Prática de Shadowing: RAG's Evolution: From Simple Retrieval to Agentic AI - Aprenda a falar inglês com vídeo
Escolha um modo de prática
Cada modo abre uma sala de prática em tela cheia.
94 frases
Se as frases estiverem muito curtas ou longas, clique em Edit para ajustá-las.
1
We've all had this experience.
2
You search for something, you get thousands of results, and somehow none of them are what you wanted.
3
Well, what if I told you search engines don't actually understand your questions?
4
At least, they didn't used to.
5
From simple keyword search to present -day agentic rag, information retrieval has seen an evolution.
6
And search engines didn't get smarter overnight.
7
They grew up one step at a time.
8
Let's start from the beginning.
9
The earliest search systems were designed around the question of, where does this word appear?
10
Documents were indexed using what's called inverted indices, aka a mapping of keywords to documents.
11
When a user asks a question,
12
the search system will look up these words and quickly return the matching documents.
13
These documents may then be ranked using TF -IDF or BM -25 to measure how important or frequent different terms were.
14
This powerful keyword matching approach still powers a lot of the internet today.
15
But there's a fundamental limitation.
16
It doesn't understand language.
17
It treats words as symbols, not meaning.
18
Synonyms, ambiguity, and any complex intents were essentially invisible.
19
For example, is the search help
20
Python related to coding or did I just get a pet snake?
21
It was on the user to be asking the right questions with the exact right words.
22
The next major leap was semantic search.
23
Instead of treating text as words, we began representing them as language.
24
This is done using vectors or high dimensional number representations that can understand meaning.
25
For example, coffee might be represented as 010 versus house might be represented as 100.
26
These embeddings don't just come out of nowhere.
27
They are learned by large neural networks trained on massive text corpora.
28
By encountering words in context, Over time, these similar concepts will end up close together, even if they use different words.
29
If this is coffee, maybe espresso is represented here.
30
Very close in concept to coffee, but not anywhere close to house.
31
Semantic search turns your words into a kind of map, so the system knows espresso and coffee are pointing to a very similar place.
32
It's essentially your friend who knows what you mean even if you don't say it perfectly every time.
33
This allowed search systems to understand intent.
34
Even if the exact keywords were not used, you could still find relevant documents.
35
And this didn't replace keyword search.
36
It actually complemented it.
37
Hybrid systems began to emerge, bridging the precision of keyword search with semantic recall.
38
For the first time, instead of just matching text, search was able to approximate understanding.
39
Then the world shifted.
40
Large language models were born.
41
These are models trained on a large corpora of text to learn patterns in the data.
42
LLMs don't retrieve facts. When prompted, they will predict the most likely next token
43
or words for an answer based on those patterns that they learned from the training data.
44
The user asks a question to the LLM and it will return a text answer.
45
These are super powerful and revolutionized the business world.
46
However, they had a problem.
47
LLMs only use specific knowledge they learned during a long and expensive training process.
48
Realistically, that means any knowledge is locked to only the documents
49
that that specific LLM was trained on before a certain point in time.
50
LLMs don't know today's information and certainly don't know your specific documents.
51
So what's the solution?
52
Well, it's actually search.
53
Retrieval Augmented Generation or RAG was born.
54
The idea is very simple.
55
The user asks a question.
56
The system does a search for relevant documents using an external knowledge base.
57
This retrieval is used to augment
58
The LLM's prompt and a final answer is generated.
59
This gave LLMs a form of external memory.
60
Now they could cite sources, adapt to new information, and even operate in specialized domains without the costly retraining.
61
These original RAG pipelines were very linear.
62
Documents were embedded offline into these vector databases.
63
They were retrieved once at query time and passed straight into the model.
64
It was simple but effective.
65
This massive improvement significantly dropped hallucinations and enabled LLM adoption across a multitude of new domains.
66
But traditional RAG is nowhere near perfect.
67
It cannot adapt to new scenarios.
68
And suddenly, we are back at the problem of traditional search.
69
The answer is only as good as the search itself.
70
Within such a short period, countless advancements were made to RAG, developing the simple concept into a sophisticated power to be reckoned with.
71
Instead of a single retrieval step, pipelines added re -rankers to reorder results to be more relevant.
72
User queries were rewritten or expanded upon to improve recall.
73
Similar to before, hybrid retrieval became the norm,
74
leveraging the precision of keyword search with semantic vector search.
75
These systems were far more accurate, but still fundamentally static.
76
The pipeline was predetermined, and retrieval was smarter, but still not intelligent.
77
Enter the next disruptor, agents.
78
Agents are systems that use LLMs and tools to perform tasks autonomously.
79
Suddenly, we shifted from simple pipelines to complex decision -making systems.
80
Agents have a variety of tools, such as LLMs, memory, planning,
81
critics, tactics, retrievers, and many more.
82
Agents have become autonomous decision makers planning and executing complex tasks.
83
Now, instead of linear rag retrieval, when the user asks a question, an AI agent will decide whether retrieval is needed,
84
where to search, what questions should be asked, when enough information is obtained, and then generate a final answer.
85
Agents can compare sources, validate claims, refine queries, and iterate.
86
It can invoke APIs, pull data from many knowledge bases, and incorporate multimodal data.
87
Retrieval is no longer fixed.
88
It's a tool invoked as part of reasoning.
89
This opens up a world of possibilities.
90
Now, agentic rag systems are capable of multi -step research, cross -document synthesis, and general adaptive behavior.
91
The system doesn't just answer questions, it reasons and figures out how to answer them.
92
From simple search to current agentic rag, we have learned time and time again that the next big step isn't better answers.
93
It's systems that know how to find them.
94
And the hardest part of AI isn't generation, it's deciding what to look at.
Baixar aplicativo
Tudo o que você precisa para falar fluentemente
Pronúncia com IANota em cada frase
Prática de IPADomine cada som
VocabulárioMonte seu banco de palavras
Jogo de vocabulárioAprenda brincando
Mesmo canal
B21
Every Generation Thinks This Time Is Different: The AI Job Myth
Daily ConversationWorking Adults
C12
What is Event Driven Architecture (EDA)?
Daily ConversationUniversity StudentsWorking Adults
C11
Agentic AI Frameworks Explained: Workflows, Multi-Agent, & Production
Daily ConversationUniversity StudentsWorking Adults
C11
Spec-Driven Development: AI Assisted Coding Explained
Daily ConversationUniversity StudentsWorking Adults
C15
AI Agents vs Business Rules: Which Should Make Decisions?
Daily ConversationUniversity StudentsWorking Adults
C13
AI & Education: Generative AI & the Future of Critical Thinking
Business EnglishUniversity StudentsWorking Adults
C12
EDR vs. EPP vs. NGAV
Business EnglishWorking Adults
C13
MCP vs ADK: How Modern AI Agents Connect and Work Together
University StudentsWorking Adults
C11
Cybersecurity Architecture: Data Security
Business EnglishUniversity StudentsWorking Adults
C14
What Is Jev? The AI Model That Doesn't Generate Text
Daily ConversationUniversity StudentsWorking Adults
C118
How Smart Organizations Will Use AI: Jevons Paradox & Workforce Impact
Business EnglishUniversity StudentsWorking Adults
C111
What are AI Agents?
Business EnglishUniversity StudentsWorking Adults
C16
Cybersecurity Architecture: Five Principles to Follow (and One to Avoid)
Business EnglishUniversity StudentsWorking Adults
C19
AI, Machine Learning, Deep Learning and Generative AI Explained
Daily ConversationUniversity StudentsWorking Adults
C11
Zero Trust Explained in 4 mins
Business EnglishUniversity StudentsWorking Adults
C114
The 7 Skills You Need to Build AI Agents
Business EnglishUniversity StudentsWorking Adults
C1What is Responsible AI? A Guide to AI Governance
Daily ConversationUniversity StudentsWorking Adults
C1What is AI Ethics?
Daily ConversationUniversity StudentsWorking Adults
FOR YOU
Vídeo recomendado
C11
197 frases
How I use AI to learn engineering (anti brain rot studying)
Daily ConversationUniversity Students
C11
847 frases
The Complete Guide to Hybrid Search in RAG (BM25 + Embeddings + Reranker)
Daily ConversationUniversity Students
C11
199 frases
十年後的電腦會長什麼樣？黃仁勳：就像今天的手機，最少拿來打電話｜中英字幕
Daily ConversationUniversity StudentsWorking Adults
C11
173 frases
AI Agent Memory Architecture: Structured SQL, Semantic Vector Search & Memory Bank
Daily ConversationUniversity Students
C11
198 frases
Jack Dorsey’s New Dessa AI Voice Sounds Scary Human
Daily ConversationUniversity StudentsWorking Adults
C12
54 frases
Living with Machines: The language of mechanisation
Daily ConversationUniversity Students
C11
34 frases
What Is Chain-of-Thought Prompting in Generative AI?
Daily ConversationUniversity Students
C11
712 frases
Master AI in 60 Minutes with Vector Databases!
Daily ConversationUniversity Students
C18
55 frases
When AIs act emotional
Daily ConversationUniversity StudentsWorking Adults
C12
11 frases
Why Is There So Much AI Slop?
Daily ConversationUniversity StudentsWorking Adults
Vocabulário e dicas de fala para esta lição
Esta aula de conversação usa o vídeo “RAG's Evolution: From Simple Retrieval to Agentic AI”. As palavras que mais se repetem: search, document, Llm, retrieval, word. Este vídeo tem 79 frases e 1029 palavras para praticar shadowing. A fala dura 8:36. O falante mantém um ritmo estável de cerca de 120 palavras por minuto, confortável para o shadowing. Apenas 77% das palavras estão entre as 3.000 mais comuns do inglês, por isso o vocabulário é exigente.
Vocabulário principal deste vídeo
15 palavras do vídeo que vale a pena aprender, com pronúncia e significado:
| Palavra | Pronúncia | Significado |
|---|---|---|
| pipeline substantivo | /ˈpaɪpˌlaɪn/ | ducto, duto |
| vector substantivo | /ˈvɛktɚ/ | vector, vetor |
| query substantivo | /ˈkwɪɹ.i/ | pergunta |
| essentially advérbio | /ɪˈsɛnʃəli/ | em essência, essencialmente |
| nowhere advérbio | /ˈnoʊ.(h)wɛɹ/ | em nenhum lugar, em lugar nenhum |
| recall verbo | /ɹɪˈkɔl/ | reconvocar |
| domain substantivo | /ˌdəʊˈmeɪ̯n/ | domínio |
| intent substantivo | /ɪnˈtɛnt/ | intenção |
| generate verbo | /ˈd͡ʒɛn.ə.ɹeɪt/ | gerar |
| linear adjetivo | /ˈlɪn.i.əɹ/ | linear |
| hybrid substantivo | /ˈhaɪ.bɹɪd/ | híbrido |
| precision substantivo | /pɹɪˈsɪʒ.ən/ | precisão |
| adapt verbo | /əˈdæpt/ | adaptar |
| prompt verbo | /pɹɑmpt/ | incitar, impelir |
| retrieve verbo | /ɹɪˈtɹiːv/ | recuperar |
Phrasal verbs que você vai ouvir
| Palavra | Significado |
|---|---|
| end up verbo | terminar, acabar |
| figure out verbo | descobrir, deduzir |
| grow up verbo | crescer |
Frases que vale a pena repetir
Frases curtas e completas do vídeo que você pode usar na conversa do dia a dia:
- We've all had this experience.
- Let's start from the beginning.
- Related to coding, or did I just get a pet snake?
Gramática neste vídeo
As estruturas que o falante mais usa, com as palavras exatas do vídeo:
| Estrutura | No vídeo |
|---|---|
| Voz passiva be + particípio passado — o foco está no que acontece, não em quem faz | were designed · were indexed · be ranked |
| Present perfect have/has + particípio passado — uma ação passada que ainda importa agora | has seen · have become · have learned |
Pronúncia para ficar de olho
O falante usa 14 contrações e formas reduzidas, como don't, didn't, doesn't. Diga-as na forma curta, do jeito que você ouve.
- Os sons de “sh” e “zh”: essentially /ɪˈsɛnʃəli/, precision /pɹɪˈsɪʒ.ən/, evolution /ˌiːvəˈluːʃ(ə)n/, representation /ˌɹɛp.ɹə.zɛnˈteɪ.ʃən/, adoption /əˈdɑp.ʃən/
- Palavras longas — acerte a sílaba tônica: essentially /ɪˈsɛnʃəli/, evolution /ˌiːvəˈluːʃ(ə)n/, representation /ˌɹɛp.ɹə.zɛnˈteɪ.ʃən/, fundamental /ˌfʌn.dəˈmɛn.təl/, intelligent /ɪnˈtɛlɪd͡ʒənt/
Sons difíceis para falantes de português:
- Consoante final — sem acrescentar um “i” depois: intent /ɪnˈtɛnt/, generate /ˈd͡ʒɛn.ə.ɹeɪt/, hybrid /ˈhaɪ.bɹɪd/, adapt /əˈdæpt/, prompt /pɹɑmpt/
- /l/ final — a língua toca o céu da boca, não vira “u”: recall /ɹɪˈkɔl/, fundamental /ˌfʌn.dəˈmɛn.təl/, symbol /ˈsɪmbəl/, enable /ɪˈneɪ.bəl/, invisible /ɪnˈvɪzəb(ə)l/
- /h/ — um sopro suave, diferente do “r”: nowhere /ˈnoʊ.(h)wɛɹ/, hybrid /ˈhaɪ.bɹɪd/, hallucination /həˌluːsɪˈneɪʃən/
Como praticar com este vídeo
- Ouça o vídeo inteiro uma vez sem falar e anote as palavras que você não conhece.
- Faça shadowing frase por frase na velocidade normal, repetindo cada uma até o seu ritmo ficar igual ao do falante.
- Grave a sua voz e compare com o original, prestando atenção a palavras como pipeline, vector, query.
O que é a Técnica de Shadowing?
Shadowing é uma técnica de aprendizado de idiomas com base científica, originalmente desenvolvida para o treinamento de intérpretes profissionais. O método é simples, mas poderoso: você ouve áudio em inglês nativo e repete imediatamente em voz alta — como uma sombra seguindo o falante com 1-2 segundos de atraso. Pesquisas mostram melhora significativa na precisão da pronúncia, entonação, ritmo, sons conectados, compreensão auditiva e fluência na fala.