They don't capture a light field like Lytro did, they capture a regular image with a very deep depth of field, extract a depth map (usually with machine learning, but some phones augment it with stereoscopy or even LIDAR on high end iPhones) and then selectively blur based on depth.
They don't capture a light field like Lytro did, they capture a regular image with a very deep depth of field, extract a depth map (usually with machine learning, but some phones augment it with stereoscopy or even LIDAR on high end iPhones) and then selectively blur based on depth.