on Neuroscience, AI, Machine Learning, Reinforcement Learning, Robotics, Brain-Computer Interfaces, and other things

My core interests lie on the intersection of artificial intelligence and neuroscience, the use of artificial neural networks to help us understand how the brain works, neurotechnolgy in general, and brain-computer interfaces in particular. Since 2023 I am heading up a neurotechnology and AI advisory company Neurotech Lab that provides consultations and research services on AI techniques and data analysis for neuroscience and neurotechnology sectors.

Sometime during my PhD studies I moved from Tartu to Sydney, where I co-led the data science team of Omniscient Neurotechnology, an FDA-approved neurotechnology company, to apply machine learning and deep learning methods to human fMRI recordings and develop techniques to discover biomarkers of neurological conditions in human connectomes.

The beginning of my PhD coincided with the “AI spring” of 2012, when Deep Neural Networks have first revolutionized the field of computer vision, and then several other fields of computer science followed. One particular field that was propelled by DNNs was Reinforcement Learning – a framework for building intelligent agents that learn continuously from interactions with the surrounding environment. Robotics is one of the applications that naturally fits the Reinforcement Learning framework. This jumpstarted my involvement with a fascinating robotics company OffWorld, Inc. based in Pasadena, CA. Here, as employee #1, I have architected the long-term strategy on the application of machine learning to autonomous robotic agents, and then, as the Head of Machine Learning Strategy and Research, oversaw the development of ML and RL methods that are applicable to real-world robots, increasing their autonomy to the level necessary for operating with minimal human oversight.

During my years at University of Tartu, Estonia AI became my major focus and the main subject of study, allowing me in to join 2013 a newly formed Natural and Artificial Intelligence Lab at the Institute of Computer Science of University of Tartu, Estonia. Here both ML and Neuro knowledge were exactly the right tools to analyze a unique dataset of intracoritcal recordings from more than 100 human participants, explore the intersection of machine learning and neuroscience, and to work on my PhD thesis on “Understanding Information Processing in Human Brain by Interpreting Machine Learning Models” where I compare biological and artificial models of vision and explore the importance of explainability of machine learning models for gaining neuroscientific insights from models trained on brain data.

Looking back, what will become a career in machine learning and neurotechnology, started in 2010 as an almost science-fictional appeal of the brain-computer interface technology. The fact that the brain activity in different mental states results in measurably different signals is astonishing, as it means that our brain can be directly connected to various external systems and devices, not only to our body. The field of brain-computer interfaces or BCIs lies on the intersection of neuroscience and machine learning and explores this possibility.