My AI Model Was 89% Accurate. It Missed 99% of the State I Cared About
I trained a small CoreML classifier to estimate whether a player was bored, in flow, frustrated, or...
Tag archive
I trained a small CoreML classifier to estimate whether a player was bored, in flow, frustrated, or...
I hit a very ordinary computer-vision failure this week: the model was doing what I asked, but the...
Canonical version:...
I spent part of this week debugging an on-device livestock counter in a field-video mode. The model...

Environment for everything below: macOS 26.5 (Darwin 25.5.0), Apple Silicon (arm64), coremltools 9.0,...
Every iPhone user knows the feeling: iCloud Storage Full. When I checked my own device, 68% of...

I have been testing how far Stable Diffusion 1.5 can be pushed running locally on iPhone 17 and how...
CoreML Metal GPU delivers 9ms vs TFLite's 21ms Running MobileNetV2 inference on an iPhone...
Most on-device AI content focuses on inference — you ship a pre-trained model in your app bundle and...
Research papers on pose estimation show impressive accuracy numbers. Production apps on consumer...
For a while I'd wanted to build a Swift Package that runs Meta's Segment Anything Model (SAM)...
What it does Type text like "person, red car, coffee cup" and it detects those objects in...