Page 149 of 506 · Back to latest writing
Yejin Choi, a computer scientist and MacArthur Fellow, researches common sense and moral reasoning in AI. She rejects the belief that larger models will repair their own flaws, emphasizing algorithmic efficiency, pluralistic values, and intelligence’s unspoken rules.
Jitendra Malik, a UC Berkeley professor and computer vision researcher, developed early methods for image segmentation and object recognition. His case for physical, sensorimotor intelligence poses a reciprocal challenge: machines must move to see and see to move.
Christopher Manning, a Stanford professor, studies how computers process language. His work across statistical NLP, the GloVe embedding model, foundation models, linguistics, and deep-learning architectures offers a view of how large language models actually work.
Michael I. Jordan, a computer scientist and statistician, linked graphical models with machine learning and applies microeconomics to decentralized data systems. His work reframes AI around decision mathematics, reliability, and the need for an engineering discipline.
Joy Buolamwini, computer scientist and founder of the Algorithmic Justice League, exposed racial and gender biases in commercial facial recognition. Her “coded gaze” reveals how prejudice enters AI, making digital civil rights inseparable from data and power.
Timnit Gebru studies algorithmic bias and AI’s social impact as a computer scientist. Her co-written Gender Shades and Stochastic Parrots link facial-recognition and language-model failures to data representation, corporate accountability, and the field’s ideology.
Sanjay Ghemawat, a software engineer behind Google’s distributed systems, co-wrote MapReduce and the Google File System with Jeff Dean. His practices in pair programming, system design, and performance optimization show how abstractions enable internet-scale processing.
Quoc Le, a Google DeepMind researcher, helped develop sequence-to-sequence learning, AutoML, and EfficientNet. His work replaces hand-designed models with systems that discover their own architectures, showing how machine learning can scale across research areas.