AI科學家薪資公開｜16萬美金簽字費

What about you? I'm probably the middle class. I'm probably the medium class. I'm not particularly tall or particularly short. I'm probably 3 out of 3. Hello everyone, I'm Terry. Today we're here in San Diego. I'm at my friend Dave's house. Because my channel has always been talking about software engineering. I think there are a lot of other things in the tech field. Like data science, AI, or research. There are a lot of different fields in tech.

I'm happy to be able to find

Dr. Pai Pai, to talk about AI with you guys on this channel. If you guys are not familiar with AI, or don't know much about it, I hope this episode can give you a more basic idea. Dave is currently working in San Diego, USA. He's working at Qualcomm. What does Qualcomm do? Qualcomm mainly does AI chips, like DSPs, MCUs, those kinds of mobile chips.

Thanks for having me first

其實我是在台灣出生長大的,然後我是在台灣... 長台灣人 對,I'm Taiwanese 然後我在台灣念到大學吧 因為大學的時候大概大二、大三的時候我就想說就開始就 like career plan right 大學念什麼? 機電系 電機系 對,就 like mechanical and electrical engineering 然後那時候我就 like planning my future

I really wanted to be a professor when I was abroad. I also wanted to decide what major I wanted to study. I didn't know much about AI back then. At that time, machine learning was more popular. Like statistics, big data. Big data was popular back then. After I went abroad, I wanted to study an EC major. I started looking for professors.

有這麼多的 funding 然後可以去投資在 grad school like grad student 然後發現

Nobody wants to fund me until I find a project that they do AI. I just thought, "Okay, I need to spend time studying AI." I spent a lot of time because not everyone understands AI in college. So I spent a lot of time studying papers, articles, and related papers. Just your own interests?

-Yeah -You started your own path Yeah, and I also actively asked the professors if there were any projects I could do I also actively talked to the professors and their professors about what direction they were going to take And I was talking to them myself Yeah, that's the starting point Why did you choose to do a PhD in the US?

其實在美國讀PhD 最一開始是覺得說 在台灣我可能的選擇沒有很多 例如說大家畢業就是去什麼台積電聯發科 Like it's a norm, it's a good job 但是

I want to try something different. And then I want to know what's my potential be like. So you decided to come to the US. You came here alone. Yes. And you studied at Duke. Yes, I graduated from Duke. What's your major at Duke? I majored in ECE at Duke. You majored in ECE as well? Yes. And then I mainly did on-device machine learning. For example,

手機端或是移動端 IoT端的 Machine Learning 現在就是畢業了 去年剛畢業 去年畢業 然後現在在跨行第一份工作 對 但是也有實習對不對 之前有在Facebook實習 跨行也實習 對 對 對 其實我在念書的時候 因為其實大家知道 念書其實薪水沒有很多 然後我是靠獎學金撐過來 就是

I was a loaner and I went abroad. So I didn't have much money. So I had to find a part-time job in the summer. I was lucky to go to Kaotong first. Later I went to Facebook as a part-time researcher. And now I'm back in Kaotong. Our channel is about transparent cars. First, no. Wait, don't be nervous. Okay.

第一個我要問的就是 你當時在Facebook的實習的時候是哪一年? 薪水多少? 這個其實滿公開的 我在Facebook的時候是2020還是2021年 就一兩年前 對然後那個時候薪水是 他們是算實行 因為如果你是part-time researcher 不然你是part-time researcher還是intern 你是算實行 然後我記得是

五十出頭,五十一塊還是五十二塊? 五十出頭,我印象很深刻的是我的 monthly paycheck 是九千塊錢

喔,那還不錯。 還可以。

For students, it's very attractive. I'm more curious about software engineering background. I know a lot about software engineering, programming, coding. I'm not familiar with AI research, especially machine learning. I think it's a strange industry for most people. Maybe people are not familiar with this industry.

Sure. Can you introduce yourself in a simpler way? What is the content of your work? AI researchers are responsible for integrating AI into products and delivering them to people.

This is a very difficult job because AI scientists not only need to have the ability of software or hardware engineers, but also need to be able to develop the latest AI model. Our most familiar model is called State of the Art. This job includes researching the latest paper, publishing it in academic journals, designing the best models, and giving products to companies.

and discuss with software engineers how to optimize the AI model to the product. This includes top-level design, programming, and compression, and reasoning. This is the main job of AI scientists.

- -

然后然后 you have experience like intern experience or like PhD 目前为止我身边的人都是有at least PhD and from

top 10 top 20 schools 你從這個學校畢業的 is more likely that you are being more successful in this work and like 你已經有 accumulate publication when you in that kind of school 我覺得說是跟 PhD 還是差蠻多的 對 PhD 是就是四年以上嘛 Yes 很多人是五年是正常的 對 有沒有

Which schools in the US do you recommend to people? If you are interested in AI and research, and want to become someone like you in the future, do you have any schools or professions that you would recommend to people? I think there are many schools that have very good science and technology. For example, Stanford, MIT, or

CMU, Carnegie Mellon. Because our channel has always been about transparency. Public. And my favorite thing is to ask the guest how much he earns. That's what I do. Share your salary with everyone. Let's pause for a moment. This video is sponsored by Moomoo. Moomoo is a credit card trading software under Fudu. All accounts in the US have FDIC insurance. Fudu is a company listed on the US NASDAQ. Currently, there are more than 18 million users worldwide.

This platform has a special opening discount for the US users with a maximum of 10 stocks. Before July 31st, the opening of the US Moomoo was to send 6 stocks with the amount of interest. If the amount reaches $100, it will send one more stock. If it reaches $2000, it will send 3 more stocks. So, a total of 10 stocks with a value of 3 to 2500.

This discount event is only available for US residents with SSN account. I used to watch the trend chart and some company information on Yahoo Finance's app. But MooMoo is really professional in this area. The overall UI is also very good. And buying US stocks and ETFs in the US is free. A account can trade Hong Kong stocks and US stocks. So if you are in the US, I recommend you to try this trading platform. And free of charge, the event of 10 stocks and US stocks is really worth it. As long as you enter the gold, you will have it. Basically, it is free to send you stocks.

If you are interested in the US, I will put the link of the event registration under the video. You can go and see for yourself.

As an AI researcher, you just graduated from a PhD program. You are in this industry. Qualcomm is a good company. It's a great company. Can you share about your experience in San Diego? How was your first year? Okay, if you are a fresh grade, your first year as a doctor, I think it's not uncommon to see

250k to

300k 20,000 to 300,000 Yes, it's pretty normal. You graduated with a PhD, right? Yes, I graduated with a PhD. For example, if the market was not good that year, it would be normal to get a lower grade. If the market was better that year, you would get a 32 or 35 grade. It's normal. So the pay grade is between 25 and 30. Yes, I think it's normal. Of course, there is a difference between companies, right? Yes, of course, there is a difference between companies. And whether you have a competitive offer also affects it. What about you?

I'm probably the middle rank I'm probably the median I'm not particularly high or particularly low I'm probably the top 3 or top 3 I don't know what the structure of Qualcomm Pay is To be honest, it's what you call Sign-on Bonus It's a pretty high signature fee If you don't mind sharing with us your signature fee Most of my signature fee is 160,000 US dollars Yeah, almost 160,000 160,000 is okay, so it's 160,000 US dollars

Yeah, you just have to sign the name and it's $160,000. I sold my body for two years. Then you just have to write the name. Okay, Dr. Pi Pi. Just sign the name and it's $160,000. Yeah, that's about it. I think this is pretty cool. Because as far as I know, as a software engineer, I've heard that the highest fee is only $100,000. When you told me that $160,000 was a pretty high fee. I've never heard of such a high fee.

I was like, "Wow, that's awesome." Do you think this is normal for you? Or is it just a better negotiation fee? I think it's actually the negotiation component of myself. Just like software engineers, we are also being labeled. No matter how good you are, they're going to label you. When you're looking for a job, especially in North America, you have to work hard to invest.

各種不同的公司 不管他好或不好 或是你想不想去 你就是 have to put yourself out there 你可能會 I mean 很正常你可能不會over 或是不會有interview but you just have to try right 然後

But as you try, you will know your market value. And you will know how much you are worth. Yes. And then, maybe there is a company somewhere in the world, they pay you 200 grand, something like that. And then you will know that you are worth this price. And then you can compete with this price. And then you can raise your total package. Right.

So at that time, because I wanted to invest together, maybe you don't know, Terry is actually my business partner. So we, you know, investing means cash, so cash is a sign-on bonus. That's right. So...

Like I got a bunch of non-profit stocks. If you want to consider this industry, if you are interested in this industry, do you have any suggestions for them? Whether it's the path of learning or the direction of preparing for an interview, do you have any suggestions for them? Actually, I think everyone's success or path is unlikely to be reproduced 100%.

So if you just want to make money, I can guarantee you that studying to become a doctor is definitely not the fastest way. I think people who choose to study to become a doctor, or a researcher, or a research scientist, they all have some obsession with engineering, they have a passion. I believe you're not doing it for money. Yes, I really didn't think about how much money I would make in the future.

I was just like a normal college student at the beginning, and I was just fascinated by engineering. I saw that I could program, and I saw the LED light being lit up, and changing that, which was really cool. So I think I'm more like passionately driven.

Because if you want to make money, you go to Wall Street or do trading yourself. You can start a startup faster. It's way faster. There are a lot of restaurants in China. It's a lot more than us. Right. What if I want to be a...

AI scientist 嗯 What direction can I take to go this way? OK, so some people might think that I might need to get a degree Or go to Course Zero for a lesson I think this is helpful, it depends on how much money you want to spend How much money and how much time you spend on this But you need to know that AI is part of the industry Part of the tech industry

它有一個波,它有一個上一個下 所以很有可能如果你規劃五年十年

因為你畢業後這個浪潮已經不在了 Right 你可能想說 OK Dave 他要念一個博士 我也要念一個博士 但是你可能你博士出來 AI 已經沒有當年這麼紅了 這麼火的 對 所以我覺得你要對市場 跟整個 Tech industry 有一些基本的了解 By following Terry Chen's YouTube channel 沒錯 訂閱分享按讚

So you can only know how much time you want to invest after you have all these knowledge. You don't have to get a degree. You can find a company intern and do related projects.

You can go to some AI-related companies. They may not necessarily pay that much, but they may have very related projects. And it makes you feel very potential. It's worth trying. There's another thing I think is quite interesting. Like I'm a software engineer now. I write code. I don't know much about AI. But suppose I went to an AI company today. And then, although I still write, for example, I still write server. I still write my own code. But because I have a little contact with AI.

然后我如果跟比如做AI的那些research team 关系比较好的话 我可以慢慢的说 我会帮你们做一些东西 然后跟你们学一些东西 对 就通常这种方式其实还蛮好 我个人认为是一个蛮好切入的一个方式 就是你可以slowly去加入他们 然后可能在旁边

You can go home and take some courses for yourself. Or you can learn online. I think these are all ways to develop yourself in a new field. It doesn't mean that you have to go to a PhD in Chongqing. Of course, you have money and time. I think this is probably the best way. But the PhD takes five years.

不好讀,而且這五年來要花很多錢, depend on what kind of PhD, 有些學校會給你錢。 對。 對,那據我所知的話, Dave 以前也是蠻省的, 就是也是跟我一樣,我們都是窮學生過來的。 對。

還有一點我覺得就 echo 一下 Dave 剛剛講的那個我覺得就是

I think passion is very important. Especially in this industry, you can't say to yourself, "I'm going to make money today." Because if you really want to make money, you can invest 4 or 5 years to get a PhD. This is a very stupid choice. You really need to be interested. I think you will succeed. Thank you very much Dave for coming to our channel today. I will stay in San Diego for a few days. Is Dave single?

If you are interested in this cool AI scientist, you can find his IG here.

Thank you.