Sandboxed - On-Device AI for iOS

Logicbridge
  • 0.0 (0)
  • TECNOLOGÍA
  • CADA SEMANA

Sandboxed is a podcast for iOS developers who want to add AI and machine learning features to their apps—without needing a PhD in ML. Each episode, we take one practical topic—like Vision, Core ML, or Apple Intelligence—and walk through how it works, what you can build with it, and how to ship it.

Episodios

  1. HACE 3 DÍAS

    Neural Network Architecture: The 10,000-Foot View

    We strip away the biological metaphors of AI to reveal the mechanical reality: differentiable decision trees. We explore the Forward Pass, the role of ReLU as a logic gate, and why "Deep" hierarchies work. Finally, we break down the hardware reality—why CPUs are Ferraris, NPUs are Bullet Trains, and why memory bandwidth is the only metric that matters.

    39 min

  2. 11 FEB

    Machine Learning Engineering: Training vs. Inference

    As iOS engineers, we respect the science of training models, but we live in the trenches of inference. In this episode, we explore the "Great Divide" between creating a model and running it. We break down the memory mechanics of O(N) training vs O(1) inference, dissect compiler optimizations like kernel fusion, and explain exactly how the Apple Neural Engine cheats bandwidth physics using quantization.

    23 min

  3. 8 FEB

    The Frozen Artifact: Understanding Core ML Models

    We deconstruct the .mlmodelc bundle to understand exactly what we're shipping. Does a model have logic? No. It has settings. We explore the "Factory" analogy, the difference between architecture and weights, and why we must treat ML models as frozen assets, not readable code.

    25 min

  4. 1 ENE

    Machine Learning Primitives: The Shape of the Answer

    We stop treating AI as a "magic black box" and start treating it as a function with a strict return type. Sam and Alex break down the three fundamental primitives of Machine Learning—Classification, Regression, and Clustering—and explain how to map vague product requirements into rigorous engineering specifications.

    29 min

  5. 28/12/2025

    The Universal Function Approximator

    We dismantle the mystical "black box" of AI to reveal simple mechanical gears: Function Approximation and Optimization. We explain why code is dying, how "learning" is just curve-fitting, and why your iPhone is built to run these models, not build them.

    33 min
  6. The Fork in the Road: Edge vs. Cloud

    24/12/2025

    The Fork in the Road: Edge vs. Cloud

    We explore the strategic shift from cloud-dependent dominance to local intelligence. Why latency is the enemy, how Apple's Unified Memory Architecture "cheated" the laws of PC physics, and why the future of AI economics favors the device in your pocket.

    28 min

  7. 19/12/2025

    The End of "If-Else"

    We explore the paradigm shift from deterministic programming (Software 1.0) to probabilistic machine learning (Software 2.0). We break down how ML models work using the "Factory" analogy, explain the "Semantic Gap" that kills traditional code, and dive deep into how Apple Silicon (ANE) handles these massive computational loads with "Zero Copy" efficiency.

    25 min
  • 7 episodios

Acerca de

Sandboxed is a podcast for iOS developers who want to add AI and machine learning features to their apps—without needing a PhD in ML. Each episode, we take one practical topic—like Vision, Core ML, or Apple Intelligence—and walk through how it works, what you can build with it, and how to ship it.

Información