Community Building and Teaching in AI & Tech

Links:

Did you like this episode? Check other episodes of the podcast, and register for new events.

Transcript

The transcripts are edited for clarity, sometimes with AI. If you notice any incorrect information, let us know.

Erum's Background
Omdena Academy and Erum's Role There
Omdena's Community and Projects
Course Development and Structure at Omdena Academy
Student and Instructor Engagement
Engagement and Motivation
The Role of Teaching in Community Building
The Importance of Communities for Career Building
Advice for Aspiring Instructors and Freelancers
DS and ML Talent Market Saturation
Resources for Learning AI and Community Building
Erum's Resource Recommendations

Alexey: This week, we'll talk about community building. We have a special guest today, Erum. Erum is a Lead Machine Learning Engineer and PhD researcher at Justus Liebig University. Did I pronounce it correctly? (1:25)

Erum: Yes. (1:44)

Alexey: Erum specializes in AI for education. She instructs at the Woman in AI Academy and leads Omdena Academy. Previously, she taught deep learning internationally and organized data science boot camps. You'll probably tell us about them. Thanks for joining us. Welcome. (1:46)

Erum: Thank you for having me. I'm excited to share what I've been doing. (2:15)

Alexey: The questions for today's interview were prepared by Johanna Bayer. Thanks, Johanna, for your help. Let's start. (2:24)

Erum's Background

Alexey: Before we dive into community building, let's start with your background. Can you tell us about your career journey so far? (2:32)

Erum: I'm Erum from Pakistan. Currently, I'm based in Germany, as Alexey mentioned. I'm doing my PhD here. I come from a less-developed area of Pakistan where females are often not allowed to get an education. I'm lucky to have received an education, moved to the city, and started working with the community. Instead of just attending lessons, I began delivering workshops and courses. (2:42)

Erum: Through various networks, I got my research position at Justus Liebig University in Giessen. I was involved in a project in Germany, which this university applied for and was selected. I mentored that project and suggested offering a PhD position. They agreed and offered it to me, so I joined. (2:42)

Alexey: Which city are you based in? (4:35)

Erum: Giessen, near Frankfurt. (4:37)

Alexey: Okay, near Frankfurt. (4:42)

Erum: Yes. (4:45)

Alexey: I've never been to that Frankfurt. (4:46)

Erum: You should visit sometime. [chuckles] (4:50)

Alexey: You work at Omdena. What's your title there? (4:54)

Omdena Academy and Erum's Role There

Erum: I'm currently Head of Omdena Academy. I started as a Lead Machine Learning Engineer, then we began delivering courses and transformed Omdena School into Omdena Academy. (5:03)

Alexey: It's amazing. It's unfortunate that women in your area can't get an education. It's fortunate you managed to do this. Hopefully, the situation changes. I'm interested to learn more about Omdena and your role there. (5:28)

Erum: Omdena is a global community that started in 2019. Every participant is called a "collaborator." We solve real-world challenges with AI. We started with a mental assistive application in 2019, then projects like finding anomalies on Mars. Our major learning is through diverse, global collaboration. (6:04)

Erum: There are groups of 50-60 people from different backgrounds, learning together. We tackle projects with real-world datasets, which is rare in AI. Networking and learning from diverse perspectives are key. What do I do there? (6:04)

Erum: I started as a collaborator on the first project. Later, I helped with graphics and certificates. We aimed to make AI and data science courses accessible to underdeveloped communities by building courses from our projects. This helps both the learners and instructors. (6:04)

Erum: At Omdena, we provide a platform for instructors and students. After completing a project, participants can join courses and future projects. This training increases their competency and allows them to join Omdena as collaborators. (6:04)

Alexey: Was it driven by necessity when you saw participants lacking skills? Or was it a demand from students wanting to improve in specific areas? (9:40)

Erum: It's not 100% aligned with requirements, but we address necessary skill sets. Courses are delivered after project completion to teach how Omdena works, giving people an idea of how to handle projects. (10:19)

Omdena's Community and Projects

Alexey: From what I understood, it's about the platform and community interaction. Are there courses on basic data science, like Pandas or NumPy? (11:13)

Erum: Yes, we now offer basic courses. Initially, it was project-based, like developing a solution for vision-impaired people in Germany. Now we have courses on natural language processing and interview preparation. (11:32)

Erum: We also had agricultural AI projects. Initially, we focused on advanced projects, but eventually, we started foundational courses due to organizational demand. Now we have a foundational data science course covering basics like Pandas and Python. (11:32)

Alexey: When you said courses start after the project, is it to teach what was learned from the project? Like understanding better by teaching? (13:12)

Erum: Exactly. You learn more by teaching and filling gaps in your understanding. (13:40)

Alexey: That's awesome. (13:46)

Erum: Yes, that's the idea. (14:20)

Course Development and Structure at Omdena Academy

Alexey: How do you decide what content to teach and what courses to offer? (14:24)

Erum: I work as a facilitator and teach a few courses. I create courses based on my learning experiences. Instead of taking paid courses, I learn on my own, make notes, and then deliver the course. This way, I effectively use my knowledge. (14:32)

Erum: When delivering, I learn more because I have to solve problems to teach them. We have a registration page where collaborators apply to be instructors. We review their content, make adjustments, and then launch the course. (14:32)

Alexey: And the courses are free, correct? (15:50)

Erum: Yes, free for both instructors and students. Instructors don't get paid. (15:52)

Alexey: This is how DataTalks.Club works too. Free courses mean wider reach, which is amazing, right? (16:02)

Erum: True. We often discuss selling courses, but my CEO believes education should be free for students. We charge organizations, not students. (16:27)

Alexey: When organizations ask for courses, you provide them for a fee, correct? (16:56)

Erum: Yes, that's how we work. (17:19)

Alexey: That's the business model. (17:22)

Erum: True. (17:25)

Student and Instructor Engagement

Alexey: Do I have to be in the Omdena community to join a course? Do I need to do a project first? (17:26)

Erum: No. Organizations can develop courses through Omdena without doing projects. For companies, courses come first, then projects. Graduates can join projects later. (17:38)

Alexey: Let's say I'm a student interested in a foundational data science course. Can I join without being part of Omdena or a project? (19:32)

Erum: Yes, you can join an Omdena course without being part of the community. Once you graduate, you become part of the Omdena community and can join any project. (20:03)

Alexey: How do people usually become instructors? Do they approach you, or do you seek them out? (20:32)

Erum: It's a mix. The link to apply as an instructor is always open. Sometimes, we reach out to interesting project teams and suggest creating a course on a specific topic. (20:48)

Engagement and Motivation

Alexey: At DataTalks.Club, we have free courses with a low completion rate. Do you see something similar? (21:36)

Erum: Yes. Our courses are live, not recorded. We invite 100 out of 400-500 registrants due to capacity and financial constraints. Our graduation rate is about 33%. Many attend but don't submit assignments, so they don't get a certificate. (22:29)

Alexey: So, you do pre-screening to select participants. What criteria do you use? (23:49)

Erum: We consider prerequisites, motivation, and availability. We also offer free subscriptions to learning platforms like DataCamp, which motivates participants. (24:18)

Alexey: That's interesting. How do you see the connection between teaching and creating and maintaining a community? (25:22)

Erum: Creating the online community isn't my main role. We have a community manager and others who handle it. My focus is on teaching. (26:40)

Alexey: You've been involved in various communities like Women Who Code and Neuromatch. What motivated you to take on leadership roles? (29:21)

Erum: For Omdena, I joined a project and met people who referred me to other opportunities. It was gradual. I took on leadership roles to stay connected with the community. (29:48)

Alexey: The motivation was gradual. You built your network, and more opportunities appeared. Communities have frameworks for motivated individuals to take on roles. (30:58)

Erum: True. (31:39)

The Role of Teaching in Community Building

Alexey: As a community manager, I want to empower people to take initiative. How do you create an environment where people feel motivated to lead? (31:43)

Erum: In our community, we built AI Wonder Girl, starting with seven-eight girls and growing to 100+. We meet online, discuss topics, and motivate each other. Starting small and gradually growing is key. (33:03)

Erum: Small groups meeting regularly to discuss and explore topics is effective. Don't start with big aims that seem unattainable. Begin with doable tasks. (33:03)

Alexey: Like a small WhatsApp group. (35:33)

Erum: Yes. (35:38)

Alexey: We interviewed Sarah, a co-founder of AI Wonder Girls. She talked about the group and winning hackathons. (35:39)

Erum: Yes, she's a good friend. (35:54)

Alexey: For those listening, check out that episode. It was an interesting chat. (36:19)

Erum: I'll check it out. Thank you. (36:34)

The Importance of Communities for Career Building

Alexey: I'm a big fan of communities. They helped me a lot in my career. What do you think about their importance for career building? (36:37)

Erum: I'm also a big fan of using communities for career building. They provide updates and help you understand what skills are in demand. Joining communities gives you an idea of needed skill sets and helps you develop them. (37:26)

Erum: Starting your own community and teaching others is a fast track to learning. It helps you and others grow. (37:26)

Alexey: How can communities foster the creation of sub-communities and empower people to take initiative? (39:14)

Erum: As a bigger organization, allowing people to use the community's name can help. It provides authenticity and reach. Ethical guidelines are important, but empowering people to represent the community is beneficial. (40:00)

Erum: WhatsApp groups are limited. A public page or social network can help people from the same region connect and represent the larger community. (40:00)

Advice for Aspiring Instructors and Freelancers

Alexey: I should write a blog encouraging people to start local chapters of DataTalks.Club. Anyone in our Slack can create a channel if one is missing for their area. (41:53)

Erum: Yes, and use the name for reachability and authenticity. (42:24)

Alexey: Any other tips for good community management? (42:35)

Erum: Engagement and providing value are crucial. People need to know what they will gain from spending time in the community. Define clear takeaways for each session or activity. (42:50)

Alexey: How to compete with Netflix? Show the value of attending meetups versus watching recorded content. People need to know they'll miss out if they don't attend. (44:09)

Erum: Offer sessions with clear takeaways, like learning how to deploy a machine learning model. This can trigger more interest and attendance. (45:07)

Alexey: Any other tips? (46:28)

Erum: Categorize offerings into basic, intermediate, and advanced levels. This makes it easier for people to find relevant content and know what to expect. (46:33)

DS and ML Talent Market Saturation

Alexey: We have a question from the community: Do you feel that the market is saturated with data scientists and ML engineers? (47:50)

Erum: I don't think so. There is a shortage of unique skills. Everyone knows data science, but competition is high. You need to distinguish yourself with unique skills to increase your value. (48:20)

Alexey: It's strange that there's a shortage, yet many people struggle to find jobs. We need to make our skills more applicable and not rely on shortcuts. (49:54)

Erum: Sometimes, job applicants use plagiarized tests. Companies can't hire people who don't show originality. The more access to resources we have, the lazier we've become. (50:56)

Alexey: With ChatGPT, I'm so lazy now. (51:39)

Erum: Yes. (51:41)

Erum: It's an awesome tool, but you can't learn by just copying. You need to use it smartly and do the work yourself. (51:52)

Alexey: How can someone become an instructor at Omdena? (52:11)

Erum: They can apply on OmdenaAcademy.com. (52:27)

Alexey: How do you spell it? (52:33)

Erum: I can share the link. (52:36)

Alexey: Google Omdena Academy. The first link is Omdena.com/Omdena-Academy. You'll find options to apply as a student or instructor there. (52:41)

Resources for Learning AI and Community Building

Alexey: What can you recommend to a freelancer wanting to start teaching in AI? Is it possible to apply to Omdena? (53:11)

Erum: For Omdena, we focus on capacity building for collaborators who have completed projects. I recommend joining an Omdena project first, learning the process, and then applying to teach. (53:23)

Erum: Teaching is mostly volunteer-based since we offer free courses. However, it helps you build a good structure and potentially sell your course on other platforms like Coursera or Udemy. (53:23)

Alexey: Any resources that help women access education on AI? (54:29)

Erum: Join AI communities. There are courses available, and sometimes scholarships on platforms like Coursera. Learning through projects on GitHub is also valuable. (54:49)

Alexey: Teaching friends can also help you become proficient. This advice applies to any gender. There are many courses on Coursera, Omdena, and DataTalks.Club. (55:48)

Erum's Resource Recommendations

Alexey: Before we finish, do you have any resources on teaching, community building, or anything else you find interesting? (56:16)

Erum: I prefer newsletters. I'm part of the Stanford University Law Department's AI-powered law school newsletter. It's important to understand how to ethically implement AI. I'll share the link later for the description. (56:39)

Alexey: We'll include it in the description. Thanks for sharing your knowledge and experience. Thanks to everyone for joining and asking questions. That's all for today. (57:46)

Erum: Thank you for having me. (58:25)

Alexey: Enjoy the rest of the week and goodbye. (58:28)