The Missing Half of Science That AI Can't See | Serge Didenko | TEDxBoston
What if the biggest obstacle to scientific progress is the research we never see? Inspired by the mysteries of dark matter and dark energy, this talk explores a surprising parallel between the universe and modern science. While published discoveries represent only a fraction of scientific knowledge, millions of failed experiments, abandoned hypotheses, and negative results remain invisible—even though they hold invaluable lessons for future breakthroughs. Drawing from research at Harvard Medical School, the speaker argues that if we want to build the next generation of AI scientists, we must teach them not only from successful papers, but also from the unseen experiments that shape how science is truly done. The future of artificial intelligence may depend less on perfect answers and more on learning from failure. Serge Vasylechko, who has also published as Serge Didenko Vasylechko, is a postdoctoral research fellow at Harvard Medical School and Boston Children's Hospital, where he develops deep learning methods for medical imaging. He advances physics-aware AI for clinical challenges in magnetic resonance imaging (MRI) and builds educational and community platforms to teach AI development skills. Concurrently, he is a founder and educator in the field of AI entrepreneurship, having co-founded the Sundai Foundation and holding teaching positions at the Massachusetts Institute of Technology (MIT) and INSEAD. Since October 2019, Vasylechko has been a postdoctoral research fellow in the Quantitative Intelligent Medical Imaging (QUIN) Research Group, which is part of the Computational Radiology Laboratory (CRL) at Boston Children's Hospital. His research focuses on generative modeling and representation learning for quantitative parameter mapping in MRI. He works under the supervision of Sila Kurugol and Onur Afacan, both Associate Professors of Radiology at Harvard Medical School This talk was given at a TEDx event using the TED conference format but independently organized by a local community. Learn more at https://www.ted.com/tedx
