocacile ukuthi, ku-EEG, ukufunda ngokusobala ubudlelwano besikhashana phakathi kwezingcezu zesignali kunikeza inzuzo engokoqobo phezu kokumane kwakhiwe kabusha noma ukufihla ubuso.
Amasethi edatha asetshenzisiwe: kusukela ekulaleni kwe-EEG kuya ekutholweni kwesithuthwane
Ukuqinisekisa i-PARS, kusetshenziswe i-Apple nabahlanganyeli bayo amasethi edatha e-EEG amane aziwayoehlanganisa izimo ezihlukene zokusetshenziswa: ukulala, izifo, umsebenzi wezimoto ngisho nokulungiswa kwama-electrode endlebeni.
Isethi yedatha yokuqala ithi Isiteji Sokulala Esigqokekayo (EESM17), egxile ekuqaphelweni kokulala ngemishini egqokekayo. Kufaka phakathi ukurekhodwa kwasebusuku kwezihloko ezingu-9 ngesistimu yokuqapha ukulala 12-channel ear-EEG kanye ne-6-channel scalp EEGLe dathasethi inentshisekelo ngokukhethekile ngoba ibonisa ukuthi ama-electrode abekwe endlebeni angathwebula ingxenye ebalulekile yomsebenzi wobuchopho ohlobene nokuhlukanisa izigaba zokulala.
Owesibili uthi I-TUAB (Temple University Abnormal EEG Corpus), ikhophasi eyenzelwe i ukutholwa kwe-EEG okungavamileIqoqa amarekhodi alebulwe njengokujwayelekile noma okugulayo, okuwusizo kumamodeli okuqeqesha athola ukushintshwa okuvamile kwemizwa, ngale kwesimo esithile.
Owesithathu, I-TUSZ (I-Temple University Seizure Corpus), igxile ku- ukutholakala kwesifo sokuwaKuhlanganisa izichasiselo ezimaka ukuqala nokuphela kokuquleka, kanye namasegimenti ahlanganayo. Ingelinye lamasethi edatha ayizethenjwa esithuthwaneni ukuze kuhlolwe ama-algorithms e-AI.
Ekugcineni, idathasethi yesine ithi I-PhysioNet-MI, kugxilwe ku imisebenzi motor imaginationKulokhu, ababambiqhaza bacabanga ukunyakaza (isibonelo, ukuhambisa isandla) kuyilapho i-EEG irekhodwa, okuvumela ukuqeqeshwa kwamamodeli aqaphela amaphethini ahlotshaniswa nenjongo yokunyakaza, okuthile okubalulekile ekuxhumaneni kobuchopho nomshini.
Ukuqeqeshwa kwangaphambili kwe-PARS kwenziwa kulawa nakwamanye amasethi edatha achazwe kusithasiselo sobuchwepheshe, kuyilapho ukulungisa kahle kuhambisana nemisebenzi ethile ngaphakathi kwaleyo naleyo. Ukukhethwa kwezilinganiso ezinjalo ezihlukene kubonisa indlela yokwenza Ayikhawulelwe esimweni sokusetshenziswa esisodwa futhi lokho kungasebenza njengesisekelo esijwayelekile samamodeli we-EEG azibhekayo.
I-Ear-EEG ne-AirPods: Ucwaningo lwe-PARS luxhuma kanjani namahedfoni e-Apple
Esinye isici esiphawuleka kakhulu salesi sihloko sonke ukusetshenziswa kwe indlebe-EEG, okungukuthi, ukuthunjwa kwezimpawu zobuchopho ezivela endlebeniI-dataset ye-EESM17 isivele isebenzisa izinhlelo ezibeka ama-electrode ku-canal indlebe nendlebe, esikhundleni se-scalp, okunciphisa kakhulu umthelela obonakalayo futhi kuthuthukise ukunethezeka.
Ngaleso sikhathi, imibhalo yomphakathi kanye namalungelo obunikazi akwa-Apple akhombisa ukuthi inkampani isinesikhathi ihlola ama-headphone akwazi ukukala ama-biosignals kusukela endlebeni. Kusicelo selungelo lobunikazi sango-2023, inkampani ichaza “insiza kagesi egqokekayo” eklanyelwe ukurekhoda umsebenzi wobuchopho isebenzisa ama-electrode atholakala ngaphakathi noma eduze kwendlebe, njengenye indlela engabonakali kancane kumasistimu e-EEG ekhanda ekhanda.
Ilungelo lobunikazi ngokwalo liyavuma ukuthi izixazululo ezijwayelekile ze-EEG zendlebe ngokuvamile zidinga amadivayisi enzelwe umsebenzisi ngamunye (kufakwe ngokwezifiso usayizi nokuma kwendlebe yakho, umsele wendlebe, njll.), okumba eqolo futhi okungenakwenzeka. Ngaphezu kwalokho, ngisho nedivayisi eyenziwe ngokwezifiso ingalahlekelwa ukuthintana nesikhumba ngokuhamba kwesikhathi, okokwehlisa ikhwalithi yesignali.
Ukubhekana nalezi zinselelo, i-Apple iphakamisa isixazululo kulowo mbhalo ngokusekelwe ekubekeni ama-electrode amaningi kunokudingekile ngokuqinile, asakazwa kuyo yonke indlebefuthi uvumele imodeli ye-AI inqume ukuthi yiziphi ezinikeza ukufunda okuhle kakhulu nganoma yisiphi isikhathi. Ukuze wenze lokhu, amamethrikhi afana ne- i-impedance, izinga lomsindo, ikhwalithi yokuxhumana kwesikhumba, noma ibanga phakathi kwama-electrode asebenzayo nayireferensi.
Uma lawa mamethrikhi esebaliwe, isistimu iyabela izisindo ezahlukene electrode ngayinye futhi ihlanganisa amasignali ayo ibe i-waveform eyodwa elungiselelwe. Ilungelo lobunikazi lifaka ngisho nokuthinta okulula, okufana nokucindezela noma ukukhama indlebe, ukuqala noma ukumisa ukulinganisa, kanye nokuhluka okuhlukile kwedizayini nokuhlanganisa okungenza hardware.
Ukusuka kumbono kuye emkhiqizweni: izinzwa kuma-AirPods nokuqapha kobuchopho kwansuku zonke
Inhlanganisela yalo mkhakha wocwaningo ne-PARS yenza kube lula ukukucabanga Ama-AirPod anezinzwa ezikwazi ukukala i-EEG kusuka endlebeniEqinisweni, sekuvele kunentuthuko kuleyo ndlela: i-AirPods Pro 3 ifake inzwa ye-photoplethysmograph (PPG) ukukala ukushaya kwenhliziyo, futhi i-Apple ibilokhu yengeza izici zezempilo ezintweni ezigqokekayo eminyakeni embalwa edlule.
Uma sengeza kulawa ma-electrode e-ear-EEG kanye nemodeli ezigadile efana ne-PARS ekwaziyo ahumushe amasignali ngaphandle okukhulu yolwazi kuphawuliweUmphumela uzoba idivayisi ekwazi ukubona izigaba zokulala, izinguquko ekunakeni, noma izimpawu zokuqala zezifo ezithile zemizwa, konke ngokusobala kumsebenzisi.
Ezivivinyweni ezichazwe ephepheni locwaningo, i-algorithm ye-PARS ithatha izingxenye ezingahleliwe zesignali yobuchopho futhi ufunde ukubikezela ibanga lesikhashana phakathi kwazoNgokusekelwe kuleli khono, imodeli ithuthukisa ukuqonda okuthuthukisiwe kokuthi umsebenzi wobuchopho uguquka kanjani ngokuhamba kwesikhathi, okuholela emiphumeleni engcono lapho ihlukanisa izigaba zokulala, ithola izehlakalo zesithuthwane, noma ihlukanisa okuvamile kuma-EEG angajwayelekile.
Isikhalo esikhulu sale ndlela ukuthi ingasebenza esimweni lapho amalebula eyivelakancane. Emkhiqizweni wezohwebo, lokho kungasho ukuthi ama-AirPod afakwe lobu buchwepheshe azokwazi zivumelane nomsebenzisi ngamunye ngolwazi oluncane olugadiwe, ukusebenzisa inani elikhulu lesiginali eluhlaza ethathwe ngesikhathi sokusetshenziswa kwansuku zonke.
Ngokwemiphumela yocwaningo, amamodeli aqeqeshwe kusengaphambili ane-PARS ayafinyelela ukufanisa noma ukweqa ukunemba kwezindlela zangaphambilini emisebenzini ehlukahlukene, futhi lokhu kuvula umnyango wezinto ezisetshenziswa kakhulu ezisetshenziswa ngabathengi, njengama-headphones, ukuze ziqale ukunikeza izilinganiso ebezigcinelwe izinto zokusebenza zasesibhedlela ezinkulu nezikhethekile.
Yiqiniso, konke lokhu kuza nemibuzo enengqondo mayelana ubumfihlo bedatha nokuvikeleka kanye nemingcele yokuziphatha. Umqondo wokuthi i-AI ayikwazi nje ukushaya kwenhliziyo yakho noma izinyathelo zakho zansuku zonke, kodwa namaphethini akho okusebenza kobuchopho, udala inani elithile lokukhathazeka. Okwamanje, kokubili iphepha kanye namalungelo obunikazi kuhlala kumkhakha wocwaningo kanye nomklamo womqondo, ngaphandle kosuku oluthile lomkhiqizo wokuthengisa.
Izinhlelo zokusebenza ezingaba khona: ezempilo, ukuphila kahle, ukushayela, nokusebenza kwengqondo
Uma i-PARS kanye nobuchwepheshe obusekelwe endlebeni be-EEG buguquka kuma-AirPods azayo noma ezinye izinto ezigqokekayo, ububanzi bezinhlelo zokusebenza bungaba banzi kakhulu. Okokuqala, kungaba ekahle for qapha ukulala ngokuqhubekayo nangokunethezeka, ihlukanisa ngokuzenzakalelayo izigaba ze-REM ne-NREM (NREM 1, NREM 2 kanye ne-NREM 3) futhi inikeze imininingwane enemininingwane ngekhwalithi yokuphumula.
Ngaphezu kwalokho, ukutholwa kwe amazinga okunaka, iziqephu zokucindezeleka, noma izimo zokuqaphela Lokhu kungaba usizo kakhulu ezimweni ezinjengokushayela, umsebenzi onzima ngokwengqondo, noma ukufunda. Idivayisi ehlonza ukwehla okungazelelwe kokunaka ingase ixwayise umsebenzisi lapho esengozini yokulala ehamba ngesondo noma enze amaphutha amakhulu ngenxa yokukhathala.
Esimeni somtholampilo, ukuqapha kobuchopho okuhlakaniphile nokuqhubekayo kungasiza ukutholwa kusenesikhathi kwezinkinga ezifana nesifo sokuwa, izinkinga zokulala, noma izifo ze-neurodegenerativeAkukhona mayelana nokushintsha udokotela wezinzwa noma isibhedlela, kodwa mayelana nokunikeza idatha ebalulekile engasebenza njengesixwayiso sangaphambi kwesikhathi noma umphelelisi ekuxilongweni.
Omunye umugqa othakazelisayo yi- i-biofeedback igxile enhlalakahleni yengqondoUma idivayisi ikwazi ukuhlobanisa amaphethini athile e-EEG nezimo zokuphumula, ukugxilisa ingqondo okujulile, noma ingcindezi, ingase iqondise umsebenzisi ekuzilolongeni ukuphefumula, ukuzindla, noma ukuqeqeshwa kwengqondo, inikeze izinkomba zesikhathi sangempela zokuthi ingabe lezi zinqubo zinomphumela oyifunayo.
Bekungeke kume nje ebuchosheni. Imibhalo yasesidlangalaleni yakwa-Apple ikhuluma ngokungezwa okungenzeka kwe izinzwa zokulinganisa umthamo wegazi, ukusebenza kwemisipha yobuso, nokunyakaza kwamehlo kusuka kuma-earphone uqobo. Ihlanganiswe nesignali ye-EEG futhi icutshungulwe ku-a iPhone noma enye idivayisi, le datha ingaphakela amamodeli e-AI akwazi ukunikeza isithombe esiphelele kakhulu sesimo somzimba nesomzwelo somsebenzisi.
Konke lokhu kusetshenziswa kuzofanela kuhambisane ukulawula okuqinile kwemvume, ukuphathwa kwedatha kanye nokufinyelela kwabantu besithathuEzibonelweni ezethulwa, kucatshangwa ukuthi ulwazi lungabiwa nezingcweti zokunakekelwa kwezempilo kuphela uma umsebenzisi elugunyaza, nokuthi ukucubungula okuningi kwenziwa endaweni ukuze kuncishiswe izingozi.
Empeleni, lokho okubonisa kokubili iphepha le-PARS namalungelo obunikazi kanye ne-prototypes kuwukuhlangana okucace kakhulu: I-Apple iyahlola, ngakolunye uhlangothi, indlela engcono kakhulu yokuqoqa izimpawu zobuchopho endlebeni futhi, ngakolunye uhlangothi, indlela engcono kakhulu yokuwatolika nge-AI ngaphandle kokuncika ezichasiselweni zomuntuUma zombili izingcezu zihlangana ndawonye, ama-headphone angayeka ukuba izidlali zomsindo ezilula futhi abe amathuluzi athuthukile ezempilo nokusebenza, inqobo nje uma izici zokuziphatha nezobumfihlo ziphathwa kahle.
Konke kukhomba eqinisweni lokuthi sisekuqaleni kwesizukulwane esisha sezinto ezigqokwayo lapho Izindlela ezifana ne-Apple PARS (Parwise Relative Shift) kanye nezinzwa ezihlukene njenge-ear-EEG Bangaguqula indlela esikuqonda ngayo ukulala kwethu, ukunaka kwethu, kanye nempilo yethu yemizwa, baqale ngokuthile kwansuku zonke njengama-AirPods.
Umbhali oshisekayo ngomhlaba wamabhayithi nobuchwepheshe ngokujwayelekile. Ngiyathanda ukwabelana ngolwazi lwami ngokubhala, futhi yilokho engizokwenza kule bhulogi, ngikubonise zonke izinto ezithakazelisayo kakhulu ngamagajethi, isofthiwe, ihadiwe, izitayela zobuchwepheshe, nokuningi. Inhloso yami ukukusiza ukuthi uzulazule emhlabeni wedijithali ngendlela elula nejabulisayo.