HPC kitchen [how-to-learn]: Don't give up in learning computing

Duration: 5:13 868 transcript words EN summary

Video summary

The video begins by challenging the common assumption that reading textbooks or completing courses is sufficient to become an expert in any field, using cooking as a relatable analogy. The speaker argues that while following a recipe allows one to produce a basic result, true mastery involves adapting to specific conditions, troubleshooting unexpected issues, and acquiring local knowledge that cannot be found in books. This same principle applies deeply to computing; simply finishing a curriculum does not automatically grant the ability to handle real-world complexities or fill in the gaps between theoretical instruction and practical application. To bridge this gap between formal education and professional competence, the speaker emphasizes the critical importance of learning through collaboration rather than isolation. Working within teams, research groups, or with friends allows individuals to focus on their current strengths while relying on others to guide them toward areas they have not yet mastered. The transcript illustrates this with examples of how different team members can cover complementary skills, such as one person mastering cluster usage while another provides expertise in storage systems, command lines, and data transfer. This collective approach prevents learners from feeling overwhelmed by trying to understand every topic simultaneously and accelerates the path to self-sufficiency. The speaker shares personal experiences that reinforce the value of mentorship and community engagement, noting how early interactions with a friend who helped troubleshoot Linux installations were instrumental in their own growth. Similarly, when new members joined research groups, the speaker adopted a supportive role until they became capable independently, at which point they encouraged those individuals to seek answers on their own to foster deeper understanding. Beyond direct mentorship, the video highlights the benefits of observing others by following open-source projects on platforms like GitHub or joining mailing lists, where witnessing development discussions provides practical insights that can be immediately applied to one's own work. Ultimately, the core message is an encouragement for learners not to give up when faced with difficulties but instead to actively seek out communities and mentors who can support their journey. By forming teams and finding colleagues willing to teach both technical skills and the nuances of the development process, individuals can make their computing careers more productive, enjoyable, and resilient. The speaker concludes by urging everyone to avoid the trap of trying to learn everything alone and to embrace the power of shared knowledge, ensuring that the path forward is smoother and more rewarding for all involved.

Read the full video transcript

Hello. For the next edition of HPC Kitchen, we're not in the kitchen. It's more of a little fireside chat thing. So, I'll start with this initial question. If I got a bunch of cooking books and gave them to you and said, "Read them all and then you went and read them, do you think you would be an expert cook?" Well, probably not. I had a friend that once said that anyone that can follow a recipe can be a cook. And I mean, yeah, I guess if you follow the recipe, you can get something done. But isn't there a lot more that's there? Like what you can do when something isn't quite right or what happens when um like how to know really when something is done, how to adjust the recipe to your own ovens and stoves and stuff like that. There's lots of little local knowledge that you can get by doing the cooking which you just don't find in books. And I think the same is through true of computing. So yeah, we have all kinds of courses, reading material, things like that that you can take and give to someone and say, "Okay, you've done this. Now you can go and do your computing work." But I think it's not so simple because there's all kinds of this little like folk knowledge kind of stuff that is really helpful and can make your career much more productive, much more fun and so on that you just don't pick up by reading. You pick up by doing and more important by watching other people do. So what can we do about this? So I think that the way of doing this is by working in groups. So basically a person can't learn everything they need to know by themselves or by reading. Instead the way society works is that we work in teams. So there's a research group or friends or colleagues and so on. And by working with these people you learn a lot more. You learn all this little knowledge. you can focus on what you're learning now and the things you haven't learned yet. Then someone else can help you with, can point you in the right directions and so on. So that way you can more directly go to your path instead of trying to um figure out every topic at the same time. For example, as I've said before, using a computing cluster isn't that hard if you're good at the Linux command line stuff, the storage system, and so on. But you know those aren't taught in the courses where you learn to use the cluster. So you sort of have to like go together. So maybe that someone will take the cluster course, learn that, but then they work with someone that sort of fills in the gaps about the storage, the command line, the data transfer, all of these kinds of things. When I began using Linux long ago, the way it worked, I had a friend and the friend basically sort of got me started, helped me figure out what the um the initial installation kind of things. And as I was doing stuff, there'd be often be something I try to do and I just don't know how it works. So I asked the friend, "Hey, what about this?" And they'd say, "Oh, you do it this way. What about that?" Uh, well, search this for more information. And this little bits of help would allow me to move much faster and get to a point where I was self-sufficient. I did the same thing when other people would start working in my research groups. Also, so I was there and basically would help them with the stuff until they basically ask questions and I didn't know the answer. And then I told them, well, okay, I get think that you're actually kind of self-sufficient now, and I'm not answering because I don't want to answer. I'm not answering because, well, I don't even know where to look, but you need to go look yourself. And that worked pretty well. Another place that I learned a lot of stuff was just by following other projects. I'd, for example, go and subscribe to other open- source projects on GitHub or join their mailing list. And just by seeing the way people talked about this open- source development, I would learn lots of stuff that I would slowly go and start using myself in my own work. Anyway, that's basically it. So my goal here is to encourage you. Don't give up if things seem difficult because yeah, there is a lot to learn. Don't try to go by yourself. Instead, try to work with people. Find friends, find colleagues, and form a team. Get people to teach you all these things. Get a good mentor that mentors you not just on whatever science you might be doing, if that's the thing, but the development process and so on. And hopefully this will make your career much more enjoyable. So, with that being said, thanks a lot and see you next time. Bye.