"The people are everything you think about AI": why AI/ML Community is necessary, helpful, and inspiring
Roosh Circle
We are a tech community of founders, engineers, investors, and entrepreneurs across Europe
What is particular about our Papers Club is that we always reach out to our community members and seek feedback about our events and activities. We also ask about other communities, events, initiatives they tried and enjoyed, people they subscribe to, channels they follow, and what they find essential in community building. But the story below is spoken from the heart. What started as a standard user interview with a community member, Michael Yahaya , Senior Data Professional, has grown into an excellent article about perspective on the community.
How to stay updated with AI/ML domain?
Michael: It's pretty much impossible to know everything that's going on. So, the way I see it is simply that there are some personalities you mentioned in the post, and we always go to their website, blog, or emails to find out what's happening. So, for example, deep learning AI has a newsletter they send out, and then you can sort of have a recap. If you were to follow every new development, you would probably spend all your time doing that and nothing else. So I think it's beneficial to have these newsletters or these weekly digests that can say this is what happened in AI this week and things like that. So that, for me, really helps.
Check the list of the newsletters our community members are subscribed to:
So, I'm a big believer in multichannel updating. Right. And I believe that there are many ways you get information. Sometimes, you read a paper, and then this paper introduces you to another paper or another author, then takes you to another paper, etc. Then you find out a little bit more about who wrote the paper, and then you understand, oh, this is an interesting person in this area. That happens as well. The community also works as you invite some speakers, and they present something very useful. You can also get insights via email or by watching a new video on YouTube. So, every single channel is helpful for information. I would not neglect any if I could.
The next thing that will help you stay updated is attending online and offline events. Recently, I went to a Google developer group. It's for Google Cloud here in Zurich. And it was enjoyable, first of all, because of a real group of very committed people. The topic was what Google has now called Gemini. And for me, it was nice to speak to specific community members about some of the issues that they've encountered, talk about other things that they're finding fascinating, that are coming out in terms of the functionality that AI can do, in terms of some of the frontiers of research, but also terms of day to day jobs.
Even though it was a Google event, many people compare Google tools to Azure and AWS tools. Some people were convinced that their tool was the best, and others were almost on the bane of being confident that the new developments would make the tool the new tool. Sorry, new developments in an alternative tool will make that tool more attractive than the current tool. So you have these back-and-forth conversations. I found that quite stimulating because the market seemed somewhat segmenting back in the day. You would probably have one prominent provider, and this provider would give you an application, and you would be subject to vendor lock-in.
And now, thanks to the cloud, the commoditization, and the fact that even those who are not that technical can use no-code solutions, many people are trying to patch together the best of breed, and this, of course, has its problems as well. So, there is no perfect solution. But it's just interesting to talk to people out there. Then, you learn about some of the ways in which they design their architectures and some of the ways in which there are problems using the available tools. And it's pretty fascinating.
Why is the community important?
Michael: It's crucial for me to emphasize just how essential a community is because the people are everything you think about AI. AI is nothing more than a way to make computers (machines) more helpful to people. Some of the things which people do if you think about the whole concept of AI and how it was developed over the past hundred or so years, people were saying, how can we give machines the ability to do the things that we do as people? So, first of all, the people aspect starts with trying to copy what people do and give that intelligence to "machines." But also, even when you look at the state of AI today, it's all about the needs of people.
When you have a product, you're a product owner or a product manager trying to solve people's problems. And there's sometimes this dichotomy between technologists who love technology for technology's sake and those who are more pragmatic. They're looking at problems and trying the different tools that are available to solve the problems. When you have a community, you get to understand the people fully, what they're doing, their problems, and how technology is helping them. At the same time, of course, you do get some people, and this is the irony of it: You get some people who are also so passionate about technology that they begin to introduce you to new tools. You're still determining how they will be applied or if they're any better than other tools.?
But knowing that they're available expands your horizon, and it really makes you aware of not only how vast the landscape is but also the different perspectives that people are coming at regarding specific problems. And I think that your toolkit becomes richer both in terms of the problems that you've seen people solve, but also both in terms, but also in terms of the tools which exist, which you begin to think about how they could solve some problems, which you have.
Is community about networking and sharing experiences?
Michael: 100%, as you probably know, in this space. And also it's more of a Swiss thing as well. Everything is network-driven, so a lot of the time, people are aware of good people or opportunities via the network. The network also serves as a social validation mechanism. Because if you have somebody who is well known in the network for doing good work, it makes everything a lot easier, especially if they are a specialist in a particular domain where you have a need.?
So, for example, imagine you had a task that involved computer vision, and you knew somebody in the network who was well-known and validated by others. As a computer vision expert, it sort of makes your process a lot easier. You don't have to spend so much money looking for someone because you can find that very quickly through the network if it's the kind of person who a) you can work with and b) has the skills that can do the job that you need. So, the network is critical.
In terms of the sharing, we share two things. We share different use cases or different applications of tools to the same use case. So, we look at different ways we can approach it. But one thing we've found most valuable is being able to share potential problems and potential solutions, either because we have a “lessons learned” log or because we have, I would say, a bird's eye view of the landscape, and we could potentially foresee issues because they're related to other similar issues. So either we have direct experience of these issues that we've solved, or we have some indirect exposure to some of these issues and have seen how others have solved them. For me, learning is essential because, as you can imagine, there are many ways in which things can go wrong.
And it's very important for people to be able to find ways in which they can solve these problems. And this is very important for us because you have two problems (with AI):
领英推荐
So that's important for us (because even if AI can have a business impact, understanding where and how to apply it, deploying it at scale in Production, and getting people to adopt the tech have to be addressed)
What about mentorship in AI/ML?
Michael: I think mentorship is crucial for many reasons. The first reason is that some people need someone to give them that initial push. It's beneficial for other people to have somebody who can provide them with a roadmap of the terrain as a mentor because there's a lot of information. People often look at all the things available and need help figuring out where to go.?
Second, you probably know that when you research most high achievers, people who made it into the C suite, those CTOs, and CFOs, you almost always find a mentor behind them. There was somebody who saw their talent and put them in a fast-track lane or someone who believed in them and gave them an opportunity to do something that changed their career. So, I mentioned that community is essential.?
Third, I have always believed in the socio-technical aspect of technology. It's socio-technical computers and technology in almost every organization. Mentoring is one of the bedrocks of the social aspects because it really allows you to get the most out of people. I think there are some scrum masters today who help enable people to do their best, and that's good. They say you have the strength.
Also, a mentor gives you the tools to help you climb the corporate ladder. So it's essential. From a mentor perspective, I think something in us compels us to want to give back because we were helped along the way, almost all of us, if not all of us. And so giving back means that we help others. And I felt this obligation, this depth of gratitude to the world, to those who helped me. I may or may not be able to pay the person who helped me directly, but at least I could help somebody else, and therefore, I can pay it forward. Therefore, these motivations and reasons are compelling.
Check the recording of the session with our moderator, Cristina Gurguta , Senior ML Operations Strategy Lead at 拉法基 , about Data Science Mentorship here:
How else can the community help you?
Michael: I find that having a community helps some people focus on the research because it could be that you have ten papers that are on your desk that you want to read, and now you're trying to write a paper of your own based on a research gap you store in maybe one or two of the papers, and sometimes you come to the community and you find somebody else who's tried to answer or fill that same research gap that you've seen, or somebody else has another way in which you can solve a problem that you're grappling with. So I think it's vital.
I find that the big problem with many people is that the term research sometimes makes people nervous because they feel that it's all about doing things in the most academically rigorous academia, which I believe is important. But I also think that there's practical innovation, which happens and can happen and doesn't have to be overly formalized. The key is to ensure that you respect other people's property rights and give them credit where credit is due. But sometimes you find that if you just ask the question and just put it out there, you might find that nobody else has asked the question, and that's an excellent place to start.
You could also find somebody who would say to you, I will take your finding, and I will put it in an academically presentable way, or somebody who would say, I take your academic paper, and I can commercialize it for you. So, the research spectrum must be more than just purely academic or not. It's as many grades in between because we're all trying to find. I remember the first day I walked into a research class with my professor, and she said something very simple. Yes, we do academic research, but whenever you're trying to solve a problem, you're going out there. You're putting forward suitable hypotheses and trying to solve those problems; that's research. It's just that you have to publish within specific academic guidelines and give credit to those whose ideas you are building upon.
But even though everyone knows that there are newspapers, books, and magazine articles in which the audience is not always academic, it's not technical. It's a result of people asking questions, investigating those questions, and finding answers in a collaborative way. And that is also a beautiful side of community.?
Many thanks to Michael Yahaya, Senior Data Professional who is leading end-to-end data and AI projects for this inspiring talk!?
Feel free to join our next event:
Also, please don't hesitate to follow us on Discord, Twitter [X], and Meetup.