بهترین کتاب برای یادگیری مفاهیم تئوری AI Engineering

[00:00:01]Hello, we 're back with a book, that is, introducing a book.

[00:00:13]Let's see why we brought a book all at once. Let's introduce it.

[00:00:19]Look, based on my observations, something happened a while ago.

[00:00:30]What happened?

[00:00:32]Some people have come and told the nation that if you go and use the LMs, for example, you are fooled, you should go and learn from the roots. I mean from the roots of some of them.

[00:00:53]I'm seeing that when someone comes to form a team, they create doubts, they create stress, which is useless. You have to learn from the roots. Then one of them comes and asks, "Excuse me, what should I do?" He says, "Let's go to the courses of a professor, for example, Bilbilkiano. See what he's doing." This guy has been a university professor for 20 years. Now he's scared. He says, " Oh, dad, I'm not going to teach regression now. I do n't know. I don't know the very basic things about deep learning. I don't know.

[00:01:31]These are very good things. But now suppose you become one of the gods of deep learning. You can't do that at all. Now, can you go to OpenAI and work there, for example? It's useless. Now, you can put together a bunch of libraries with Python, like 10 years ago or 17 years ago, and then launch your webcam, hold your fingers in front of it, and show it, for example.

[00:02:03]Write it on the page, for example, "Well, now you can fix it with JavaScript in 20 minutes, so what's the point? What are you going to do with it now?"

[00:02:15]As a result, they're just fooling the public because they're afraid of losing their jobs, and these are good things, but what's the point?

[00:02:26]Now, you'll know what it's like tomorrow morning. It's like someone wants to learn to drive and I know it quickly, he knows a few things, he wants to get a license, sit in the back of a car, become a taxi driver, he wants money, we say no, you have to come for 2 years, we'll give you a class on engine combustion, then we'll go and show you electric cars, we'll show you the engine, then we'll go and see if this engine is made of, for example, I don't know what kind of steel it is, and in terms of steel, for example, titanium is used in some places, we have to bring you titanium and the battery is something that Tesla made, we'll talk to you about the battery for, for example, 10 years, and so this unfortunate person died of hunger, a lot of them give up, they regret it, in general, so this is where I object to it, not that these are bad things, theological discussions [snoring sound] At the same time, now that you suddenly jump into the middle of, for example, LM, you'll be left with some lame places.

[00:03:40]Well, it means there are some topics that you should know.

[00:03:45]Now, this book in 2026 has been one of the most talked about books, and it's a really good book. This book doesn't have code in it. You can't do practical work with it. It's not Python code. I don't know if it's TypeScript. It's not about the cloud. You ca n't create an agent with it. But it has opened up a number of things for you that are interesting.

[00:04:16]Now, there are some places where it has code. Look, there are some places where it has done something. Now, it says, " Where is the sample code?"

[00:04:25]If you need my code somewhere, go there.

[00:04:27]Maybe you're in it. But look at the point it has and the things that this book can enlighten us. For example, I see that you are very excited about this. They take pictures from LinkedIn and I don't know if they make an Instagram account. Then suddenly it gets 70,000 likes, and then, wow, it's like this.

[00:04:59]Then, Prep Enir, for example, these are very strange to you. Well, it's in this book.

[00:05:08]Look, it's explained. Look, look at the application development model, the infrastructure development model.

[00:05:15]Look, it explains all of these in detail.

[00:05:20]Or, for example, someone comes and says, " What is AI engineering?

[00:05:25]What is machine learning engineering?" Then everyone is excited, wow, what an important thing you said, well, if you like these things, and well, that's fine, at least these are here as general information, you can see this. Now, look, he talked a little about models.

[00:05:51]Look, the first line is so beautiful. He wrote in the Foundation application, "need." That's right, he says, " need."

[00:06:07]That's what I'm saying. You do n't need to go and make a giant model, can you? Those days are over.

[00:06:19]Now, you have to read these things yourself. He says, "Have a high level of knowledge, not too deep. It's not bad." Then, he says, "Training Fund Foundation Model from Incredible Complex." He says, "It's hard work. It takes money. Now, "cost" does n't just mean money. There are many things. So, in short, look at this book.

[00:06:47]For example, he talked a little about training models. He gave some examples. Maybe someone who is seeing this. I think a very professional person is seeing this right now. He says, " Shut up, Dad, what are you talking about?" But a novice wants to say, "Sir, this is what I'm saying." What is model optimization?

[00:07:07]Well, here are some small things.

[00:07:13]See everything in the title, at least it has been skipped. See modeling, post-training, the steps to produce a model.

[00:07:30]See modeling, sampling, etc. These are good things. Now, this was about the model and this, and now this is a little deeper. From here, it goes into the engineering process. These are good things. See the user property system. Maybe you know the cloud very well now, you work with it, you bring a project, go up, but suddenly in the interview they ask you what the property system is?

[00:07:59]What is user pram? Don't you know, if you want to learn this, instead of going to TikTok and Instagram and LinkedIn, I don't know, someone can come and turn this into an infographic and let you guys like it, hey, tick tick tick, and wait for someone else to post it again. He comes to this book and tells you these things. Again, you want to know what a vein is. Look, he said, look at this. He explained it completely. What is a memory agent? Then, fine tuning?

[00:08:52]I got a lot of comments on YouTube. Mr. Fine Tuning, tell me what fine tuning is.

[00:08:58]Many times, instead of fine tuning, you can use vein in 85 times.

[00:09:04]And he didn't have all those rewards and GPW and I don't know, these stories. Well, when you know, look at it, not in your father's name, may God forgive you, the words that we wanted to say, he brought this here, he put it here, Dataset Engineering, see these are things that if you want to have a high level of understanding, as he himself said, I think this is a good book, get it, and in general, in my opinion, in my opinion, especially in Iran, now if you are outside of Iran, buy it, if you are in Iran, there are places that print them, cover them, put the same book inside, have 10 to 20 of these, 8 books, always have them, because even draw some places with a marker, for example, because these are good things. This is one of those books that was recommended, I have it too, and it's interesting. In short, I like it. It's very original. It's thrown everything out without going into the code and I don't know if it's an editor or anything like that.

[00:10:27]Now I said I'd recommend it, if you liked it, you guys should take a look at it too, if anyone knows of a good book. Comment and I'll introduce them to you. Thank you, goodbye.