Categories
Artificial Intelligence Education Reading

Latest Read: AI and Machine Learning for On-Device Development

AI and Machine Learning for On-Device Development: A Programmer’s Guide by Laurence Moroney.

AI and Machine Learning for On-Device Development: A Programmer's Guide by Laurence Moroney

Laurence holds a Bachelor of Science in Physics and Computer Science from Cardiff University, PgD, Microelectronics Systems Design from Birmingham City University and Graduate Certificate in Artificial Intelligence from Stanford University.

Today he is Chief AI Scientist at VisionWorks Studios. He previously was an AI Advocate at Google for 10 years and served as a Senior Developer Evangelist at Microsoft. He wrote this book in 2021 and previously published AI and Machine Learning for Coders in 2020.

While OpenAI’s ChatGPT kick started the LLM surge, AI will simply be a component installed upon our mobile devices. The OpenAI/Microsoft partnership is certainly enterprise focused, Google Gemini and Apple will drive their Android and iOS devices to simply adopt AI as part of their mobile ecosystem. Phones will simply remain the go to device.

So, there was a lot of buzz regarding new AI devices including Humane’s AI Pin or the Rabbit R1. Their rush to market to capitalize on the AI hype cycle leads to critical mistakes.

ML to your smartphone

Yet ask anyone who is forced to carry a work phone in addition to their personal phone. How happy are they really? So here, Google and Apple can press their advantage with Android Studio and Swift. It will be critical for their developers to adopt AI services into their existing IDEs. But hardware upgrades to support NPUs are needed. I would highly recommend Sean Gerrish’s book How Smart Machines Think and Luis Serrano’s Grokking Machine Learning.

October 2021 Review
October 2021 Review

This is a very enlightening look at creating AI models on mobile devices. I found the refresher on Machine Learning concepts concise for those not aware of this subset of AI. The tools programmers can use to write AI apps include computer vision and text recognition. There is of course a nod to his role at Google in recommending ML Kit, TensorFlow Lite, and Core ML.

In addition, TensorFlow Lite can run on a Raspberry Pi with Google’s Coral. This USB accelerator which allows AI tools to be created on the small and very affordable maker device. Perfect for both K12 and Higher education. Furthermore, Google has made TensorFlow Lite free to download and offers a rich set of support documents to empower students of all ages.

In conclusion, Laurence brings this book to closure with chapter 15: Ethics, Fairness, and Privacy with Responsible AI. Well positioned, he shares Google’s AI Principles.


George Zoto | A chat with Laurence Moroney, AI Lead at Google
Sierra Circuits | AI and ML Talk with Google Laurence Moroney
Google for Developers | Machine Learning Foundations