A precise vocabulary can distinguish the discipline that builds intelligent systems from the intelligent computational systems that operate in the world.
Machine Intelligence (MI) is the operational instantiation of cognitive capabilities—such as learning, reasoning, planning, adaptation, and goal achievement—within engineered computational substrates rather than biological organisms.
A historical and continuing field of research and engineering concerned with creating systems that exhibit intelligent behavior.
A machine-based intelligence operating through computational hardware, software, state, memory, models, and interfaces.
Because artificial describes engineered origin, while machine describes the substrate in which the intelligence is instantiated. The distinction is intended to be taxonomic: it separates the operating system from the field that produced it.
No. Intelligence, consciousness, sentience, agency, autonomy, moral status, and legal personhood are separate questions. The terminology change does not settle any of them.
Yes. MachineIntelligences.org retains AI for the scientific field, established standards and law, historical references, direct quotations, interoperability, and search discoverability.
No. It is this project’s preferred terminology and has historical use in computing. The site presents the argument for the term rather than claiming that every institution has adopted it.
MachineIntelligences.org deliberately retains “Artificial Intelligence” and “AI” in metadata, quotations, historical discussion, legal mappings, and transitional explanations. Precision should improve discoverability, not destroy it.
The visual campaign condenses the site’s central distinction: Artificial Intelligence can name the field while Machine Intelligence names the instantiated computational intelligence.
Read the full visual essay