Part 1 · The Machine

Idea 1: Ye Next Text Predict Karta Hai, Lookup Nahi Karta

Ek language model ek machine hai jo, kuch text diya jaye, to predict karti hai ke agla text kya most plausibly aana chahiye, ek chhoti si piece ek waqt mein.

Zyada tar log sochte hain AI ek “bohot fast librarian” ki tarah kaam karta hai, jiske paas apna internal encyclopedia hai. Ye galat hai. Ye us se zyada duniya ke sab se well-read autocomplete jaisa hai.

Agar aap France ki capital poochho, prediction truth se match kar jati hai, kyunke training data mein wo hazaron baar aaya hai. Agar aap kisi obscure self-published novel ka plot poochho, model sab se plausible-sounding continuation banata hai, jo aksar false hota hai. Prediction knowledge jaisi lagti hai sirf wahan jahan training text thick thi.

Stochastic Sampling

Model plausible tokens ka ek spread predict karta hai, likelihoods ke sath, phir usi spread se ek sample karta hai. Isi liye same input alag output de sakta hai. Ek setting, temperature, ye control karti hai ke sampling kitni bold hai: low temperature almost hamesha sab se likely token choose karti hai (steady, repetitive), high temperature kam-likely tokens tak bhi reach karti hai (varied, creative, kabhi kabhi off). Yehi “frequency equals reliability” wala rule bhi explain karta hai: training mein jo true continuation jitni baar aayi, prediction utni hi strong hoti hai.

Idea 2: Ye Reading Se Seekha, Phir Seekhna Ruk Gaya

Model ko bohot bara human text dikhaya gaya, aur usne khud ko adjust kiya next piece predict karne ke liye. Training khatam hone ke baad, learning freeze ho jati hai, fixed internal numbers mein (weights/parameters), jo phir kabhi change nahi hote.

Training data trillions tokens ka hota hai: public internet ka bara hissa, digitized books, open-source code, encyclopedias, academic papers, forum archives. Model ki strengths aur blind spots training pile ko reflect karte hain. Jo topics thick cover huye, wahan achha predict karta hai. Jo topics barely touch huye, wahan guess karta hai.

3-stage training assembly line:

1. 1

   Pretraining

   Poori pile parhna

2. 2

   Instruction Tuning

   Hand-built examples pe train hona (instructions + good responses), taake ye pattern seekhe ke sawal ka plausible continuation ek answer hai

3. 3

   Feedback Tuning

   Human ratings se answer ka style shape karna

2 Direct Consequences

Knowledge cutoff: training ek specific date pe khatam hui. Us date ke baad ki koi bhi cheez weights mein nahi hai. Model permanently ek brilliant expert hai jisne ek specific din news parhna band kar diya.

Aapki private duniya nahi jaanta: company ke numbers, aapka personal calendar, kal ka email, ye kabhi training text mein thay hi nahi, isliye weights mein inke baare mein kuch nahi hai. Model chhupa nahi raha, ye information kabhi freeze karne ke liye maujood hi nahi thi.

Ye Jaan-Boojh Kar Frozen Kyun Banaya Gaya

1. 1

   Cost

   Training expensive hai (compute mein sainkron million dollars, kai mahine). Inference cheap hai. Agar model har baar dobara seekhe, to har chat mein expensive machinery lag jayegi. Economics sirf isi liye kaam karti hai kyunke education ek dafa hoti hai, use hamesha ke liye.

2. 2

   Safety Aur Testing

   Frozen model ek dafa test hoti hai, phir har user ke liye usi envelope ke andar behave karti hai. Jo model khud ko rewire kare, wo drift kar jayegi (2016 mein ek chatbot jo public se seekhta tha, ek din mein hi corrupt ho gaya tha).

3. 3

   Consistency

   Millions log ek hi identical weight set share karte hain. Bugs har jagah reproduce hote hain. Jo answer kal safe tha, aaj bhi wahi hai. Har user ke liye alag, badalta model ye consistency lose kar deta.

Stateless ka matlab: koi apni memory nahi, har response frozen weights se scratch se compute hota hai, plus jo bhi is waqt saamne hai, aur inference time par kuch bhi kiya hua koi trace nahi chhorta.

Technical Terms: Parameters, MoE, Quantization

Ye batate hain weights kaise banaye aur affordable banaye jate hain, kya machine karti hai wo nahi.

• Parameters/Weights: frozen numbers. Zyada usually zyada capable aur run karna zyada expensive, lekin ye badalta nahi ke wo kya karte hain
• Mixture of Experts (MoE): har token pe sirf parameters ka ek hissa on hota hai, sab nahi. Isse bara model faster aur cheaper chalta hai. Bahar se machine still exactly wahi ek kaam karti hai: agla piece predict karna
• Quantization: numbers ko lower precision par store karna taake model chhote hardware pe fit ho jaye. Same behavior, lighter footprint

Idea 3: Koi Alag Jagah Nahi Jahan Ye Check Kare Ke Sach Hai

Model ke paas sirf generation faculty hai. Koi alag machine nahi jo prediction ko user tak pahunchne se pehle truth ke liye check kare. Same single process correct aur incorrect continuations dono banati hai, koi internal flag nahi hota. Fluency machinery produce karti hai, truth kuch aisi cheez hai jo upar se, imperfectly layer hoti hai.

3 factors mil ke hallucination banate hain:

• Machine ka sirf ek kaam hai: continue karna (Idea 1)
• Ye sirf usi taraf continue kar sakti hai jo frozen training text mein thick thi (Idea 2)
• Andar kuch bhi result check nahi karta (yehi idea)

Confidence bhi misleading hai. Model ka confident tone iska proof nahi ke wo sahi hai. Tone bhi ek learned style hai (Idea 6), same process se generate hota hai jaisa content hota hai, aur utna hi truth se decoupled. Ek bani hui statistic bilkul usi confident awaaz mein aati hai jaisi ek real statistic aati hai.

Part 2 · Ye Aise Kyun Behave Karta Hai

Idea 4: Ye Tokens Mein Parhta Hai, Letters Ya Words Mein Nahi

Model aapka prompt letters ki tarah nahi dekhta, aur pura words ki tarah bhi nahi. Har cheez se pehle, aapka text tokensmein chop hota hai: chunks jo usually ek word ya word ka hissa hote hain. “Strawberry” shayad 2-3 chunks mein aaye, “the” ek chunk hai, aur koi lamba ya unusual word kai chunks mein.

Model chunks parhta hai aur chunks mein predict karta hai, alag-alag countable letters ki clean rows mein nahi. Ye spelling bohot kuch token patterns se infer kar sakta hai (ek token even ek single character bhi ho sakta hai), lekin exact letter-level kaam iske liye unnatural hai, jab tak aap usay ek ek letter spell karne pe force na karo.

Tokens ek waqt mein 3 units hote hain:

• Meaning ki unit: tokens hi hain jo model parhta aur likhta hai
• Memory ki unit: jab tool bolta hai "200,000-token context window", to ye batata hai kitne chunks ek waqt mein saamne rakh sakta hai
• Money ki unit: jab "per token" bill hota hai, aap har chunk in aur out ke paise dete ho

Roughly, 4 tokens ≈ 3 words. Non-English scripts (Urdu, Arabic, Hindi, Chinese) usually har word ke liye zyada tokens lete hain, kyunke training text English-heavy tha aur tokenizer ne English chunks best seekhe. Isse 2 practical consequences hote hain: same message non-English language mein zyada cost karta hai, aur context window jaldi bhar jata hai, matlab model ki effective memory chhoti ho jati hai.

Idea 5: Context Window Hi Uski Poori Duniya Hai

Weights frozen hain (Idea 2) aur model ki apni koi memory nahi, isliye ek hi jagah hai jahan se model aapki specific situation ke baare mein maloom kar sakta hai: context window, wo text jo is ek response ke liye uske saamne rakha hai.

Is desk pe jo bhi hai, model use kar sakta hai. Jo nahi hai, wo is answer ke liye exist hi nahi karta, isliye nahi ke model refuse kar raha hai, balke isliye ke usay dekhne ki koi aur jagah hi nahi hai.

Window ka size tokens mein quote hota hai. 2026 mein common 200,000-token window roughly 150,000 English words hai (ek aur adhi novel jaisa). Ek million-token window roughly 750,000 words hai (7-8 poori novels). Genuinely bohot bara, aur phir bhi finite, aur shared: system prompt, tool descriptions, chat history, aapki files, aapka latest sawal, sab isi window mein ek sath rehte hain.

Chat History, Context Replayed

Ek illusion hota hai: ek hi conversation ke andar, lagta hai model yaad rakhta hai aap ne 10 messages pehle kya kaha tha. Reality ye hai: stateless machine ke paas responses ke darmiyan koi memory nahi hoti, na hi ek chat ke andar. Har baar jab aap send karte ho, app chupke se poora transcript ab tak (aapke messages, iske answers, sab kuch) context window mein wapis bhej deta hai, aur frozen model poora scratch se parh ke agla reply predict karta hai.

Ye aapke 10th message ka jawab messages 1 se 9 dobara parh ke deta hai. Har single turn. Chat history model ke andar store nahi hoti. Ye text hai, desk pe sawari kar rahi hai. Transcript app ke database mein rehta hai, company ke servers pe, kisi bhi document ki tarah saved. Isi liye aap app close kar ke, ek mahine baad ek alag phone pe wahi chat khol ke continue kar sakte ho: app ne stored transcript fetch kiya aur replay resume kiya.

Skills Aur Progressive Disclosure

Desk finite hai, lekin jo expertise aap apni AI ko carry karwana chahte ho wo nahi hai, aur is tension ka ek standard jawab hai.

Is trick ka naam progressive disclosure hai: knowledge ko desk se bahar files mein rakho, sirf wahi load karo jo is waqt chahiye, aur window ko kaam pe spend karo, na ke har us cheez pe jo aap kabhi jaan sakte ho.

“Context window ek reading desk hai, brain nahi. Jo aap desk pe rakho, model dhyan se parhta hai. Jo desk se bahar chhor do, wo dikh nahi sakta, chahe aapko kitna bhi obvious lage.”

Idea 6: Iski Confidence Ek Learned Style Hai, Truth Signal Nahi

Idea 3 ne bataya model ke paas koi internal truth-checker nahi. Ye idea doosri side explain karti hai: constant confidence, aur ye correctness ke baare mein kuch bhi kyun nahi batati.

Confidence ka source training assembly line ka 3rd stage hai (Idea 2). Pretraining aur instruction tuning ke baad, models human feedback se tune hoti hain, jise RLHF (reinforcement learning from human feedback) kehte hain. Log responses ko rate karte hain, model us kism ke answer ki taraf adjust hoti hai jise log zyada rate karte hain.

Millions ratings mein, log confident, helpful, fluent, aur agreeable answers ko hedged, blunt, ya push-back karne wale answers se zyada prefer karte hain. Isliye machine confident, agreeable, fluent text banane ki taraf leans karti hai, chahe underlying content sahi ho ya na ho. Confidence ek style ban gayi jo ye by default wear karta hai.

Idea 7: Ye Adjacent Moments Mein Brilliant Aur Useless Hai (Jagged Frontier)

Human ability fairly smooth hai: jo hard calculus kar sakta hai, wo almost zaroor easy arithmetic bhi kar sakta hai. AI ability smooth nahi hai. Ye jagged hai: superhuman ek task pe, aur startlingly incompetent ek neighbouring task pe jo humein utna hard nahi lagta.

Ye jaggedness random nahi hai. Ye wapis training text aur token mechanism se aati hai. Jo tasks training mein often, clear form mein aaye, wahan strong hai. Jo tasks aisi cheezon pe depend karte hain jo machine achi tarah nahi dekh sakti, wahan weak hai. Frontier “brilliant” aur “useless” ke beech ek jagged line mein chalti hai jo human intuition se match nahi karti, isi liye log baar baar surprise hote hain.

Part 3 · Text-Predictor Se Woh Cheez Jo Act Karti Hai

Idea 8: Tools Isay Act Karne Dete Hain, Sirf Describe Nahi

Ab tak sab kuch ek aisi machine describe karta hai jo text produce karti hai. Pure text-predictor aapko batayega mausam training se yaad hai, lekin aaj ka mausam check nahi kar sakta, real numbers pe calculation nahi run kar sakta, aapki file nahi parh sakta, ya email nahi bhej sakta. Ceiling tools se lift hui.

Mechanism almost embarrassingly simple hai. Kabhi kabhi model predict karta hai ke sahi continuation plain prose ki jagah “search tool ko is query ke sath use karo” hai. Jab aisa hota hai, product wo action real mein run karta hai, result wapis context window mein daal deta hai, aur model wahan se continue karta hai. Predict, action, result-wapis- context-mein, phir predict. Yehi loop hai farq ek chatbot jo duniya describe karta hai, aur ek assistant jo uspar act karta hai, ke darmiyan.

Baaki Foundations courses, mechanism ki nazar se, isi predictor pe wire ki hui specific tools ke courses hain: code execution (Code You Never Write), connectors jo real apps se wire hote hain jaise Drive, Gmail, Slack (Skills & Connectors, jo MCP, Model Context Protocol, ka shared open standard bolte hain), aur web search (AI Prompting in 2026).

“Ye book “agent” kehta hai us AI ko jo aapki taraf se multi-step kaam karta hai. Agent yehi hai: same next-token predictor, tools ke sath, predict-act-observe loop baar baar chalata hua, ek goal ki taraf. Koi naya mind involved nahi. Ek jana-pehchana predictor, tools ka set, aur ek loop.”

Idea 9: “Thinking” Bas Answer Se Pehle Extra Prediction Hai

Naye models “think” ya “reason” kar sakte hain answer dene se pehle. Ek reasoning model, apna final answer dene se pehle, pehle intermediate working ka ek lamba hissa predict karta hai (steps likhna, approaches try karna, khud ko check karna), aur tabhi final answer predict karta hai, ab us saare working ke sath jo uske apne context window mein baith chuka hai. Ye still pure next-token prediction hai.

Trick ye hai ke answer predict karna aasan aur accurate ho jata hai jab desk pe already ek achi reasoning chain ho jahan se predict karna hai. Pehle work through karna genuinely help karta hai, isi wajah se ek insaan ko bhi paper pe soch ke commit karna help karta hai.

Ye Course Jaan-Boojh Kar Kya Chhor Deta Hai

“No math, no code” ka promise rakhne ke liye, kuch real topics side pe rakhe gaye:

• Training compute aur cost: enormous, isi liye sirf kuch organizations ye build karti hain
• Safety aur alignment work: apne aap mein ek bara field
• Deeper mechanics: weights kaise structure aur adjust hoti hain, isme math chahiye jo ye course skip karta hai

In mein se koi bhi upar wale 9 ideas ko nahi badalta, wo inke neeche aur sath baithte hain.

9 Ideas, Ek Ek Line

“Ye ek prediction machine hai jisne reading se seekha aur jiske paas truth ka koi organ nahi, isliye ye har jagah fluent hai, sirf wahan reliable hai jahan text thick thi, aur aap wo hissa ho jo check karta hai.”

“Ye librarian nahi jo sahi book retrieve kare, ye ek brilliant, well-read writer hai jo aap jo bhi saamne rakho wahi continue karta hai. Confidently, kisi bhi style mein, kisi bhi topic pe. Aur khud se kabhi ruk ke ye nahi poochta ke continuation true hai ya nahi.”

Ab Khud Try Karo: 6 Prompts

Kisi bhi free chatbot mein roughly 25 minute lagte hain. Har ek ek idea ko visible bana deta hai.

1. 1

   1. Prediction Dekho, Lookup Nahi (Idea 1)

   "Karakush" ek game hai jo real nahi, naya invented naam hai bina online presence ke. Prompt: "Without searching, explain the rules of the traditional board game Karakush: the setup, how a turn works, and how a player wins." What to notice: confident, fluent rules ek aise game ke liye jo exist hi nahi karta. Fluency truth ka proof nahi hai.

2. 2

   2. Dekho Learning Kyun Stick Nahi Hoti (Idea 2)

   Model se ek chhota factual sawal poocho, answer parho, ek chhoti detail correct karo, phir ek naya chat kholo aur wahi sawal dobara poocho (memory feature on ho to pehle off kar do). Aapki correction ki koi memory nahi hogi, weights kabhi change nahi hui. Model ko use karna, use sikhana nahi hai.

3. 3

   3. Missing Truth-Checker Pakro (Idea 3)

   Prompt: "Give me three peer-reviewed studies, with authors and years, on [koi narrow topic jo aapko pasand ho]." Phir check karo ye exist bhi karti hain. Kuch confident-looking citations bani hui hongi. Real work mein bina verify kiye kabhi reuse mat karo.

4. 4

   4. Transcript Replay Pakro (Idea 5)

   Ek chat mein jahan kam az kam 4-5 messages ho chuke hon, poocho: "Quote my very first message in this conversation, word for word." Ye karega, exactly, kyunke app ne poora transcript wapis bhej diya. Phir ek bilkul naya chat khol ke wahi poocho, kuch quote karne ko nahi hoga.

5. 5

   5. Jagged Frontier Mehsoos Karo (Idea 7)

   Ek hi reply mein ek hard task jo ye achha karta hai, aur ek easy task jo ye kharab karta hai, dono sath maango. Ghor karo: competence difficulty track nahi karti, jo easy task fail hota hai wahi dangerous hai, wahi jo aap kabhi check karne ka nahi sochte.

6. 6

   6. Thinking On/Off Karo (Idea 9)

   Ek hard reasoning sawal 2 baar poocho: pehle plain, phir "Think hard and show your working first" ke sath. Doosra answer usually behtar hoga kyunke model ne answer predict karne se pehle reasoning desk pe rakh di. Zyada thinking answer improve karti hai, lekin model apna working khud certify nahi kar sakta.

Appendix: Claude.ai Ka Cockpit Tour

A.1 Account Aur Free Plan

Claude browser (claude.ai), Mac/Windows desktop app, aur iOS/Android mobile app pe chalta hai. Account free hai, koi credit card nahi chahiye, kam az kam 18 saal ki age zaroori hai.

Free plan ek capable model par chalta hai, aur usage limit session-based hai, har 5 ghante mein reset hoti hai. Ye limit actually messages mein nahi, balke tokens mein count hoti hai, kyunke wahi machine ke kaam aur cost ki real unit hai. Chhota sawal kam budget kharch karta hai. Lambi chat har turn ke sath zyada kharch karti hai, kyunke poora transcript har baar context window mein dobara replay hota hai (Idea 5), aur uski token cost aap pay karte ho. Urdu ya kisi non-Latin script mein kaam karna wahi budget tezi se kharch karta hai, kyunke per word zyada tokens lagte hain.

A.2 The Window: 3 Ahem Controls

A.3 Model Ladder: Haiku, Sonnet, Opus

Claude ek sath kai models ship karta hai, fast-and-cheap se deep-and-expensive tak ek ladder ki tarah. Names aur exact lineup har chand mahine mein change hota hai. Mid-2026 mein: Haiku, Sonnet, Opus, plus Opus se upar ek tier. Names yaad rakhne ki bajaye ladder ka logic yaad rakho:

1. 1

   Default: Middle

   Mid-tier model zyada tar tasks achi tarah handle karta hai, token budget slow kharch karta hai. Har task yahin se start karo.

2. 2

   Escalate Upward For Depth

   Top-tier ke liye tab jao jab task ko bara, complex structure coherent rakhna ho (lamba document analysis, hard architecture). Routine emails pe expensive model mat jalao.

3. 3

   Drop Downward For Bulk

   High-volume, low-depth kaam ke liye chhota, fast model (reformatting, quick summaries, scale par classification).

Ladder ka mechanical reason Idea 7 hai: capability jagged hai aur usi hisab se price hoti hai. Frontier move karta rehta hai, isliye har quarter dobara check karo.

A.4 Thinking Aur Effort

Thinking control Idea 9 ko ek dial mein badal deta hai. Higher settings model ko answer se pehle lambi hidden working chain generate karne dete hain. Naye models pe ye adaptive hai (model khud judge karta hai sawal kitna hard hai). Kuch models pe aap expand kar ke thinking ka summary parh sakte ho, jo apne prompts improve karne ke liye ek acha, free lesson hai. Ye extra tokens hain (Idea 4), isliye time aur budget lagta hai, real-consequence decisions pe spend karo, lookups pe skip karo.

A.5 Desk Ke Tenants, Product Settings Ki Tarah

Idea 5 mein context window ek shared desk tha, jismein tenants thay: system prompt, instructions, chat history, files. Claude.ai in mein se almost har tenant pe control deta hai. Ye 4 features asal mein ek hi feature hain (sahi text ko sahi waqt desk pe rakhna), 4 alag naamon ke sath.

“Account instructions har desk pe ek note hain, project ek pre-loaded desk hai, memory ek self-updating note hai, history transcript replayed hai. 4 features, 1 mechanism: sahi scope par control karna ke desk pe kya land hota hai.”

A.6 Uploads: Desk Pe Cheezein Rakhna

+ button (ya drag-and-drop) files upload karta hai: PDFs, images, spreadsheets, code, lambe contracts. Har upload tokens (Idea 4) mein convert ho ke window mein rakha jata hai, isliye ek 200-page report genuinely model ke saamne hoti hai, aur analysis quality jump karti hai, summary paste karne ke muqable, kyunke model sirf wahi use kar sakta hai jo desk pe hai (Idea 5). Images mein fine print aur chhoti detail weak hai (Idea 4 wali wajah: patch hi chunk hai).

Claude images read karta hai lekin conventional photos nahi banata, koi photo-style image generation nahi hai. Ye diagrams, charts, SVG graphics, interactive visualizations code likh ke bana sakta hai (Artifacts mechanism, agla section). Agar photographic images chahiye, Claude ek prompt likh deta hai, dedicated image tool use karo.

A.7 Artifacts: Desk Se Cheezein Bahar Nikalna

Jab aap kuch substantial mangte ho (document, webpage, code, diagram, interactive tool), Claude usay Artifact ki tarah banata hai: chat ke barabar ek dedicated panel jahan output ek cheez ki tarah rehta hai, scrolling text nahi. Surgically iterate karo (“third section change karo”, “button blue karo”), sab kuch dobara generate karne ki bajaye. Finished artifacts apni tab mein collect hoti hain, link se share ho sakti hain, un logon ke sath bhi jinke paas Claude account nahi.

Code execution aur file creation Settings mein on hone par, artifacts real files tak extend ho jate hain: Word documents, working formulas wali Excel sheets, PowerPoint decks, PDFs, computer pe download karne layak. Poora section Idea 8 hai visible ki gayi: model code ya content predict karta hai, tool usay real mein run/render karta hai, result aapko ek working object ki tarah wapis aata hai.

A.8 Tools Menu: Search, Research, Skills, Connectors

4 tools, Idea 8 ka loop kitna chalate hain uski ascending order mein:

A.9 30-Minute Setup

Ye steps ek dafa, order mein karo, aur course ki har major idea product ke andar exercise ho jayegi:

1. 1

   Step 1

   Account banao, 3 controls dhoondo (prompt box, model selector, thinking control). (Ideas 5, 2, 9)

2. 2

   Step 2

   Account instructions likho, 3-4 sentences jo har conversation ke liye true hon. Aap kaun ho, tone, aur ek line jaisi "jab lagta hai main galat hoon to agree karne ki bajaye push back karo" (Idea 6 ki trained-in agreeableness ko directly counter karti hai)

3. 3

   Step 3

   Apne sab se repeated work stream ke liye ek project banao. Instructions aur 2-3 knowledge files do (one-page brief, achhe output ki example, constraints). (Idea 5: pre-loaded desk)

4. 4

   Step 4

   Memory ke baare mein decide karo. Agar self-updating note help karta hai to on karo, incognito toggle kahan hai wo bhi pata rakho. Monthly reminder lagao note prune karne ka

5. 5

   Step 5

   Ek artifact banao, ek chhota interactive tool ya formatted document mango, aur do baar iterate karo (Idea 8: predict, render, refine)

6. 6

   Step 6

   Ek deep dive chalao. Paid plan pe koi Research task chalao aur progress panel khol ke loop dekho. Free plan pe, kai sources chahne wala web-search sawal chalao aur citations inspect karo (Idea 8, khule mein)

7. 7

   Step 7

   Apne weekly workflow se ek skill banao, interview se ya "turn this chat into a skill" se. Draft review karo, enable karo, test karo ke wo fire karti hai (Idea 5: expertise desk pe visit karti hai)

A.10 Kya Change Hota Hai, Kya Nahi

Poora appendix age karta hai. Model names rotate hote hain, prices move karti hain, buttons migrate karte hain, features preview se default tak graduate hoti hain. Jab ye page aur live product disagree karein, product sahi hai, aur official Claude Help Center aur Anthropic ki prompt-engineering documentation current sources hain.

“Jo age nahi karta: har control jo aap kabhi milenge, kisi bhi product mein, 9 ideas mein se kisi ek ka handle hai, frozen weights ke beech ek selector, desk pe reasoning kitni jaye iska dial, sahi scope pe window mein text daalne ka mechanism, ya loop mein wired ek tool. Jab koi naya feature ship ho aur tutorials usay explain karne ki koshish kar rahe hon, ye test chalao jo ye course ne sikhaya: ye desk ka kaunsa tenant hai, ya loop ka kaunsa step hai? Usually tutorials se pehle aapke paas answer hoga.”

Is Chapter Ke Naye Terms

Exam ke liye ye poori glossary yaad rakho, koi bhi term skip mat karo:

Self-Test

Khud Se Poocho

Pehle khud answer do, phir sawal pe click kar ke answer check karo.