AI Simplified (previously published as Simply AI, shown above), is a primer on Artificial Intelligence. It covers:
- History of AI
- Classical AI
- Statistical AI
- Using AI
- Philosophy of AI
- Living with AI
As a primer, it skims the surface of many topics within this arena. It is short (155 pages with large text), and highly visual (there is a graphic on each page).
AI is obviously a hot topic, and I wanted to learn more. Two key takeaways for me:
- "Machine-learning networks are only as good as the data on which they are trained." That means that incorrect, incomplete, or biased data will create predictable problematic outputs. And "hidden behind many of today's most powerful AIs is a vast amount of human labor. Millions of workers are used to train AIs"—workers that are faulty (and may be poorly compensated). And generative AI "relies on data scraped from the internet," which creates a few problems (including consent from and compensation for the creators).
- As AI is increasingly incorporated into many functions and can be 'agentic' (acting on behalf of a human or institution), lots of questions arise, including responsibility (who is responsible if AI does something that harms?) and ethics ("the only way to ensure that AIs think ethically is to program them with ethical principles, although then the question becomes whose ethics?").
DK Publishing does a good job with overviews, and this is no exception. This book provides helpful definitions, history, basic concepts, and more related to AI. Recommended for those who want to understand the radical transformation going on in the world around us (with much more to come).
Rating: A
