Philosophy › Contemporary
Philosophy of artificial intelligence
1972 onward (roots in Turing, 1950)· School or movement· Knowledge, logic & science
Mostly agreedGood evidence; some details uncertain.
The study of whether machines can think, understand or be conscious, and of how artificial intelligence should behave.
See it on the Philosophy map →The idea
Idea: building thinking machines tests our theories of mind, meaning and ethics. Turing asked in 1950 whether machines can think. Hubert Dreyfus's What Computers Can't Do (1972) argued that human skill rests on bodily know-how no rule system captures. John Searle's Chinese room (1980) argued symbol manipulation is not understanding. Later questions: can AI be conscious, who is responsible for its decisions, and how can its goals be aligned with ours? Main dispute: whether thinking is just intelligent behavior or requires something more inside.
Why it matters
AI systems now write, advise and drive, so questions about understanding, responsibility and risk that were once hypothetical have become practical and urgent.
Where it comes from
Turing's 'Computing Machinery and Intelligence' (1950), Dreyfus (1972), Searle's 'Minds, Brains, and Programs' (1980), Bostrom's Superintelligence (2014), and a fast-growing literature on large language models.
Still debated
Whether intelligence requires consciousness, a body, or only the right computations remains open; there is no agreed test for machine understanding.
Inside this (4)
Connections (3)
▸Turing's imitation game ("Can machines think?") launched Philosophy of artificial intelligenceWell documented
The connection
Turing's 1950 paper turned 'Can machines think?' into a testable question and answered nine objections, from theology to consciousness. It set the agenda: later philosophers either defended behavior as the mark of mind or, like Dreyfus and Searle, argued that something more is needed.
When and how
'Computing Machinery and Intelligence', published in the philosophy journal Mind in October 1950, while Turing was at the University of Manchester.
How we know
Its appearance in a philosophy journal and its list of objections make it the standard starting point; nearly every textbook in the field begins with it.
Debates and open questions
Whether Turing meant the test as a definition of thinking, or only as good evidence for it, is still argued.
▸The trolley problem resurfaced in debates about Philosophy of artificial intelligenceMostly agreed
The connection
Self-driving cars made the thought experiment concrete: if a crash is unavoidable, should the car protect its passengers or pedestrians, the many or the few? MIT's Moral Machine experiment gathered public verdicts worldwide. The case now frames broader questions about building ethical choices into machines.
When and how
MIT's Moral Machine website (2016), reported in Nature in 2018: about 40 million decisions from 233 countries and territories. Germany issued ethics rules for automated driving in 2017.
How sure are we?
The Nature paper and government guidelines document the link, though engineers often say real crashes rarely pose clean trolley-style choices.
Debates and open questions
Many argue trolley cases are the wrong model for AI ethics, which is more about bias, everyday risk and responsibility than dramatic dilemmas.
▸Nick Bostrom: superintelligence and the simulation argument popularized AI risk in Philosophy of artificial intelligenceMostly agreed
The connection
Bostrom's Superintelligence argued that a machine smarter than humans could gain decisive power, and that its goals might not match ours: the 'control problem'. This moved philosophy of AI from 'Can machines think?' toward 'What if they outthink us?' and helped make AI alignment a field.
When and how
Superintelligence: Paths, Dangers, Strategies (Oxford University Press, 2014); the statistician I. J. Good had predicted an 'intelligence explosion' in 1965.
How sure are we?
The book was a bestseller and prompted public warnings from figures such as Elon Musk and Bill Gates, though others, like Eliezer Yudkowsky, developed the ideas in parallel.
Debates and open questions
AI researchers remain split on whether superintelligence is near, distant, or a distraction from present problems like bias and misuse.