A debate “The Superpowers of Human and Machine Intelligence: a Bumpy Road toward ever Greater Intelligence” between Dr. Stefan Buijsman (TU Delft) and Mark Saroufim (ML engineer at Facebook AI)
Profile
Stefan Buijsman studied computer science and philosophy in Leiden and completed his Ph.D. on the philosophy of mathematics at Stockholm University when he was only 20 years old. Then, he conducted research at the intersection of philosophy of mathematics and cognitive science at Stockholm University and the Institute for Futures Studies. Aside from research, he dedicates time to popular science writing, with now three books to his name.
Research
His research focuses on the explainability of AI algorithms, and more generally, how we can design AI such that it can be used responsibly and beneficially. This includes work on fairness and transparency. He is interested in combining work from a philosophy of mathematics, philosophy of science, and artificial intelligence, primarily to further our theoretical understanding of AI.
Profile
Mark Saroufim is a Machine Learning Engineer on Pytorch at Facebook AI. Previously, Mark has worked at Graphcore, http://yuri.ai, Microsoft, and NASA’s JPL. Mark regularly writes opinion pieces on the ML landscape on https://marksaroufim.substack.com and is the author of an ebook on robotics http://robotoverlordmanual.com.
A deep dive into the topic:
Korteling, J., van de Boer-Visschedijk, G., Blankendaal, R., Boonekamp, R., & Eikelboom, A. (2021). Human-versus Artificial Intelligence. Frontiers in Artificial Intelligence, 4. [link]
Lake, B. M., Ullman, T. D., Tenenbaum, J. B., & Gershman, S. J. (2017). Building machines that learn and think like people. Behavioral and Brain Sciences, 40. [pdf]
Shneiderman, B. (2020). Design lessons from AI’s two grand goals: Human emulation and useful applications. IEEE Transactions on Technology and Society, 1(2), 73-82. [pdf]