I love when computation moves beyond the screen into our lived environment, into surfaces, tools, and everyday objects, making our interaction with the physical world richer. (We eat, sleep, drive, live, pollute in the physical world – so why not?) I lead the Spatial AI initiative at Amazon Lab126, exploring how intelligence can bridge digital and physical systems – making invisible bits of sensing smart using on-device CNNs and VLMs. I led human-centred AI at Useful Sensors, a spinout with the founding team of TensorFlow/Google Brain, and earlier developed multi-projector stroke and hover-gesture recognition systems, integrating speech input to replace traditional WYSIWYG interfaces for complex design tools, at Google and HP Labs. I am a graduate and research affiliate of the Fluid In...