SELECTED PUBLICATIONS & PRESENTATIONS

| | | | |

| | | | |

Key Papers on GENERAL MACHINE INTELLIGENCE
The Future of AI Research: Ten Defeasible 'Axioms of Intelligence' Proc. Machine Learning Research, 2022 192:5-21 With Henry Minsky PDF	Abstract. What sets artificial intelligence (AI) apart from other fields of science and technology is not what it has achieved so far, but rather what it set out to do from the very beginning, namely, to create autonomous self-contained systems that can rival human cognition - machines with human-level general intelligence. To achieve this aim calls for a new kind of system that, among other things, unifies - in a single architecture - the ability to represent causal relations, create and manage knowledge incrementally and autonomously, and generate its own meaning through empirical reasoning and control. We maintain that building such systems requires a shared methodological foundation, and calls for a stronger theoretical basis than simply the one inherited directly from computer science. This, in turn, calls for a greater emphasis on the theory of intelligence and methodological approaches fo building such systems. We argue that necessary (but not necessarily sufficient) components for general intelligence must include the unification of causal relations, reasoning, and cognitive development. A constructivist stance, in our view, can serve as a good starting point for this purpose.
A Theory of Foundational Meaning Generation in Autonomous Systems, Natural and Artificial Proc. Intl. Conf. Artificial General Intelligence, 2024, 32-70 PDF Video Lecture @ AGI Day 4 August 16, 2024	Abstract.The concept of 'meaning' has long been a subject of philosophy and people use the term regularly. Theories of meaning detailed enough to serve as blueprints in the design of intelligent artificial systems have however been few. Here we present a theory of foundational meaning creation - the phenomenon proper - sufficiently broad to apply to natural agents yet concrete enough to be implemented in a running artificial system. The theory states that meaning generation is a process bound in the present now, resting on the concept of reliable causal models. By unifying goals, predictions, plans, situations and knowledge, it explains how ampliative reasoning and explicit representations of causal relations participate in the meaning generation process. According to the theory, meaning and autonomy are two sides of the same coin: Meaning generation without autonomy is meaningless; autonomy without meaning is impossible.
Seed-Programmed Autonomous General Learning Proc. Machine Learning Research, 2020 Vol. 131, 32-70 PDF Video Lecture NARS Workshop @ AGI June 23, 2020	Abstract. The knowledge that a natural learner creates based on its experience of any new situation is likely to be both partial and incorrect. To improve such knowledge with increased experience, cognitive processes must bring already-acquired knowledge towards making sense of new situations and update it with new evidence, cumulatively. For the initial creation of knowledge, and its subsequent usage, expansion, modification, unification, compaction and deletion, cognitive mechanisms must be capable of self-supervised surgical operation on existing knowledge, involving among other things self-inspection or reflection, to make possible selective discrimination, comparison, and manipulation of newly demarcated subsets of any relevant part of the whole knowledge set. Few proposals exist for how to achieve this in a single learner. Here we present a theory of how systems with these properties may work, and how cumulative self-supervised learning mechanisms might reach greater levels of autonomy than seen to date. Our theory rests on the hypotheses that learning must be (a) organized around causal relations, (b) bootstrapped from observed correlations and analogy, using (c) fine-grain relational models, manipulated by (d) micro-ampliative reasoning processes. We further hypothesize that a machine properly constructed in this way will be capable of seed-programmed autonomous generality: The ability to apply learning to any phenomenon - that is, being domain-independent - provided that the seed reference observable variables from the outset (at birth), and that new phenomena and existing knowledge overlap on one or more observables or inferred features. The theory is based on implemented systems that have produced notable results in the direction of increased general machine intelligence.
The 'Explanation Hypothesis' in General Self-Supervised Learning Proc. Machine Learning Research, 2021 Vol. 159, 5-27 PDF Video Lecture NARS Workshop @ AGI Oct. 14, 2021	Abstract. Self-supervised learning is the ability of an agent to improve its own performance, with respect to one or more goals related to one or more phenomena, without outside help from a teacher or other external aid tailored to the agent's learning progress. A general learner's learning process is not limited to a strict set of topics, tasks, or domains. Self-supervised and general learning machines are still in the early stages of development, as are learning machines that can explain their own knowledge, goals, actions, and reasoning. Research on explanation proper has to date been largely limited to the field of philosophy of science. In this paper I present the hypothesis that general self-supervised learning requires (a particular kind of) explanation generation, and review some key arguments for and against it. Named the explanation hypothesis (ExH), the claim rests on three main pillars. First, that any good explanation of a phenomenon requires reference to relations between sub-parts of that phenomenon, as well as to its context (other phenomena and their parts), especially (but not only) causal relations. Second, that self-supervised general learning of a new phenomenon requires (a kind of) bootstrapping, and that this - and subsequent improvement on the initial knowledge thus produced  relies on reasoning processes. Third, that general self-supervised learning relies on reification of prior knowledge and knowledge-generation processes, which can only be implemented through appropriate reflection mechanisms, whereby current knowledge and prior learning progress is available for explicit inspection by the learning system itself, to be analyzed for use in future learning. The claim thus construed has several important implications for the implementation of general machine intelligence, including that it will neither be achieved without reflection (meta-cognition) nor explicit representation of causal relations, and that internal explanation generation must be a fundamental principle of their operation.
Cumulative Learning Proc. 12th International Conference on Artificial General Intelligence 2019, 198-209 PDF	Abstract. An important feature of human learning is the ability to continuously accept new information and unify it with existing knowledge, a process that proceeds largely automatically and without catastrophic side-effects. A generally intelligent machine (AGI) should be able to learn a wide range of tasks in a variety of environments. Knowledge acquisition in partially-known and dynamic task-environments cannot happen all-at-once, and AGI-aspiring systems must thus be capable of cumulative learning: efficiently making use of existing knowledge while learning new things, increasing the scope of ability and knowledge incrementally — without catastrophic forgetting or damaging existing skills. Many aspects of such learning have been addressed in artificial intelligence (AI) research, but relatively few examples of cumulative learning have been demonstrated to date and no generally accepted explicit definition exists of this category of learning. Here we provide a general definition of cumulative learning and describe how it relates to other concepts frequently used in the AI literature.
Outstanding Paper Award Autonomous Acquisition of Natural Language Proc. 7th International Conference on Intelligent Systems & Agents 2014, 58-66 PDF	Abstract. Humans learn how to use language in a society of language users. No principles that might allow an artificial agents to learn language this way are known at present. We present work that addresses this challenge: Our auto-catalytic, endogenous, reflective architecture (AERA) supports the creation of agents that can learn natural language by observation. Our S1 agent learns human communication by observing two humans interacting in a realtime mock television interview, using gesture and situated language. Results show that S1 can learn multimodal complex language and multimodal communicative acts, using a vocabulary of 100 words with numerous free-form sentence formats, by observing unscripted interaction between the humans, with no grammar being provided to it a priori. The agent learns both the pragmatics, semantics, and syntax of complex sentences spoken by the human subjects on the topic of recycling of objects such as aluminum cans, glass bottles, plastic, and wood, as well as use of manual deictic reference and anaphora, all through observation.
Bounded Seed-AGI Proc. 7th International Conference on Artificial General Intelligence 2014, 85-96 PDF	Abstract. Four principal features of autonomous control systems are left both unaddressed and unaddressable by present-day engineering methodologies: (1) The ability to operate effectively in environments that are only partially known beforehand at design time; (2) A level of generality that allows a system to re-assess and re-define the fulfillment of its mission in light of unexpected constraints or other unforeseen changes in the environment; (3) The ability to operate effectively in environments of significant complexity; and (4) The ability to degrade gracefully — how it can continue striving to achieve its main goals when resources become scarce, or in light of other expected or unexpected constraining factors that impede its progress. We describe new methodological and engineering principles for addressing these shortcomings, that we have used to design a machine that becomes increasingly better at behaving in under- specified circumstances, in a goal-directed way, on the job, by modeling itself and its environment as experience accumulates. The work provides an architectural blueprint for constructing systems with high levels of operational autonomy in underspecified circumstances, starting from only a small amount of designer-specified code — a seed. Using value-driven dynamic priority scheduling to control the parallel execution of a vast number of lines of reasoning, the system accumulates increasingly useful models of its experience, resulting in recursive self-improvement that can be autonomously sustained after the machine leaves the lab, within the boundaries imposed by its designers. A prototype system named AERA has been implemented and demonstrated to learn a complex real-world task — real-time multimodal dialogue with humans — by on-line observation. Our work presents solutions to several challenges that must be solved for achieving artificial general intelligence.
Kurzweil Prize Resource-Bounded Machines are Motivated to be Effective, Efficient & Curious Proc. 6th International Conference on Artificial General Intelligence 2013, 119-129 PDF	Abstract. Resource-boundedness must be carefully considered when designing and implementing artificial general intelligence (AGI) algorithms and architectures that have to deal with the real world. But not only must resources be represented explicitly throughout its design, an agent must also take into account their usage and the associated costs during reasoning and acting. Moreover, the agent must be intrinsically motivated to become progressively better at utilizing resources. This drive then naturally leads to effectiveness, efficiency, and curiosity. We propose a practical operational framework that explicitly takes into account resource constraints: activities are organized to maximally utilize an agent’s bounded resources as well as the availability of a teacher, and to drive the agent to become progressively better at utilizing its resources. We show how an existing AGI architecture called AERA can function inside this framework. In short, the capability of AERA to perform self-compilation can be used to motivate the system to not only accumulate knowledge and skills faster, but also to achieve goals using less resources, becoming progressively more effective and efficient.
Kurzweil Prize On Attention Mechanisms for AGI Architectures: A Design Proposal Proc. 4th International Conference on Artificial General Intelligence 2012, 89-98 PDF	Abstract. Many existing artificial general intelligence (AGI) architectures are based on the assumption of infinite computational resources, as researchers ignore the fact that real-world tasks have time limits, and managing these is a key part of the role of intelligence. In the domain of intelligent systems the management of system resources is typically called “attention”. Attention mechanisms are necessary because all moderately complex environments are likely to be the source of vastly more information than could be processed in realtime by an intelligence’s available cognitive resources. Even if sufficient resources were available, attention could help make better use of them. We argue that attentional mechanisms are not only nice to have, for AGI architectures they are an absolute necessity. We examine ideas and concepts from cognitive psychology for creating intelligent resource management mechanisms and how these can be applied to engineered systems. We present a design for a general attention mechanism intended for implementation in AGI architectures.
A New Constructivist AI: From Manual Construction to Self-Constructive Systems P. Wang & B. Goertzel (eds.), Theoretical Foundations of Artificial General Intelligence, 4:145-171, 2012 PDF	Abstract. The development of artificial intelligence (AI) systems has to date been largely one of manual labor. This constructionist approach to AI has resulted in systems with limited-domain application and severe performance brittleness. No AI architecture to date incorporates, in a single system, the many features that make natural intelligence general-purpose, including system-wide attention, analogy-making, system-wide learning, and various other complex transversal functions. Going beyond current AI systems will require significantly more complex system architecture than attempted to date. The heavy reliance on direct human specification and intervention in constructionist AI brings severe theoretical and practical limitations to any system built that way. One way to address the challenge of artificial general intelligence (AGI) is replacing a top-down architectural design approach with methods that allow the system to manage its own growth. This calls for a fundamental shift from hand-crafting to self-organizing archi- tectures and self-generated code – what we call a constructivist AI approach (CAIM), in reference to the self-constructive principles on which it must be based. Methodologies employed for constructivist AI will be very different from today’s software development methods; instead of relying on direct design of mental functions and their implementation in a cog- nitive architecture, they must address the principles – the “seeds” – from which a cognitive architecture can automatically grow. In this paper I describe the argument in detail and examine some of the implications of this impending paradigm shift.
Anytime Bounded Rationality Proc. 8th International Conference on Artificial General Intelligence 2015, 121-130 PDF	Abstract. Dependable cyber-physical systems strive to deliver anticipative, multi-objective performance anytime, facing deluges of inputs with varying and limited resources. This is even more challenging for life-long learning rational agents as they also have to contend with the varying and growing know-how accumulated from experience. These issues are of crucial practical value, yet have been only marginally and unsatisfactorily addressed in AGI research. We present a value-driven computational model of anytime bounded rationality robust to variations of both resources and knowledge. It leverages continually learned knowledge to anticipate, revise and maintain concurrent courses of action spanning over arbitrary time scales for execution anytime necessary.
Towards a Programming Paradigm for Control Systems With High Levels of Existential Autonomy Proc. 8th International Conference on Artificial General Intelligence 2015, 78-87 PDF	Abstract. Systems intended to operate in dynamic, complex environments – without intervention from their designers or significant amounts of domain- dependent information provided at design time – must be equipped with a sufficient level of existential autonomy. This feature of naturally intelligent systems has largely been missing from cognitive architectures created to date, due in part to the fact that high levels of existential autonomy require systems to program themselves; good principles for self-programming have remained elusive. Achieving this with the major programming methodologies in use today is not likely, as these are without exception designed to be used by the human mind: Producing self-programming systems that can grow from first principles using these therefore requires first solving the AI problem itself – the very problem we are trying to solve. Advances in existential autonomy call for a new programming paradigm, with self-programming squarely at its center. The principles of such a paradigm are likely to be fundamentally different from prevailing approaches; among the desired features for a programming language designed for automatic self-programming are (a) support for autonomous knowledge acquisition, (b) real-time and any-time operation, (c) reflectivity, and (d) massive parallelization. With these and other requirements guiding our work, we have created a programming paradigm and language called Replicode. Here we discuss the reasoning behind our approach and the main motivations and features that set this work from apart from prior approaches.

SELECTED VIDEOS
A Theory of Foundational Meaning Generation in Autonomous Systems — Natural & Artificial AGI Conference, Seattle, August 16, 2024 Related Paper: A Theory of Foundational Meaning Generation in Autonomous Systems -- Natural & Artificial ,
Ísland, Fjórða iðnbyltingin og nýsköpunarhringrásin Grand Hotel Reykjavik, March 1, 2019 Related Paper: Iceland & the 4th Industrial Revolution Office of the Prime Minister of Iceland ,
The 'Explanation Hypothesis' in Autonomous General Learning NARS Workshop @ AGI-21, Oct. 14, 2021 Related Paper: The 'Explanation Hypothesis' in General Self-Supervised Learning Proc. Machine Learning Research, 159:5-27
Seed-Programmed Autonomous General Learning NARS Workshop @ AGI-20, June 23, 2020 Related Paper: Seed-Programmed Autonomous General Learning Proc. Machine Learning Research, 159:32-70
The Road To 'Artificial' Understanding - Interview with Kristinn R. Thórisson by Adam Ford Online interview @ Dec. 11, 2019 Related Paper: About Understanding 9th International Conference on Artificial General Intelligence (AGI-19), 106-117
Why An AI Lab Needs an Ethics Policy Public presentation, AI Festival, Reykjavik University, October 23, 2015 Related: Ethics Policy of the Icelandic Institute for Intelligent Machines
Towards True AI: Artificial General Intelligence Public presentation, CAIDA/IIIM AI Festival, Reykjavik University, October 31, 2014 Related Papers: Anytime Bounded Rationality Autonomous Acquisition of Natural Situated Communication Bounded Seed-AGI A New Constructivist AI: From Manual Construction to Self-Constructive Systems
Why Progress in AI is Not What We Had Hoped For (and Why This May Change) Public presentation, Reykjavik University, March 20, 2012 Related Papers: Reductio ad Absurdum: On Oversimplification in Computer Science & its Pernicious Effect on Artificial Intelligence Research What Should AGI Learn from AI and CogSci?
Introduction to the Icelandic Institute for Intelligent Machines Public presentation, Reykjavik University, May 28, 2010
Multiparty Turntaking – Using a Cognitive Model of Multimodal Dialogue Skills (2010) Related Paper: A Multiparty Multimodal Architecture for Realtime Turntaking Proceedings of Intelligent Virtual Agents 2010
MIRAGE – An Embodied Agent in Augmented Reality (2002) Related Paper: Constructionist Design Methodology for Interactive Intelligences A.I. Magazine, 25(4):77-90. Menlo Park, CA: American Association for Artificial Intelligence
Gandalf – The Conversive Humanoid Agent (1996) Related Paper: A Mind Model for Multimodal Communicative Creatures and Humanoids International Journal of Applied Artificial Intelligence, 13(4-5): 449-486
ICONIC – Iconic and pantomimic gesture at the interface (1993) Related Paper: Integrating Simultaneous Input from Speech, Gaze and Hand Gestures In M. T. Maybury (Ed.), Intelligent Multimedia Interfaces, Ch. 11, 257-276. Cambridge, Massachusetts, U.S.A.: AAAI Press / M.I.T. Press

Selected JOURNAL ARTICLES, BOOK CHAPTERS, CONFERENCE PAPERS
A Theory of Foundational Meaning Generation in Autonomous Systems — Natural and Artificial K. R. Thórisson & G. Talevi (2024) Proc. Artificial General Intelligence Conference, 188-198			PDF
Argument-Driven Planning & Autonomous Explanation Generation L. Eberding, J. Thompson & K. R. Thórisson (2024) Proc. Artificial General Intelligence, 73-83	AGI Society Award		PDF
Causal Generalization via Goal-Driven Analogy A. Sheikhlar & K. R. Thórisson (2024) Proc. Artificial General Intelligence, 165-175	Best AGI Paper Prize		PDF
High-Level Conceptual Design Automation Requires Ampliative Reasoning K. R. Thórisson & C. Shaff (2024) Proc. NordDesign			PDF
Explicit Goal-Driven Autonomous Self-Explanation Generation K. R. Thórisson, H. Rörbeck, J. Thompson & H. Latapie (2023) Proc. Artificial General Intelligence Conference (AGI-23), 286-295			PDF
Addressing the Unsustainability of Deep Neural Networks with Next-Gen AI A. Vallentin, K. R. Thórisson & H. Latapie (2023) Proc. Artificial General Intelligence Conference (AGI-23), 296-306			PDF
Causal Reasoning over Probabilistic Uncertainty L. M. Eberding & K. R. Thórisson (2023) Proc. Artificial General Intelligence Conference, 74-84			PDF
The Future of AI Research: Ten Defeasible 'Axioms of Intelligence' K. R. Thórisson & H. Minsky (2022) Proc. Machine Learning Research, 192:5-21			PDF
Explicit General Analogy for Autonomous Transversal Learning A. Sheikhlar, K. R. Thórisson, J. Thompson (2022) Proc. Machine Learning Research, 192:48-62			PDF
Artificial intelligence and auditing in small- and medium-sized firms: Expectations and applications P. Rikhardsson, K. R. Thórisson, G. Bergthorsson & C. Batt (2022) A.I. Magazine, 43(3):323-336			PDF
The 'Explanation Hypothesis' in General Self-Supervised Learning K. R. Thórisson (2021) Proc. Machine Learning Research, 159:5-27			PDF
Comparison of Machine Learners on an ABA Experiment Format of the Cart-Pole Task L. M. Eberding, K. R. Thórisson, A. Prabu, S. Jaroria, A. Sheikhlar (2021) Proc. Machine Learning Research, 159:49-63			PDF
About the Intricacy of Tasks L. M. Eberding, M. Belenchia, A. Sheikhlar and K. R. Thórisson (2021) Proc. Artificial General Intelligence, 65-74			PDF
Causal Generalization in Autonomous Learning Controllers A. Sheikhlar, L. M. Eberding and K. R. Thórisson (2021) Proc. Artificial General Intelligence (AGI-21), 228-238			PDF
Elements of Task Theory M. Belenchia, K. R. Thórisson, L. M. Eberding and A. Sheikhlar (2021) Proc. Artificial General Intelligence (AGI-21), 19-29			PDF
Seed-Programmed Autonomous General Learning K. R. Thórisson (2020) Proc. Machine Learning Research, 131:32-70			PDF
Introduction to the JAGI Special Issue "On Defining Artificial Intelligence": Commentaries and Author's Response D. Monett, C. W. P. Lewis & K. R. Thórisson (2020) Editorial, Special Issue: On Defining Artificial Intelligence, Journal of Artificial General Intelligence, 11(2):1-4			PDF
Autonomous Cumulative Transfer Learning A. Sheikhlar, K. R. Thórisson & L. Eberding (2020) Proc. Artificial General Intelligence (AGI-20), 306-316			PDF
SAGE: Task-Environment Platform & For Evaluating a Broad Range of Learners L. M. Eberding, K. R. Thórisson, A. Sheikhlar & S. P. Andrason (2020) Proc. Artificial General Intelligence (AGI-20), 72-82			PDF
Error-Correction for AI Safety N.-M. Aliman, P. Elands, W. Hürst, L. Kester, K. R. Thórisson, P. Werkhoven, R. Yampolskiy & S. Ziesche (2020) Proc. Artificial General Intelligence (AGI-20), 12-22			PDF
Cumulative Learning K. R. Thórisson, J. Bieger, X. Li & P. Wang (2019) Proc. 12th International Conference on Artificial General Intelligence (AGI-19), Shenzhen, China, Aug. 6-9, 198-209			PDF
Cumulative Learning With Causal-Relational Models K. R. Thórisson & A. Talbot (2018) Proc. 11th International Conference on Artificial General Intelligence (AGI-18), Prague, Czech Republic, Aug. 22-15, 227-238			PDF
Abduction, Deduction & Causal-Relational Models K. R. Thórisson & A. Talbot (2018) IJCAI-18 Workshop on Artchitectures for Generality, Autonomy & Progress in AI, International Joint Conference on Artificial Intelligence, Stockholm, Sweden, Jul. 15			PDF
Task Analysis for Teaching Cumulative Learners J. Bieger & K. R. Thórisson (2018) Proc. 11th International Conference on Artificial General Intelligence (AGI-18), Prague, Czech Republic, Aug. 22-15, 21-32			PDF
Requirements for General Intelligence: A Case Study in Trustworthy Cumulative Learning for Air Traffic Control J. Bieger & K. R. Thórisson (2018) IJCAI-18 Workshop on Artchitectures for Generality, Autonomy & Progress in AI, International Joint Conference on Artificial Intelligence, Stockholm, Sweden, Jul 15			PDF
Understanding & Common Sense: Two Sides of the Same Coin? K. R. Thórisson & D. Kremelberg (2017) Proc. Artificial General Intelligence (AGI-17), August 15-18, Melbourne, Australia, 201-211			PDF
Do Machines Understand? K. R. Thórisson & D. Kremelberg (2017) Understanding Understanding Workshop, 10th International Conference on Artificial General Intelligence (AGI-17), August 18, Melbourne Australia			PDF
Evaluating Understanding J. Bieger, K. R. Thórisson (2017) IJCAI-17 Workshop on Evaluating General-Purpose Intelligence, International Joint Conference on Artificial Intelligence, August 20, Melbourne, Australia			PDF
The Pedagogical Pentagon: A Conceptual Framework for Artificial Pedagogy J. Bieger, K. R. Thórisson & B. R. Steunebrink (2017) Proc. 10th International Conference on Artificial General Intelligence (AGI-17), 212-222			PDF
Machines With Autonomy & General Intelligence: Which Methodology? K. R. Thórisson (2017) IJCAI-17 Workshop on Architectures for Generality & Autonomy, International Joint Conference on Artificial Intelligence, August 19, Melbourne, Australia			PDF
About Understanding K. R. Thórisson, D. Kremelberg, B. R. Steunebrink, E. Nivel (2016) B. Steunebrink et al. (eds.), Proc. Artificial General Intelligence (AGI-16), July, New York City, 106-117			PDF
Evaluation of General-Purpose Artificial Intelligence: Why, What & How J. Bieger, K. R. Thórisson, B. R. Steunebrink, T. Thorarensen, J. S. Sigurdardóttir (2016) EGPAI 2016 - Evaluating General-Purpose A.I., Workshop @ the European Conference on Artificial Intelligence, The Hague, The Netherlands, Tuesday 30th Aug. 2016.			PDF
FraMoTEC: Modular Task-Environment Construction for Evaluating Adaptive Control Systems T. Thorarensen, K. R. Thórisson, J. Bieger, J. S. Sigurdardóttir (2016) EGPAI 2016 - Evaluating General-Purpose A.I., Workshop held in conjuction with the European Conference on Artificial Intelligence, The Hague, The Netherlands, Tuesday 30th Aug. 2016.			PDF
Growing Recursive Self-Improvers B. R. Steunebrink, K. R. Thórisson, J. Schmidhuber (2016) B. Steunebrink et al. (eds.), Proc. 9th International Conference on Artificial General Intelligence (AGI-16), July, New York City, 129-139			PDF
Why Artificial Intelligence Needs a Task Theory – And What It Might Look Like K. R. Thórisson, J. Bieger, T. Thorarensen, J. S. Sigurdardottir & B. R. Steunebrink (2016) B. Steunebrink et al. (eds.), Proc. 9th International Conference on Artificial General Intelligence (AGI-16), July, New York City, 118-128			PDF
Towards Flexible Task Environments for Comprehensive Evaluation of Artificial Intelligent Systems & Automatic Learners K. R. Thórisson, J. Bieger, S. Schiffel & D. Garrett (2015) J. Bieger, B. Goertzel & A. Potapov (eds.), Proc. 8th International Conference on Artificial General Intelligence (AGI-15), July, Berlin, Germany, 187-196			PDF
Safe Baby AGI J. Bieger, K. R. Thórisson & P. Wang (2015) J. Bieger, B. Goertzel & A. Potapov (eds.), Proc. 8th International Conference on Artificial General Intelligence (AGI-15), July, Berlin, Germany, 46-49			PDF
Anytime Bounded Rationality E. Nivel, K. R. Thórisson, B. Steunebrink & J. Schmidhüber (2015) J. Bieger, B. Goertzel & A. Potapov (eds.), Proc. 8th International Conference on Artificial General Intelligence (AGI-15), July, Berlin, Germany, 121-130			PDF
On Applicability of Automated Planning for Incident Management L. Chrpa & K. R. Thórisson (2015) The International Scheduling and Planning Applications Workshop (SPARK-2015), June, Jerusalem, Israel.			PDF
Autonomous Acquisition of Situated Natural Communication K. R. Thórisson, E. Nivel, B. R. Steunebrink., H. P. Helgason, G. Peluzzo, R. Sanz, J. Schmidhuber, H. Dindo, M. Rodriguez, A. Chella, G. Jonsson, D. Ognibene & C. Hernandez (2014) International Journal of Computer Science & Information Systems, 9(2):115-131			PDF
Autonomous Acquisition of Natural Language E. Nivel, K. R. Thórisson, B. R. Steunebrink.,H. Dindo, G. Pezzulo, M. Rodriguez, C. Hernandez, D. Ognibene, J. Schmidhuber, R. Sanz, H. P. Helgason, A. Chella & G. Jonsson (2014) A. P. dos Reis, P. Kommers & P. Isaías (eds.), Proc. IADIS International Conference on Intelligent Systems & Agents 2014 (ISA-14), Jul., Lisbon, Portugal, 58-66	Outstanding Paper Award		PDF
Towards a General Attention Mechanism for Embedded Intelligence Systems Helgason, H. P., K. R. Thórisson, D. Garrett & E. Nivel (2014) International Journal of Computer Science and Artificial Intelligence, 4(1): 1-7			PDF
Tunable & Generic Problem Instance Generation for Multi-objective Reinforcement Learning D. Garrett, J. Bieger & K. R. Thórisson (2014) IEEE Symposium on Adaptive Dynamic Programming and Reinforcement Learning, Orlando, Florida, 1-8			PDF
Bounded Seed-AGI Nivel, E., K. R. Thórisson, B. R. Steunebrink.,H. Dindo, G. Pezzulo, M. Rodriguez, C. Hernandez, D. Ognibene, J. Schmidhuber, R. Sanz, H. P. Helgason & A. Chella (2014) B. Goerzel, L. Orseau & J. Snaider (eds.), Proc. Artificial General Intelligence (AGI-14), Aug., Quebec, Canada, 85-96			PDF
Raising AGI: Tutoring Matters Bieger, J., K. R. Thórisson & D. Garrett (2014) B. Goerzel, L. Orseau & J. Snaider (eds), Proc. Artificial General Intelligence (AGI-14), Aug., Quebec, Canada, 1-10			PDF
What Should AGI Learn from AI and CogSci? P. Wang, B. Steunebrink & K. R. Thórisson B. Goerzel, L. Orseau & J. Snaider (eds), Proc. Artificial General Intelligence (AGI-14), Aug., Quebec, Canada			PDF
Resource-Bounded Machines are Motivated to be Efficient, Effective, & Curious Steunebrink, B. R., J. Koutnik, K. R. Thórisson, E. Nivel & J. Schmidhuber (2013) K-U Kühnberger, S. Rudolph & P. Wang (ed.), Proc. Artificial General Intelligence (AGI-13), Jul., Beijing, China, 119-129	Kurzweil Award		PDF
Predictive Generative Heuristics for Decision-Making in Real-World Environments Helgason, H. P., K. R. Thórisson, E. Nivel & P. Wang (2013) K-U Kühnberger, S. Rudolph & P. Wang (eds), Proc. Artificial General Intelligence (AGI-13), Jul., Beijing, China, 50-59			PDF
Towards a Programming Paradigm for Control Systems With High Levels of Existential Autonomy Nivel, E. & K. R. Thórisson (2013) K-U Kühnberger, S. Rudolph & P. Wang (eds), Proc. Artificial General Intelligence (AGI-13), Jul., Beijing, China, 78-87			PDF
Reductio ad Absurdum: On Oversimplification in Computer Science & its Pernicious Effect on Artificial Intelligence Research Thórisson, K. R. (2013) A. H. M. Abdel-Fattah &K.-U. Kühnberger (eds), Proceedings of the Workshop Formalizing Mechanisms for Artificial General Intelligence and Cognition (Formal MAGiC), Jul., Beijing, China, July 31st, 31-35. Institute of Cognitive Science, Osnabrück			PDF
A Distributed Architecture for Real-Time Dialogue & On-Task Learning of Efficient Cooperative Turn-Taking Jonsdottir, G.R., & K. R. Thórisson (2013) M. Rojc and N. Campbell (eds), Speech, Gaze and Affect, Ch. 12, 293-324. Boca Raton, Florida, US: Taylor & Francis			PDF
Approaches & Assumptions of Self-Programming in Achieving Artificial General Intelligence Thórisson, K. R., E. Nivel, R. Sanz & P. Wang (2012) Editorial, Special Issue of Journal of Artificial General Intelligence on Self-Programming, 3(3):1-10			PDF
On Attention Mechanisms for AGI Architectures: A Design Proposal Helgason, H. P., E. Nivel & K. R. Thórisson (2012) J. Bach, B. Goertzel & M. Ilké (eds.), Proceedings of Artificial General Intelligence (AGI-12), Dec. 8-11, Oxford U., 89-98	Kurzweil Award		PDF
A New Constructivist AI: From Manual Construction to Self-Constructive Systems Thórisson, K. R. (2012) P. Wang and B. Goertzel (eds), Theoretical Foundations of Artificial General Intelligence. Atlantis Thinking Machines, 4:145-171			PDF
Attention Capabilities for AI Systems Helgason, H. P. & K. R. Thórisson (2012) Proc. 9th International Conference on Informatics in Control, Automation & Robotics, Rome, Italy, July			PDF
Cognitive Architectures & Autonomy: A Comparative Review Thórisson, K. R. & H. P. Helgason (2012) Journal of Artificial General Intelligence, 3(2):1-30	Commentary & Author Response PDF		PDF
Learning Problem Solving Skills from Demonstration: An Architectural Approach H. Dindo, A. Chella, G. La Tona, M. Vitali, E. Nivel & K. R. Thórisson (2011) Proc. Artificial General Intelligence (AGI-11)			PDF
A Multiparty Multimodal Architecture for Realtime Turntaking K. R. Thórisson, O. Gislason, G. R. Jonsdottir & H. Th. Thorisson (2010) Proc. Intelligent Virtual Agents (IVA '10)			PDF
Evaluating Multimodal Human-Robot Interaction: A Case Study of an Early Humanoid Prototype Jonsson, G. K. & K. R. Thórisson (2010) A.J. Spinks, F. Grieco, O.E. Krips, I.W.S. Loijens, I.P.J.J. Noldus and P.H. Zimmerman (eds), Measuring Behavior 2010: Proceedings of the 7th International Conference on Methods and Techniques in Behavioral Research, 273-276. ACM New York, NY, USA			PDF
Laughter Detection in Noisy Settings Felkin, M., J. Terrien & K. R. Thórisson (2010) J. Griffith, C. Hayes, M. Madden, D. O'Hora & C. O'Riordan (eds), AICS - Proc. 21st National Conference on Artificial Intelligence & Cognitive Science, 102-111			PDF
The Semantic Web: From Representation to Realization Thórisson, K. R., N. Spivack, J.M. Wissner (2010) Transactions on Computational Collective Intelligence II, Lecture Notes in Computer Science, 6450:90-107, Ngoc Thanh Nguyen and Ryszard Kowalczyk (eds.), DOI: 10.1007/978-3-642-17155-0			PDF
From Constructionist to Constructivist A.I. Thórisson, K. R. (2009) AAAI Fall Symposium Series: Biologically Inspired Cognitive Architectures, Washington D.C., Nov. 5-7, 175-183. AAAI Tech Report FS-09-01, AAAI press, Menlo Park, CA	Keynote		PDF
Applying Constructionist Design Methodology to Agent-Based Simulation Systems Thórisson, K. R., R. J. Saemundsson, G. R. Jonsdottir, B. Reynisson, C. Pedica, P. R. Thrainsson and P. Skowronski (2009) 3rd International KES Symposium on Agents and Multi-agent Systems – Technologies and Applications			PDF
Teaching Computers to Conduct Spoken Interviews: Breaking the Realtime Barrier With Learning Jonsdottir, G. R. & K. R. Thórisson (2009) Proc. Intelligent Virtual Agents (IVA '09), Springer Lecture Notes in Artificial Intelligence 5773, 446-459	Best Conference Paper Nomination		PDF
SemCards: A New Representation for Realizing the Semantic Web Thórisson, K. R., N. Spivack & J. M. Wissner (2009) N.T. Nguyen, R. Kowalczyk and S.M. Chen (Eds.), Proc. First International Conference on Computational Collective Intelligence, Wroclaw, Poland, Oct. 5-7, Springer Lecture Notes in Artificial Intelligence 5796, 425-436			PDF
A YARP-based Architectural Framework for Robotic Vision Applications Stefánsson, S. F., B. Th. Jonsson & K. R. Thórisson (2009) Proc. of International Conference on Computer Vision Theory and Applications (VISAPP), Lisboa, Portugal, Feb. 5-8, 1:65-68			PDF
Cognitive Map Architecture: Facilitation of Human-Robot Interaction in Humanoid Robots Ng-Thow-Hing, V., K. R. Thórisson, R. K. Sarvadevabhatla, J. Wormer & T. List (2009) IEEE Robotics & Automation Magazine, March, 16(1):55-66			PDF
Holistic Intelligence: Transversal Skills & Current Methodologies Thórisson, K. R. & E. Nivel (2009) Proc. of the Second Conference on Artificial General Intelligence (AGI-09), 220-221. Arlington, VA, USA, March 6-9			PDF
Achieving Artificial General Intelligence Through Peewee Granularity Thórisson, K. R. & Nivel, E. (2009) Proc. of the Second Conference on Artificial General Intelligence (AGI-09), 222-223. Arlington, VA, USA, March 6-9			PDF
Self-Programming: Operationalizing Autonomy Nivel, E. & K. R. Thórisson (2009) Proc. of the Second Conference on Artificial General Intelligence (AGI-09), 150-155. Arlington, VA, USA, March 6-9			PDF
Towards a Neurocognitive Model of Realtime Turntaking in Face-to-Face Dialogue Bonaiuto, J. & K. R. Thórisson (2008) I. Wachsmuth, M. Lenzen, G. Knoblich (eds.), Embodied Communication in Humans And Machines. U.K.: Oxford University Press			PDF
A Brief History of Function Representation from Gandalf to SAIBA Vilhjálmsson, H. & K. R. Thórisson (2008) Proc. of the 1st Function Markup Language Workshop at AAMAS, Portugal, June 12-16, 2008			PDF
Methods for Complex Single-Mind Architecture Designs Thórisson, K. R., G. R. Jonsdottir & E. Nivel (2008) Padgham, Parkes, Müller & Parsons (eds.), Proceedings of the Autonomous Agents & Multiagent Systems (AAMAS 2008), May, 12-16, Estoril, Portugal, 1273-1276			PDF
Learning Smooth, Human-Like Turntaking in Realtime Dialogue Jonsdottir, G. R., K. R. Thórisson & E. Nivel (2008) Proc. of Intelligent Virtual Agents (IVA), Tokyo, Japan, September 1-3			PDF
Modeling Multimodal Communication as a Complex System Thórisson, K. R. (2008) I. Wachsmuth, M. Lenzen, G. Knoblich (eds.), Springer Lecture Series in Computer Science: Modeling Communication with Robots and Virtual Humans, 143-168. New York: Springer			PDF
A Granular Architecture for Dynamic Realtime Dialogue Thórisson, K. R. & G. R. Jonsdottir (2008) Proc. of Intelligent Virtual Agents (IVA '08), Tokyo, Japan, September 1-3			PDF
Fluid Semantic Back-Channel Feedback in Dialogue: Challenges & Progress Jonsdottir, G. R., J. Gratch, E. Fast, & K. R. Thórisson (2007) Proc. of 7th International Conference on Intelligent Virtual Agents (IVA '07), 154-160, September. Paris, France			PDF
Design and Evaluation of Communication Middleware in a Humanoid Robot Architecture Ng-Thow-Hing, V., T. List, K. R. Thórisson, J. Lim & J. Wormer (2007) IROS 2007 Workshop on Measures and Procedures for the Evaluation of Robot Architectures and Middleware, Oct. 29, San Diego, CA.			PDF
Avatar Intelligence Infusion—Key Noteworthy Issues Thórisson, K. R. (2007) 10th International Conference on Computer Graphics and Artificial Intelligence (3IA), Athens, Greece, May 30-31, 123-134	Keynote		PDF
Integrated A.I. Systems Thórisson, K. R. (2007) Invited paper at The Dartmouth Artificial Intelligence Conference: The Next 50 Years — Commemorating the 1956 Founding of AI as a Research Discipline, July 13-15, 2006, Dartmouth, New Hampshire, U.S.A. Minds & Machines, 17:11-25			PDF
Towards a Common Framework for Multimodal Generation: The Behavior Markup Language Kopp, S., B. Krenn, S. Marsella, A. N. Marshall, C. Pelachaud, H. Pirker, K. R. Thórisson & H. Vilhjálmsson (2006) Proc. of Intelligent Virtual Agents (IVA '06), August 21-23 Also published in Springer Lecture Notes in Computer Science	Best Conference Paper Nomination		PDF
Modular Simulation of Knowledge Development in Industry: A Multi-Level Framework Saemundsson, R. J., K. R. Thórisson, G. R. Jonsdottir, M. Arinbjarnar, H. Finnsson, H. Gudnason, V. Hafsteinsson, G. Hannesson, J. Ísleifsdóttir, Á. Th. Jóhannsson, G. Kristjánsson & S. Sigmundarson (2006) Proc. of WEHIA – 1st International Conference on Economic Sciences with Heterogeneous Interacting Agents, 15-17 June, University of Bologna, Italy			PDF
Whiteboards: Scheduling Blackboards for Semantic Routing of Messages & Streams Thórisson, K.R., T. List, C. Pennock & J. DiPirro (2005) K. R. Thórisson, H. Vilhjalmsson, S. Marsella (eds.), Proc. of AAAI-05 Workshop on Modular Construction of Human-Like Intelligence, Pittsburgh, Pennsylvania, July 10, 8-15. Menlo Park, CA: American Association for Artificial Intelligence			PDF
Two Approaches to a Plug-and-Play Vision Architecture – CAVIAR and Psyclone List, T., J. Bins, R. B. Fisher, D. Tweed & K. R. Thórisson (2005) K. R. Thórisson, H. Vilhjalmsson, S. Marsella (eds.), AAAI-05 Workshop on Modular Construction of Human-Like Intelligence, Pittsburgh, Pennsylvania, July 10, 16-23. Menlo Park, CA: American Association for Artificial Intelligence			PDF
On the Nature of Presence Thórisson, K. R. (2005) Proc. Joint Symposium on Virtual Social Agents, AISB-05 / SSAISB Convention: Social Intelligence and Interaction in Animals, Robots and Agents, 12-15 April, University of Hertforshire, Hatfield, U.K., 15-21			PDF
Constructionist Design Methodology for Interactive Intelligences Thórisson, K. R., H. Benko, A. Arnold, D. Abramov, S. Maskey and A. Vaseekaran (2004) A.I. Magazine, 25(4): 77-90. Menlo Park, CA: American Association for Artificial Intelligence			PDF
Artificial Intelligence in Computer Graphics: A Constructionist Approach Thórisson, K. R., C. Pennock, T. List & J. DiPirro (2004) Computer Graphics Quarterly, 38(1):26-30. New York: ACM SIGGRAPH			PDF
Natural Turn-Taking Needs No Manual: Computational Theory and Model, from Perception to Action Thórisson, K. R. (2002) B. Granström, D. House, I. Karlsson (eds), Multimodality in Language and Speech Systems, 173-207. Dordrecht, The Netherlands: Kluwer Academic Publishers			PDF
Machine Perception of Multimodal Natural Dialogue Thórisson, K. R. (2002) P. McKevitt, S. Ó Nulláin, C. Mulvihill (Eds.), Language, Vision & Music, 97-115. Amsterdam: John Benjamins			PDF
Dragons, Bats & Evil Knights: A Three-Layer Design Approach to Character-Based Creative Play Bryson, J. & K. R. Thórisson (2000) Virtual Reality, Special Issue on Intelligent Virtual Agents, 5(2):57-71. Heidelberg: Springer-Verlag			PDF
A Mind Model for Multimodal Communicative Creatures & Humanoids Thórisson, K. R. (1999) International Journal of Applied Artificial Intelligence, 13(4-5):449-486			PDF
The Power of a Nod and a Glance: Envelope vs. Emotional Feedback in Animated Conversational Agents Cassell, J. & K. R. Thórisson (1999) International Journal of Applied Artificial Intelligence, 13(4-5):519-538			PDF
Real-Time Decision Making in Multimodal Face to Face Communication Thórisson, K. R. (1998) Second ACM International Conference on Autonomous Agents, Minneapolis, Minnesota, May 11-13, 16-23			PDF
Layered Modular Action Control for Communicative Humanoids Thórisson, K. R. (1997) Proc. of Computer Animation '97, Geneva, Switzerland, June 5-6, 134-143			PDF
Gandalf: An Embodied Humanoid Capable of Real-Time Multimodal Dialogue with People Thórisson, K. R. (1997) Proc. of the First ACM International Conference on Autonomous Agents, Mariott Hotel, Marina del Rey, California, 536-7			HTML
Why Put an Agent in a Body: The Importance of Communicative Feedback in Human-Humanoid Dialogue Thórisson, K. R. & J. Cassell (1996) Proc. of Lifelike Computer Characters '96, Snowbird, Utah, October 5-9			HTML
Multimodal Interaction with Humanoid Computer Characters Thórisson, K. R. (1995) Proc. of Lifelike Computer Characters '95, Snowbird, Utah, September 26-29, 45			HTML
Computational Characteristics of Multimodal Dialogue Thórisson, K. R. (1995) AAAI Fall Symposium on Embodied Language and Action, Massachusetts Institute of Technology, Cambridge, Massachusetts, November 10-12, 102-108			PDF
Simulated Perceptual Grouping: An Application to Human-Computer Interaction Thórisson, K. R. (1994) Proc. of the Sixteenth Annual Conference of the Cognitive Science Society, Atlanta, Georgia, August 13-16, 876-81			PDF
Face-to-Face Communication with Computer Agents Thórisson, K. R. (1994) Proc. of AAAI Spring Symposium on Believable Agents, Stanford University, California, March 19-20, 86-90			PDF
Estimating Three-Dimensional Space from Multiple Two-Dimensional Views Thórisson, K. R. (1994) Presence: Teleoperators and Virtual Environments, 2(1):44-53			PDF
Dialogue Control in Social Interface Agents Thórisson, K. R. (1993) InterCHI Adjunct Proc., Amsterdam, Holland, April 24-29, 139-40. New York: ACM Press			PDF
Synthetic Synesthesia: Mixing Sound with Color Thórisson, K. R. & K. Donoghue (1993) InterCHI Adjunct Proc., Amsterdam, Holland, April 24-29, 65-6. New York: ACM Press			PDF
Integrating Simultaneous Input from Speech, Gaze & Hand Gestures Koons, D. B., C. J. Sparrell & K. R. Thórisson (1993) In M. T. Maybury (Ed.), Intelligent Multimedia Interfaces, Ch. 11, 257-276. Cambridge, Massachusetts, U.S.A.: AAAI Press / M.I.T. Press	1	2	PDF
Estimating Direction of Gaze in Multi-Modal Context Koons, D. B. & K. R. Thórisson (1993) Presented at 3Cyberconf - The Third International Conference on Cyberspace, Austin, Texas, May 15-16			PDF
Multimodal Natural Dialogue Thórisson, K. R., D. B. Koons & R. A. Bolt (1992) SIGCHI '92 Proc., 139-140			PDF
The Effects of ESP-Belief & Distorted Feedback on a Computerized Clairvoyance Task Thórisson, K. R., F. Skulason and E. Haraldsson (1991) Journal of Parapsychology, 55:45-58
Short Term Memory Demands in Processing Synthetic Speech by Old and Young Listeners Smither, J. A., R. Gilson, M. Thomas & K. R. Thórisson (1990) The Gerontologist, 30, Special Issue 309A

INVITED LECTURES & PANELS

Framtidarnefnd Althingis (Icelandic Parlament's Committee on the Future)
Reykjavík, February 28, 2023
Ísland og gervigreind (Iceland & Artificial Intelligence)