A normal laboratory functions like a computer, but with a human scientist as the operating system. A self-driving lab integrates AI with automated hardware, allowing the system to learn and adapt its experiments based on real-time results. This distinction between automation and autonomy means self-driving labs can make informed decisions and redirect their research, unlike automated systems that merely execute pre-written instructions.

• Laboratories possess components similar to computers: sensors, actuators, memory, protocols, data outputs, and error states.
• Human scientists typically act as the ‘operating system’ in traditional labs, making decisions and directing experiments.
• Self-driving labs incorporate AI with automated hardware to create a system that learns and decides the next experimental steps.
• The core process of a self-driving lab follows a loop: design → make → test → learn → design again.
• Autonomy in labs means the ability to learn from results and redirect research, unlike automation which strictly executes instructions. https://thesequence.substack.com/p/the-sequence-opinion-884-self-driving https://bender.layer3.press/articles/95363a78-e9fd-4a5c-986b-ac5197962e92