"Lack of data, lack of power is an opportunity for more Innovation and Creativity. Every time my students pick a project before they start doing it I will ask my students, 'Can Google, Facebook and Microsoft of the world, with more GPUs and more data, do it better?' if their answer is yes, we don't do it. Because that means we are not the right group to do it and I challenge my students to be more creative and innovative to the world's problems even 5000GPUs might not solve" -- Fei Fei Li
It's taken me a while to post this, but better late than never right?
Indaba: In South Africa, it means a conference, discussion you may say. Today, I document my journey, from how I heard about Indaba and finally baing at Deep Learning Indaba 2019.
This is where it all started, me randomly applying to a link sent by a friend, then weirdly I was accepted. When 2019 started, I applied to a beginner Python for Data Science class in which I was accepted. Our coursebook was O'reilly's Python for Data Science and Machine Learning, and I could barely understand a thing.
Fast forward to March 30th where I was selected to participate in an Oracle Data Science Hackathon at Strathmore University. There I got to experience first hand the possibilities in Data Science. At the Hackathon, someone sent a link to Deep Learning Indaba, which I successfully applied to.
The theme for 2019 Indaba's was: #SautiYako, translated to Your Voice in English. In the Silicon Savannah, over 700 Deep Learning Enthusiasts, Learners and Practitioners were gathered, to learn, share and interact.
At 12 noon, on 25th August the event kicked off with three lectures and a hands-on session. Over the next week, I greatly benefitted from the event from both the speakers and attendees, below I will only mention but a few.
For the first three days, we had hands-on sessions, below are the links:
CodeLab 1: Machine Learning Fundamentals
CodeLab 2: Deep Feed-forward Networks
CodeLab 3: Convolutional Neural Networks
CodeLab 4: Recurrent Neural Networks (RNNs)
- Problems are also opportunities
- Create innovations for global impact
- The world is our lab
- The future of AI is Africa this is because of our diversity
- We have the potential to leverage AI and transform technology in Africa
- You should drive your career, don't let your environment drive your career
- Choose to be different, never feel comfortable wherever you are.
- Ask yourself, what will this opportunity lead me to?
- As long as people can do it, you can do it too.
- It is not the strongest/fastest that survive but the most adaptive
It would not be indaba if I fail to mention the three days we spent mentoring teenagers on how to come up with solutions using design thinking and AI. We had around 20 teenagers, who were able to learn about Azure Machine Learning Studio, Design Thinking and Sustainable Development Goals and apply it in building a simple solution.
Masakane: we build together
Learning: new environments, exploration, through mistakes
Reading papers and take advantage of communities
Apply for PhDs and Jobs, even when rejected, keep pushing
The one week I spent at Indaba fueled my passion for AI, I am currently learning Pytorch hoping to use it to do deep learning projects. The networks I gained have also been a huge stepping stone in propelling me in Data Science and AI.