Tina Huang on AI, Education, Freelancing, and Boosting Personal Productivity

The What's AI Podcast episode 29: Tina Huang

Tina Huang on AI, Education, Freelancing, and Boosting Personal Productivity

In this new episode of the What's AI podcast, I had the pleasure to receive Tina Huang, a renowned figure in the realm of data science and technology, to share her insights and experiences about freelancing, getting into the field in 2024 and tips for better leveraging AI for learning anything and boosting your productivity.

The conversation begins with a focus on the role of Artificial Intelligence in education. Tina talks about how AI is revolutionizing learning and teaching methodologies, making education more accessible and personalized.

Tina also discusses the benefits and challenges of freelancing in the tech industry. She offers valuable advice on navigating the freelance market, highlighting the importance of continuous learning and adaptation, mainly leveraging ChatGPT! For those aspiring to carve a niche in tech freelancing, Tina's guidance in this episode is a goldmine of information.

Another significant part of our discussion revolves around personal productivity. Both Tina and I are fans of self-improvement books and resources and leveraging ChatGPT and other AI tools to the maximum. In an era where technology is both a tool and a distraction, Tina shares practical strategies to boost productivity and reduce procrastination.

Throughout the episode, Tina emphasizes the need for adaptability and continuous learning in the continuously (and rapidly) evolving tech world. Her insights are not theoretical, they are real-world experiences, making them particularly valuable for professionals and enthusiasts in the field of technology.

This episode is a must-listen for anyone interested in the intersections of AI, education, freelancing, and personal productivity. Learn more in the full episode on YouTube, Spotify or Apple podcast:

Full podcast transcript:

Tina Huang: [00:00:00] Absolutely, you should be using these new technologies. It gives you an edge. Anything that gives you an edge over everything else, that's a good thing. Like you should not be like, this is cheating if I'm using Chattopadhyay. No, it's not. That's like saying like 20 years ago, you're like, oh, I can't Google stuff because that's cheating.

Like who's not Googling stuff now, right? Working for a large company, it's consistent, but not safe. You get the idea of feeling like. you're safe because you always get the same paycheck every month, but you don't even know when you're going to suddenly get fired and then have the rug pulled under your feet.

Actual thing was like, what's the ultimate question? So it's like, you know, you can know the ultimate answer, but the whole thing is like, you have to be asking the right question to get the right answer. So I think prompt engineering, understanding these abstractions and stuff, that's going to be what really sets people apart because you're asking the After you know what the right question is.

AI tools will tell you what the right answer is.[00:01:00] 

Louis-Francois Bouchard: This is an interview with Tina Huang. Tina has a YouTube channel of now over 600,000 subscribers in the AI space, where she shares all her tips on better leveraging AI tools for both. Productivity, but also for finding freelancing work. Speaking of freelancing, she built a platform called Lonely Octopus, where she teaches artificial intelligence by working on real projects from companies and learning through practice.

In this discussion, we dive into the education space, but also artificial intelligence and how to better leverage ChatGPT and other tools for getting into the field and boost productivity. We also talk about YouTube, freelancing, building an audience, and more. If you enjoyed this episode, please remember to leave a like and a five star review depending where you are listening.

This helps my work a lot. This is Louis François Bouchard from the What's AI podcast, and let's get right into it. [00:02:00] How did you get into AI and into YouTube? And which one came first? And why did you, did you start both? 

Tina Huang: Yeah, so, in terms of YouTube, well, I was trained as a computer science in computer science, so worked as an engineer for a bit.

And I also worked as a data scientist. As you and I both know, a lot of like these technical roles are pretty interchangeable. Right, like between data science, like AI engineer, where like software engineer, a lot of these, like people kind of go like back and forth in a lot of them. skill sets are pretty interchangeable too.

So YouTube definitely came first and in the beginning I focused on data stuff a lot of time because I was working as a data scientist at that time, although I also talked about computer science and software engineering. with the advent of AI, I started talking about how to use AI, which again, it's, it's honestly very similar software engineering because we're not trying to develop our own [00:03:00] models in this case, usually we're just using the models that are there, like fine tuning it, maybe something like that. So, there's so many opportunities available in AI right now. So I started incorporating those topics into my YouTube channel and my other forms of social media as well. 

Louis-Francois Bouchard: What's your ultimate goal to become an expert?

In, in your field or more to focus on YouTube and to find like content to share and be useful on YouTube, what, what's the, the main goal or, or your main like mission? 

Tina Huang: Well, from the very beginning, I did not think I would be doing YouTube. Like, if you asked me that a few years back, I would never have said my goal.

My, the reason I did a master's degree and went to go work for these companies is to become a YouTuber. that was definitely not the case. Like I wanted to get good in my field. but then when I started doing YouTube, I started applying and being able to talk about all this information. I would say now my primary role [00:04:00] is to be able to learn certain aspects and then like apply my skill sets and then be able to present that in a way that's more readily available for other people.

Louis-Francois Bouchard: Yeah, I've seen you have a lot of different videos in the sense that like a lot are on self improvement, I would say, and others on how to learn, obviously, which is linked to the first one I said. Lots are on learning, but also on our, on data science. I wonder if from. What we've seen on YouTube and from like all the bigger YouTubers or the people that, that teach to start a new venture, they also, they always say to focus on the niche.

And I wonder if, are you also aiming for a specific niche as in the, like the self improvement niche, or are you just sharing what you, what you like to share? Because they, as I said, like it goes from AI to leveraging to the ChatGPT to some other things [00:05:00] that are not. I don't related to both, so I wonder, are you just sharing videos that, that you enjoy doing and that's it, or are you aiming for a specific audience and to reach more people?

Tina Huang: Yeah, I think that advice, I starting off a niche is definitely really, really important because I started off a niching very heavily into data science specifically. as time went on, though, I started talking about other topics that were interesting to me, and they also related. there actually is an overarching theme between all of them as well.

which may not be like super obvious, just looking at the videos themselves, because my whole kind of thing, is learning new things. I'd be able to learn quickly and then be able to apply these skills in a way that improves your life. Just in my specific case, usually the things that I'm trying to improve, that I'm learning is usually on the technical side, as opposed to other things that I'm learning.

[00:06:00] So, yeah, and in terms of like productivity based projects that I do as well, all of that kind of ties into this sense of helping people find a career and then finding a role that they truly enjoy in order to craft the life that they want to have. so I would say that's the role that I put myself in, and it's definitely not as specific as say, like, I make data science content.

Tutorials, right? Definitely not like that, but I like to think about it more in terms of, like holistically helping people improve their lives in order to create a career that they want with a technical tilt. 

Louis-Francois Bouchard: Yeah, that's, that's really cool. And you've actually created as well, Lonely Octopus, and I wanted to dive into the education side a bit in this.

And first, to talk about the formal education, I personally have some thoughts on it. [00:07:00] And I recently left a PhD to do online, to create online education. So I guess that illustrates my thoughts, but I would love to know your opinion on the current graduate studies and formal education, like going through university.

is a master's still relevant, should you should do a PhD, what are your thoughts on education, especially if we talk about the AI or data science space? 

Tina Huang: So in regards to higher education, I would say it depends a lot on the person. So for if you're someone who is a self starter who prefers self learning, the ability to do that is so much higher now.

There are so many ways, especially customized learning methods with using AI, even like ChatGPT. It offers you a lot of opportunity to learn things by yourself. With that being said, though, I wouldn't discredit having going for a degree program. Like, for example, if I was 17 years old and then I [00:08:00] wanted to be an engineer or a data scientist or an AI engineer.

I would not have the discipline. It's not that there's not enough information available to me. I just wouldn't have the discipline in order to actually learn this stuff and then get myself to, to the point of getting a job. So I would still go for a degree. I think if you're going for like a master's degree or a PhD, degree, then you really got to think like, is it because I want to just learn more things?

if that's the case, you generally don't need to go get another degree, but save your aiming for an academic role. If you want to go be a professor, correct me if I'm wrong. You can't like not have a degree and go be a professor, generally speaking. so yeah, those are, those will be things that I think is important to keep in mind.

Louis-Francois Bouchard: So to you, it would mainly be to. like a guidance or some specific tasks to achieve and in order, in order to learn and to keep learning, but do you think the paper or the like diploma at the end is [00:09:00] also something that will keep having value in the near future or is. Online learning, something that will be just, or like self learning, something that is or will be just as equivalent as having like a formal degree with a paper at the end.

Tina Huang: It's, it's not as useful as it used to be for sure. Like before you're like, I have a PhD, I have a master's degree. People are like, wow, you know, but nowadays it's like, Oh, like a lot of people can have it. But you can like, even look at the software engineering industry specifically, there's a lot of people working.

Really great companies who don't even have a degree at all So I would definitely say the value of it is not as important and just because you have a degree doesn't mean you're gonna be Guaranteed to actually get a job either anymore Yeah, so it used to be like if you have a degree then you're automatically in a really good position in terms of like, where it is that you can go for work.

I, I really don't think that's the case anymore. And on top of [00:10:00] that, I think a lot of people are pursuing more like freelance careers. I think that's like what's popular, these days with people going like, Hey, like, why do I need to work at a job and have to go to the office when I could? work remotely from anywhere in the world, and I can work, different freelancing jobs and be traveling at the same time.

So in that case, it definitely does not depend on your degree. Nobody's going to hire a freelancer because they have a cool degree. it's very, very much going to be based upon your skill level, and all the other things that you have done and your self marketing. 

Louis-Francois Bouchard: if, if A person wants to get into the space like data science or just artificial intelligence in general.

How would you recommend them to start by themselves online and to finally land some kind of freelance work? Like what skills should one go look for? And you mentioned ChatGPT, should they just ask ChatGPT and Go [00:11:00] only use that, or should they go on some platforms like, like your own with Lonely Octopus or start on, I saw you mentioned that you, you and you like deep learning that AI and their courses, how would you recommend them to start?

For example, now we are seeing a lot of people. That don't know anything about programming that want to get into AI because of language models and just the power it has for analytics and other stuff. So for someone with no background at all, what would you suggest them to do and to start with? 

Tina Huang: I'm going to have like a really dumb unsophisticated answer to this.

Just choose something. Like, choose anything, right? Go do a certificate program on Coursera, if you can sit through that. If you're like, I have a short attention span, go do a short course on it. I think people put too much emphasis on which course to choose or which program to take. Look at a course, you like the [00:12:00] instructor, it has good reviews and you feel like you vibe with it.

then just choose it. Like a lot of these introductory courses in particular, they talk about the same thing, really. It's just presented in a different type of way. And some ways stick with some people and stuck with some ways stick with other people. my general recommendation is choose a course that you like the instructor of.

It has good reviews about somebody that you trust. Like somebody that you're like, okay, this person is trustworthy to me for whatever reason it is, something that's very hands on is the best because, if it's something that you already work on hands on, you would have portfolio projects already.

Like, that's for me, like, the reason why I made Lonely Octopus is not because I think I can make a better course than, than like the experts out there, right? I would never claim that. so I actually aggregate a lot of their courses and I put them in such a way that, you're learning these courses specifically.

In order to land a freelance projects, I actually match you [00:13:00] with a company because in my head, you can learn so many courses, but that's not gonna matter if you don't actually use it. So everything that you learn, you will use at that company. And at the end of the program, you will have. a actual freelance project on your resume, which I think speaks more than doing 500 courses even.

So yeah, that's kind of like how I see things. Like if that's something that you vibe with, definitely check out Lonely Octopus. I do, I take the best courses. I also create some of my own, but really the highlight is for you to actually get that freelance project, the real freelance project from the real company.

If you want to do a certificate program and you're like, I really like Google's certificate program. It has a capstone project. Great. Choose that as well. generally speaking, if you don't know which course, which next course it is for you to take, it probably means that you should be doing a project.

Louis-Francois Bouchard: But I think it can be also very intimidating to start when You don't know anything about even just a [00:14:00] Python environment or what thing to open. So I guess a course, like a very basic course, as you said, something not even on AI, but on Python or just basics of programming could be super helpful. And then I would also actually start right on projects.

On my end, I learned coding during the summer in the During my college, I don't remember when, but I learned coding by programming like a dumb game on mobile, just because I wanted to try to make some money and to create and to apply an idea that I had, and it just taught me a lot about Java and other programming.

So it was, it's definitely better and it's also super motivating. So you just work and learn a lot more. But I had another question about online learning. We are seeing. Lots of certifications like Google certified or Intel or whatever. Do you think [00:15:00] those are relevant for companies or for people that, that try to hire someone?

Should, should the students be looking for certifications or just look for interesting projects that you can show? And, and build on your own, like, are, are these certifications, do they have any values? 

Tina Huang: Yeah, so I actually know the answer to this because I talked to a technical recruiter about this before.

So, what she told me is that if you absolutely have no background at all, like say you're fresh out of high school, you only have a high school diploma, and you want to work at a tech company or something like that, And, and try to get a job there, right? As like an AI engineer, data scientist or something like that.

then it would be worth having at least just having one certificate because it shows that you at least did something that is relatively, like normal courses. So you didn't just not have any, not do anything. but after [00:16:00] that one certificate that you have having 10 certificates does not add any value to your resume after that.

And if you do have a degree, even if it's not related to computer science or engineering or data science, like that, it's not worth having like an additional certificate. I would say if you have anything kind of related, if you have a stats degree, if you have like, a science degree which generally have stats in it or something like that.

and you list out some of like the anything related to engineering or statistics or computer science or programming courses as some of the courses that you took. If you have that, you don't need a certificate. If you have like a fine arts major, maybe, and you have to take like zero of these courses, then maybe get that certificate as well.

but after that, not useful just work. What they're looking at is the projects that you've already done your skill level. Can you pass the interview? and then they see like, Oh, this person has used these technologies. a lot of times what [00:17:00] these companies are looking for is also not necessarily how many things that, you know, it's about your ability to learn new things.

Because why do I need to know how many things, you know, if I could just, you know, everybody can just go to ChatGPT and like ask for stuff now. It's more about, do you have the ability of learning new things, of keeping up, of being able to problem solve using the tools that are available? That's what they're looking for.

Louis-Francois Bouchard: I guess now what's pretty much what's important is to, to use ChatGPT and to better leverage. Those generative tools that just make you much more efficient and allow you to learn a lot more stuff. So it's, it's really weird how it's just like math, I guess. Like you, you don't have to remember how to derivate something precise.

You can just Google it or, or use a calculator. And now it's the same thing. Like you don't have to remember how to do a bubble sort or whatever. You can just ask ChatGPT to print it right away. So it's, it's just very, I'm [00:18:00] just excited to see what it will look like. It will basically, like, people will be looking for interesting people that.

Are willing to learn and do more, but I feel like the skills themselves will become less and less essential or, or looked into, I don't know if you, if you would agree to that, or if you think that, for example, mastering Python is still super relevant and will stay super relevant other than compared to mastering.

Those language models are like programming aids. 

Tina Huang: Yeah, I think I actually really agree with you on, on that front. I think your ability to know how to implement a bubble sort is, is becoming increasingly irrelevant. Yeah. is. Yeah, like it's becoming very irrelevant at this point. Implementation in general is becoming a lot more irrelevant.

They say the future coders are going to be more managers or editors [00:19:00] as opposed to actually people who write code. I think there's even more of an emphasis on problem solving, big picture thinking, when you should be using that bubble sort as opposed to not using that bubble sort. higher level abstractions.

I think it's going to be the most crucial thing. What you talked about in terms of technologies, absolutely, you should be using these new technologies. It gives you an edge. Anything that gives you an edge over everything else, that's a good thing. Like, you should not be like, this is cheating if I'm using ChatGPT.

No, it's not. That's like saying, like, 20 years ago, you're like, oh, I can't Google stuff. Because that's cheating. Like, who's not Googling stuff now, right? So it just means that you're an early adopter, so you have an edge over other people, and you should take advantage of that edge in order to create and do whatever it is that you want to be doing.

In the future, mastering Python, I think it's not I think it's easier to master Python, way easier. Like before, if you're trying to master Python, you're like, [00:20:00] run into a problem, you don't know what it is, you go look at Stack Overflow for like two hours until you find something that helps you, probably.

or, you know, it's very hard to know how to improve on something unless you have a mentor telling you, like, how do I improve this code? you generally don't know what you don't know with ChatGPT. And and like other software out there, it's a lot easier to master these these languages. So I still think it is important to learn these things right now, especially if you're trying to join together like different APIs and stuff together.

It's still important. I think it's a lot easier to do so moving into the future. I do believe it's going to be less and less relevant, your ability to spit out code is going to be less and less relevant. it's more about, do you understand how the language works? Like, do you understand the fundamentals?

More the computer science principles, right? Or the software engineering principles that govern writing code. Those will become more important. 

Louis-Francois Bouchard: Yeah, I completely agree. And it also, a [00:21:00] bit loops back to school where. It can actually stay useful because it teach, it teaches you like all the basic stuff and the, just to understand the whole system to then be able to know what, what you need to do next, or at least to know what to ask ChatGPT.

Tina Huang: Yes, that's the thing. Prompt engineering is something that is so crucial. you know, like that, that movie, Hitchhiker's Guide to the Galaxy, the analogy I wanted to give is in that. movie, I'm not going to spoil it, but it was posed, somebody posed, they wanted to know the answer to the universe and everything.

And then they gave him the answer. The answer was 42. And then the actual thing was like, what's the ultimate question? So it's like, you know, you can know the ultimate answer, but the whole thing is like, you have to be asking the right question to get the right answer. So I think prompt engineering, understanding these like abstractions and stuff, that's going to be [00:22:00] what really sets people apart because you're asking the right questions after you know what the right question is, AI tools will tell you what the right answer is.

Louis-Francois Bouchard: Yeah, I completely agree. 

Hey, this is a quick interruption of this episode to remind you to leave a like or a 5 star review depending on where you are listening this episode from. It helps the channel a lot. And speaking of learning, please do share the knowledge with your friends by sending them this episode.

Thank you for watching and I will let you enjoy the rest of this episode. I know that you really like to use ChatGPT for learning and for doing like improving productivity and you also did a lot of videos on productivity itself and on like self help I wonder what would be your recommendation for The people that currently have a full time job, for example to pay their bills and etc, but that they don't like And it is not in programming.

Like, do you have any advice for them to switching [00:23:00] fields while staying sane mentally? 

Tina Huang: I think I have something I can say to this. Cause for Lonely Octopus, over 90 percent of our students are actually, they are full time employees. So they're, they're trying to switch their career, start a freelancing gig, like things like that without actually quitting their jobs.

so the way that we designed that program is, you gotta do your full time job. You can't like not do your full time job. I don't want to be doing a bad job. so what you do is like you create your study plan and a study schedule that's suitable towards you. I always say it's not about trying to get as much done as possible.

It's about consistency. Say you're going to do like 3 hours a week. That's fine. If you can manage to do 3 hours a week, you don't need to do more. You don't need to like, that's okay. And you can even take breaks. And as long as you focus on learning the right things and then applying those things in the correct way.

that's all it is that. You need to do like, I [00:24:00] think people have this assumption where they're like, Oh my God, I need to learn like 20 hours a week. And like, and then they burn out because, you know, they're like, why can't I learn 20 hours a week? Well, maybe because you have a full time job and a child and a family.

so you just need to be a lot more realistic about what it is that you're doing. Another big thing for me personally is like, I have no self discipline. I got my Before, if you tell me just to learn something, I'd probably like give up after like a week. So what I always have to do for myself is I need to build in accountability.

So I either like announce it socially, like and make sure I have friends, we're a Lonely Octopus. We put people into groups of people, so they're keeping each other accountable and I try to get a job. Or like some sort of job that's associated with what I'm learning because I can't just randomly quit, right?

Like, for example, if you're in Lonely Octopus and you have a freelance project with a company, you're like, you're kind of like forced to learn these things and implement it because you said that you would do it. And then you're going to [00:25:00] have a presentation in like two weeks. So you're like, shit. Like I better, sorry about that, but you're like, I better like learn everything, so that I'm able to do that.

It's a lot more motivating than if you're just trying to learn something yourself, because then you could be like, Oh, you know, it's okay. I'll just do it tomorrow. I'll just do it tomorrow. And then tomorrow never comes. 

Louis-Francois Bouchard: Yeah. And that's actually how I started into AI in general. I, I got my first class like towards the end of my engineering degree, and then I just really loved it.

And I don't know why, but in order to learn more, I decided to create YouTube videos about artificial intelligence and to explain what I was currently learning. So that was like, I, I guess it's funny because what you mentioned is basically to have a deadline and like not a force to work, but something that, that you have to do, whereas on my end.

It still felt this way, but I didn't have to do anything. It was just like me that wanted to put out a video every [00:26:00] week. So I had to learn and I had to create the video, but I didn't have any hard deadline per se or anything obligated to do. And I wonder if, if you did the same with YouTube, like leveraging this platform to force yourself to do things you wouldn't have done otherwise.

Tina Huang: Oh yeah, absolutely. Even with my live streams, right? I do weekly live streams where I'll be like rating AI courses, something like that, right? How am I supposed to rate an AI course if I didn't take the AI course, right? So like that actually forces me to like Okay, I gotta like learn a lot more things like I'm sampling different stuff or for example I would be doing like a video like we're like a live stream on like building an AI tool, right?

Well, like I gotta go build the AI tool so I can make a live stream video on it So those keep me like very very accountable Because I can't just be like, oh just kidding. Like I don't know how to do it Right. Where I can't just be like, I'm just going to leave and not do it. Cause I [00:27:00] said I was going to do it.

And there's like thousands of people who are going to see that video. so to not embarrass myself, I have to make sure that it's actually good. Recommend live streams for sure. if you want to be super accountable, cause like if you're live, you really can't fuck up. 

Louis-Francois Bouchard: Yeah, that's for sure. Yeah. What would you suggest for someone to motivate themselves to have like fine deadlines or like, because it's really cool that Lonely Octopus allows you to have a freelance project with an actual company. But if you cannot manage to find that, how would you suggest people to have commitment and what kind of commitments would you suggest them to take?

Tina Huang: First, identify what you care about, right? What I care about a lot of times is I get very, I don't want to disappoint people because I'm a people pleaser. So I care about that. Maybe like you don't care about that. You're like, I don't care what you think, then that's not going to work for you. But like for me, I'm people pleaser.

So if I tell, I think about it, like [00:28:00] what would make me not like, what would make it so that it would be very hard for me to not do something. If I tell a lot of people that I'm going to do something, then I have to do it because I'm going to be too embarrassed if I don't do it. the whole idea with the job as well is, is like very similar.

The accountability is because like the idea of having to confront my boss about not doing it scares me so much that I do it. Maybe you don't care. That won't work for you then. So you gotta think about like what you actually care about and try to implement that. For some people, like, like what else do people care about?

Like, like social accountability is a really big one. Like there's a lot of cognitive dissonance. So usually like telling People that you love and you care about that you're going to do something will make you do it because you don't want to break your word right to to doing something like that, it could be like paying the bills.

If you put your I'm not recommending that you quit your job. I just want to put this out there. Okay. But like money is a big motivator. Haven't you noticed like, [00:29:00] say, if you I don't know, like, if you've had this experience, maybe as a PhD student, like, if you don't have that much money as a student, you're, like, pretty motivated to make extra income, right?

So you're, like, okay, like, I better learn all these things. So, money is a, it can be a good motivator for some people. it's not the best for me, but for some people, the idea, like, if I do this, I can make, like, XYZ amount of money, then that's a good motivation to have as well. yeah, I would say like, generally speaking, what works for almost everybody is social accountability, because as humans, we are designed to want other people to like us, 

so. that is something that is really, really good to tap into. I also want to make one more note, like with Only Octopus, yes, we give you like a freelance, like an actual freelance project from different companies, and I think that's really, really helpful. it's, it is, I agree, it's more difficult to do that by yourself, but you can kind of like replicate this, like for example, just go to your, I don't [00:30:00] know, like your neighbor and go like, hey, I'm gonna like build this thing for you, and you told them they're going to build it, so you can actually also Go and build it out for them because generally people like free things, so they're probably not going to say no to that.

Louis-Francois Bouchard: Yeah, I think that would be a really good motivation. On my end, I just wanted to mention because I will assume some others are in this situation, but I think the social aspect doesn't really work for me. Like for example, I Started YouTube completely anonymously, and I didn't share it with any friends, any of my family.

I was just like shy and I, I didn't want to tell anybody that I was doing this in case it flopped completely and didn't work. So it was just, I was pushing and grinding, but completely anonymously. And then I started to, to use my, I started using a text to speech and not like nothing from me. Then I started to record with my own voice and then showing my face and etc.

But. Just [00:31:00] recently, I I don't want to say this on the podcast, but I said to, to some of my close ones that I wanted to do something, that is a bigger project than I usually do, and I feel like it actually puts more pressure or like, it, it makes, it makes it harder for me to work on the project. I don't know why.

Just like it's a. Maybe, yes, it's a good extra pressure, but on my end, it's maybe too much, or like, I prefer to not, not talk about anything to anyone and just work on it and then be proud of myself and share it once it's done. But yeah, social, the social aspect doesn't work for me. I really don't know why, but.

Tina Huang: There's like a bit of a distinction here, right? It's actually very small. If you tell other people that you're going to do something, it actually decreases the likelihood of you doing it because just you telling them makes you think that you already did it and your brain gives you a dopamine rush, so you're less likely to do it, but [00:32:00] social accountability is more that you force yourself into a situation where you can't not do it.

Like the cost of not doing it is too painful. Like, for example, if I sign up, like I'm, if I sign up to say, I'm going to do this live stream in front of thousands of people, right. I'm probably going to do that live stream because the idea of disappointing thousands of people is going to make me very nervous.

if you just tell your friend, you're going to do something, then. It doesn't really have that kind of thing, right? They're not gonna be like, why didn't you do it? Like, afterwards. So that's kind of the social accountability. Or like, say you have a job. That's social accountability as well. Because if you just like, decide to not show up one day.

Or like, just not do it. And then you have a meeting with your PI or something. Doesn't that make you very worried? Like, you're like, oh my god, shit, like, I have to, like, do these things, because what am I going to say to my PI if I just tell him, oh, I didn't do anything for two weeks? So, I think it's, like, [00:33:00] more, it's like a little bit of a subtle difference, in terms of, like, how it is that you do that.

I actually did have one question for you, if you don't mind me asking. Oh, of course. The fact that, you know, you've done, you've like started a PhD in these topics, where do you think is the greatest opportunities in using artificial intelligence? And where do you stand on in terms of what you're doing in order to take advantage of it?

Louis-Francois Bouchard: I see. It's just so hard to answer. It's just, it, I think it will just change all the industries. So it's very difficult to choose one. But on my end, I think education is the space that might be changed the most, especially graduate studies, but even university or college or high school, I think it will be changed a lot.

Just having, for example, here in Quebec, we, we lack teachers, especially good ones. And even for younger people and so we have like huge classrooms with lots of students that cannot learn [00:34:00] properly because they have like one teacher for 70 small children or whatever. And I can't, I can just imagine having like a personalized tutor to each of them or like with an hologram or something like it can just change everything just to have infinite teachers.

So it's, it's really cool, but, and that's also what I'm trying to, to focus on now, like with, towards AI, we are trying to build courses and, and other ways of, of learning than the traditional way. So I, I guess. Yeah, I would assume there are other industries that will be challenged even more, but on my end, I think it's very promising, and I think in a positive way for education.

Of course, I would assume some, some jobs will be lost along the way, but in the end, I think it will help teachers more than hurt them. [00:35:00] I would assume. Like, obviously, there's the, there's two sides where the students will leverage things like ChatGPT to cheat or to not learn. And on the other side, teachers and schools can leverage those language models or Any other tools to better teach the children.

So I don't know. I think in the end it will, it will be quite a positive gain for the education space. I think that would be my answer. And what, what is your answer to that question? 

Tina Huang: I think there's just so much opportunities available. yeah, there was a really good talk by Andrew Ning, where he was explaining the different layers of AI.

he says that there's like the base layer, we're developing like chips, right? The hardware. And then there's people who are developing like fundamental technologies, like, like different types of large language models and like generative AI models. And there's a layer on top of that, that are developing platforms for people.

Like for example, like the ChatGPT, who's developing like [00:36:00] ChatGPT store and things like that on the platform level. Or, there's like companies who would be like labeling data and like things like that. Right. and then on top of that, I believe. Don't quote me if I get this wrong. Then you have the application layer.

This is when you can apply, artificial intelligence to different Industries around you. So I think in that layer, there's just so much that can be done. It's like prior to these days. most of that money, like most of the attention is being focused on things like marketing, like advertisements, stuff like that, because there's a lot more return on investment, and say, like doing small apps for like healthcare or something like that wouldn't be as worth it because the ROI is just It's not good, right?

But nowadays, because of generative AI and these different models that are there, it lowers, it significantly lowers the cost of being able to improve those industries. So [00:37:00] the return on investment is actually good enough now. It can be, I think healthcare is going to be a really, really, really big one.

It's traditionally a place where there's like a bunch of data, but the data quality is like kind of not very good. It's like kind of shit. and. There's just a lot that can be done in that side. Education, 100 percent agree on that side as well. I think in terms of like climate change and like these types of things, there can be significant development in those areas as well.

yeah. And personal productivity, just like that's a no brainer as well. And productivity, corporate productivity, personal productivity, being able to live life better, and things like that. I actually, I also see that as a place of huge, huge opportunity. 

Louis-Francois Bouchard: Is there one industry that you personally want to contribute more?

Tina Huang: I do see myself like playing, like how I like to envision myself at least. Like I told you earlier about the way that I saw my channel, which is I'm trying to help people like take advantage of. New [00:38:00] technologies, try to help them develop better lives that develop better careers and to develop better lives right through the avenue of learning new things.

So, I'm trying to put myself in a position where I'm not an expert in terms of developing these fundamental technologies. I never claimed to do so, but I also do know more than most people who are, say, like, just talking about AI, like new AI things that are coming out. So I like to see myself. as someone who can help democratize the information that's available, like, here's some tools that you can be doing.

Here's really cool things that other people are doing. Give you some ideas about what you can learn so that you're able to contribute to it or it's a different field as well. Like, maybe I can tell you about. the fact that people are making a lot of discoveries like drug discoveries, like small molecule drugs, by using AI in a specific way.

And if you're someone who's interested in that field and you see that video, you see that piece of content, [00:39:00] my hope is that for you to be like, Oh, this is interesting. And then look into it and maybe even start working in that field. 

Louis-Francois Bouchard: Speaking of your, your YouTube channel, what's the best thing that your venture on YouTube bring you?

Tina Huang: What's the best thing that it has brought me? It has forced me to learn things that I would never have learned before. 

Louis-Francois Bouchard: Yeah, that's a very good answer. Do you have any anecdotes on, like, did it allow you to find, for example, freelancing work or any other opportunities that you couldn't have landed without this, this YouTube channel?

Tina Huang: Oh, yeah, absolutely. Absolutely. I say like if you don't have a if I didn't have this YouTube channel, I would not have like nearly as many opportunities available to me like these days. I like I don't need to go find freelance projects like freelance projects come and find me. Right? Yeah, because I spent so long building up a platform like this.

That's also why I also tell the people I work with from Only Octopus, like one of [00:40:00] the major things I push them on is you have to start building an audience. So you're not the one trying to get freelance clients anymore. The clients will come directly to you and be begging you to do their thing because they know that you have the ability of doing so being able to also just like being able to meet so many people like you, like, how would I have known you if I did not have a YouTube channel and like our mutual friend, Ken, like I didn't know him until. we started doing YouTube as well. And all of these connections, all the opportunities that I get to like be able to speak to people, all of this comes from YouTube.

Like who would want to listen to Tina talk? You know, like if I didn't have, if I didn't have a voice already.

Louis-Francois Bouchard: How would you suggest people to build an audience? Like that's, I assume something very difficult. For example, in our space, like data science, AI, it's becoming super saturated, like with lots of people just sharing news or.

ChatGPT related things. So how would you suggest someone starting in the field to build an audience and try to [00:41:00] find contracts and, and opportunities like that? Would you suggest them to do whatever they want or to create a YouTube channel and newsletter? Like, is there. Something you think is either easiest or like more fitting.

Tina Huang: Yeah, I think so. Think about it from two pers I like to say if you're being going to be serious about social media, kind of have to know the reason for it. Like if you're someone who, wants to use it in order to get contracting and like freelancing jobs, that's one type of, of method. The other method.

would be like someone like me who's trying to build an audience because I want to create content like content. I want content to be my business. So on the freelancing side, if that's what you want to do, show your work. Like whatever you're working on, all these things, the technologies that you're using, LinkedIn is probably the easiest.

To do that, just snippets of your work, like, Oh, I use this technology to build this, put a screen cap of it, put like a screenshot of it. That's it. And then other people [00:42:00] who are looking for people to work in that technology might just search the hashtag and then find you and be like, Oh, this person can work on this technology for me because they already did a project before, using like writing articles, like towards data science, like towards AI, like these, stuff.

Is probably like medium articles is probably like the next easiest step to do something like that. if you're into that on the content side, writing like sub stack or like hive beehive, I think is if you're into writing, that would be a good start as well as LinkedIn. If I were to start over again, knowing the fact that I'm a content based business, although I do freelancing stuff, I'm mostly like a content based business.

I would start with YouTube shorts. Because I will cross post that onto TikTok, YouTube and Instagram Reels. When I first started there were no shorts, like that was not a thing. But if I were to start today, that's definitely where I would be doing. That area is not that saturated. Especially [00:43:00] on good educational content that's entertaining, it's actually not saturated at all.

Louis-Francois Bouchard: But would you say that the conversion into a real audience is not necessarily that good with short content. Like, for example, a, a larger TikTok audience or Instagram doesn't necessarily convert as well as a, a good YouTube audience. So I don't know if you've seen that or if you, if you have any thoughts about that.

Tina Huang: No, I don't think that's the case at all. I really don't think that's the case at all. And it's not like you start out with shorts and you have to like do shorts forever. really successful shorts creators always convert to long form creating as well. And then really successful long form creators also start doing shorts.

That's what I'm trying to do. Like, I'm not that successful on YouTube, long form in itself, but I'm trying to improve that and also trying to like, try my best to expand into like shorts as well. So it's not like you do one thing and you stick with it. I think you just pick whatever [00:44:00] is, gives you the best return on investment.

And then think about the conversion a little bit later. Actually be good first and then think about conversion. 

Louis-Francois Bouchard: Yeah, that makes sense. And it's also much more accessible than trying to make, like a, whatever, 20 minute long explanation on something. Like you can just do a, a 40 second, very good and concise explanation, which is actually quite hard to do, but still seems more accessible at first.

But yeah, I've, I've also tried, I'm now trying to into shorts, but mostly podcast shorts. But otherwise the, the like real shorts where I try to explain something in less than a minute, it's so complicated. It's like so hard to have a good catch and storyline and try to. Explain, like, try to be pertinent and useful within a minute.

It's quite hard, I feel like. 

Tina Huang: I would agree. I definitely have not figured out the shorts game. But with that being said, AI technology is [00:45:00] fundamentally transforming a content game because being able to produce content, using AI, if you're not using AI, that's That's like, not good. You should definitely be using that.

Like, for me, I'm actually experimenting with, using, like, script writing, like, being able to analyze videos and see what does one, doesn't do well. I think another part that's, like, 100 percent being used now, and it's gonna be even rising later, is how, like, I'm sure you've heard, you've used, like, HeyGen and, like, things like that, right, for, like, avatars.

They're really good like really good and if it's only a one minute video people aren't going to notice that it's avatar So I think a lot of videos that we're watching are they're being created Using avatars now where they definitely will be increasingly being used by creating by being created by avatars 

Louis-Francois Bouchard: Before we end this, I just wanted to come back a bit on freelancing.

I have a few quick questions. First, are you still doing freelancing, even with all the YouTube content that we want [00:46:00] to push and Lonely Octopus? Are you still taking on contracts and working for other companies? 

Tina Huang: I am. I am. I. Started when I first started, like I quit my job and did YouTube. I definitely did more of that because I'd like freak out that I can't make enough money.

nowadays I still do cause I like, I enjoy that part, of my work. I think I'm someone that I get bored if I just do one thing continuously. So being able to switch between like technical things and. I guess like YouTube is kind of technical, but more like, you know, talking about stuff and like, storyline based stuff.

Like it's very nice to be able to switch. I think that's also, that's kind of also the beauty of freelancing, right? You get to choose how much work it is that you want to do and what type of work it is what you want to do. Yeah. 

Louis-Francois Bouchard: Yeah. That's really cool. And so my next, you almost answered the next question, but how would you differentiate, for example, freelancing versus working at a big company?

Versus entrepreneurship or building a startup, [00:47:00] like who, who should go for which? 

Tina Huang: I would say working for a large company, actually, Ken was the one that gave a really good, so credits to Ken for giving this analogy. It's consistent, but not safe. what that means, it's like, it's stable, but not safe in the sense that you always get the same paycheck every single month, but then if they fire you, you're like gone, right?

It's not safe. You get the idea of feeling like you're safe because you always get the same paycheck every month, but generally speaking, you don't even know when you're going to suddenly get fired and then have the rug pulled under your feet. Unfortunately, doing like layoffs and tech times, which is very common these days, the past like year or so.

A couple of years, that has happened to a lot of people, like completely, they just don't realize that, whoa, like suddenly I'm out of a job, right? Like what am I supposed to do now? in terms of like freelancing, I see it as, I'll kind of explain [00:48:00] why I think it's like suitable for different types of people from my opinion, because I've actually done all three at this point.

I would say freelancing is more like you are, it's not consistent, like it's not stable, but it's actually safer. Because it's not stable because you have to go find these jobs and the amount of money that you're taking that you're getting each month can vary month on month. but you, you're also diversified because if one thing doesn't work out, you know, you have like four other projects, right?

That can cover that one thing. And you can much more easily see where the market is heading. Like if you were a freelancer and you were working like a year and a half ago, you would have saw that the money was drying up. And you would have taken your precautions at that time in order to defend against that, as opposed to if you were working at a big company, you might not have known that until you got fired.

Yeah. So on the entrepreneurship side, I see it more as like, it's, it's more of like a kind of like more intense environment is how I would say it. Like if freelancing is building a business, entrepreneurs is kind of like [00:49:00] building a startup. So it's more about whether you take like a lot of people take venture capital money.

But you're trying to like build things like you're throwing in money and then turning out like something, a product and trying to grow that, you're thinking about revenue, but more. So you're thinking about growth. So very, very different type of environments and different things suit different people.

Say if you're someone who likes to just like go to work and then do that and then not think about anything else and come home, I would say full time job. Yeah. Try to diversify it though at least like maybe do some freelancing have some clients like on retainer or at least like diversify your The amount of money you're making invest that properly if you're someone who likes working who has some initiative and doing that And you like having a lot of control over your time.

I think freelancing is really good option You can have like a team of people in freelancing as well doesn't have to be just you but it's not like a lot of people that you're trying to manage you're still doing a lot of [00:50:00] the Work the technical work. And then if you're someone who, is someone who's like, very, he like, you love having a team.

You like having like a fast paced environment. and you really like being able to grow something really fast. Like you're like a super intense person. I would say then entrepreneurship is something that kind of, I think would be like, really suitable. again, this is like, kind of like a more simplified definition.

yeah. But that's kind of how I've experienced it. Like going all the way from being a full time employee to a freelancer and then taking venture capital money. That's kind of the difference that I saw. 

Louis-Francois Bouchard: Yeah. And my second to last question would be what are the skills that you would suggest people to learn more specifically, like which programming language or software or whatever to learn for freelancing in the AI space?

Tina Huang: Python, for sure, if you're going to do that, and if you're like the biggest, I feel like people have this misconception when they [00:51:00] think like AI engineer or something like that, it's actually just a rebranded software engineer, it's just that you happen to use. AI APIs and like AI things as opposed to other things.

So, but oftentimes it's like knowing how to use these new tools and like the new stacks that are there. that's going to get you really far because there's just not that many people who are taking that initiative to do that. I think. Yeah, that would be your traditional like stack of like, you know, using like reaction stuff and things like that.

It's still, it's, it's like literally the same as software engineering. You're just connecting it to like a different type of backend where you're connecting it to different types of APIs. but that being said though, I think, skill side level, I would also, I would almost say that's like 50%, if not even less than that.

Because as a freelancer, your actual role. Is not just to know things you have to let other people know that, you know, things, whereas they're not going to hire you. So, obviously like don't like pretend to know [00:52:00] things you should definitely still know some things. but I think investing heavily into your own brand is going to be setting You the apart the most like a lot of the best freelancers They're not the best engineers like not nearly the best engineers and on top of that a lot of the work over time It's not even done by them anymore.

They're more of like a managerial role where they're hiring other people and working with other people to build out parts of the work that they take on, right? But what they're really, really, really good at is marketing themselves. It's like I, as a trustworthy brand, as someone who would deliver on the product, so.

That is where I actually think you should spend even more time on. 

Louis-Francois Bouchard: Yeah, and just, I would assume communication skills as well. Like, later on, I, I, I've never done freelancing, so I couldn't say, but I, I would assume that you have meetings with somewhat important people, or at least the people that will decide if they hire you and rehire you.

So. I would say [00:53:00] that communication, explaining things that you do and keeping them up to date and etc. is super important. Documentation, like everything. Related to communication is something that is even more, more important than in a regular job, I think. 

Tina Huang: Oh yeah, absolutely. Absolutely. Because you're constantly, what you're trying to make them realize, like what you're trying to reassure them on is, what you're paying me is worth it.

Yeah. I'm bringing you value. They, you don't want them to go like, what's this person doing? Are they just doing nothing, right? You have to be like, say, like, hey, you know, here's my weekly update, like, here's what's happening right now. That makes the person paying you feel like, oh, like, this person's actually doing stuff.

Yeah. like, there's a lot of, like, psychology that goes into it. Your job, a lot of times, is to try to make your person feel comfortable and hurt, and you want them to not feel like they're stupid. Like, if you just throw technical jargon at them, they're gonna feel like they're, they feel stupid, and then they're not gonna want to work with you anymore.[00:54:00] 

Louis-Francois Bouchard: I don't know if you know Luis Serrano. He, he is active on YouTube as well, and like, he works at Cohere now, building education. Things, LLM university and other things, but he also mentioned that it never hurts to be simpler. Like it's, nobody will tell you that you, you sound dumb because you are talking with two simple words.

Like it's just better and it's, it's easier for everybody. It's more clear. It's better communication. Everybody understands and you know what to do next. It's, it's just clearer. And so, yeah, I think communication is super important. And, and. Yeah, I would love to know what's next for you and for Lonely Octopus anything, any projects or upcoming things you'd like to share to The people listening.


Tina Huang: I guess we're doing our YouTube. We're expanding that Lonely Octopus. If you're interested, maybe we can like put a link somewhere [00:55:00] for that as well. So we will be taking applications shortly and we'll be opening up our next cohort. And, in the next month or so, and on top of that, so I am also working.

I mean, I always have like different side projects and things like that, that I'm working on, but I would say like, those are my kind of like two. Those are definitely my like two major things, that I'm working on right now, but usually we'll, I do have a newsletter. So on YouTube, I generally stick to like specific topics, but on my newsletter, I talk about other things that I'm working on as well that may not be specifically related to AI data or like tech stuff.

Louis-Francois Bouchard: Awesome. Well, thank you very much for taking this almost two hours of time to, to speak with me and to teach us a lot about education and freelancing and all other topics that I really enjoy. So it was really fun. Thanks. Thanks again. It was amazing to talk with 

Tina Huang: you. [00:56:00] Thank you. It was a really fun conversation.