跟读练习: MCP vs ADK: How Modern AI Agents Connect and Work Together - 通过YouTube学习英语口语
C1
We've been hearing so much about AI agents and all of the different protocols lately, right?
149 句
如果句子过短或过长，请点击 Edit 进行调整。
1
We've been hearing so much about AI agents and all of the different protocols lately, right?
2
So if you're wondering whether you need to use MCP or to use ADK or maybe both of these or even neither for your AI agent, just know that you're not alone.
3
Yeah, and before you search what is MCP, again, we've got you covered.
4
That's right.
5
Because today we're going to be breaking down the model context protocol and the agent development kit, talking about what they actually do, how they're different, some use cases, and when you would typically reach for one versus the other.
6
And spoiler, they're not really competing, but we'll get to that.
7
Hey, whoa, no spoilers.
8
Sorry, sorry, all right, let's freak it down.
9
Well, let's begin with the world that we're living in, right?
10
Because AI agents, not just chatbots, but agents that can actually do things are having a moment.
11
Yeah, a very loud moment.
12
Right. And as developers, when we start building these things, we immediately hit two big questions.
13
So, number one, how does my agent talk to external tools and data?
14
But also, number two, how do I actually build and orchestrate the agent?
15
Yeah, so two different problems entirely.
16
And that's exactly where MCP and ADK come in.
17
So the Model Context Protocol, or MCP for short, is an open standard created by Anthropic that answers the first question.
18
It's all about connectivity.
19
How do you give an LLM access to things like APIs, and databases, and files in a way that's clean and reusable?
20
Right. And the Agent Development Kit, or the ADK, which comes from Google, is all about that second question.
21
How do you actually build, how do you structure to also run an agent?
22
But let's say that you have multiple agents that you want to run, like a team of engineers or developers.
23
Well, how you structure all of these agents together to be orchestrated?
24
Yeah, valid questions.
25
I like to think of it like this.
26
MCP is about the LLM agents and how they talk to the outside world.
27
And ADK is the framework for building the agent itself.
28
Okay, so model context protocol.
29
Let's go a little bit deeper because before MCP, if you want in your LLM or the agent that you're building, we have right here to access different resources that it needs to answer a question, which could mean, say for example, accessing a database, so Postgres or something like that, or maybe scraping the web to find new up-to-date information or maybe reading files that are on your directory or your file system.
30
You had to write this custom integration every single time.
31
And I might write this integration, but Anna might write it, and other organizations are all doing the same thing just to retrieve data from these different sources.
32
Yeah, custom glue code, I mean, really accelerating work, don't you think?
33
Fun stuff, right?
34
No, but MCP, what it did was standardize this interface and it defined a protocol, meaning a set of rules for how a client, so maybe your AI agent or an LLM host here, would talk and communicate with these different types of servers, right, so what is actually exposing the tool or the data that our agent needs in order to answer a question or perform task.
35
Yeah, and how does the MCP host communicate with the MCPs server?
36
Do they have like their own secret language?
37
Well, it's not actually much of a secret, because the message format that MCP uses is JSON RPC, which is almost like plain text that is passed from the MCP server to the host agent and the LLM that's running.
38
So it depends, though, where the actual server is running.
39
So for example, if we're running an MCP's server locally on our machine, like an IDE plugin that's accessing your file system, or maybe a desktop application that's access accessing github.
40
It's going to use standard input and output to process and pass that data back and forth.
41
Now, if we're trying to do this on the web with remote servers, it's gonna use HTTP with streaming support in order to pass that information using things like authentication tokens to authenticate that the host is actually talking to the server and back and fourth.
42
Nice! So the MCP server is like a little wrapper around your tool.
43
You write it once and any MCP compatible client can use it.
44
Exactly. And MCP has three main primitives.
45
First off are the tools, like what we've been talking about here.
46
And these are functions that the LLM can call like search the web or run this SQL query.
47
Now, secondly, there's also resources.
48
So resources are things that the LM can read.
49
So files, documentation, internal and external databases.
50
And then finally, MCP also includes the capability for prompts that are pre-built prompt templates that you can use and expose to the server so that you don't have to write them each time.
51
Okay, and what happens if you swap the model you use to build the application or maybe you build an entirely new one?
52
Well, that's a great question.
53
And actually, that is not a problem at all, because since we're not rebuilding these integrations each time, we're actually able to reuse the same MCP server.
54
And that's what makes MCP such a time saver, because there's no need for countless hours spent building these custom integrations to talk to our data sources.
55
It's just plug and play.
56
Nice. Yeah, and I think the thing that developers really appreciate about MCP is that it's model-agnostic, right?
57
So it doesn't matter if you're using Claude, GPT, Gemini, a local model, whatever.
58
If it speaks MCP, it works.
59
Exactly, and the ecosystem is growing really fast.
60
Actually, there's already MCP servers for a lot of the resources that you use every day, like GitHub or Slack, Google Drive, Postgres as a database, Jira, Figma, and The List keeps growing because the community is building them.
61
All right, so you've got your tools figured out, maybe you've a couple MCP servers running, but now you need to actually build the agent.
62
And that's where ADK comes in.
63
So ADK is an open source Python framework for building AI agents, and specifically it gives you structure.
64
Okay, so that's really interesting.
65
What do you mean about having structure for these AI agents?
66
So at its core, ADK forces you to organize your agent in a way like you'd structure real software.
67
Under the hood, it has a couple core building blocks.
68
Those are agents, tools, memory, events, and runners.
69
All the primitives you need to build predictable and testable systems.
70
Okay, that's great.
71
So, what does it actually look like in practice?
72
Yeah, so let's start with the agent itself.
73
In ADK, an agent isn't just an LLM with a prompt, it's a defined execution unit.
74
So it has a model, it has sets of instructions, tools it's allowed to use, and a controlled reasoning loop.
75
An ADK gives you different types of agents.
76
So you can have LLN-driven ones for reasoning, which are more flexible, or deterministic workflow agents.
77
So these can be sequential, parallel, and loop, when you want strict control form.
78
And we also have the flexibility of creating our own custom agents.
79
So that's really cool because ADK lets you pick between, hey, we can let the model figure it out and other situations where we need to do these steps exactly in order each time.
80
That's exactly right.
81
And behind the scenes, a user query is received by the runner and then passed to the agent.
82
And once the agent has a response, so a tool call request or a state change to report, it yields back and passes it to the runner.
83
And the runner can then process the update produced by the agent and can pass it upstream.
84
That's super cool.
85
And I'm assuming that the runner executes these tools and it commits that change to the overall agent session state, right?
86
Yeah. Exactly.
87
So in this loop here, the agent is suspended at each yield, so it's giving the runner full control to handle consequences before the agent sees what's going to happen next.
88
And that's what makes debugging and tracing behavior so much easier with ADK than with ad hoc agent code.
89
I understand that, I get that because I like predictable behavior and our agents historically don't always provide that, right?
90
Yeah, and that's why ADK also gives you proper session state and memory.
91
So state is short-term.
92
It's the working memory inside a single conversation.
93
And memory is long-term, it's what the agent remembers across sessions, such as user preferences.
94
And for anything external, ADK lets agents use tools to run code, call APIs, or even use other agents as tools.
95
That's awesome.
96
And ADK, I guess, it gives you this entire framework, right?
97
You've got structure, you've got state, the orchestration, the debugging, the whole thing.
98
Yeah, it's pretty great.
99
And that's the power of it.
100
ADK isn't just making an LLM do stuff.
101
It's a full stack system for building reliable, multi-agent architectures.
102
Right, and I've heard ADK is great for these multi-agent systems where you could have a root orchestrator agent that delegates subtasks to specialized sub-agents.
103
So instead of one general purpose agent that we have, we've got a research agent, a writing agent, a validation agent, all of these different individual roles that work together that are specialized for each use case, and ADK supports different types of agents out of the box.
104
Yeah, right, so you've got LLM agents, like we spoke about, which are the classic think and act loops, but you also have workflow agents, so those are the sequential, parallel, and loop-based agents.
105
And if you know ahead of time that certain steps need to happen in order, you can hard code that structure instead of leaving it up to the model.
106
That's huge for reliability.
107
Less LLM decision-making stuff that we don't always need and more ability to build custom workflows to fix our problems and solve what we need.
108
Yeah, exactly.
109
An ADK supports multiple model backends, so you're not locked in.
110
And the cool thing is that ADK is a tool framework agnostic at the integration layer.
111
So that means.
112
Well, it means it works with MCP servers.
113
You can plug any MCP server in as a tool source for your ADK agent.
114
Okay, let's make this concrete with a real scenario for when you should use MCP or ADK, because it depends.
115
It's a valid answer, I think, but it makes for a terrible video.
116
That's fair.
117
So let's say that you built your own coding assistant that can search your repo, open up files, run tests for you, and help you debug problems that you might have.
118
So first question, how are you actually building the agent's logic, right?
119
The memory, the tool use, the reasoning that it's going to have, and the retries when it inevitably makes something worse.
120
Yeah, so that's ADK territory.
121
You're defining what the agent's behavior is, how it decides to search the repo, for example, when to run tests, and how it should react to inevitable failures.
122
Agreed, and ADK helps you structure the cognition of the agent, so this involves the planning and the tool orchestration and the guardrails so it doesn't delete your production database for debugging purposes.
123
Yeah, that's some real nightmare fuel.
124
Tell me about it.
125
But let's say now that your assistant needs standardized access to that repository, and to your test runner, and access to your issue tracker.
126
That's where MCP comes in.
127
And so MCP is going to help us out to standardize how the model and the agent talks to those external tools in a consistent way.
128
It's that protocol layer.
129
Okay, I like this.
130
So, ADK defines what the agent should do, and MCP defines how it actually does those things by communicating with the world while doing them.
131
So different layers, but complementary roles.
132
It's not Rivals, right?
133
Yeah, only in the title of the video.
134
Well, listen, the algorithm demanded we have some conflict.
135
I get it. So, to wrap it up, if you're building AI agents, you're going to hear a lot about both MCP and ADK.
136
Don't let the names confuse you.
137
Right? Because MCP solves this connectivity problem and ADK solves the orchestration problem.
138
They're complementary and they're not competitors.
139
Yeah, and the real question isn't which one, it's what problem am I solving right now?
140
And hopefully this video gave you the context that you needed.
141
That was good.
142
I see what you did there.
143
It's corny, but I'll allow it one time.
144
You like that one?
145
I think, you know, I've got more if you'd like to hear some more.
146
Well, how about we save some for the next video?
147
Alright, I tried.
148
I tried, but if you're using ADK, MCP, or both, let us know down below.
149
And if this helped you finally make sense of these two concepts, please make sure to hit the like and subscribe to the channel for more content like this.
下载应用
AI 为你说出的每个句子打分
TRENDING
热门
关于本课
在本节课中，学习者将通过分析视频内容掌握有关人工智能代理的升级发展，包括模型上下文协议（MCP）和代理开发工具包（ADK）的特点与区别。学习者将利用所学词汇掌握与此主题相关的英语表达，进而提升雅思口语练习的能力。通过影子跟读练习，学习者将能增强发音与理解能力。
关键词汇与短语
- 人工智能代理 (AI agents)
- 模型上下文协议 (Model Context Protocol, MCP)
- 代理开发工具包 (Agent Development Kit, ADK)
- 外部工具与数据 (external tools and data)
- 自定义整合 (custom integration)
- 标准输入输出 (standard input and output)
- JSON RPC
- 权限令牌 (authentication tokens)
练习技巧
在进行英语影子跟读时，建议学习者选择轻松的语速。有意识地模仿说话者的语调和节奏，帮助提升语音连贯度和流利度。你可以先听一段，试着理解大意，然后暂停视频，跟随视频的语调进行反复练习。
视频中的内容涉及到了人工智能代理的连接和运作，这些概念有助于扩展相关专业词汇。而在进行shadow speech时，可以尝试将自己每次跟读的数据记录下来，反复对比自己和原音之间的差距，以便找出不足之处。努力提升可以让你的雅思口语练习更为有效，并帮助你在面对真实对话时能胸有成竹。
通过对于MCP与ADK的影子学习，掌握如何在上下文中使用专业术语，也能拓宽你的视野，让你在与他人交流时，自信地表达自己的观点。
什么是跟读法？
跟读法 (Shadowing) 是一种有科学依据的语言学习技巧，最初开发用于专业口译员的培训，并由多语言者Alexander Arguelles博士普及。这个方法简单而强大：您在听英语母语原声的同时立即大声重复——就像是一个延迟1-2秒紧跟说话者的影子。与被动听力或语法练习不同，跟读法强迫您的大脑和口腔肌肉同时处理并模仿真实的讲话模式。研究表明它能显着提高发音准确性，语调，节奏，连读，听力理解和口语流利度——使其成为雅思口语备考和真实英语交流最有效的方法之一。