Yesterday’s Deep Learning Indaba X in Cape Town was so stimulating – there is such a diversity of activity going on in this field in South Africa – wow!
The final debate on what we can do to strengthen machine learning in Africa really got me thinking about the role of EDUCATION and I have a few thoughts I’d like to share.
At the conference, I was very aware of the illustrious academic credentials of most of the speakers. Prof’s and Dr’s abounded, many references were made (and jokes cracked) about theses and research papers and, on a more serious note, the talent and capability of these individuals to take ML to the next level really shone through. And yet, I would have loved one or two self-taught data scientists to be featured – the kind spoken of on datasciencemasters.org:
“We’re likely to see more uncredentialed, inexperienced individuals try their hands at data science, bootstrapping their skills on the open-source ecosystem and using the diversity of modeling tools available. Just as data-science platforms and tools are proliferating through the magic of open source, big data’s data-scientist pool will as well.”
For many Africans, university educations don’t grow on trees. And sometimes academia has an absurd logic to it that keeps people out for “interesting” reasons. For example, when I was looking into data science courses I found the Masters Program in data science offered by UCT. Upon enquiry I was advised that if I went back to university and completed my Bachelor of Music Honours (really!!), then I would then be eligible to apply – needless to say I’m not considering this route :).
However, despite the academic emphasis, it does seem to be well-acknowledged that there can be multiple ways to enter this field. This is one of the aspects that I really admire about the Explore Data Science Academy from BCX which was mentioned in the debate: the entry requirements are not the usual academic ones, but rather aptitude and passion: everyone from matriculants to PhD’s can apply. Just the fact that this academy exists, and that they believe that with aptitude and passion anyone can succeed, has spurred me on to continue with my own self-taught journey!
When you attend a conference like yesterday’s Indaba, it’s easy to suffer from “impostor syndrome”: here I am with my Bachelor of Music and a Postgrad Diploma in a room full of mathematicians, statisticians and computer science gurus – what am I thinking??! And yet I do not give in – for these reasons:
- There are so many wonderful role models, who have proved that, despite barriers to entry, so much can be achieved.
I recently read the story of French mathematician Sophie Germain – in her time it was unthinkable for a woman to embark on a formal education of any kind, so she “gave herself an education using her father’s books and became a brilliant mathematician, physicist, and astronomer, who pioneered elasticity theory and made significant contributions to number theory”. She even resorted to masquerading as a man in order to correspond with fellow mathematicians like Gauss – only much later coming clean and disclosing that she was actually a woman!
She did what she did because she loved the work and was highly motivated. Let’s encourage newcomers to learn however they can – embrace less traditional education routes if they work!
- From everything I’ve seen in the IT industry, there are many different skill levels – and they all have their place.
There are those that take technology to the next level – and those are pretty special, talented individuals. But there are also those that lead, those that conceptualize, those that implement, those that do some of the “dog-work” (like data cleansing!!), and those that rollout existing technologies to communities that would otherwise have been left behind.
Just yesterday, keynote speaker Dr. Jacques Ludik mentioned his high-school son who is already getting into machine learning and using Keras to experiment with neural networks. It’s possible that this fellow is a South African Sheldon Cooper type genius (?), but I tend to think if this young man can already find meaningful problems to solve, garner sufficient knowledge and experience to conceive of potential ML solutions, and realise them using tools like Keras, it means there are likely 1000’s of others who can do the same. At this level, they may not develop ground-breaking improvements in methods and techniques – but they surely can use what they know to make an incredible contribution to individuals and communities in their vicinity.
- The data science and ML community is an incredibly generous one
There are many fields where knowledge is jealously guarded and sharing is not a thing. But so far, in all of my personal and online interactions with people working as data scientists, I have encountered nothing but enthusiasm and generosity. Advice is offered, resources are shared, code is made available, problems are answered. In this sense I find the barriers to entry are actually few. I haven’t had one person tell me “you shouldn’t be doing this”, and I have had several who offered support, coffee and encouragement! The very fact that Deep Learning Indaba X welcomed us newbies and laid on talks for a variety of levels of expertise is indicative of this kind of generosity and openness.
- I am excited by the possibilities
I think the lady who made the comment about internships in yesterday’s debate was spot on: she learned a whole bunch of stuff, but when she did an internship and saw what she could actually accomplish with all that theory she got really excited.
So…
- Let’s get people switched on to the potential of ML and AI and start generating ideas – as Prof. Amit Mishra said “ideas come from everywhere”
- Be open to non-traditional educations, and evaluate people based on aptitude, skill and passion
- Carry on being a generous, supportive community!
Thanks for a fascinating day :).
Sho't left to data science 