Xebec
Well-known member
After listening to this podcast, I bought the book, which I found an enjoyable read. (Like the podcast talked about,) The book covers a large variety of topics related to AI, in pretty plain english -- making it easy to follow. I enjoyed learning about how AI LLMs came to be, the opportunities and limitations that have come with increasing scale (i.e. GPT 1-4, and predecessors), and also the full gamut of cost considerations for AI efforts - startups, AI providers, and chip makers. It was also fun seeing in a printed book - the topics I already know about with respect to semiconductors: chip scaling, CPU vs GPU vs ASIC, etc.
Some of my favorite chapters:
Chapter 4 - Language Models (covers the history, limitations)
Chapter 5 - (How) Image Generation (Works), also Chapter 6 and 7 - Recommender and Multimodal Models
Chapter 14 - Computation and Storage Requirements
Chapter 20 - Nvidias competitors (Signs of AI market maturity here)
Chapter 39 - How far can we go?
Note that the chapters in this book are very short - they're more of a primer for a topic for the reader to dive into more deeply "offline" if they choose. Some of the chapters are also a bit repetitive, though that also serves as a reinforcing function.
Thank you Semiwiki (Daniel) for this podcast and book recommendation.
Podcast
semiwiki.com
Some of my favorite chapters:
Chapter 4 - Language Models (covers the history, limitations)
Chapter 5 - (How) Image Generation (Works), also Chapter 6 and 7 - Recommender and Multimodal Models
Chapter 14 - Computation and Storage Requirements
Chapter 20 - Nvidias competitors (Signs of AI market maturity here)
Chapter 39 - How far can we go?
Note that the chapters in this book are very short - they're more of a primer for a topic for the reader to dive into more deeply "offline" if they choose. Some of the chapters are also a bit repetitive, though that also serves as a reinforcing function.
Thank you Semiwiki (Daniel) for this podcast and book recommendation.
Podcast
Podcast EP332: How AI Really Works – the Perspectives of Linley Gwennap - Semiwiki
Daniel is joined by Linley Gwennap, technology analyst and author of the new book “How AI Really Works: The Models, Chips, and Companies Powering a Revolution.” Linley was the long-time editor of Microprocessor Report and chaired the popular Linley Processor Conferences. Dan explores what impact...
semiwiki.com