🧼 Izinkampani zase-US zisolwa 'ngokuhlanza i-AI' ngokucaphuna ubuhlakani bokwenziwa ngokulahlekelwa yimisebenzi ↗
Izinkampani zilokhu zithi ukuxoshwa emsebenzini “kungenxa ye-AI”… kodwa ukuphindisela emuva kuya ngokuya kuzwakala. Iphuzu eliyinhloko lilula: I-AI ingokoqobo, impela - kodwa futhi iyindlela yokucashela yesimanje uma unciphisa izindleko noma kunjalo.
Okuphikiswayo ukwakheka. I-“automation ikwenzile” izwakala ingenakugwenywa futhi ithambekele phambili, kuyilapho “siqashe kakhulu” noma “sicindezela imingcele” ifika ngobuqhawe obuncane. Futhi kungaba kokubili - kodwa hhayi njalo ngezilinganiso ezishiwo yisitatimende sabezindaba.
🏈 I-Crypto.com ibeka ukubheja okungu-$70M kusizinda se-AI.com ngaphambi kwe-Super Bowl ↗
Ukuthengwa kwesizinda esingu-$70M sekuvele kukhululekile ngendlela ejabulisayo - futhi manje sekubekwa njengomnyango wangaphambili "wabameli be-AI bomuntu siqu" abakwenzela izinto. Imiyalezo, ukusebenzisa izinhlelo zokusebenza, ngisho nokuhweba ngamasheya - kucacile ephepheni, futhi kuningi okuthembisayo ngokuphefumula okukodwa.
Okubalulekile ukudlala kokusakaza: ukuba ne-URL ekhumbuleka kakhulu empeleni ukuthenga ibhodi lokukhangisa ku-inthanethi. Kungakhathaliseki ukuthi umkhiqizo uyamangalisa noma nje... kulungile, ukwethulwa kuklanyelwe ngokusobala ukuphoqa ukunaka okukhulu.
📈 Indlela yokuvikela ibhamuza, uhlelo lwe-AI ↗
Isimo lapha siwukuba nethemba eliqaphile kanye ne-calculator esandleni. Ukusetshenziswa kwe-AI kukhulu kakhulu, okulindelwe kukhulu kunesikhulumi senkundla, futhi umbuzo uba ukuthi ungahlala kanjani uvikelekile ngaphandle kokupheka uma umkhuhlane uphola.
Akuyona into ethi “i-AI iwumgunyathi” - “intengo ingaba yinqaba.” Le ngxenye igxila ekuziphatheni kwabatshalizimali okusebenzayo: ukuhlukanisa, cabanga ngabawinile besigaba sesibili, futhi ungacabangi ukuthi wonke ama-ticker aseduze kwe-AI abusiswa ngokuzenzakalelayo ngonkulunkulu be-silicon.
🧬 I-ByteDance Ikhipha i-Protenix-v1: Imodeli Entsha Yomthombo Ovulekile Efinyelela Ukusebenza Kwezinga Le-AF3 Ekubikezelweni Kwesakhiwo Se-Biomolecular ↗
Ukwehla okukhulu komthombo ovulekile ohlangothini lwe-bio lwe-AI: I-Protenix-v1 ibekwa njengohlelo olubalulekile lokubikezela isakhiwo, hhayi nje idemo enhle. Isimangalo esiyinhloko ukusebenza "kwe-AlphaFold3-class" - ifulegi elinesibindi okufanele litshalwe, noma ngabe izilinganiso zihlala zinezixwayiso.
Okuthakazelisa kakhulu yi-openness angle. Uma ikhodi nezisindo zisetshenziswa ngempela endle, lokhu kungasheshisa ukuhamba komsebenzi wocwaningo ngokushesha - njengomuntu ovula izibani ngokuzumayo elabhorethri ebisebenza ngekhandlela.
🛂 Imikhawulo Emisha Yokufuduka Iyavela Njengoba I-AI Iqhuba Ama-visa E-H-1B Ezinkampanini Zobuchwepheshe ↗
I-AI ayishintshi nje imikhiqizo kuphela - ishintsha ukuthi izinkampani zizama ukuqasha bani, nokuthi zivelaphi. Lo msebenzi uhlanganisa izifiso ze-AI nesidingo semisebenzi ethile yamakhono aphezulu izinkampani ezivame ukuyigcwalisa ngezindlela ze-H-1B.
Ukungezwani kujwayelekile: izinkampani zifuna izindlela ezikhethekile kakhulu zokuthuthukisa amakhono, kuyilapho abenzi benqubomgomo bekhuluma ngokuqinisa imithetho. Ngakho-ke ugcina unalokhu kusunduza okungathandeki lapho "sidinga abantu abaningi be-AI" kungqubuzana nokuthi "sinciphisa imizila yokubathola."
Imibuzo Evame Ukubuzwa
Kusho ukuthini ukuthi “ukugeza nge-AI” lapho izinkampani zisola ukudilizwa emsebenzini ngenxa yobuhlakani bokwenziwa?
“Ukugeza nge-AI” kubhekisela endleleni ezinye izinkampani ezibheka ngayo ukudilizwa emsebenzini njengokuqhutshwa yi-AI, okwenza ukunciphisa kuzwakale kungokwesimanje, kungenakugwenywa, futhi kunengqondo. Empeleni, i-AI ingaba yingxenye yendaba, kodwa futhi ingasebenza njengesilwane esilula sokususa izindleko, ingcindezi ye-margin, noma ukuqasha ngokweqile. Ukucindezela emuva kumayelana kakhulu nezilinganiso: ukuzenzekela kungadlala indima, hhayi kakhulu njengoba ukukhishwa kwabezindaba kusikisela.
Kungani abantu begxeka ukuhlanza i-AI ezindabeni zokulahlekelwa imisebenzi?
Ukugxekwa kuhlose ukwakheka kakhulu kunokuba khona kwe-AI uqobo. Ukuthi “ukwenziwa kwe-automation” kungazwakala sengathi kuthambekele phambili, kuyilapho ukuvuma ukuthi “siqashe kakhulu” noma “sicindezela izindleko” kufundeka njengobuqhawe obuncane. Ukubuyela emuva kuvame ukukhuphuka lapho incazelo izwakala njenge-varnish yomkhiqizo kunokuba kube yindaba ecacile yalokho okushintshile. Ababukeli abaningi bafuna ukucacisa okwengeziwe kanye nenkulumo engagwemeki.
Yini engenza isimangalo esithi “ukuxoshwa emsebenzini okubangelwa yi-AI” sibe yiqiniso kakhudlwana?
Isimangalo esithembekile sivame ukufaka imininingwane ethile: ukuthi yimiphi imisebenzi eyenziwa ngokuzenzakalelayo, ukuthi yiziphi izindima ezishintshiwe, nokuthi izinqumo zokubalwa kwezisebenzi zixhumana kanjani nesikhathi sokukhishwa. Kuphinde kusize ukuhlukanisa izinzuzo zokukhiqiza eziqhutshwa yi-AI ezinhlelweni zokunciphisa izindleko ezibanzi. Ezindleleni eziningi, kokubili kungaba yiqiniso ngesikhathi esisodwa, ngakho-ke ukwabiwa okuhlanzekile kubalulekile. Ngaphandle kwemininingwane, i-"AI" ingafika njengelebula elikhazimulayo kunokuba ibe umshayeli oyinhloko.
Kungani i-Crypto.com ingachitha u-$70M esizindeni se-AI.com?
Ukuthenga i-AI.com kuwumdlalo wokusabalalisa omsulwa: i-URL ekhumbuleka emhlabeni wonke esebenza njengebhodi lokukhangisa elihlala njalo kwi-inthanethi. Inkulumo iwukuthi iba umnyango wangaphambili “wabameli be-AI bomuntu siqu,” okwenza umkhiqizo uzizwe sengathi ungumnikazi wesigaba. Ngisho noma umkhiqizo umane nje ulungile, isizinda singaphoqa ukunaka kanye nelukuluku ngezikhathi zokwethulwa.
Ayini “ama-ejenti e-AI yomuntu siqu,” futhi yini ebangela izithembiso ezinkulu?
Kulolu hlaka, ama-ejenti e-AI yomuntu siqu angabasizi abahloselwe ukukwenzela imisebenzi - ukuthumela imiyalezo, ukusebenzisa izinhlelo zokusebenza, ngisho nokuhweba ngamasheya. Inkinga ukuthi ukuhlanganisa amakhono amaningi kangaka esithembisweni esisodwa kuphakamisa imibuzo mayelana nokuthembeka, izindlela zokuvikela, nokuthi i-ejenti idinga ukufinyelela okungakanani. Ezimweni eziningi zokusetshenziswa kwangempela, ulwazi lufika endaweni ethile phakathi "kokusiza" kanye "nokulinganiselwe," hhayi okumangalisayo.
Ungalivikela kanjani ibhamuza le-AI ngaphandle kokuphuthelwa yinzuzo?
Indlela evamile ukuvezwa ngokuqapha: hlala utshale imali, kodwa uhlukanise ukuze ungathembeli ekhoneni elilodwa elinogwebu emakethe. Umqondo ukufuna abawinile be-second order futhi ugweme ukucabanga ukuthi wonke ama-ticker "aseduze kwe-AI" avuzwa ngokuzenzakalelayo. Amanani angashintshashintsha ngesikhathi semijikelezo ye-hype, ngakho-ke usayizi wesikhundla nobubanzi kubalulekile. Ukuba nethemba kusebenza kahle uma kuhlanganiswe ne-calculator.
Iyini i-Protenix-v1, futhi kungani ukusebenza kwe-“AlphaFold3-level” kubalulekile?
I-Protenix-v1 ichazwa njengemodeli yokubikezela isakhiwo se-biomolecular esivulekile, ebekwe njengethuluzi lokucwaninga elibalulekile kunokuba yi-demo. Isimangalo esiyinhloko ukusebenza kwe-"AlphaFold3-class", okubamba ukunaka kodwa kusaxhomeke kumongo wokulinganisa kanye nezixwayiso. I-engeli yokuvuleka ibalulekile: uma ikhodi nezisindo zisetshenziswa ngempela, ingasheshisa ukuhamba komsebenzi wocwaningo ngokushesha.
Isidingo se-AI sibumba kanjani ukuqashwa kwe-H-1B, futhi kungani imikhawulo yokufuduka ibalulekile?
Okuchazwe ngendlela eguquguqukayo ukuthi izifiso ze-AI zandisa isidingo sezindima ezikhethekile, ezinamakhono aphezulu izinkampani eziningi ezivame ukuzigcwalisa ngezindlela ze-H-1B. Ngesikhathi esifanayo, abenzi bezinqubomgomo abaxoxa ngemithetho eqinile badala ukusunduza phakathi kwezidingo zamathalente kanye nemikhawulo yokufuduka. Lokhu kungathinta ukuthi izinkampani zakha kuphi amaqembu, ukuthi zikhula ngokushesha kangakanani, nokuthi zingafinyelela yini ulwazi olukhethekile. Umphumela uba ukungqubuzana phakathi kwesu kanye nenqubomgomo.