When somebody asks me where to start with Physical AI, I used to send three links and an apology. The links were long and the apology was short.
So I spent some time putting the basics in order. Ten short guides, each one answers one question, and none of them needs an equation… with one exception, and it comes with a bucket.
The shelf
- What Physical AI is: the long version, with the research behind it.
- Physical AI vs generative AI: the pancake test.
- Physical AI vs embodied AI vs robotics vs edge AI: where I draw the lines.
- The glossary: 26 terms in plain words, also as a file and a printable PDF.
- The numbers: seven figures I trust, each with its source.
- Continual learning for robots: why machines forget, and what to do about it.
- The leaky integrator equation: explained with a bucket.
- A reading list: twelve papers, 1987 to 2023.
- What a Physical AI engineer does: skills, backgrounds and what I look for.
- The FAQ: 23 questions, short answers.
Start anywhere. If you are in a hurry, read the first two and the FAQ. If you are a student, read the reading list and build something that moves.
Something missing, or wrong? Tell me. I would rather fix it than defend it.