Part 1 · Cockpit

1. Do Cockpits, Ek Nazm-o-Zabt (Discipline)

Claude aur ChatGPT ko sath sath khola jaye to donon dekhne mein kaafi milte julte lagte hain. Donon mein pichli chats ki fehrist, message box, model ya reasoning control, files aur tools ka menu, projects ya workspaces, memory aur instructions, aur search-research features shamil hain.

Jo salahiyatein dono jaga kaam aati hain, wo ye hain:

• Kaam ko wazeh tareeqe se bayan karna
• Faidamand context faraham karna
• Kya delegate karna hai, ye faisla karna
• Jawabon ko check karna
• Kab zyada qawi model ki taraf jana hai, ye jaanna
• Hassas maloomat ki hifazat karna

Jo cheez farq karti hai wo hai product ke naam, layout, aur taqat, aur ye waqt ke sath badalti rehti hain.

Donon ko jaanna is liye zaroori hai kyunke Claude aur ChatGPT har kaam par ek jaisa kaarkardagi nahi dikhate. Jo model ek qisam ke kaam mein bemisaal hai, wo doosri qisam ke kaam mein kamzor ho sakta hai. Isay platform awareness kehte hain, yani ye jaanna ke har tool kis cheez mein achha hai, ek hi interface se wafadar ban jane ki bajaye.

Dono companies chat se aage bhi desktop apps, coding tools, aur browser agents pesh karti hain, lekin ye course sirf chat window tak mehdood rehta hai.

Sab Se Ahem Beginner Aadat

“Assistant se aise baat karein jaise ek qabil colleague se karte hain.”

Koi bhi ahem darkhwast bhejne se pehle teen cheezein poori karein:

1. 1

   Manzar Set Karein

   Aap kaun hain aur asal mein kya hasil karna chahte hain?

2. 2

   Kaam Define Karein

   Assistant ko theek theek kya karna chahiye?

3. 3

   Qawaid Set Karein

   Kaunsa lehja, format, hadood, ya misaalein follow karni chahiyen?

“Pattern seekhein, button ki jagah nahi.”

2. Models Aur Thinking Modes

Claude aur ChatGPT, donon aapko ye chunne dete hain ke kisi kaam par kitni salahiyat kharch karni hai. Ghalat shuruaati nuqta model ke naam hain, kyunke wo tezi se badalte hain. Sahi shuruaati nuqta teen darjon wala pattern hai.

Claude ke model naam: Haiku (sab se tez aur sasta), Sonnet (roz-marra ka intekhab), aur Opus (heavy flagship). Haiku aur Sonnet fast darje mein aate hain, Opus flagship darje mein. Thinking ek alag switch hai, model ka naam nahi.

ChatGPT ke model naam: Instant (roz-marra) se lekar Pro (sab se mushkil kaam) tak, aur beech mein thinking ke darje.

Do Aasan Usool

1. 1

   Usool 1 · Tez Se Shuru Karein

   Zyada tar roz-marra ke kaam ke liye normal fast mode istemal karein. Chhote rewrites, khulasay, ya aasan wazahaton ke liye heavy reasoning ka intezar na karein.

2. 2

   Usool 2 · Mushkil Kaam Par Barhein

   Kai marahil wali logic, ghor se muwazna, riyazi (mathematics), mushkil debugging, ya sahi hona zaroori ho aisi code ke liye thinking mode ya zyada qawi model istemal karein.

Wo usool jo har naye model release ke baad bhi qaim rehta hai: model ko kaam ke mutabiq chunein, aadat ke mutabiq nahi. Roz-marra ke liye fast, mushkil reasoning ke liye thinking, aur waqai mushkil kaam ke liye flagship. Pehle reasoning ka darja chunein, phir current model naam seekhein.

3. Wo Context Jo Aap Attach Karte Hain

Ek language model sirf usi maloomat par kaam kar sakta hai jo uske saamne ho. Isi liye attachments matter karti hain. Support ki jane wali files mein PDFs, Word documents, spreadsheets, CSV files, images, screenshots, aur code files shamil hain.

Jab aap koi file attach karte hain, to aap usi guftagu ke liye assistant ko zyada context de rahe hote hain. Iska muwazna dekhein:

Alfaaz taqreeban ek jaise hain. Doosra darkhwast is liye faidamand hai kyunke assistant waqai contract dekh sakta hai. Agar aapko kisi dashboard, error message, chart, ya interface mein madad chahiye, to usay dikhana usay yaddasht se bayan karne se behtar rehta hai. Donon products live web search bhi kar sakte hain, jab model ki apni maloomat kaafi na ho to search taaza maloomat le aati hai.

Lambi Guftagu Ko Sambhalna

Donon products current guftagu ka record rakhte hain, jo baad ke jawabon ko shakl deta hai. Is record ki bhi ek hadd hoti hai, aur kisi lambe kaam mein aap us hadd tak pahunch jayenge.

Iski ek misaal: aap ne poora din ek hi chat mein quarterly board report likhne mein guzara. Teen alamaat batati hain ke chat purani (stale) ho chuki hai:

• Subah di gayi koi hidayat (jaise "bullet points na likho") bhool jati hai aur bullets wapis aa jate hain
• Pehle diye gaye kisi figure se tazaad hota hai
• Jawab aam mashware ki taraf drift ho jate hain, jo kisi bhi report par lagoo ho sakte hain

Aisa hone par na to zabardasti chalte rahein, aur na hi chat chhor dein. Pehle ek state summary mangwayein, yani ek mukhtasar record ke ab tak kya kaam ho chuka hai, kya abhi baaqi hai, aur di gayi har hidayat:

Summarize what we have decided so far, what is still open, and every
rule I gave you. Keep it short enough to paste into a new chat.

Summary parhein, ghaltiyan theek karein, phir usay files ke sath ek naye chat mein paste kar ke aage barhein. Agar aap khud ko har hafte wahi summary paste karte hue payein, to yehi ishara hai ke ab ek project ki zaroorat hai, jo Concept 4 hai.

Behtar context aksar ek chaalaak prompt se zyada jawab behtar banati hai.

Part 2 · Workspace Ko Apna Banana

4. Projects, Ek Kaam Ke Liye Ek Kamra

AI ko kuch hafte istemal karne ke baad, sidebar bikhar jata hai. Ek hi client ya mauzu ke baare mein kai chats sidebar mein alag alag pari hoti hain, kaam ki files sirf ek chat mein hoti hain, aur background baar baar dobara samjhana parta hai.

Ek project teen cheezein rakhta hai:

• Chats: usi kaam se juri saari guftagu
• Knowledge: usi project ke liye upload ki gayi files
• Instructions: project ke andar assistant kis tarah kaam kare, is ki pukhta rehnumai

Iski ek misaal, Quarterly Board Reporting Project mein ye shamil ho sakta hai: pichle quarter ka board deck, current financial model, reporting ka style guide, tone aur audience ke baare mein hidayaat, aur board report se juri har guftagu. Ab aapko wahi files dobara upload karne ya wahi hidayaat dohrane ki zaroorat nahi. Yehi poori baat hai: ek project ek silsile wale kaam ke liye context ko pukhta (persistent) bana deta hai. Is kitab ki zabaan mein, ek project ek chhota System of Context hai.

Claude Aur ChatGPT Ka Farq

Claude ka tareeqa (retrieval): jab project knowledge context window se bari ho jati hai, to Claude khud ba khud project knowledge ko search karna shuru kar deta hai, is se gunjaish taqreeban das guna barh jati hai. Ye switch khud ba khud hota hai, aur ye ek paid-plan feature hai.

ChatGPT ka tareeqa (memory): project memory shamil karta hai, jo do tarah se kaam kar sakti hai: default memory mein wasee (wider) memory bhi shamil ho sakti hai, jabke project-only memory mein project ek asal hadd (boundary) ban jata hai, aur bahar ki memory andar nahi aati. Project-only memory hi asal mein isolation banati hai.

Yani ek project hamesha kaam ko tarteeb dene mein madad karta hai. Project-only memory hi wo cheez hai jo usay isolation mein badalti hai.

Project Kab Banayein

“Projects usi kaam ke liye hain jo jari rehta hai.”

5. Memory Aur Pukhta Instructions

Naye log in teenon cheezon ko aksar aapas mein khalt-malt kar dete hain.

Standing Instructions: Pukhta Qawaid

Standing instructions wo qawaid hain jo ek dafa likhe jate hain aur wasee paimane par lagoo hone chahiyen, jaise “wazahat se pehle jawab do”, “jab tak main technical tafseel na maangoon, aasan zabaan istemal karo”, ya “main business leaders ke liye likhta hoon, software engineers ke liye nahi.” Claude mein ye settings aur styles mein rakhi jati hain, ChatGPT mein Custom Instructions mein (Settings aur Personalization ke andar). Ye un cheezon ke liye hain jo pukhta (stable) hain.

Memory: Aapke Baare Mein Badalti Hui Context

Memory wo cheez hai jo assistant aapki guftaguon se uthata hai aur baad mein istemal kar sakta hai, jaise zahir ki hui pasandeed cheezein, baar baar aane wale maqasid, aapke kaam karne ke tareeqe se juri facts, ya koi bhi cheez jo aap ne yaad rakhne ko kaha ho. Donon products aapko memory review, tarmeem, aur delete karne dete hain.

Donon products ek aisa mode bhi dete hain jo guftagu ko history aur memory se bahar rakhta hai, Claude mein incognito chat, ChatGPT mein temporary chat. Lekin dono companies safety aur ghalat-istemal ke jaiza ke liye taqreeban tees din ke liye ek copy phir bhi mehfooz rakhti hain. Isliye inhein memory-free samjhein, trace-free nahi.

Projects: Ek Silsile Wale Kaam Ki Context

Ye donon se mukhtalif hai. Projects kisi khaas client, course, tehqeeqi mauzu, product, ya jari kaam ke silsile ke gird context rakhte hain.

Mehfooz Context Ek Zimmedari Hai

Do fikarein hain, aasani aur hassasiyat. Mehfooz context mein hassas ya niji maloomat bhi ho sakti hai. Ye purani bhi hoti rehti hai, aur purani context khamoshi se nakaam hoti hai, koi error message nahi aata, sirf ek aisa jawab aata hai jo bharosemand lagta hai aur ghalat hota hai.

Misaal ke taur par, agar pichle quarter ka financial model abhi bhi Quarterly Board Reporting project mein pada hai, to is quarter ka board deck pichle quarter ke numbers ko sunwar kar quote kar dega. Assistant ye nahi bata sakta ke aap ne jo di hai wo purani ho chuki hai.

Mehfooz context is liye faidamand hai kyunke wo waqt ke sath barhta hai. Ye ek aisi jagah bhi hai jahan ehtiyat matter karti hai. Ek achi tarah configure ki hui assistant hafton ke istemal ke baad zyada kaam ki ho jati hai. Model nahi badla, uske gird ki context behtar ho gayi.

“Pukhta context tabhi madad karti hai jab sahi maloomat sahi jagah ho.”

6. Artifacts Aur Writing Blocks: Wo Kaam Jo Aap Sath Le Ja Sakte Hain

Chat guftagu ke liye achi hai, lekin mukammal kaam rakhne ke liye ek kamzor jagah hai. Agar aap 2,000 lafzon ki report, ek chalta hua calculator, ek diagram, ya ek chhoti application mangte hain, to natija ek alag cheez ki tarah chahiye hota hai, scroll mein bikhri ek message ki tarah nahi.

Claude Mein: Artifacts

Artifact ki qismein: document, code, web page, diagram, vector image, calculator, dashboard, aur ek chhoti interactive application. Aap Claude se baat karte rehte hain jabke artifact wahin maujood rehta hai, jise banaya aur tarmeem kiya ja raha ho. Claude ke paas Word, Excel, PowerPoint, aur PDF files banane ki alag salahiyatein bhi hain.

ChatGPT Mein: Writing Blocks Aur Code Blocks

Purani tutorials “canvas” aur ek sath-sath edit karne wale panel ka zikr karti hain. 2026 mein OpenAI ne apne current models se canvas hata diya aur wahi kaam khud guftagu ke andar le aaya, writing blocks aur code blocks ki soorat mein, yani thread ke andar hi maujood, tarmeem ke qabil hisse. Agar koi tutorial aapko canvas kholne ko kahe aur wo na mile, to yehi wajah hai.

Design mukhtalif hai. Khayal wahi hai: kaam ke natije ko kaam ke baare mein guftagu se alag rakhna.

Behtar Natije Kaise Hasil Karein

“ek dashboard” ya “ek report” mat maangein. Bataein ke mukammal cheez ko asal mein kya karna chahiye.

Behtareen:

Build a monthly budget tracker. I should be able to enter expenses
by category, see a pie chart, and get a warning when I go over budget.

Kamzor:

Build a budget tracker.

Ye bhi batayein ke ye kis ke liye hai, kyunke naye employees ke liye ek flowchart, tajurbekaar engineers ke liye flowchart jaisi cheez nahi hoti. Format bhi bata dein: agar jawab data hai, to table ya spreadsheet file mangein, paragraph nahi. Data se murad barah vendors, unki qeematein aur lead times, ya koi bhi aisi cheez jise colleague Excel mein daalega. Barah vendors ka ek paragraph sort, filter, ya sheet mein paste nahi ho sakta. Ek table ho sakta hai. Agar aap maangein, donon assistants spreadsheet file bana denge, jise download kiya ja sakta hai. Phir ek waqt mein ek tabdeeli kar ke tarmeem karein.

“Chat guftagu hai, artifacts, writing blocks, aur code blocks hi asal kaam rakhte hain.”

7. Skills Aur Plugins: Kaam Karne Ke Bandhe Hue Tareeqe

Projects is sawal ka jawab dete hain ke maloomat ko ek sath kaise rakha jaye, lekin dohrahat (repetition) ek aur sawal khari karti hai: assistant se har baar wahi tareeqa kaise dohrwaya jaye? Jawab hai: Skills.

Is mein hidayaat, misaalein, madadgar wasail, aur kabhi kabhi code shamil hota hai. Jab koi milta julta kaam saamne aata hai, assistant khud us skill ko load kar leta hai.

Aap ek skill in cheezon ke liye bana sakte hain:

• Quarterly business reviews likhna
• Ek contract ko checklist ke against check karna
• Customer meeting ka brief tayar karna
• Kachay notes ko pasandeeda report format mein badalna

Har chat mein tareeqa dobara sikhane ki bajaye, aap tareeqe ko ek hi dafa pack kar dete hain. Claude ke paas skills hain, ChatGPT ke paas bhi skills hain. Products is mein farq rakhte hain ke skills kahan available hain aur kaise manage hoti hain, lekin bunyadi khayal ab dono mein sanjha hai.

Agent Skills Standard

Donon vendors ek khuli standard, Agent Skills (agentskills.io), par kaam karte hain. Ek skill Markdown files ka ek folder hoti hai, jiske peeche na koi server hota hai na koi runtime, aur isi liye ye ek jaga se doosri jaga safar kar sakti hai. Kai companies ke dus-hazaar tools ab isi format ko parh sakte hain, isliye ek product ke liye likhi hui skill doosre product mein bhi install ho sakti hai. Ye ek workflow ko ek prompt se zyada pukhta cheez bana deta hai, jo kisi ek vendor mein qaid na ho. Yehi is course ka shuruati sabaq bhi hai: nazm-o-zabt hamesha rehta hai, tool badalta rehta hai.

Plugin Kya Hai?

Is mein kai skills, bahar ki apps se connections, aur kabhi kabhi commands ya sub-agents shamil hote hain. Ye ek install hone wala package hota hai jo kisi job function ke liye banaya jata hai.

Custom GPTs Ke Baare Mein

ChatGPT ke paas custom GPTs bhi hain. Ek custom GPT ek alag se configure ki hui assistant hai jise aap jaan-boojh kar khulte hain, aur iski apni hidayaat, persona, knowledge, aur tools hote hain. Ye ek skill se mukhtalif cheez hai.

“Jab koi tareeqa baar baar dohraya jaye, to us tareeqe ko pack kar dein.”

Part 3 · Pahunch Barhana

8. Connectors, Apps, Aur Sawal Ko Route Karna

Ab tak assistant zyada tar usi maloomat par kaam karti rahi hai jo aap ne uske saamne rakhi. Lekin aapki asal maloomat zyada tar kahin aur rehti hai: email, calendar, cloud storage, project-management tools, chat systems, aur company ke knowledge bases. Iska hal connectors aur apps hain.

Claude is ke liye connector lafz istemal karta hai, ChatGPT app (pehle connector bhi kehta tha). Dono naam ek hi qisam ki cheez ke liye hain, sirf product ke naam farq hai. Ek connector ya app assistant ko doosre nizaam mein dhoondhne, parhne, aur kabhi kabhi amal karne dete hain, jitni ijazat aap dein.

• "Wo email dhoondo jahan hum ne vendor contract par baat ki thi."
• "Kal mere kaunse meetings hain?"
• "Pichle hafte ke project notes ka khulasa do."
• "Mere sab se ahem priority tasks kaunse hain?"

Ab assistant aapke asal systems se jawab de sakta hai, har cheez chat mein copy karne ki bajaye.

MCP: Sanjha Integration Standard

Dono ecosystems ek standard, Model Context Protocol (MCP), ko support karte hain. Isay aam taur par “AI tools ke liye USB-C” kaha jata hai, yani ek hi standard plug, jisse ek dafa banaya gaya tool kai mukhtalif assistants istemal kar sakte hain. Claude ke custom connectors MCP par chalte hain, ChatGPT ke apps bhi Apps SDK ke zariye MCP par bante hain.

Connections Permission Ke Faisle Hain

Kisi tool ko jorna sirf productivity ka faisla nahi, ye security aur governance ka faisla hai, yani ye faisla ke kis ko kya dekhne aur karne ki ijazat hai. Jorne se pehle ye poochein:

• Ye kya parh sakta hai?
• Kya ye bahar ki duniya mein likh sakta hai, bhej sakta hai, mita sakta hai, khareed sakta hai, ya koi aur amal kar sakta hai?
• Kya mujhe is source par bharosa hai?
• Kya mujhe is account ya nizaam ko jorne ki ijazat hai?

Connectors aur apps ko waise hi samjhein jaise software install karne ko samajhte hain: faidamand, aur jaan-boojh kar permission-gated.

Ek connector ya app sirf wahi kar sakta hai jiske liye wo bana ho. Agar aapke kaam ko assistant se koi cheez bhejwani hai, to pehle taeed karein ke bhejna us tool ke actions mein shamil hai. Permissions un actions mein se chuntein jo tool ke paas pehle se hain, wo nayi actions add nahi karte.

Search, Thinking, Ya Research?

Naye log aksar kaam ke liye ghalat mode chun lete hain.

Pehla sawal ye poochna chahiye: main assistant se kis qisam ka kaam maang raha hoon? Ek fact par jo search second mein de sakta hai, deep research zaya na karein. Jab asal masla reasoning ho to search ki taraf na jayein. Jab kaam ko kai-source tehqeeq chahiye ho to sirf aam chat istemal na karein.

Research Mode Ek Alag Cheez Hai

Research mode “behtar search” nahi hai. Assistant ek tehqeeq ka mansooba banata hai, kai searches chalata hai, raabton (leads) ko follow karta hai, sources parhta hai, aur hawalon (citations) ke sath ek tarteeb-yaafta report deta hai. Is mein second nahi, minute lagte hain. Aap ek tehqeeq ka kaam de rahe hote hain, sirf ek chhota sawal nahi.

“Sawal bhejne se pehle usay route karein.”

9. Us Kaam Par Sabit Karein Jo Aap Pehle Se Jaante Hain

Aapko kaise pata chalega ke assistant waqai aapke kaam mein achhi hai? Sahi paimana koi benchmark ya demo nahi, balke aapka apna kaam, aapka apna data, aapka apna maiyaar (standard) hai.

Misaal: Program Director

Ek program director har quarter attendance aur employment outcomes ka tajziya karta hai. Wo agle report ke liye AI se madad chahta hai. Usay naye data par pehle se bharosa nahi karna chahiye. Iski jagah, wo assistant ko pichle quarter ka raw data deta hai, kyunke usay pehle se maloom hai ke sahi tajziya kaisa dikhta hai. Phir wo assistant se wo kaam dobara karwata hai. Ab uske paas muwazna karne ke liye ek asal cheez hai. Agar assistant koi ahem pattern miss kar de, to wo hidayaat behtar bana sakta hai. Agar data mein hi koi masla nikle jo tajziya ke liye zaroori hai, to usay ek data ka masla mil gaya. Agar assistant kaam ka koi hissa bharosemand tareeqe se nahi kar pata, to wo hissa insaan ke paas rehta hai. Teenon natije faidamand hain.

Paanch Marahil Ka Sabit Karne Wala Chakkar

1. 1

   Ek Dohraya Jane Wala Kaam Chunein

   Theek theek wazeh rahein.

2. 2

   Ek Purana Case Dhoondein

   Aisa jiska sahi natija aapko pehle se maloom ho.

3. 3

   Assistant Se Dobara Banwayein

   Usay wahi context dein jo aapke paas aam taur par hoti.

4. 4

   Muwazna Karein

   Kya match hua? Kya nakaam hua? Kaunsi hidayat missing thi?

5. 5

   Behtar Banayein Aur Dobara Chalayein

   Jab tak kaafi achha na ho, ya ye faisla ho jaye ke ye kaam delegate nahi hona chahiye.

Pass Hona Asal Mein Kya Sabit Karta Hai?

Ye aapko milte julte mustaqbil ke kaam ke liye taeed-shuda bharosa deta hai. Ye sabit nahi karta ke assistant har agle case mein bhi sahi hoga. Aur ye zimmedari transfer nahi karta.

Ab bhi in cheezon ki zaroorat rehti hai:

• Naye natijon ka jaiza lein ke kya wo mantiqi (make sense) hain
• Aakhri kaam ke zimmedar khud bane rahein
• Jahan munasib ho, AI ki madad zahir karein
• Bare-khatray (high-risk) faislon ko theek insaani nigrani mein rakhein

Ye chhoti si mashq aapki pehli evaluation suite hai.

“Bharosa maloom kaam ke against kamana chahiye, kisi achhe lagne wale jawab se farz nahi kar lena chahiye.”

Chhota Khulasa, Prompts Se Pehle

Ab aapke paas donon products ka naqsha (map) hai. Claude aur ChatGPT alag alag tools hain, lekin kaafi hisson mein ek jaisi soch par bane hain.

• Models aur thinking modes, kitni salahiyat kharch honi hai, ye control karte hain
• Attachments, is waqt zaroori context faraham karte hain
• Projects, ek kaam ke silsile ke liye context ek sath rakhte hain
• Standing instructions, pukhta qawaid rakhti hain
• Memory, aapke baare mein badalti hui context rakhti hai
• Artifacts, writing blocks, aur code blocks, mukammal kaam ko chat scroll se bahar rakhte hain
• Skills, dohraye jane wale tareeqe pack karti hain
• Plugins, poore kaam ki qisam ke liye salahiyatein bandhte hain
• Claude mein connectors aur ChatGPT mein apps, bahar ke systems tak pahunchate hain
• MCP, donon ke neeche sanjha integration standard hai
• Search, thinking, aur research, teen mukhtalif sawalon ke teen mukhtalif rastay hain
• Jaani-pehchani kaam par test karna, bharosa kamane ka tareeqa hai

Products badalte rahenge. Ye zehni naqsha current interface se kahin zyada der tak qaim rahega.

Ab Khud Aazmayein: Chhe Prompts

Cockpit ke baare mein parhna, usme baith kar udane jaisa nahi. In chhe mashqon mein taqreeban pachees minute lagte hain.

1. 1

   1. Ek Rehnuma Daura (Guided Tour) Maangein

   Chalane se pehle, do features ka andaza laga lein jo assistant ke bataye jane ki tawaqqo hai. Phir Claude ya ChatGPT se poochein ke wo aaj ke workspace ka daura kar wa de, models badalne, file upload, projects, memory, aur research ke baare mein. Ghor karne wali baat: apni pehle se lagayi qayas se muwazna karein, phir jo screen par nazar aata hai usse. Jahan assistant apne hi interface ke baare mein ghalat ho, wahi sab se dilchasp hissa hai.

2. 2

   2. Ek Kaam Donon Assistants Mein Chalayein

   Apne hafte ka koi chhota, asal kaam chunein. Chalane se pehle andaza lagayein ke kaunsa assistant behtar karega, aur kyun. Phir dono se wahi kaam karwayein aur poochein ke unhein behtar karne ke liye kaunsi context chahiye thi. Ghor karne wali baat: outputs ka muwazna karein, phir dono ke jawab ka muwazna karein ke unhein kya chahiye tha.

3. 3

   3. Apna Pehla Project Banayein

   Kaam ka koi silsila chunein jahan aap ne pehle se wahi background ya files dohrayi hon. Assistant se poochein ke wo aapse mukhtasar interview kare, phir do cheezein de: project instructions, aur upload karne layak documents ki tarteeb-shuda fehrist. Ghor karne wali baat: jo sawalat wo poochta hai, wahi context hai jo aap har hafte hath se type kar rahe the.

4. 4

   4. Jo Yaad Rakha Gaya Hai Uska Jaiza Lein

   Assistant se poochein ke wo aapke baare mein jo bhi yaad rakhta hai, ek fehrist ki soorat mein dikhaye. Har item ke liye khud faisla karein: rakhein, theek karein, ya mita dein. Phir poochein ke settings mein memory kahan manage ki ja sakti hai. Ghor karne wali baat: kya kisi cheez se aapko sharmindagi hoti, aur kya kuch purana ho chuka hai.

5. 5

   5. Teen Asal Sawal Route Karein

   Apne kaam se teen sawal chunein: ek jise taaza fact chahiye, ek jise sakht reasoning chahiye, aur ek jo poori report ka mustahiq ho. Har sawal poochne se pehle route chunein, search, thinking, ya research. Ghor karne wali baat: ghalat routing, jaise research mode jise search hona chahiye tha, dono products mein sab se aam roz ki barbadi hai.

6. 6

   6. Assistant Ko Apni Purani Tareekh Par Aazmayein

   Assistant ko batayein ke aap ek dohraye jane wale kaam ko test karna chahte hain, jiska sahi natija aapko pehle se maloom hai. Purana input material dein, aur assistant se wahi kaam dobara karwayein jaise aap khud karte. Apna asal jawab tab tak roke rakhein jab tak kaam mukammal na ho. Baad mein teen cheezein likh lein: kya match hua, kaunsi extra tafseel chahiye thi, aur kya insaan ke paas rehna chahiye.

Is Chapter Ki Nayi Istilaahaat (Terms)

Imtihaan ke liye ye poori fehrist yaad rakhein, koi bhi istilaah chhorein nahi:

Self-Test

Khud Se Poochein

Pehle khud jawab dein, phir sawal par click kar ke jawab check karein. Peechhe dekhe baghair jawab dene ki koshish karein.