Blog

How I Became a Professional Problem Solver

I have always been fascinated by how we think and learn. I find beauty in the irony that we know more about what is “out there” than what is in our heads. Becoming a professional problem-solver, i.e., a scientist and software engineer, was a natural progression from this wonder.

Specifically, I knew I wanted to study the brain and yet also pursue computer science. Serendipitously, I managed to combine the two with computational neuroscience. To reach my childhood dream of becoming a neuroscientist, I started with a BSc in Biochemistry, Computer Science, and Psychology. I transitioned to Neuroscience for my BSc Hons, joined Dr. Joseph Raimondo’s electrophysiology lab for my Masters, and did my Ph.D. neuroscience using computational and theoretical neuroscience techniques him. Being in Joe’s lab allowed me to visit Prof Tim Vogels’ group in Oxford and Prof Henning Sprekeler’s group in Berlin, each for 3 months.

Along the way, I’ve been doing consulting for machine learning engineering solutions, mainly using computer vision. That is, putting into production a real-time, robust, and resilient object detection model supported by loads of software engineering to go from notebook to client. Now, I’m a postdoc in Prof Tim Vogels’ group at the Institute of Science and Technology Austria, analyzing and modeling neurons derived from human induced pluripotent stem cells (hiPSCs), specifically from patients with epilepsy and an autism spectrum disorder.

The study of the brain (i.e., neuroscience) is so wonderfully complex. I have yet to find a background that cannot provide insight into how we experience the world around us or within our minds. Crucially, the standout requirement is to ask questions and seek solutions – the cornerstone of science. I have managed to find an angle that speaks to my strengths and passions.

Computational neuroscience, specifically, has special ties to machine learning. Both fields attempt to understand networks’ computations; each field chooses which constraints are more important – biological or technological. That is, how do neurons talk to each other to solve tasks like vision and decision-making? Also, how does the communication breakdown lead to neuropathologies like epilepsy? These questions have persisted in neuroscience and flowed over into machine learning, steady if frustratingly slow progress. Still, it provides the chance to be the absolute expert in the world on a particular subject. If this prospect excites you, go for it! Just remember that Google may not have the answers: it will be your turn to generate them.

I have been driven forward by excellent mentors for whom I aspire to impress rather than fear to talk. This has, in turn, motivated me to spread my passion for computational neuroscience and machine learning. To that effort, I have been heavily involved in the IBRO-Simons Computational Neuroscience Imbizo – http://imbizo.africa – and the Deep Learning IndabaX South Africa – https://indabax.co.za. Laser-like, these efforts promote people with disadvantaged backgrounds and push the fields forward by making them more diverse and inclusive. Furthermore, these are team-driven approaches to systemic challenges. I cannot imagine working on these alone. Finding like-minded individuals has been a tremendous boon, sometimes by the fortune of meeting the right people at the right time and sometimes by sending out a call about a shared passion into the internet.

Building communities and strengthening connections through these initiatives have been some of the most rewarding aspects of my career. Along the way, they have provided humility, insight, and valuable experience of working with people. I am fortunate to have fostered my vision early and the right people to support me. We take so many things for granted at the time that we later realize is critical for success. There have been many turns and uncertainty along the way. I have learned to be adaptable: investigating the mind is hard, and the brain seems to provide many roadblocks to unlocking its mysteries. Still, perseverance and pausing for reflection have trenched a path forward.

There’s no “trick” to becoming a computational neuroscientist or machine learning engineer. The key is to acknowledge where one is, where one wants to go and figure out the steps in-between. With so much room to flail rapidly and the ease with which imposter syndrome sets in, take a pause. Reflection can focus one’s thoughts and hone actions. I ponder the progress we shall make in the next 2, 5, 10, etc., years in understanding how we think and how to make machines learn. I hope to contribute just a little more to the body of knowledge and, crucially, to enable people to achieve more than I possibly could.

Originally publish https://howigotjob.com/career-advice/how-i-became-a-professional-problem-solver/

Christopher Currin is a NOMIS Fellow at the Institute of Science and Technology (IST) Austria, working closely with the research groups of Tim Vogels (Computational Neuroscience and Neurotheory) and Gaia Novarino (Genetic and Molecular Basis of Neurodevelopmental Disorders). Born in South Africa, Currin received a BS in biochemistry, computer science, and psychology in 2013 from Rhodes University (South Africa) and a Ph.D. in neuroscience from the University of Cape Town (South Africa) 2020. He has received numerous awards, including the Alfred Beit award for best undergraduate performance at Rhodes University, a prestigious DAAD-NRF Joint Scholarship for masters and doctoral research, as well as visits to the University of Oxford and Technische Universität Berlin. In parallel, Currin has played an active leadership role in growing computational neuroscience and machine learning in Africa through the IBRO-Simons Computational Neuroscience Imbizo and Deep Learning Indaba X initiatives.

Currin’s doctoral studies focused on experimentally informed computational models of brain disorders like epilepsy that explore better pharmacological treatment regimes. Currin also has extensive industry experience in data analysis and systems architecture, creating engineering solutions for various problems. As a NOMIS Fellow, Currin will use software engineering, machine learning, and computational models to study the emerging dynamics of human neural networks from healthy subjects and people with epilepsy and autism spectrum disorder (ASD). His project, Unlocking Crucial Cortical Connections in Human Neural Dynamics for Health and Disorder, in collaboration with the Vogels lab, will utilize recordings from high-density multi-electrode arrays (HD-MEAs), monitoring thousands of neurons simultaneously from a human-induced pluripotent stem cell (iPSC)-derived cultures that are developed in the Novarino lab at IST Austria. Using these recordings, Currin and colleagues will guide the future of model development by building human data-driven biological neural network models. Employing advanced genetic and computational techniques, the researchers aim to determine how neurons connect and maintain their connections to form functional networks. This collaborative work will reveal the differences between natural and disease dynamics and contribute to effective lifelong clinical treatments for “plasticity pathologies” such as epilepsy and ASD.

Bio taken from https://nomisfoundation.ch/people/christopher-currin/