Yintoni kufuneka uqhagamshelane ngokuzenzekelayo

Ukucinga

TL;DR: Qhagamshelana nathi xa usebenzisa izixhobo ze-AI. Ukukhusela ngokufanelekileyo kunokufumana iingxaki ezininzi ezininzi.

Ukucinga

TL;DR: Qhagamshelana nathi xa usebenzisa izixhobo ze-AI. Ukukhusela ngokufanelekileyo kunokufumana iingxaki ezininzi ezininzi.

Ukuphendula

Ndicinga iingcebiso ezininzi ngeveki kwaye ndingathanda ukuvelisa iingcebiso ezininzi kwinkqubo yam.

Ukukhangisa

Kwiintsuku, i-memo ye-internal ye-CEO ye-Shopify, i-Tobias Lutke, yaba i-viral.

Ukucinga

TL;DR: I-memo inikezela abasebenzi we-Shopify ukuba ziyafumaneka kwi-AI tools.

Ukucinga

TL;DR: I-memo inikezela abasebenzi we-Shopify ukuba ziyafumaneka kwi-AI tools.

Umgangatho we-AI ngoku kufanelekileyo

Ukusetyenziswa ngempumelelo kwe-AI ayikho kwakhona kwi-Shopify; kuyinto imibuzo ye-base-line kumadlaleli bonke, ngaphandle kokufanelekileyo. Lokhu kubandakanya ukuguqulwa kwe-cultural kunye ne-operational.

Ukusetyenziswa kwe-AI kufuneka isizathu

Abasebenzi kufuneka zibonise ukuba i-AI ayikwazi isetyenziswe ngaphambi kokubiza izixhobo ezininzi (isib. ukwandisa abasebenzi okanye ixesha). I-Stagnation ifumaneka njenge-Failure, kwaye abasebenzi kubandakanya ukuvelisa kunye nokuvavanya nge-AI ngokuqhelekileyo.

AI kwi-Development Product

I-AI kufuneka ifakwe kwiinyanga ezincinane (i-prototype phase) zeeprojekthi ze-GSD (get shit done). Le nqakraza yenzelwe ukukhuthaza kakhulu ukufundisa, ukuhlaziywa, kunye ne-team collaboration.

Ukusabela Ukusabela

I-Shopify ibandakanya ukusetyenziswa kwe-AI kwi-performance kunye ne-peer review criteria zayo, yenza ukusetyenziswa kwe-AI ingxenye yindlela yokubhalisa kunye nokukhwabanisa abasebenzi, kubandakanya i-leadership kunye ne-executive teams.

Ndingathanda kwi-Shopify, kwaye ndingathanda kwi-company.

Umgangatho

Ndingathanda izixhobo ze-AI njengabaxhasi kunye nezixhobo ezininzi njenge-developer ye-software kunye ne-writer, kwaye ndingathanda kakhulu kubo.

Ndingathanda i-AI kwiminyaka engama-10, ubhalisa malungaIntelligent IntelligentUkwenzaI-AI-Assisted Refactoring, kwaye usebenzisa izixhobo ezisemthethweni ukudalaikhowudi CleanNgathi

Ndiya kwakhona umnqweno weVibe Coding professional.

KwimekoinkxasoUkucingaiimveliso, kunye neephepha ndiyibhale, ndiyabakhokela abantu ukuthatha izixhobo ze-AI kwaye usebenzise ngokubanzi (kucacisa).

Ndiyathanda iiyure ezimbini (ayikwazanga yayo):

Ukucinga

Uya kufuneka uqhagamshelane ngokuzenzekelayo.

Ukucinga

Uya kufuneka uqhagamshelane ngokuzenzekelayo.

iimveliso

Ukucinga

I-AI ayikwazi ukuthatha umsebenzi wakho, kodwa umntu abasebenzisa uya kufumana; kufuneka uye kwenziwe ukuba umntu.

Ukucinga

I-AI ayikwazi ukuthatha umsebenzi wakho, kodwa umntu abasebenzisa uya kufumana; kufuneka uye kwenziwe ukuba umntu.

(Ukuba unako ukufumana oku kwi53:40 ngexeshaUkusuka kwi-conference)

Ngaba ushiye izinto

Kwiintsuku, ndingathanda ukwenza ezinye izifundo zendawo usebenzisa iPython.

Ndicinga iimveliso kwi-25 iilwimi ezahlukileyo ze-programming kwaye awukwazi ukuyifaka yonke into kwi-local.

Ngamanye amaxesha ndisebenzisa iintlobo ze-virtual machines, i-codelabs kunyeUkubalaUkusebenza kwiwebhusayithi.

Kwiintsuku, ndingathanda ukufaka i-Python ngokukhawuleza kwi-locally (iingxaki ye-accidental).

Iqala nge-idea elula: ukuvelisa izinto ezininzi. Ukucinga ixesha. Yenza smart.

Kwaye ndiye ndingayifumana into ezininzi abantu abasebenza ngoku.

AI wabelana ngokukhawuleza. I-Confident. Njengoko kwenziwa ezininzi iiyure ezidlulileyo.

SPOILER: Ndingathanda ngentloko. Enye isicelo emva koko, inkqubo yam yokusebenza iye itshintshe.

Yintoni iingxaki

KwimekoUkucinga CodeIzixhobo ze-10 zokusetyenziswa kakhulu kwi-AI programming.

Ndingathanda nje (Ukucinga) yaye yandisa ukuba banikeze ezinye imiyalelo ukufaka i-Python kwi-Ubuntu yam.

Ukucinga

Nceda ufumane i-command ukufaka i-Python kwi-Ubuntu

Ukucinga

Nceda ufumane i-command ukufaka i-Python kwi-Ubuntu

Kwiintsuku ezidlulileyo, ndicinga:

Ukucinga

hello, nceda ushiye umyalezo ukufaka i-Python kwi-Ubuntu

Ukucinga

hello, nceda ushiye umyalezo ukufaka i-Python kwi-Ubuntu

Kodwa ngoku ndingathanda ukuba ama-tokens ezininzi zithunyelwe izinto ezininzi (kwamanzi, umzekelo), kwaye ndingathanda kumatshini ngakumbi.

Ukusabela ngokucacileyo kwaye ngokucacileyo:

Ukucinga

sudo apt update && sudo apt upgrade -y && sudo apt install python3 python3-pip -y

Ukucinga

sudo apt update && sudo apt upgrade -y && sudo apt install python3 python3-pip -y

Ndingathanda ukucacisa.

Ngaba usebenzisa i-Linux iminyaka engaphezu kwama-20.

Ndingathanda i-password yam mechanically (ngokusebenzisa i-sudo ukufaka i-pip), kwaye waya ukutya.

Xa ndibuyela, i-terminal ibonisa iindidi ezininzi ezininzi zeengxaki. Yinto emangalisayo. Ndingathanda nje ukufaka i-Python (kuquka i-pip, nto leyo i-package manager).

I-problem ( njengoko ungayifumana ngexesha elide) yaba le ngxaki olungabikhoyo:

Ukucinga

sudo apt ukuhlaziywa -y

Ukucinga

sudo apt ukuhlaziywa -y

Yintoni i-upgrade (and bricked) i-computer yam.

I-ChatGPT ilungise ukuba kuqala ukuhlaziywa kwinkqubo yokuVavanyelwa.

Ndingathanda ukuba ukugcina inkqubo yokuVavanyelwa iyiphi imiphumo efanelekileyo ukufumana iinkcukacha ezintsha, kodwa oku kufuneka uqhagamshelane, ukuba i-backup efanelekileyo, kwaye akuyona into efanelekileyo.

Umgangatho wahlala, kwaye ndiya kwakhona.

Ukucinga ngokugqithisileyo kwi-command (and a question to other IAs) iya kukunika isibuyekezo (and safer):

Ukucinga

sudo apt ukucaciswa

Ukucinga

sudo apt ukucaciswa

Yintoni oku kutyelela kwakhona?

Ndingathanda kakhulu kwi-Ubuntu. Ndingathanda ngokupheleleyo ukufumana i-forum okanye i-googling (ngaba u-googles nayiphi na into namhlanje, BTW?).

Ndingathanda kwakhona ukuba awufunayo i-AI ngokufanelekileyo, kwaye ndiyifunayoZonke iimvelisoNdicinga.

Ndingathanda ukuba oku kuthatha umntu efana nami.

Yintoni i-Reptiles

Umlilo wethu umlawuli we-efficiency.

Yonke imibuzo ethatha i-energy, ngoko i-part ye-brain yethu, i-so-called "reptilian brain", iye yandiswa ukuze yenzelwe ukhuseleko, ngaphandle kokufunda.

Yintoni kufuneka imiphumo eshushu, inxaxheba ezincinane, kunye ne-zero i-sugar waste.

Yintoni ndiyabandakanya i-delegation.

Kodwa oku kuthatha ukuba into ukuba uxhumane kuyinto umatshini?

Ngaba usebenzise ngosuku zonke: Uyafuna imiyalezo ze-GPS ngaphandle kokufunda, ukhuseleko kwi-autocomplete ukufinyelela iingcebiso zakho, kwaye uqhagamshelane iinjini ye-recommendation ukuba uqhagamshelane, ukuthenga, okanye ukucinga.

Ukusuka kwi-surface, oku kubonakala njenge-comfort.

Phantsi, yinto i-biological shortcut.

Izixhobo zayo zokusungula iingcebiso zayo zihlanganisa. Ukunciphisa ukunyanga, iingcebiso zihlanganisa iingcebiso apho kunokwenzeka.

I-Routines kunye ne-heuristics ibonelela ukuba usebenza kwihlabathi elifutshane ngaphandle kokuphumelela.

Kodwa xa le ye-reflexe ye-primitive ithatha i-control, unako ukutshintsha ukuba "ngoko ke?" - yaye ngexesha leyo iingxaki zithuba.

I-Illusion ye-Competence

Enye yeziphumo ezinzima ze-AI ye-delegation yi-illusion ye-competence. A machine that speaks eloquently, formats correctly, or generates working code appears smart. Sometimes smarter than you.

Kodwa phantsi kwePolish kungabangela isisombululo engapheliyo, ukucaciswa kwangaphambili, okanye kwakhona iinkcukacha ezincinane.

Ngaba awukwazi ukufumana, ngenxa yokuba i-cognitive laziness yakho isixeko.

Uyakwazi ukufikelela ngokugqithisileyo ngenxa yokuba "ngokunxibelelana ngokugqithisileyo." I-part ye-brain yakho ye-analytical - i-prefrontal cortex yakho - ayikho.

Yintoni ngexesha iinkcukacha zithunyelwe, iingxaki zithunyelwe, iingxaki zithunyelwe, kunye neenkqubo zithunyelwe.

I-delegation ngaphandle kwe-accounting

I-cost ye-human of faulty AI decisions is already here: abantu abalandeli ukhuseleko lwezempilo, abalandeli njenge-fraudulent, okanye abalandeli ngokusekelwe iingcebiso ze-algorithmic. Kwiimeko ezininzi, ingozi elidlulileyo ayikho kuphela i-AI – kuyinto ukusetyenziswa okuzenzakalelayo kwimveliso yayo.

Xa uqhagamshelane isisombululo kwi-system ungenza kwaye ungenza, ungenza i-agentship. Indawo engaphezulu: ungenza i-accountability.

"I-algorithm said so" yenza i-shield. Kodwa kungekho.

Indlela yokuthintela i-reptile ngaphakathi kwakho

Delegation ayikho ngokwemvelo evamile. Kodwa kufuneka uyenze ngokufanelekileyo. Oku kuthetha:

Ukucinga
• Ukubuyekeza i-output ye-AI ye-AI esebenzayo, ikakhulukazi iingxaki ezincinane
Ukucinga
• Ukunyaniseka iinkcukacha kwinkqubo ufumane
Ukucinga
• Ukuvumelana nenkxaso, nangona izixhobo zithembisa
Ukucinga

Nceda uxhomekeke:

Ukucinga

Ngaba ndingathanda oku ngenxa yokusebenza - okanye ngenxa yokuzonwabisa?"

Ukucinga

Ngaba ndingathanda oku ngenxa yokusebenza - okanye ngenxa yokuzonwabisa?"

Oku kuphela ingxaki inokukwazi ukuqhuba ubunzulu wakho olusebenzayo.

Ukucinga

Umzimba wakho ufuna ukuhambisa.

Ngaba kuthetha ukuba i-software yakho isebenza, i-model yakho iyathembekile, okanye i-report yakho iyathembekile.

Yintoni ndifuna kuphela ukunceda.

Oku kwenziwa kakuhle kwi-jungle.

Kodwa akukho kwihlabathi apho izixazululo ziquka iimiphumo ze-digital.

I-delegation ngaphandle kwe-screening ayikho yi-efficiency. Yinto i-abdication.

Kwaye uma ungenza ukulawula izixazululo zakho, umntu - okanye into - uya.