Quantum cognition

From Wikipedia, the free encyclopedia

Quantum cognition is an emerging field which applies the mathematical formalism of quantum theory to model cognitive phenomena such as information processing by the human brain, language, decision making, human memory, concepts and conceptual reasoning, human judgment, and perception.^[1][2][3][4] The field clearly distinguishes itself from the quantum mind as it is not reliant on the hypothesis that there is something micro-physical quantum mechanical about the brain. Quantum cognition is based on the quantum-like paradigm^[5][6] or generalized quantum paradigm^[7] or quantum structure paradigm^[8] that information processing by complex systems such as the brain, taking into account contextual dependence of information and probabilistic reasoning, can be mathematically described in the framework of quantum information and quantum probability theory.

Quantum cognition uses the mathematical formalism of quantum theory to inspire and formalize models of cognition that aim to be an advance over models based on traditional classical probability theory. The field focuses on modeling phenomena in cognitive science that have resisted traditional techniques or where traditional models seem to have reached a barrier (e.g., human memory^[9]), and modeling preferences in decision theory that seem paradoxical from a traditional rational point of view (e.g., preference reversals^[10]). Since the use of a quantum-theoretic framework is for modeling purposes, the identification of quantum structures in cognitive phenomena does not presuppose the existence of microscopic quantum processes in the human brain.^[11]

Main subjects of research

Quantum-like models of information processing ("quantum-like brain")

The brain is definitely a macroscopic physical system operating on the scales (of time, space, temperature) which differ crucially from the corresponding quantum scales. (The macroscopic quantum physical phenomena such as e.g. the Bose-Einstein condensate are also characterized by the special conditions which are definitely not fulfilled in the brain.) In particular, the brain is simply too hot to be able perform the real quantum information processing, i.e., to use the quantum carriers of information such as photons, ions, electrons. As is commonly accepted in brain science, the basic unit of information processing is a neuron. It is clear that a neuron cannot be in the superposition of two states: firing and non-firing. Hence, it cannot produce superposition playing the basic role in the quantum information processing. Superpositions of mental states are created by complex networks of neurons (and these are classical neural networks). Quantum cognition community states that the activity of such neural networks can produce effects which are formally described as interference (of probabilities) and entanglement. In principle, the community does not try to create the concrete models of quantum (-like) representation of information in the brain.^[12]

The quantum cognition project is based on the observation that various cognitive phenomena are more adequately described by quantum information theory and quantum probability than by the corresponding classical theories, see examples below. Thus the quantum formalism is considered as an operational formalism describing nonclassical processing of probabilistic data. Recent derivations of the complete quantum formalism from simple operational principles for representation of information supports the foundations of quantum cognition. The subjective probability viewpoint on quantum probability which was developed by C. Fuchs and collaborators^[13] also supports the quantum cognition approach, especially using of quantum probabilities to describe the process of decision making.

Although at the moment we cannot present the concrete neurophysiological mechanisms of creation of the quantum-like representation of information in the brain,^[14] we can present general informational considerations supporting the idea that information processing in the brain matches with quantum information and probability. Here, contextuality is the key word, see the monograph of Khrennikov^[1] for detailed representation of this viewpoint. Quantum mechanics is fundamentally contextual.^[15] Quantum systems do not have objective properties which can be defined independently of measurement context. (As was pointed by N. Bohr, the whole experimental arrangement must be taken into account.) Contextuality implies existence of incompatible mental variables, violation of the classical law of total probability and (constructive and destructive) interference effects. Thus the quantum cognition approach can be considered as an attempt to formalize contextuality of mental processes by using the mathematical apparatus of quantum mechanics.

Decision making

Suppose a person is given an opportunity to play two rounds of the following gamble: a coin toss will determine whether the subject wins $200 or loses $100. Suppose the subject has decided to play the first round, and does so. Some subjects are then given the result (win or lose) of the first round, while other subjects are not yet given any information about the results. The experimenter then asks whether the subject wishes to play the second round. Performing this experiment with real subjects gives the following results:

1. When subjects believe they won the first round, the majority of subjects choose to play again on the second round.
2. When subjects believe they lost the first round, the majority of subjects choose to play again on the second round.

Given these two separate choices, according to the sure thing principle of rational decision theory, they should also play the second round even if they don’t know or think about the outcome of the first round.^[16] But, experimentally, when subjects are not told the results of the first round, the majority of them decline to play a second round.^[17] This finding violates the law of total probability, yet it can be explained as a quantum interference effect in a manner similar to the explanation for the results from double-slit experiment in quantum physics.^[2][18][19] Similar violations of the sure-thing principle are seen in empirical studies of the Prisoner's Dilemma and have likewise been modeled in terms of quantum interference.^[20]

The above deviations from classical rational expectations in agents’ decisions under uncertainty produce well known paradoxes in behavioral economics, that is, the Allais, Ellsberg and Machina paradoxes.^[21][22][23] These deviations can be explained if one assumes that the overall conceptual landscape influences the subject’s choice in a neither predictable nor controllable way. A decision process is thus an intrinsically contextual process, hence it cannot be modeled in a single Kolmogorovian probability space, which justifies the employment of quantum probability models in decision theory. More explicitly, the paradoxical situations above can be represented in a unified Hilbert space formalism where human behavior under uncertainty is explained in terms of genuine quantum aspects, namely, superposition, interference, contextuality and incompatibility.^{[24][25][26][19]}

Considering automated decision making, quantum decision trees have different structure compared to classical decision trees. Data can be analyzed to see if a quantum decision tree model fits the data better.^{[citation needed]}

Human probability judgments

Quantum probability provides a new way to explain human probability judgment errors including the conjunction and disjunction errors.^[27] A conjunction error occurs when a person judges the probability of a likely event L and an unlikely event U to be greater than the unlikely event U; a disjunction error occurs when a person judges the probability of a likely event L to be greater than the probability of the likely event L or an unlikely event U. Quantum probability theory is a generalization of Bayesian probability theory because it is based on a set of von Neumann axioms that relax some of the classic Kolmogorov axioms.^[28] The quantum model introduces a new fundamental concept to cognition—the compatibility versus incompatibility of questions and the effect this can have on the sequential order of judgments. Quantum probability provides a simple account of conjunction and disjunction errors as well as many other findings such as order effects on probability judgments.^[29][30][31]

The liar paradox - The contextual influence of a human subject on the truth behavior of a cognitive entity is explicitly exhibited by the so-called liar paradox, that is, the truth value of a sentence like "this sentence is false". One can show that the true-false state of this paradox is represented in a complex Hilbert space, while the typical oscillations between true and false are dynamically described by the Schrödinger equation.^[32][33]

Knowledge representation

Concepts are basic cognitive phenomena, which provide the content for inference, explanation, and language understanding. Cognitive psychology has researched different approaches for understanding concepts including exemplars, prototypes, and neural networks, and different fundamental problems have been identified, such as the experimentally tested non classical behavior for the conjunction and disjunction of concepts, more specifically the Pet-Fish problem or guppy effect,^[34] and the overextension and underextension of typicality and membership weight for conjunction and disjunction.^[35][36] By and large, quantum cognition has drawn on quantum theory in three ways to model concepts.

1. Exploit the contextuality of quantum theory to account for the contextuality of concepts in cognition and language and the phenomenon of emergent properties when concepts combine^{[11][37][38][39][40]}
2. Use quantum entanglement to model the semantics of concept combinations in a non-decompositional way, and to account for the emergent properties/associates/inferences in relation to concept combinations^[41]
3. Use quantum superposition to account for the emergence of a new concept when concepts are combined, and as a consequence put forward an explanatory model for the Pet-Fish problem situation, and the overextension and underextension of membership weights for the conjunction and disjunction of concepts.^[29][37][38]

The large amount of data collected by Hampton^[35][36] on the combination of two concepts can be modeled in a specific quantum-theoretic framework in Fock space where the observed deviations from classical set (fuzzy set) theory, the above-mentioned over- and under- extension of membership weights, are explained in terms of contextual interactions, superposition, interference, entanglement and emergence.^{[29][42][43][44]} And, more, a cognitive test on a specific concept combination has been performed which directly reveals, through the violation of Bell’s inequalities, quantum entanglement between the component concepts.^[45][46]

Human memory

The hypothesis that there may be something quantum-like about the human mental function was put forward with the quantum entanglement formula which attempted to model the effect that when a word’s associative network is activated during study in memory experiment, it behaves like a quantum-entangled system.^[9] Models of cognitive agents and memory based on quantum collectives have been proposed by Subhash Kak.^[47][48] But he also points to specific problems of limits on observation and control of these memories due to fundamental logical reasons.^[49]

Semantic analysis and information retrieval

The research in (iv) had a deep impact on the understanding and initial development of a formalism to obtain semantic information when dealing with concepts, their combinations and variable contexts in a corpus of unstructured documents. This conundrum of natural language processing (NLP) and information retrieval (IR) on the web – and data bases in general – can be addressed using the mathematical formalism of quantum theory. As basic steps, (a) the seminal book "The Geometry of Information Retrieval" by K. Van Rijsbergen^[50] introduced a quantum structure approach to IR, (b) Widdows and Peters utilised a quantum logical negation for a concrete search system,^[40][51] and Aerts and Czachor identified quantum structure in semantic space theories, such as latent semantic analysis.^[52] Since then, the employment of techniques and procedures induced from the mathematical formalisms of quantum theory – Hilbert space, quantum logic and probability, non-commutative algebras, etc. – in fields such as IR and NLP, has produced significant results.^[53]

Human perception

Bi-stable perceptual phenomena is a fascinating topic in the area of perception. If a stimulus has an ambiguous interpretation, such as a Necker cube, the interpretation tends to oscillate across time. Quantum models have been developed to predict the time period between oscillations and how these periods change with frequency of measurement.^[54] Quantum theory and an appropriate model have been developed by Elio Conte to account for interference effects obtained with measurements of ambiguous figures.^{[55][56][57][58]}

Gestalt perception

There are apparent similarities between Gestalt perception and quantum theory. In an article discussing the application of Gestalt to chemistry, Anton Amann writes: "Quantum mechanics does not explain Gestalt perception, of course, but in quantum mechanics and Gestalt psychology there exist almost isomorphic conceptions and problems:

• Similarly as with the Gestalt concept, the shape of a quantum object does not a priori exist but it depends on the interaction of this quantum object with the environment (for example: an observer or a measurement apparatus).
• Quantum mechanics and Gestalt perception are organized in a holistic way. Subentities do not necessarily exist in a distinct, individual sense.
• In quantum mechanics and Gestalt perception objects have to be created by elimination of holistic correlations with the 'rest of the world'."^[59]

Amann comments: "The structural similarities between Gestalt perception and quantum mechanics are on a level of a parable, but even parables can teach us something, for example, that quantum mechanics is more than just production of numerical results or that the Gestalt concept is more than just a silly idea, incompatible with atomistic conceptions."^[59]

Quantum-like models of cognition in economics and finance

The assumption that information processing by the agents of the market follows the laws of quantum information theory and quantum probability was actively explored by many authors, e.g., E. Haven, O. Choustova, A. Khrennikov, see the book of E. Haven and A. Khrennikov,^[60] for detailed bibliography. We can mention, e.g., the Bohmian model of dynamics of prices of shares in which the quantum(-like) potential is generated by expectations of agents of the financial market and, hence, it has the mental nature. This approach can be used to model real financial data, see the book of E. Haven and A. Khrennikov (2012).

Application of theory of open quantum systems to decision making and "cell's cognition"

An isolated quantum system is an idealized theoretical entity. In reality interactions with environment have to be taken into account. This is the subject of theory of open quantum systems. Cognition is also fundamentally contextual. The brain is a kind of (self-)observer which makes context dependent decisions. Mental environment plays a crucial role in information processing. Therefore, it is natural to apply theory of open quantum systems to describe the process of decision making as the result of quantum-like dynamics of the mental state of a system interacting with an environment. The description of the process of decision making is mathematically equivalent to the description of the process of decoherence. This idea was explored in a series of works of the multidisciplinary group of researchers at Tokyo University of Science.^[61][62]

Since in the quantum-like approach the formalism of quantum mechanics is considered as a purely operational formalism, it can be applied to the description of information processing by any biological system, i.e., not only by human beings.

Operationally it is very convenient to consider e.g. a cell as a kind of decision maker processing information in the quantum information framework. This idea was explored in a series of papers of the Swedish-Japanese research group using the methods of theory of open quantum systems: genes expressions were modeled as decision making in the process of interaction with environment.^[63]

History

Here is a short history of applying the formalisms of quantum theory to topics in psychology. Ideas for applying quantum formalisms to cognition first appeared in the 1990s by Diederik Aerts and his collaborators Jan Broekaert, Sonja Smets and Liane Gabora, by Harald Atmanspacher, Robert Bordley, and Andrei Khrennikov. A special issue on Quantum Cognition and Decision appeared in the Journal of Mathematical Psychology (2009, vol 53.), which planted a flag for the field. A few books related to quantum cognition have been published including those by Khrennikov (2004, 2010), Ivancivic and Ivancivic (2010), Busemeyer and Bruza (2012), E. Conte (2012). The first Quantum Interaction workshop was held at Stanford in 2007 organized by Peter Bruza, William Lawless, C. J. van Rijsbergen, and Don Sofge as part of the 2007 AAAI Spring Symposium Series. This was followed by workshops at Oxford in 2008, Saarbrücken in 2009, at the 2010 AAAI Fall Symposium Series held in Washington, D.C., 2011 in Aberdeen, 2012 in Paris, and 2013 in Leicester. Tutorials also were presented annually beginning in 2007 until 2013 at the annual meeting of the Cognitive Science Society. A Special Issue on Quantum models of Cognition appeared in 2013 Topics in Cognitive Science.

Related theories

It was suggested by theoretical physicists David Bohm and Basil Hiley that mind and matter both emerge from an "implicate order".^[64] Bohm and Hiley's approach to mind and matter is supported by philosopher Paavo Pylkkänen.^[65] Pylkkänen underlines "unpredictable, uncontrollable, indivisible and non-logical" features of conscious thought and draws parallels to a philosophical movement some call "post-phenomenology", in particular to Pauli Pylkkö's notion of the "aconceptual experience", an unstructured, unarticulated and pre-logical experience.^[66]

The mathematical techniques of both Conte's group and Hiley's group involve the use of Clifford algebras. These algebras account for "non-commutativity" of thought processes (for an example, see: noncommutative operations in everyday life).
However, an area that needs to be investigated is the concept lateralised brain functioning. Some studies in marketing have related lateral influences on cognition and emotion in processing of attachment related stimuli.

Artificial consciousness

From Wikipedia, the free encyclopedia

Artificial consciousness^[1] (AC), also known as machine consciousness (MC) or synthetic consciousness (Gamez 2008; Reggia 2013), is a field related to artificial intelligence and cognitive robotics. The aim of the theory of artificial consciousness is to "Define that which would have to be synthesized were consciousness to be found in an engineered artifact" (Aleksander 1995).
Neuroscience hypothesizes that consciousness is generated by the interoperation of various parts of the brain, called the neural correlates of consciousness or NCC, though there are challenges to that perspective. Proponents of AC believe it is possible to construct systems (e.g., computer systems) that can emulate this NCC interoperation.^[2]

Artificial consciousness concepts are also pondered in the philosophy of artificial intelligence through questions about mind, consciousness, and mental states.^[3]

Philosophical views

As there are many hypothesized types of consciousness, there are many potential implementations of artificial consciousness. In the philosophical literature, perhaps the most common taxonomy of consciousness is into "access" and "phenomenal" variants. Access consciousness concerns those aspects of experience that can be apprehended, while phenomenal consciousness concerns those aspects of experience that seemingly cannot be apprehended, instead being characterized qualitatively in terms of “raw feels”, “what it is like” or qualia (Block 1997).

Plausibility debate

Type-identity theorists and other skeptics hold the view that consciousness can only be realized in particular physical systems because consciousness has properties that necessarily depend on physical constitution (Block 1978; Bickle 2003).^[4][5]

In his article "Artificial Consciousness: Utopia or Real Possibility" Giorgio Buttazzo says that despite our current technology's ability to simulate autonomy, "Working in a fully automated mode, they [the computers] cannot exhibit creativity, emotions, or free will. A computer, like a washing machine, is a slave operated by its components."^[6]

For other theorists (e.g., functionalists), who define mental states in terms of causal roles, any system that can instantiate the same pattern of causal roles, regardless of physical constitution, will instantiate the same mental states, including consciousness (Putnam 1967).

Computational Foundation argument

One of the most explicit arguments for the plausibility of AC comes from David Chalmers. His proposal, found within his article Chalmers 2011, is roughly that the right kinds of computations are sufficient for the possession of a conscious mind. In the outline, he defends his claim thus: Computers perform computations. Computations can capture other systems' abstract causal organization.

The most controversial part of Chalmers' proposal is that mental properties are "organizationally invariant". Mental properties are of two kinds, psychological and phenomenological. Psychological properties, such as belief and perception, are those that are "characterized by their causal role". He adverts to the work of Armstrong 1968 and Lewis 1972 in claiming that "[s]ystems with the same causal topology…will share their psychological properties".

Phenomenological properties are not prima facie definable in terms of their causal roles. Establishing that phenomenological properties are amenable to individuation by causal role therefore requires argument. Chalmers provides his Dancing Qualia Argument for this purpose.^[7]

Chalmers begins by assuming that agents with identical causal organizations could have different experiences. He then asks us to conceive of changing one agent into the other by the replacement of parts (neural parts replaced by silicon, say) while preserving its causal organization. Ex hypothesi, the experience of the agent under transformation would change (as the parts were replaced), but there would be no change in causal topology and therefore no means whereby the agent could "notice" the shift in experience.

Critics of AC object that Chalmers begs the question in assuming that all mental properties and external connections are sufficiently captured by abstract causal organization.

Ethics

If it were suspected that a particular machine was conscious, its rights would be an ethical issue that would need to be assessed (e.g. what rights it would have under law). For example, a conscious computer that was owned and used as a tool or central computer of a building or large machine is a particular ambiguity. Should laws be made for such a case, consciousness would also require a legal definition (for example a machine's ability to experience pleasure or pain, known as sentience). Because artificial consciousness is still largely a theoretical subject, such ethics have not been discussed or developed to a great extent, though it has often been a theme in fiction (see below).

The rules for the 2003 Loebner Prize competition explicitly addressed the question of robot rights:

61. If, in any given year, a publicly available open source Entry entered by the University of Surrey or the Cambridge Center wins the Silver Medal or the Gold Medal, then the Medal and the Cash Award will be awarded to the body responsible for the development of that Entry. If no such body can be identified, or if there is disagreement among two or more claimants, the Medal and the Cash Award will be held in trust until such time as the Entry may legally possess, either in the United States of America or in the venue of the contest, the Cash Award and Gold Medal in its own right.^[8]

Research and implementation proposals

Aspects of consciousness

There are various aspects of consciousness generally deemed necessary for a machine to be artificially conscious. A variety of functions in which consciousness plays a role were suggested by Bernard Baars (Baars 1988) and others. The functions of consciousness suggested by Bernard Baars are Definition and Context Setting, Adaptation and Learning, Editing, Flagging and Debugging, Recruiting and Control, Prioritizing and Access-Control, Decision-making or Executive Function, Analogy-forming Function, Metacognitive and Self-monitoring Function, and Autoprogramming and Self-maintenance Function. Igor Aleksander suggested 12 principles for artificial consciousness (Aleksander 1995) and these are: The Brain is a State Machine, Inner Neuron Partitioning, Conscious and Unconscious States, Perceptual Learning and Memory, Prediction, The Awareness of Self, Representation of Meaning, Learning Utterances, Learning Language, Will, Instinct, and Emotion. The aim of AC is to define whether and how these and other aspects of consciousness can be synthesized in an engineered artifact such as a digital computer. This list is not exhaustive; there are many others not covered.

Awareness

Awareness could be one required aspect, but there are many problems with the exact definition of awareness. The results of the experiments of neuroscanning on monkeys suggest that a process, not only a state or object, activates neurons. Awareness includes creating and testing alternative models of each process based on the information received through the senses or imagined, and is also useful for making predictions. Such modeling needs a lot of flexibility. Creating such a model includes modeling of the physical world, modeling of one's own internal states and processes, and modeling of other conscious entities.

There are at least three types of awareness:^[9] agency awareness, goal awareness, and sensorimotor awareness, which may also be conscious or not. For example, in agency awareness you may be aware that you performed a certain action yesterday, but are not now conscious of it. In goal awareness you may be aware that you must search for a lost object, but are not now conscious of it. In sensorimotor awareness, you may be aware that your hand is resting on an object, but are not now conscious of it.

Because objects of awareness are often conscious, the distinction between awareness and consciousness is frequently blurred or they are used as synonyms.^[10]

Memory

Conscious events interact with memory systems in learning, rehearsal, and retrieval.^[11] The IDA model^[12] elucidates the role of consciousness in the updating of perceptual memory,^[13] transient episodic memory, and procedural memory. Transient episodic and declarative memories have distributed representations in IDA, there is evidence that this is also the case in the nervous system.^[14] In IDA, these two memories are implemented computationally using a modified version of Kanerva’s Sparse distributed memory architecture.^[15]

Learning

Learning is also considered necessary for AC. By Bernard Baars, conscious experience is needed to represent and adapt to novel and significant events (Baars 1988). By Axel Cleeremans and Luis Jiménez, learning is defined as "a set of philogenetically [sic] advanced adaptation processes that critically depend on an evolved sensitivity to subjective experience so as to enable agents to afford flexible control over their actions in complex, unpredictable environments" (Cleeremans 2001).

Anticipation

The ability to predict (or anticipate) foreseeable events is considered important for AC by Igor Aleksander.^[16] The emergentist multiple drafts principle proposed by Daniel Dennett in Consciousness Explained may be useful for prediction: it involves the evaluation and selection of the most appropriate "draft" to fit the current environment. Anticipation includes prediction of consequences of one's own proposed actions and prediction of consequences of probable actions by other entities.

Relationships between real world states are mirrored in the state structure of a conscious organism enabling the organism to predict events.^[16] An artificially conscious machine should be able to anticipate events correctly in order to be ready to respond to them when they occur or to take preemptive action to avert anticipated events. The implication here is that the machine needs flexible, real-time components that build spatial, dynamic, statistical, functional, and cause-effect models of the real world and predicted worlds, making it possible to demonstrate that it possesses artificial consciousness in the present and future and not only in the past. In order to do this, a conscious machine should make coherent predictions and contingency plans, not only in worlds with fixed rules like a chess board, but also for novel environments that may change, to be executed only when appropriate to simulate and control the real world.

Subjective experience

Subjective experiences or qualia are widely considered to be the hard problem of consciousness. Indeed, it is held to pose a challenge to physicalism, let alone computationalism. On the other hand, there are problems in other fields of science which limit that which we can observe, such as the uncertainty principle in physics, which have not made the research in these fields of science impossible.

Role of cognitive architectures

The term "cognitive architecture" may refer to a theory about the structure of the human mind, or any portion or function thereof, including consciousness. In another context, a cognitive architecture implements the theory on computers. An example is QuBIC: Quantum and Bio-inspired Cognitive Architecture for Machine Consciousness. One of the main goals of a cognitive architecture is to summarize the various results of cognitive psychology in a comprehensive computer model. However, the results need to be in a formalized form so they can be the basis of a computer program. Also, the role of cognitive architecture is for the A.I. to clearly structure, build, and implement it's thought process.

Symbolic or hybrid proposals

Franklin's Intelligent Distribution Agent

Stan Franklin (1995, 2003) defines an autonomous agent as possessing functional consciousness when it is capable of several of the functions of consciousness as identified by Bernard Baars' Global Workspace Theory (Baars 1988, 1997). His brain child IDA (Intelligent Distribution Agent) is a software implementation of GWT, which makes it functionally conscious by definition. IDA's task is to negotiate new assignments for sailors in the US Navy after they end a tour of duty, by matching each individual's skills and preferences with the Navy's needs. IDA interacts with Navy databases and communicates with the sailors via natural language e-mail dialog while obeying a large set of Navy policies. The IDA computational model was developed during 1996–2001 at Stan Franklin's "Conscious" Software Research Group at the University of Memphis. It "consists of approximately a quarter-million lines of Java code, and almost completely consumes the resources of a 2001 high-end workstation." It relies heavily on codelets, which are "special purpose, relatively independent, mini-agent[s] typically implemented as a small piece of code running as a separate thread." In IDA's top-down architecture, high-level cognitive functions are explicitly modeled (see Franklin 1995 and Franklin 2003 for details). While IDA is functionally conscious by definition, Franklin does "not attribute phenomenal consciousness to his own 'conscious' software agent, IDA, in spite of her many human-like behaviours. This in spite of watching several US Navy detailers repeatedly nodding their heads saying 'Yes, that's how I do it' while watching IDA's internal and external actions as she performs her task."

Ron Sun's cognitive architecture CLARION

CLARION posits a two-level representation that explains the distinction between conscious and unconscious mental processes.

CLARION has been successful in accounting for a variety of psychological data. A number of well-known skill learning tasks have been simulated using CLARION that span the spectrum ranging from simple reactive skills to complex cognitive skills. The tasks include serial reaction time (SRT) tasks, artificial grammar learning (AGL) tasks, process control (PC) tasks, the categorical inference (CI) task, the alphabetical arithmetic (AA) task, and the Tower of Hanoi (TOH) task (Sun 2002). Among them, SRT, AGL, and PC are typical implicit learning tasks, very much relevant to the issue of consciousness as they operationalized the notion of consciousness in the context of psychological experiments.

Ben Goertzel's OpenCog

Ben Goertzel is pursuing an embodied AGI through the open-source OpenCog project. Current code includes embodied virtual pets capable of learning simple English-language commands, as well as integration with real-world robotics, being done at the Hong Kong Polytechnic University.

Connectionist proposals

Haikonen's cognitive architecture

Pentti Haikonen (2003) considers classical rule-based computing inadequate for achieving AC: "the brain is definitely not a computer. Thinking is not an execution of programmed strings of commands. The brain is not a numerical calculator either. We do not think by numbers." Rather than trying to achieve mind and consciousness by identifying and implementing their underlying computational rules, Haikonen proposes "a special cognitive architecture to reproduce the processes of perception, inner imagery, inner speech, pain, pleasure, emotions and the cognitive functions behind these. This bottom-up architecture would produce higher-level functions by the power of the elementary processing units, the artificial neurons, without algorithms or programs". Haikonen believes that, when implemented with sufficient complexity, this architecture will develop consciousness, which he considers to be "a style and way of operation, characterized by distributed signal representation, perception process, cross-modality reporting and availability for retrospection." Haikonen is not alone in this process view of consciousness, or the view that AC will spontaneously emerge in autonomous agents that have a suitable neuro-inspired architecture of complexity; these are shared by many, e.g. Freeman (1999) and Cotterill (2003). A low-complexity implementation of the architecture proposed by Haikonen (2003) was reportedly not capable of AC, but did exhibit emotions as expected. See Doan (2009) for a comprehensive introduction to Haikonen's cognitive architecture. An updated account of Haikonen's architecture, along with a summary of his philosophical views, is given in Haikonen (2012).

Shanahan's cognitive architecture

Murray Shanahan describes a cognitive architecture that combines Baars's idea of a global workspace with a mechanism for internal simulation ("imagination") (Shanahan 2006). For discussions of Shanahan's architecture, see (Gamez 2008) and (Reggia 2013) and Chapter 20 of (Haikonen 2012).

Takeno's self-awareness research

Self-awareness in robots is being investigated by Junichi Takeno^[17] at Meiji University in Japan. Takeno is asserting that he has developed a robot capable of discriminating between a self-image in a mirror and any other having an identical image to it,^[18][19] and this claim has already been reviewed (Takeno, Inaba & Suzuki 2005). Takeno asserts that he first contrived the computational module called a MoNAD, which has a self-aware function, and he then constructed the artificial consciousness system by formulating the relationships between emotions, feelings and reason by connecting the modules in a hierarchy (Igarashi, Takeno 2007). Takeno completed a mirror image cognition experiment using a robot equipped with the MoNAD system. Takeno proposed the Self-Body Theory stating that "humans feel that their own mirror image is closer to themselves than an actual part of themselves." The most important point in developing artificial consciousness or clarifying human consciousness is the development of a function of self awareness, and he claims that he has demonstrated physical and mathematical evidence for this in his thesis.^[20] He also demonstrated that robots can study episodes in memory where the emotions were stimulated and use this experience to take predictive actions to prevent the recurrence of unpleasant emotions (Torigoe, Takeno 2009).

Aleksander's impossible mind

Igor Aleksander, emeritus professor of Neural Systems Engineering at Imperial College, has extensively researched artificial neural networks and claims in his book Impossible Minds: My Neurons, My Consciousness that the principles for creating a conscious machine already exist but that it would take forty years to train such a machine to understand language.^[21] Whether this is true remains to be demonstrated and the basic principle stated in Impossible Minds—that the brain is a neural state machine—is open to doubt.^[22]

Thaler's Creativity Machine Paradigm

Stephen Thaler proposed a possible connection between consciousness and creativity in his 1994 patent, called "Device for the Autonomous Generation of Useful Information" (DAGUI),^[23][24][25] or the so-called "Creativity Machine", in which computational critics govern the injection of synaptic noise and degradation into neural nets so as to induce false memories or confabulations that may qualify as potential ideas or strategies.^[26] He recruits this neural architecture and methodology to account for the subjective feel of consciousness, claiming that similar noise-driven neural assemblies within the brain invent dubious significance to overall cortical activity.^[27][28][29] Thaler's theory and the resulting patents in machine consciousness were inspired by experiments in which he internally disrupted trained neural nets so as to drive a succession of neural activation patterns that he likened to stream of consciousness.^{[28][30][31][32][33][34]}

Michael Graziano's attention schema

In 2011, Michael Graziano and Sabine Kastler published a paper named "Human consciousness and its relationship to social neuroscience: A novel hypothesis" proposing a theory of consciousness as an attention schema.^[35] Graziano went on to publish an expanded discussion of this theory in his book "Consciousness and the Social Brain".^[2] This Attention Schema Theory of Consciousness, as he named it, proposes that the brain tracks attention to various sensory inputs by way of an attention schema, analogous to the well study body schema that tracks the spatial place of a person's body.^[2] This relates to artificial consciousness by proposing a specific mechanism of information handling, that produces what we allegedly experience and describe as consciousness, and which should be able to be duplicated by a machine using current technology. When the brain finds that person X is aware of thing Y, it is in effect modeling the state in which person X is applying an attentional enhancement to Y. In the attention schema theory, the same process can be applied to oneself. The brain tracks attention to various sensory inputs, and one's own awareness is a schematized model of one's attention. Graziano proposes specific locations in the brain for this process, and suggests that such awareness is a computed feature constructed by an expert system in the brain.

Testing

The most well-known method for testing machine intelligence is the Turing test. But when interpreted as only observational, this test contradicts the philosophy of science principles of theory dependence of observations. It also has been suggested that Alan Turing's recommendation of imitating not a human adult consciousness, but a human child consciousness, should be taken seriously.^[36]

Other tests, such as ConsScale, test the presence of features inspired by biological systems, or measure the cognitive development of artificial systems.

Qualia, or phenomenological consciousness, is an inherently first-person phenomenon. Although various systems may display various signs of behavior correlated with functional consciousness, there is no conceivable way in which third-person tests can have access to first-person phenomenological features. Because of that, and because there is no empirical definition of consciousness,^[37] a test of presence of consciousness in AC may be impossible.

In 2014, Victor Argonov suggested a non-Turing test for machine consciousness based on machine's ability to produce philosophical judgments.^[38] He argues that a deterministic machine must be regarded as conscious if it is able to produce judgments on all problematic properties of consciousness (such as qualia or binding) having no innate (preloaded) philosophical knowledge on these issues, no philosophical discussions while learning, and no informational models of other creatures in its memory (such models may implicitly or explicitly contain knowledge about these creatures’ consciousness). However, this test can be used only to detect, but not refute the existence of consciousness. A positive result proves that machine is conscious but a negative result proves nothing. For example, absence of philosophical judgments may be caused by lack of the machine’s intellect, not by absence of consciousness.

In fiction

Characters with artificial consciousness (or at least with personalities that imply they have consciousness), from works of fiction:

• AC – created by merging two AIs in the Sprawl trilogy by William Gibson
• Agents – in the simulated reality known as "The Matrix" in The Matrix franchise
  • Agent Smith – began as an Agent in The Matrix, then became a renegade program of overgrowing power that could make copies of itself like a self-replicating computer virus
• A.L.I.E. – Sentient genocidal AI from the TV series The 100
• AM (Allied Mastercomputer) – the antagonist of Harlan Ellison's short novel "I Have No Mouth, and I Must Scream"
• Amusement park robots (with pixilated consciousness) that went homicidal in Westworld and Futureworld
• Annalee Call – an Auton (android manufactured by other androids) from the movie Alien Resurrection
• Arnold Rimmer – computer-generated sapient hologram aboard the Red Dwarf
• Ava – a humanoid robot in Ex Machina
• Ash – android crew member of the Nostromo starship in the movie Alien
• The Bicentennial Man – an android in Isaac Asimov's Foundation universe
• Bishop – android crew member aboard the U.S.S. Sulaco in the movie Aliens
• The uploaded mind of Dr. Will Caster, which presumably included his consciousness, from the film Transcendence
• C-3PO – protocol droid featured in all the Star Wars movies
• Chappie – CHAPPiE
• Cohen (and other Emergent AIs) – Chris Moriarty's Spin Series
• Commander Data – Star Trek: The Next Generation
• Cortana (and other "Smart AI") – from the Halo series of games
• Cylons – genocidal robots with resurrection ships that enable the consciousness of any Cylon within an unspecified range to download into a new body aboard the ship upon death, from Battlestar Galactica
• Erasmus – baby killer robot that incited the Butlerian Jihad in the Dune franchise
• Fal'Cie – Mechanical beings with god-like powers from the Final Fantasy XIII series
• The Geth, EDI and SAM – Mass Effect
• HAL 9000 – spaceship USS Discovery One's onboard computer, that lethally malfunctioned due to mutually exclusive directives, from the 1968 novel 2001: A Space Odyssey and in the film
• Holly – ship's computer with an IQ of 6000, aboard the Red Dwarf
• Hosts in the Westworld franchise
• Jane – Orson Scott Card's Speaker for the Dead, Xenocide, Children of the Mind, and "Investment Counselor"
• Johnny Five – Short Circuit
• Joshua – WarGames
• Keymaker – an "exile" sapient program in The Matrix franchise
• "Machine" – android from the film The Machine, whose owners try to kill her when they witness her conscious thoughts, out of fear that she will design better androids (intelligence explosion)
• Mike – The Moon Is a Harsh Mistress
• Mimi – humanoid robot in Real Humans, (original title – Äkta människor) 2012
• The Minds – Iain M. Banks' Culture novels
• Omnius – sentient computer network that controlled the Universe until overthrown by the Butlerian Jihad in the Dune franchise
• Operating Systems in the movie Her
• The Oracle – sapient program in The Matrix franchise
• Professor James Moriarty – sentient holodeck character in the "Ship in a Bottle" episode from Star Trek: The Next Generation
• In Greg Egan's novel Permutation City the protagonist creates digital copies of himself to conduct experiments that are also related to implications of artificial consciousness on identity
• Puppet Master – Ghost in the Shell manga and anime
• R2-D2 – exciteable astromech droid featured in all the Star Wars movies
• Replicants – bio-robotic androids from the novel Do Androids Dream of Electric Sheep? and the movie Blade Runner which portray what might happen when artificially conscious robots are modeled very closely upon humans
• Roboduck – combat robot superhero in the NEW-GEN comic book series from Marvel Comics
• Robots in Isaac Asimov's Robot series
• Robots in The Matrix franchise, especially in The Animatrix
• The Ship – the result of a large-scale AC experiment, in Frank Herbert's Destination: Void and sequels, despite past edicts warning against "Making a Machine in the Image of a Man's Mind"
• Skynet – from the Terminator franchise
• "Synths" are a type of android in the video game Fallout 4. There is a faction in the game known as "The Railroad" which believes that, as conscious beings, synths have their own rights. The Institute, the lab that produces the synths, mostly does not believe they are truly conscious and attributes any apparent desires for freedom as a malfunction.
• TARDIS – time machine and spacecraft of Doctor Who, sometimes portrayed with a mind of its own
• Terminator cyborgs – from the Terminator franchise, with visual consciousness depicted via first-person perspective
• Transformers – sentient robots from the various series in the Transformers robot superhero franchise of the same name
• Vanamonde – an artificial being that was immensely powerful but entirely child-like in Arthur C. Clarke's The City and the Stars
• WALL-E – a robot and the title character in WALL-E
• Gideon – An interactive artificial consciousness made by Barry Allen shown in DC comics and shows like The Flash and Legends of Tomorrow

Multiple drafts model

From Wikipedia, the free encyclopedia

Daniel Dennett's multiple drafts model of consciousness is a physicalist theory of consciousness based upon cognitivism, which views the mind in terms of information processing. The theory is described in depth in his book, Consciousness Explained, published in 1991. As the title states, the book proposes a high-level explanation of consciousness which is consistent with support for the possibility of strong AI.

Dennett describes the theory as first-person operationalism. As he states it:

The Multiple Drafts model makes [the procedure of] "writing it down" in memory criterial for consciousness: that is what it is for the "given" to be "taken" ... There is no reality of conscious experience independent of the effects of various vehicles of content on subsequent action (and hence, of course, on memory).^[1]

The thesis of multiple drafts

Dennett's thesis is that our modern understanding of consciousness is unduly influenced by the ideas of René Descartes. To show why, he starts with a description of the phi illusion. In this experiment, two different coloured lights, with an angular separation of a few degrees at the eye, are flashed in succession. If the interval between the flashes is less than a second or so, the first light that is flashed appears to move across to the position of the second light. Furthermore, the light seems to change colour as it moves across the visual field. A green light will appear to turn red as it seems to move across to the position of a red light. Dennett asks how we could see the light change colour before the second light is observed.

Dennett claims that conventional explanations of the colour change boil down to either Orwellian or Stalinesque hypotheses, which he says are the result of Descartes' continued influence on our vision of the mind. In an Orwellian hypothesis, the subject comes to one conclusion, then goes back and changes that memory in light of subsequent events. This is akin to George Orwell's Nineteen Eighty-Four, where records of the past are routinely altered. In a Stalinesque hypothesis, the two events would be reconciled prior to entering the subject's consciousness, with the final result presented as fully resolved. This is akin to Joseph Stalin's show trials, where the verdict has been decided in advance and the trial is just a rote presentation.

[W]e can suppose, both theorists have exactly the same theory of what happens in your brain; they agree about just where and when in the brain the mistaken content enters the causal pathways; they just disagree about whether that location is to be deemed pre-experiential or post-experiential. [...] [T]hey even agree about how it ought to "feel" to subjects: Subjects should be unable to tell the difference between misbegotten experiences and immediately misremembered experiences. [p. 125, original emphasis.]

Dennett argues that there is no principled basis for picking one of these theories over the other, because they share a common error in supposing that there is a special time and place where unconscious processing becomes consciously experienced, entering into what Dennett calls the "Cartesian theatre". Both theories require us to cleanly divide a sequence of perceptions and reactions into before and after the instant that they reach the seat of consciousness, but he denies that there is any such moment, as it would lead to infinite regress. Instead, he asserts that there is no privileged place in the brain where consciousness happens. Dennett states that, "[t]here does not exist [...] a process such as 'recruitment of consciousness' (into what?), nor any place where the 'vehicle's arrival' is recognized (by whom?)"^[2]

Cartesian materialism is the view that there is a crucial finish line or boundary somewhere in the brain, marking a place where the order of arrival equals the order of "presentation" in experience because what happens there is what you are conscious of. ... Many theorists would insist that they have explicitly rejected such an obviously bad idea. But [...] the persuasive imagery of the Cartesian Theater keeps coming back to haunt us—laypeople and scientists alike—even after its ghostly dualism has been denounced and exorcized. [p. 107, original emphasis.]

With no theatre, there is no screen, hence no reason to re-present data after it has already been analysed. Dennett says that, "the Multiple Drafts model goes on to claim that the brain does not bother 'constructing' any representations that go to the trouble of 'filling in' the blanks. That would be a waste of time and (shall we say?) paint. The judgement is already in so we can get on with other tasks!"

According to the model, there are a variety of sensory inputs from a given event and also a variety of interpretations of these inputs. The sensory inputs arrive in the brain and are interpreted at different times, so a given event can give rise to a succession of discriminations, constituting the equivalent of multiple drafts of a story. As soon as each discrimination is accomplished, it becomes available for eliciting a behaviour; it does not have to wait to be presented at the theatre.

Like a number of other theories, the Multiple Drafts model understands conscious experience as taking time to occur, such that percepts do not instantaneously arise in the mind in their full richness. The distinction is that Dennett's theory denies any clear and unambiguous boundary separating conscious experiences from all other processing. According to Dennett, consciousness is to be found in the actions and flows of information from place to place, rather than some singular view containing our experience. There is no central experiencer who confers a durable stamp of approval on any particular draft.

Different parts of the neural processing assert more or less control at different times. For something to reach consciousness is akin to becoming famous, in that it must leave behind consequences by which it is remembered. To put it another way, consciousness is the property of having enough influence to affect what the mouth will say and the hands will do. Which inputs are "edited" into our drafts is not an exogenous act of supervision, but part of the self-organizing functioning of the network, and at the same level as the circuitry that conveys information bottom-up.

The conscious self is taken to exist as an abstraction visible at the level of the intentional stance, akin to a body of mass having a "centre of gravity". Analogously, Dennett refers to the self as the "centre of narrative gravity", a story we tell ourselves about our experiences. Consciousness exists, but not independently of behaviour and behavioural disposition, which can be studied through heterophenomenology.

The origin of this operationalist approach can be found in Dennett's immediately preceding work. Dennett (1988) explains consciousness in terms of access consciousness alone, denying the independent existence of what Ned Block has labeled phenomenal consciousness.^[3] He argues that "Everything real has properties, and since I don't deny the reality of conscious experience, I grant that conscious experience has properties". Having related all consciousness to properties, he concludes that they cannot be meaningfully distinguished from our judgements about them. He writes:

The infallibilist line on qualia treats them as properties of one's experience one cannot in principle misdiscover, and this is a mysterious doctrine (at least as mysterious as papal infallibility) unless we shift the emphasis a little and treat qualia as logical constructs out of subjects' qualia-judgments: a subject's experience has the quale F if and only if the subject judges his experience to have quale F. We can then treat such judgings as constitutive acts, in effect, bringing the quale into existence by the same sort of license as novelists have to determine the hair color of their characters by fiat. We do not ask how Dostoevski knows that Raskolnikov's hair is light brown.^[4]

In other words, once we've explained a perception fully in terms of how it affects us, there is nothing left to explain. In particular, there is no such thing as a perception which may be considered in and of itself (a quale). Instead, the subject's honest reports of how things seem to them are inherently authoritative on how things seem to them, but not on the matter of how things actually are.

So when we look one last time at our original characterization of qualia, as ineffable, intrinsic, private, directly apprehensible properties of experience, we find that there is nothing to fill the bill. In their place are relatively or practically ineffable public properties we can refer to indirectly via reference to our private property-detectors — private only in the sense of idiosyncratic. And insofar as we wish to cling to our subjective authority about the occurrence within us of states of certain types or with certain properties, we can have some authority — not infallibility or incorrigibility, but something better than sheer guessing — but only if we restrict ourselves to relational, extrinsic properties like the power of certain internal states of ours to provoke acts of apparent re-identification. So contrary to what seems obvious at first blush, there simply are no qualia at all.^[4]

The key to the multiple drafts model is that, after removing qualia, explaining consciousness boils down to explaining the behaviour we recognise as conscious. Consciousness is as consciousness does.

Critical responses to multiple drafts

Some of the criticism of Dennett's theory is due to the perceived tone of his presentation. As one grudging supporter admits, "there is much in this book that is disputable. And Dennett is at times aggravatingly smug and confident about the merits of his arguments [...] All in all Dennett's book is annoying, frustrating, insightful, provocative and above all annoying." (Korb 1993)

Bogen (1992) points out that the brain is bilaterally symmetrical. That being the case, if Cartesian materialism is true, there might be two Cartesian theatres, so arguments against only one are flawed.^[5] Velmans (1992) argues that the phi effect and the cutaneous rabbit illusion demonstrate that there is a delay whilst modelling occurs and that this delay was discovered by Libet.^[6]

It has also been claimed that the argument in the multiple drafts model does not support its conclusion.^[7]

"Straw man"

Much of the criticism asserts that Dennett's theory attacks the wrong target, failing to explain what it claims to. Chalmers (1996) maintains that Dennett has produced no more than a theory of how subjects report events.^[8] Some even parody the title of the book as "Consciousness Explained Away", accusing him of greedy reductionism.^[9] Another line of criticism disputes the accuracy of Dennett's characterisations of existing theories:

The now standard response to Dennett's project is that he has picked a fight with a straw man. Cartesian materialism, it is alleged, is an impossibly naive account of phenomenal consciousness held by no one currently working in cognitive science or the philosophy of mind. Consequently, whatever the effectiveness of Dennett's demolition job, it is fundamentally misdirected (see, e.g., Block, 1993, 1995; Shoemaker, 1993; and Tye, 1993).^[10]

Unoriginality

Multiple drafts is also attacked for making a claim to novelty. It may be the case, however, that such attacks mistake which features Dennett is claiming as novel. Korb states that, "I believe that the central thesis will be relatively uncontentious for most cognitive scientists, but that its use as a cleaning solvent for messy puzzles will be viewed less happily in most quarters." (Korb 1993) In this way, Dennett uses uncontroversial ideas towards more controversial ends, leaving him open to claims of unoriginality when uncontroversial parts are focused upon.

Even the notion of consciousness as drafts is not unique to Dennett. According to Hankins, Dieter Teichert suggests that Paul Ricoeur's theories agree with Dennett's on the notion that "the self is basically a narrative entity, and that any attempt to give it a free-floating independent status is misguided." [Hankins] Others see Derrida's (1982) representationalism as consistent with the notion of a mind that has perceptually changing content without a definitive present instant.^[11]

To those who believe that consciousness entails something more than behaving in all ways conscious, Dennett's view is seen as eliminativist, since it denies the existence of qualia and the possibility of philosophical zombies. However, Dennett is not denying the existence of the mind or of consciousness, only what he considers a naive view of them. The point of contention is whether Dennett's own definitions are indeed more accurate, whether what we think of when we speak of perceptions and consciousness can be understood in terms of nothing more than their effect on behaviour.

Information processing and consciousness

The role of information processing in consciousness has been criticised by John Searle who, in his Chinese room argument,^[12] states that he cannot find anything that could be recognised as conscious experience in a system that relies solely on motions of things from place to place. Dennett sees this argument as misleading, arguing that consciousness is not to be found in a specific part of the system, but in the actions of the whole. In essence, he denies that consciousness requires something in addition to capacity for behaviour, saying that philosophers such as Searle, "just can't imagine how understanding could be a property that emerges from lots of distributed quasi-understanding in a large system" (p. 439).