# Wikipedia overview >_Computing Machinery and Intelligence_ is a seminal paper written by Alan Turing on the topic of artificial intelligence. The paper, published in 1950 in _Mind_, was the first to introduce his concept of what is now known as the Turing test to the general public. > >Turing's paper considers the question "Can machines think?" Turing says that since the words "think" and "machine" cannot be clearly defined we should "replace the question by another, which is closely related to it and is expressed in relatively unambiguous words." To do this, he must first find a simple and unambiguous idea to replace the word "think", second he must explain exactly which "machines" he is considering, and finally, armed with these tools, he formulates a new question, related to the first, that he believes he can answer in the affirmative. # Summary ## 1. The Imitation Game Turing opens by proposing the question, "Can machines think?" Then, questions the utility, and calls out the ambiguity, of the terms "machine" and "think." Turing proposes the **imitation game**: >It is played with three people, a man (A), a woman (B), and an interrogator (C) who may be of either sex. The interrogator stays in a room apart front the other two. The object of the game for the interrogator is to determine which of the other two is the man and which is the woman. He knows them by labels X and Y, and at the end of the game he says either "X is A and Y is B" or "X is B and Y is A." >In order that tones of voice may not help the interrogator the answers should be written, or better still, typewritten. The ideal arrangement is to have a teleprinter communicating between the two rooms. Alternatively the question and answers can be repeated by an intermediary. >We now ask the question, "What will happen when a machine takes the part of A in this game?" Will the interrogator decide wrongly as often when the game is played like this as he does when the game is played between a man and a woman? These questions replace our original, "Can machines think?" ## 2. Critique of the New Problem Turing hypothetically asks, "Is \[the imitation game\] a worthy one to investigate?" The narrow focus on written communication, avoiding physical attributes, has the advantage of focusing on intellectual capacity rather than physical ability. One might criticize the game as unfairly favoring humans. After all, if a human were try to imitate a machine, the human would surely be revealed on certain tasks. Nonetheless, if a machine can convincingly play the imitation game, we should not worry about these differences. Someone might object that a machine’s optimal strategy might not be to closely imitate human behavior, but to employ a different strategy to achieve success in the game. However, Turing doesn’t think this would significantly alter the game’s outcome or its effectiveness. >It might be urged that when playing the "imitation game" the best strategy for the machine may possibly be something other than imitation of the behaviour of a man. This may be, but I think it is unlikely that there is any great effect of this kind. In any case there is no intention to investigate here the theory of the game, and it will be assumed that the best strategy is to try to provide answers that would naturally be given by a man. ## 3. The Machines Concerned in the Game Turing then wants to clarify what we mean by "machine." Turing considers several parameters such as possible engineering techniques and the exclusion of biological entities. In the end, Turing specifies we shall only permit "electronic computers," also known as "digital computers." The primary question isn't whether current digital computers can succeed in the imitation game, but whether it's possible to imagine or develop computers that could. ## 4. Digital Computers A digital computer can be regarded as consisting of three parts: 1. Store (Memory): Stores data and instructions. 2. Executive unit (Processor): Executes operations, such as arithmetic calculations. 3. Control (Control unit): Ensures that the instructions stored in the memory are executed correctly and in the proper sequence. Digital computers of this kind have been constructed and can mimic the actions of a human computer very closely. "Programming" machines is the creation of instruction tables that direct the computer to perform specific operations, akin to translating a human computer's process into a machine-readable format. A variant of the digital computer is a "digital computer with a random element." These unpredictably modify state (akin to throwing a die) and are described as having free will. Though Turing refrains from using the term. Charles Babbage conceived of the Analytical Engine, a digital machine. It was slow and storage was purely mechanical, using wheels and cards. This illustrates how digital computers don't need to use electricity. Rather, what matters are the mathematical principles and underlying operations. ## 5. Universality of Digital Computers Digital computers are discrete-state machines, which transition between defined, discrete states, rather than moving continuously. This characteristic allows for precise control and predictable operation, essential for computational accuracy and reliability. Unlike the chaotic unpredictability of the universe, where tiny changes can have massive effects, discrete-state machines like digital computers maintain a stable causality that allows future states to be predicted from current ones. Turing proposes that given the state table of any discrete-state machine, a digital computer could, in theory, calculate and mimic the behavior of that machine. This is contingent on the digital computer being programmed specifically for that task and having adequate storage and processing speed. >Given the table corresponding to a discrete-state machine it is possible to predict what it will do. There is no reason why this calculation should not be carried out by means of a digital computer. Provided it could be carried out sufficiently quickly the digital computer could mimic the behavior of any discrete-state machine. The imitation game could then be played with the machine in question (as B) and the mimicking digital computer (as A) and the interrogator would be unable to distinguish them. Of course the digital computer must have an adequate storage capacity as well as working sufficiently fast. Moreover, it must be programmed afresh for each new machine which it is desired to mimic. This ability to mimic any discrete-state machine by calculating its behavior and replicating its outputs qualifies digital computers as universal machines. In light of the universality of digital computers, the question "Can machines think?" might be more precisely posed as whether a digital computer, with adequate storage, suitable speed, and appropriate programming might satisfactorily participate in the imitation game. >We may now consider again the point raised at the end of §3. It was suggested tentatively that the question, "Can machines think?" should be replaced by "Are there imaginable digital computers which would do well in the imitation game?" If we wish we can make this superficially more general and ask "Are there discrete-state machines which would do well?" But in view of the universality property we see that either of these questions is equivalent to this, "Let us fix our attention on one particular digital computer C. Is it true that by modifying this computer to have an adequate storage, suitably increasing its speed of action, and providing it with an appropriate programme, C can be made to play satisfactorily the part of A in the imitation game, the part of B being taken by a man?" ## 6. Contrary Views on the Main Question Turing has done the groundwork and will now precede to debate the question, "Can machines think?" First, Turing shares his own beliefs: - In about fifty years (the year 2000), computers will perform convincingly in the imitation game about 70% of the time after five minutes of questioning. - "Can machines think?" is too meaningless to warrant discussion. - By the end of the century, language and perceptions of educated people will have altered such that people will be able to speak of machines thinking without being corrected. - Scientific progress is not a linear journey from one confirmed fact to another. Instead, of conjectures, or hypotheses, can guide research directions. Provided it's clear which are proven facts and which are conjectures, no harm can result. ### (1) The Theological Objection >Thinking is a function of man's immortal soul. God has given an immortal soul to every man and woman, but not to any other animal or to machines. Hence no animal or machine can think. Turing argues that an omnipotent, divine power could bestow souls upon non-human animals. This same argument could be extended to machines. Creating thinking machines might not be an act of hubris or a challenge to divine authority but rather could be seen as fulfilling a divine will, akin to the procreation of children. Turing expresses skepticism about theological arguments, citing historical evidence such as the churches refutation of Copernican theory. ### (2) The "Heads in the Sand" Objection >The consequences of machines thinking would be too dreadful. Let us hope and believe that they cannot do so." This argument stems from the belief that humans are somehow special and superior to the rest of creation. Turing sees this objection as emotionally driven. Turing dismisses this objection as not meriting serious refutation but rather, perhaps, consolation for those troubled by the notion of machines that can think. ### (3) The Mathematical Objection >There are a number of results of mathematical logic which can be used to show that there are limitations to the powers of discrete-state machines. The best known of these results is known as Godel's theorem (1931) and shows that in any sufficiently powerful logical system statements can be formulated which can neither be proved nor disproved within the system, unless possibly the system itself is inconsistent. There are other, in some respects similar, results due to Church (1936), Kleene (1935), Rosser, and Turing (1937). These theorems show that there are questions that a computer cannot answer correctly or at all, due to the constraints of its programming and logical structure. Turing acknowledges these limitations but points out that the argument assumes, without proof, that humans are not subject to similar constraints. While machines may fail to answer some questions, this doesn't inherently elevate human intellect above machine capabilities, as humans too can err. Further, even if human intellect triumphs over a particular machine, there might always be other machines more clever. ### (4) The Argument from Consciousness The argument posits that for a machine to be truly considered as thinking, it must not only perform tasks (like writing a sonnet) but also experience thoughts and emotions associated with these tasks—capabilities presumed to be beyond mechanical processes. In the extreme form of this view, the only way to be sure that a machine thinks is to be the machine and to feel oneself thinking. It leads to solipsistic thinking where even the minds of other humans is doubted. While this is logically sound, it's too impractical, with everyone doubting everyone else. In reality, we all adopt the polite convention that everyone thinks. >In short then, I think that most of those who support the argument from consciousness could be persuaded to abandon it rather than be forced into the solipsist position. They will then probably be willing to accept our test. Turing admits the mystery surrounding consciousness but argues that solving these mysteries isn't necessary to address whether machines can think or exhibit behaviors indistinguishable from human intelligence. ### (5) Arguments from Various Disabilities >These arguments take the form, "I grant you that you can make machines do all the things you have mentioned but you will never be able to make one to do X." Numerous features X are suggested in this connexion I offer a selection: >Be kind, resourceful, beautiful, friendly, have initiative, have a sense of humour, tell right from wrong, make mistakes, fall in love, enjoy strawberries and cream \[...\] Inductive Reasoning Limits: These objections arise from people inductively inferring that because machines they've dealt with have limitations, then all machines are necessarily limited. But this is wrongheaded because it doesn't consider the potential for future advancements. Machine Mistakes: He differentiates between "errors of functioning" (mechanical failures) and "errors of conclusion" (logical or reasoning errors), arguing that machines can indeed make "mistakes" in the latter sense, which could be intentionally programmed to mimic human error patterns in tasks like the imitation game. Machine as Subject of Its Own Thought: Turing argues that a machine can be "the subject of its own thought," meaning it can operate on, analyze, and modify its own programming or behavior, a concept that's increasingly feasible with advances in computing. Storage Capacity and Behavior Diversity: He counters the criticism regarding the lack of behavioral diversity by pointing to advancements in storage capacity, which directly influences a machine's potential for varied behavior. ### (6) Lady Lovelace's Objection Ada Lovelace noted that Babbage's machine, a precursor to modern computers, could not "originate" actions or ideas; it could only execute tasks it was explicitly instructed to carry out. >Our most detailed information of Babbage's Analytical Engine comes from a memoir by Lady Lovelace (1842). In it she states, "The Analytical Engine has no pretensions to originate anything. It can do whatever we know how to order it to perform" (her italics). Turing references D.R. Hartree, who acknowledged Lovelace's statement but suggested that future electronic machines might possess the ability to "think for themselves" or learn through a process akin to conditioned reflexes. Turing concurs with Hartree, noting the distinction between Lovelace's observations based on the technology of her time and the potential capabilities of future machines. Turing counters a variant of Lovelace's objection—that a machine cannot do anything truly new or surprising—by sharing his personal experiences with machines surprising him. He attributes these surprises not to his own creativity but to his underestimation of the machines' behaviors based on the calculations or assumptions he made. The view that machines cannot surprise us is due to a common fallacy: the belief that understanding a principle or fact instantly grants insight into all its consequences. He asserts that this overlooks the value and complexity of deriving outcomes from given data and principles, a process machines are well-equipped to perform. ### (7) Argument from Continuity in the Nervous System >The nervous system is certainly not a discrete-state machine. A small error in the information about the size of a nervous impulse impinging on a neuron, may make a large difference to the size of the outgoing impulse. It may be argued that, this being so, one cannot expect to be able to mimic the behaviour of the nervous system with a discrete state system. The difference between continuous and discrete systems is real but inconsequential in the context of the imitation game. The interrogator would still have difficulty distinguishing between a continnous machine from a discrete one. ### (8) The Argument from Informality of Behaviour >It is not possible to produce a set of rules purporting to describe what a man should do in every conceivable set of circumstances. One might for instance have a rule that one is to stop when one sees a red traffic light, and to go if one sees a green one, but what if by some fault both appear together? One may perhaps decide that it is safest to stop. But some further difficulty may well arise from this decision later. To attempt to provide rules of conduct to cover every eventuality, even those arising from traffic lights, appears to be impossible. With all this I agree. Turing challenges this argument by distinguishing between "rules of conduct," which are guidelines for action (such as stopping at a stop light), and "laws of behavior," which are descriptive of natural phenomena (like someone squeaking if pinched.) He argues that the lack of complete rules of conduct does not imply machines cannot behave in complex, human-like ways, as machines can be designed to operate under principles that mimic human "laws of behavior." ### (9) The Argument from Extra-Sensory Perception This speculative argument suggests that because certain phenomena like telepathy and clairvoyance (collectively referred to as ESP) appear to operate outside the known laws of physics, they present a realm of human experience that machines could not replicate or engage with. Turing acknowledges the challenge ESP poses to scientific understanding but argues that the practical impact of ESP on the functioning of scientific theories is minimal. He humorously speculates that if ESP influences were real, they might affect machines equipped with random number generators, potentially allowing machines to "guess" telepathically in the imitation game. However, he also notes the necessity of isolating the game from such influences to maintain its integrity. ## 7. Learning Machines