|
The
2010 North American
Conference on
Computing
and Philosophy
NA-CAP@CMU
2010: Simulations and Their Philosophical Implications
Click on Any Talk in Red to View Its Abstract
| Saturday, July 24 |
|
| |
|
|
| |
3:30p – 5:30p |
Registration (Lounge adjacent to the Erwin Steinberg Auditorium) |
| |
|
|
| |
5:30p – 6:30p |
IACAP Presidential Address (Erwin Steinberg Auditorium)
Introduction by Tony Beavers, The University of Evansville |
| |
|
|
| |
|
“The Craft of Computing and Philosophy”
Robert Cavalier, President Ex-Oficio, Carnegie Mellon University
|
| |
|
|
| |
6:30p – 7:00p |
2010 Covey Award Presentation (Erwin Steinberg Auditorium)
Introduction and Award Presentation by James Fetzer, University of Maryland |
| |
|
|
| |
|
The International Association for Computing and Philosophy is pleased to announce that John Rogers Searle is the winner of the IACAP 2010 Covey Award for Excellence in Research in the Area of Computing and Philosophy. |
| |
|
|
| |
7:00p - 8:00p |
Reception (Lounge adjacent to the Erwin Steinberg Auditorium) |
| |
|
|
| Sunday, July 25 |
|
| |
|
|
| |
8:30a – 9:00a |
Continental Breakfast (Lounge adjacent to the Erwin Steinberg Auditorium) |
| |
|
|
| |
|
Special Graduate Student Session (Erwin Steinberg Auditorium)
Session Chair: Mara Harrell, Carnegie Mellon University |
| |
|
|
| |
9:15a – 10:15a |
Goldberg Award Honorable Mention |
| |
|
|
| |
|
“Automated Construction of Causal Concepts”
Stephen Fancsali, Carnegie Mellon University
Abstract: Substantial literature in computer science, philosophy, and statistics over the last 20+ years has focused on the problem of causal inference from observational data. A standard formalism for this work is that of causally-interpreted Bayesian networks. Provably correct algorithms have been developed for learning the structure of these graphical models from observational data with given measured variables. I extend causal search to the problem of constructing causally-appropriate concepts or variables in situations in which the investigator is faced with a large number of “raw” variables upon which the investigator has no obvious means by which to suitably intervene. Given a set of “raw” variables and a fixed target variable, we seek to find constructed variables such that predictive models specified according to the causal structure learned over the constructed variables maximize predictability of the target for two distinct purposes: simple prediction (forecasting) and causal inference. After briefly summarizing extant work on algorithmic causal inference, I propose means of assessing constructions for predictive and causal purposes: “Markov blanket predictability” and “causal predictability.” A concrete example is then explored in some detail with data from the 1990 United States Decennial Census and voting dating for a referendum in Allegheny County, Pennsylvania, in 1997. As this is largely a sketch of a novel research program, avenues for future work are then briefly discussed.
|
| |
|
|
| |
|
“Measuring Representation”
Lars Marstaller, Macquarie Centre for Cognitive Science, Sydney
Abstract: Representation is a core and contested concept in cognitive science. It is notoriously hard to define let alone quantify. We offer a quantitative examination of the main intuitions behind it, viz. that intelligent systems are only able to compute complex behavior by taking into account properties of their environment that cannot be relied upon to be immediately available in the input. In artificial neural networks (ANN) it is highly problematic to conclusively prove the existence of representational vehicles because their activations are not neatly decomposable into symbols and computational processes. We therefore suggest an alternative operational definition of representation based on mutual information, viz. our measure R. Our hypothesis that R captures the intuition behind the concept of representation is tested in an artificial life simulation. We used a genetic algorithm to evolve an ANN controlling a simple simulated robot performing a categorization task and measured its representation R. Our results show that over the course of evolution representation increases proportional to fitness, i.e. any solution to the categorization problem that the genetic algorithm could find involved representation as measured by R. We conclude with a re-interpretation of the term ‘representation’ that renders it applicable to non-classical computational architectures.
|
| |
|
|
| |
|
Commentator: Marcello Guarini, University of Windsor |
| |
|
|
| |
10:15a – 10:45a |
2010 Goldberg Graduate Student Award Presentation (Erwin Steinberg Auditorium)
The impact of research in Computational Modeling, Artificial Intelligence, Machine Learning, Formal Models of Learning, and Agent-based Simulations on the discipline of Philosophy has been profound. Contemporary discussions of epistemology, ethics, theory of mind, and philosophy of language have all benefited from lively, interdisciplinary debates over the relation between computational and formal models, and traditional philosophical questions. These debates have found their way into scholarly publications and textbooks, as well as into a growing number of Masters and Ph.D. theses.
In order to recognize outstanding achievements by Graduate Students in this area of research and scholarship, the International Association for Computing and Philosophy is proud to offer the “Brian Michael Goldberg Memorial Award” for presentations in any category listed above. This Award, which carries a $500 USD stipend, will be presented each year at one of the North American Computing and Philosophy conferences. Nominees and applicants are welcome from around the world.
The department of philosophy at Carnegie Mellon is the sponsor of this award and will serve as the site for submissions. The department will establish an international committee to review applications and, in conjunction with NA-CAP, will announce the yearly winner. Each year's winner will be expected to make a presentation at a NA-CAP conference as part of the Award Ceremony.
This Award was made possible by a generous gift from Dr. Gerald and Nancy Goldberg in memory of their son, Brian Michael Goldberg. In their words:
Brian was a twenty-two year old student who was admitted to Carnegie Mellon University in 1991 to the doctoral program in philosophy. He died unexpectedly before he could realize his dream of attending Carnegie Mellon. Brian was an independent thinker who loved competition and a good challenge. Throughout his life, he found it exciting to enter and win contests. He loved challenging his mind, especially by studying philosophy, mathematics and logic. He loved challenging his creativity through photography, painting and theatre arts. He loved challenging his body by learning such diverse sports as wrestling, fencing and scuba diving. He loved debating and challenging others to think in new ways and had seriously considered becoming a university professor. To honor who he was and what he loved, this Goldberg Memorial Award is offered to challenge and motivate other graduate students in Brian's chosen field of study.
|
| |
|
|
| |
|
“Animat Liberation”
Derek Jones, Indiana University, Bloomington
Abstract: In this paper I identify a central epistemic presupposition behind animat modeling research, namely the idea that modelers can gain new knowledge about biological systems without modeling the specific underlying mechanisms of those systems. The critique of animat modeling presented by Webb (2009) and Bechtel (2009) can be viewed as presenting a skeptical challenge to this presupposition. I will attempt to offer a defense of this thesis and consequently of animat research, highlighting several ways in which animat models can provide knowledge about biological systems in the absence of Webb-style biological grounding.
|
| |
|
|
| |
10:45a – 11:00a |
Break (Lounge adjacent to the Erwin Steinberg Auditorium) |
| |
|
|
| |
11:00a – 12:30p |
The Herbert A. Simon Keynote Address (Erwin Steinberg Auditorium)
Introduction by James Moor, Dartmouth College |
| |
|
|
| |
|
“The Turing Test”
Hugh Loebner, The Loebner Prize
Abstract: What is the Turing Test? What are some common misconceptions regarding it? What are its implications for the future of A.I. and society? I will discuss these and others questions based upon my 20 years of association with The Loebner Prize “The First Turing Test.”
|
| |
|
|
| |
|
Panel Discussion |
| |
|
|
| |
12:30p – 1:15p |
Boxed Lunches |
| |
|
|
| |
1:15p – 2:15p |
Special Invited Speaker on Computer Simulations (Erwin Steinberg Auditorium)
Introduction by Mara Harrell, Carnegie Mellon University
|
| |
|
|
| |
|
"Modeling the Dynamics of Belief Networks: Epistemology, Epidemiology, and Scientific Optimization"
Patrick Grim, Stony Brook University
Abstract: This paper introduces three ongoing projects in a form of social epistemology that focuses on belief networks. In each case computational modeling takes philosophy into new territory, often with a promise of practical application.
1. A first project uses dynamic network models in an attempt to understand some of the very different ways in which (a) germs, (b) genes, and (c) memes can be expected to propagate across natural and social populations.
2. A second project constructs agent-based simulations of community belief convergence and polarization dynamics using networks abstracted from data on information and trust in the African American and White communities. The results carry implications for both epidemiological public policy and the attempt to address health care disparities.
3. A third project is fully normative, focused in philosophy of science. What form of social organization can be expected to be optimal in the exploration of what kind of question? What kind of territory does big science best explore? For what kind of issues is garage science a better bet?
Computer simulation offers new tools for both traditional philosophical questions and for non-traditional philosophical work: work in which disciplinary boundaries tend to blur, and in which abstract models often carry quite practical implications.
|
| |
|
|
| |
2:15p – 2:30p |
Break (Lounge adjacent to the Erwin Steinberg Auditorium) |
| |
|
|
| |
2:30p – 4:30p |
Panel Session: The Limits of Simulations of Human Actions (Erwin Steinberg Auditorium)
Session Chair: James Fetzer, University of Maryland
Abstract: The speaker contends that basic differences between digital
machines and human beings--including the static difference, the dynamic difference, and the affective difference--render stronger forms of simulation impossible, in principle; and that ontic and epistemic difficulties relative to
the complex interaction of distinct variables and relevant conditions to which each of us has been subjected our unique lives makes non-trivial explanations and predictions--those not involving stereotyped or scripted behavior--impossible, in practice, for even the weakest form. The commentators divide, where two contend the speaker's objections can be overcome but the third agrees that these problems pose insuperable obstacles to successful simulations of
actions except in the simplest cases.
|
| |
|
|
| |
|
"Limits to Simulations of Thought and Action"
James Fetzer, University of Maryland
Abstract: Searle’s “Chinese Room” argument establishes that the behavioristic Turing Test does not afford a criterion that is sufficient for the purpose of discriminating between the causal properties of systems of different kinds. It supports the need to differentiate between relations of simulation that display the same input/output behavior, of replication by simulations that are brought about by the same or similar processes, and of emulation, where those very replications are produced by systems that are composed of the same kind of stuff.
Since simulation is the weakest similarity relationship between animate and inanimate systems, the questions addressed here concern whether an inanimate system, such as a robot, can simulate non-trivial behavior that is displayed by humans as the effects of their internal states of motives, beliefs, ethics, abilities. and capabilities, relative to those systems' opportunities (the historical situations in which specific behaviors take place). It arises because humans have minds, but these machines do not.
While scripted or stereotypical behaviors pose no problems for simulations, the success of simulation in non-trivial cases is problematical. The argument posed maintains there are both (a) ontic problems, including the complexity of causal interactions between the relevant variables, and (b) epistemic problems, in establishing their specific values. The enduring but subtle effects of unique events in the a person's history, which are exerted through the influence of unconscious or subconscious causal factors, for example, illustrate the problems involved.
While simulating past behavior and predicting future behavior in scripted or stereotypical instances offers no problems, in principle, the anticipation of non-trivial behaviors that—as actions produced by motives, beliefs, ethics, abilities, and capabilities—in the future appears to require kinds of knowledge of histories of personal experience and of the influence of those experiences upon internal states that, in principle, are only accessible to systems that are in stronger relations of replication or even emulation to produce those simulations.
|
| |
|
|
| |
|
“Salvaging Computationalism: How Cognition Could Be Computing”
William Rapaport, SUNY Buffalo
Abstract: Elsewhere, I have argued that the proper treatment of computationalism is that it is the claim that cognition is computable, not that cognition is computation. And I have presented “syntactic semantics”—that (1) cognitive agents have direct access only to internal representatives of external objects, (2) so, words, meanings, and semantic relations between them are all syntactic, and (3) understanding is recursive--as a theory supporting the computability of cognition. Here, I argue against Fetzer that syntax suffices for semantic cognition and that these two principle theses raise problems for his theory of semiotic systems; his alleged “static,” “dynamic,” and “affective” differences between minds and computers; and his argument that minds cannot be simulated by computers.
|
| |
|
|
| |
|
“Can a Computer Simulate a Logician?”
Selmer Bringsjord, Rensselaer Polytechnic Institute
Abstract: Speaker is right: computational simulations, relative to the human
case, fall utterly short. I attempt to show this by considering the
challenge of engineering a computer capable of simulating a logician. I
consider three spheres of activity: a logician (confessedly not operating
as much of a logician) "computing" a primitive recursive function; finding,
crafting, and evaluating proofs in quantified modal logic; and proving
Godel's incompleteness theorems. In each of these three cases, what
computers have done to this point, and what they can do in the future in
the context of our present knowledge, is in no way a non-trivial simulation
of the human case. The first fatal flaw infecting computer-based simulations
of these three cases is that they are bereft of emotion, consciousness, and
understanding (Speaker’s "affective difference"). The second fatal flaw --
revealed in the Godelian case, and in proofs of theorems independent of
Peano Arithmetic -- plaguing such simulations is that they cannot capture
the infinitary nature of human reasoning.
|
| |
|
|
| |
|
“The Virtual Reality Turing Test”
James Moor, Dartmouth College
Abstract: Speaker is wrong: the question of whether computers could have real minds
is in part conceptual but it is largely empirical. Not a question to be decided
a priori or on the basis of today’s technology. How might we know? I believe
there are many tests that would provide evidence, but on this occasion, the sixtieth anniversary of Alan Turing’s seminal article, “Computing Machinery
and Intelligence”, I would like to defend the merits of some variations of the
Turing test. The test is often dismissed as a superficial simulation and therefore
not something that could establish that a computer has intelligence. Elsewhere
I have argued in favor of the original Turing test (------, ------) but here I propose two alternative versions of the Turing test that also could establish, inductively
of course, that a computer has genuine intelligence. If computer passed these alternative Turing tests, it would not merely simulate intelligence, it would have it.
|
| |
|
|
| |
4:30p – 5:30p |
Ethics and Simulations (Erwin Steinberg Auditorium)
Session Chair: Mara Harrell, Carnegie Mellon University |
| |
|
|
| |
|
“On the Creation of Virtuous Machines”
Ryan Tonkens, York University
Abstract: The emerging discipline of Machine Ethics is concerned with developing autonomous machines that behave morally out in the world. The main focus of machine ethicists thus far has been towards the engineering and computational barriers that stand in the way of the successful development of such machines. Yet, there are also other ethical concerns that need to be attended to, prior to the implementation and development stages. I argue that one important ethical constraint has gone unnoticed, and that meeting this constraint represents a serious challenge for Machine Ethics: To identify a moral framework that can be successfully implemented into machines, in such a way so that machines act ethically out in the world, and whose own tenets permit the creation of those machines in the first place. The worry is that implementing a moral framework into robots that does not allow the creation of those robots is ethically inconsistent (at best). Recently, Wallach & Allen (2009) have argued that a virtue-ethical framework is a promising framework for putting the ethic into ethical machines. Here I examine whether the creation of virtuous machines is consistent with a virtue-ethical framework. It is argued that, although the creation of virtuous machines is consistent with some of the tenets of virtue ethics, it violates certain others, most notably the virtues associated with social justice. Because of this, despite the potential computability of virtue ethics, the creation of virtuous machines is self-contradictory, Wallach & Allen’s claims notwithstanding.
|
| |
|
|
| |
|
“Learning, Simulations, and the Particularism-Generalism Debate in Moral Philosophy”
Marcello Guarini, University of Windsor
Abstract: At one point in the debate between moral particularists and generalists, Jackson, Petit, and Smith (2000), argued that particularists who reject entirely the role of rules, principles, or standards would not be able to explain how we learn the difference between right and wrong. In his reply, Jonathan Dancy (1999) suggested that artificial neural networks might be able to show how it is possible to learn right from wrong without principles. This paper will outline a few of the options available to particularists and generalists. A neural network simulation of moral case classification will be discussed, but it will be argued that it vindicates neither purely particularist nor purely generalist approaches to learning. Nothing in this paper should be read as an attempt to administer knock down blows to any of the positions under consideration. I quite deliberately bill this work as exploratory. There are empirical assumptions at work in discussions between particularists and generalists, and it is early days still in understanding the strengths and weaknesses of various computational models and in empirical research on human cognition. Clarifying what some of the options are and showing how neural network simulations may help us to see options that have gone otherwise unconsidered are the main goals of this work.
|
| |
|
|
| |
|
Commentator: Paul Bello |
| |
|
|
| Monday, July 26 |
|
| |
|
|
| |
8:45a – 9:30a |
Continental Breakfast (Lounge adjacent to the Erwin Steinberg Auditorium) |
| |
|
|
| |
9:30a – 10:30a |
Concurrent Sessions – Group 1 |
| |
|
|
| |
1A |
What Is a Simulation? (Erwin Steinberg Auditorium)
Session Chair: Mara Harrell, Carnegie Mellon University
|
| |
|
|
| |
|
“What Does a Computer Simulation Have to Reproduce? The Case of VMWare”
Keith Douglas, Statistics Canada
Abstract: Some questions one might have about computer simulations are: when do they occur; what makes a simulation a simulation (of something)? The present paper is an attempt to answer an instance of this question (namely, what counts as a simulation of a computer system by a computer system) through the resources of contemporary metaphysics.
|
| |
|
|
| |
|
“Simulating Divine Action: Hierarchical Causation via Networked Linguistic Intermediaries”
Ed Haymond, AVC
Abstract: The further progress of contemporary divine action theory is partially dependent on resolution of divergent perspectives regarding the ontological status of the laws of nature. Currently the discussion founders because the laws are interpreted either as metaphysically independent entities or as cognitive constructs. Howard Kainz has presciently outlined a way forward by drawing attention to the power of cybernetic modeling and, by implication, simulation of divine action. Taking Aquinas as a leading representative of scholars who investigate the properties and activities of separate substances, Kainz suggests that points of concentration in medieval metaphysics are reflected in the architecture of information and communication technologies (ICTs). Kainz’s observations are here extended in order to bridge back to the relevance of the ancient and medieval background of divine action theory. Although detrimentally neglected now, linguistic intermediaries play a vital role in the philosophy of Philo of Alexandria, Moses Maimonides, and Gregory Palamas. Each of these thinkers articulated characteristics of dynamic formal principles that, if recovered, may clarify the problematic relationship between immanent and transcendent causes. These logoi can also legitimate simulation, and by means of it the field of computing and philosophy can contribute to divine action theory by mapping the objects and processes of virtual worlds to their ancient counterparts. In particular, there is isomorphism between the laws of nature, angelic agency, and virtual-world design plans that, as Kainz might suspect, hints at the conditions for cosmic viability.
|
| |
|
|
| |
1B |
Computing, Philosophy and Pedagogy (Erwin Steinberg Auditorium)
Session Chair: Robert Cavalier, Carnegie Mellon University |
| |
|
|
| |
|
“Simulation, Search and Pedagogy: Lessons from Logic and Proofs”
Davin Lafon and Dawn McLaughlin, Carnegie Mellon University
Abstract: The ‘ProofTutor’ Intelligent Tutoring System is a Simulated Teaching Assistant that has been developed as part of the Automated Proof Search (AProS) project at Carnegie Mellon. A first version of this Tutor went ‘live’ roughly 1 and 1/2 years ago and has been used in the instruction of approximately 1,000 students to date. This talk will first introduce aspects of the AProS Project necessary to understand the ProofTutor, followed by demonstrations of the adaptive interaction system it employs. I will also explore uses for this interaction system beyond the AProS project, in terms of producing dialog-rich pedagogical tools. This talk will touch on several key themes, including Simulation, Automated Theorem Proving and Machine Learning.
|
| |
|
|
| |
|
“Ontological and Epistemic Questions about Computer Simulation: Toward a Philosophy of Mind and Simulation”
James Stieb, Drexel University
Abstract: This paper addresses epistemic and ontological questions preparatory to a philosophy of simulation. It argues the logical superiority of constructivist theories over reductively computationalist or realist theories. Computationalism such as Dennett’s and realism such as Searle’s are inevitably question-begging and unproductive. Only constructivist theories can adequately respond and interpret practice. Examples of each and of the philosophical implications for validating models with simulations within the physical sciences are addressed as are implications for cognitive science and the sociology of teaching methods.
|
| |
|
|
| |
10:30a – 10:45a |
Break (Lounge adjacent to the Erwin Steinberg Auditorium) |
| |
|
|
| |
10:45a – 12:00p |
The Douglas C. Engelbart Keynote Address (Erwin Steinberg Auditorium)
Introduction by Mara Harrell, Carnegie Mellon University |
| |
|
|
| |
|
“How I Learned to Reduce My Incoherence”
Teddy Seidenfeld, Carnegie Mellon University |
| |
|
|
| |
12:00p – 12:45p |
Boxed Lunches |
| |
|
|
| |
12:45p – 2:15p |
Panel Session: Using Computer Simulations to Facilitate Philosophical Exploration (Erwin Steinberg Auditorium)
Panelists: Tony Beavers (chair), Marvin Croy, Patrick Grim, Ryan Muldoon, Kevin Zollman
Abstract: Computer simulations are regularly used in both the natural and social sciences, where they have enjoyed a degree of predictive success. This success has raised several questions about the nature of science: what makes a method scientific; has science always been a matter of model construction; what does it mean to explain something, and so on. Of late, model construction has been making its way into philosophy, where it is also raising several questions about the nature of philosophy: what makes a method philosophical; is philosophy, traditionally conceived, already a matter of building conceptual models, as Simon Blackburn has suggested; what other kind of models might be appropriate for philosophical exploration, and so on. This panel will explore these questions by offering some examples of recent attempts at philosophical modeling involving computers while raising foundational questions about the nature of philosophy.
|
| |
|
|
| |
|
"Connectionist Modeling of Student Pattern Matching Data"
Marvin Croy, The University of North Carolina, Charlotte
Abstract: We present the results of a connectionist network performing pattern matching with propositional rule forms and instantiated expressions. The network is designed to model the judgments of students when presented with the same task and training instances. Then, the network faces a variety of different problem sets and its performance is noted. The aim is to observe the results of various training set characteristics on test performance for the purpose of designing more effective problem sets for students learning propositional rule application.
|
| |
|
|
| |
|
"More Fun with Jets and Sharks: Typicality Effects and the Search for the Perfect Attractors"
Anthony F. Beavers, University of Evansville
Abstract: Feature detection can be a bit of a tricky business. Various techniques can be used to find information among features that might not be immediately apparent. Here, I present a dynamic system to treat the dataset used in McClelland's 1981 famous connectionist model, Jets and Sharks. I will show that there is more information available in this dataset than McClelland's model actually makes explicit and show that a dynamic approach to setting weights in a network can draw this information out. I will then address the implications of the model and whether or not adding to the dataset will improve the accuracy and resilience of the system or whether it will collapse as toy models often do when they are extended beyond their microworlds. I will attempt to do all of this using a dynamic associative network (DAN) model designed in an ordinary spreadsheet.
|
| |
|
|
| |
|
“The Need for Simulation in Social Epistemology”
Kevin Zollman Carnegie Mellon University
Abstract: The burgeoning field of social epistemology directly addresses problems in epistemology that arise from social interaction. One particular area of social epistemology, called "systems oriented social epistemology" by Alvin Goldman, compares different features of social groups and evaluates groups' ability to achieve particular epistemic ends (like true belief, justified belief, etc.). I argue that computer simulation is a critical tool in systems oriented social epistemology – one simply cannot answer many questions in this field without simulation. The necessity of simulations arises because many of these systems exhibit the properties grouped under the general heading of "complexity theory" such as non-linearity, sensitivity to initial conditions, and cyclic or chaotic behavior. While not always required, computer simulations are being widely used to understand these types of systems. Social epistemology presents one important area where the use of computer simulations can and do make important philosophical advancements.
|
| |
|
|
| |
|
“Replacing Thought Experiments with Computer Simulation in Social and Political Philosophy”
Ryan Muldoon, Joseph L. Rotman Institute of Science and Values, University of Western Ontario - Attending via Skype
Abstract: Computer simulation provides three distinct advantages to thought experiments in political philosophy and studies of social dynamics. First, as a simulation requires an algorithm, it forces the theorist to uncover the hidden choices that she must make to fully articulate the decision-making procedure that the agents in the simulation follow. Second, it allows the theorist the opportunity to see which of these choices matter, by constructing multiple models that systematically test those choices. Third, it allows the theorist to examine situations that are more complex than what could reasonably be done in a thought experiment. In investigating these advantages, we can further find that computer simulations lend themselves to a different style of argument. As Dretske has argued, “If you can’t make one, you don’t know how it works.” Computer simulations are a kind of constructive proof: they provide an opportunity to test whether or not we know how our social and political theories actually work.
|
| |
|
|
| |
|
“Collaborative Research in Philosophical Computer Modeling”
Patrick Grim, Stony Brook University
Abstract: Philosophy has been traditionally conceived as either a solitary pursuit, lonely and isolated, or as a form of combative intellectual gunslinging. Our experience has been that philosophical computer modeling invites a very different and more collaborative form of research, much more like that common in the sciences. We have been particularly pleased with the way computational philosophy allows one to cross disciplinary boundaries, and in which the variety of techniques required allows teams in which the standard academic hierarchy is softened or dissolved. This presentation will explore these themes.
|
| |
|
|
| |
2:15p – 2:30p |
Break (Lounge adjacent to the Erwin Steinberg Auditorium) |
| |
|
|
| |
2:30p – 4:30p |
Science and Simulations (Erwin Steinberg Auditorium)
Session Chair: Mara Harrell, Carnegie Mellon University |
| |
|
|
| |
|
“Gamers, Are You Learning? Using Prompts to Learn through Games”
Seolim Kwon, Indiana University
Abstract: Digital simulations have the potential to promote learning by providing players with an authentic situation on a scenario that resembles the real life. They also allow players to participate in activities that are difficult or even impossible to do with traditional materials (Shaffer, 2006). Studies conducted on simulations in diverse fields have demonstrated the effectiveness of using simulations as part of the learning process (Anderson, 2008; Castaneda, 2008; Faria, 2001; Kuriger, Wan, Mirehei, Tamma, & Chen, 2009; Squire, Barnett, Grant, & Higginbotham, 2004). The Diffusion Simulation Game (DSG) provides a virtual scenario for players to put into practice their knowledge and skills about change management concepts and diffusion of innovation strategies.
|
| |
|
|
| |
|
“Simulating Time with Computers”
Michael Nicolaidis, TIMA Laboratory
Abstract: In a famous article entitled “Simulating Physics with Computers”, Richard Newman discusses the possibility of simulating quantum systems. The question of simulating time is also addressed. It is stated that, in simulations performed by cellular automata, time is not simulated but is rather imitated by being hidden behind the state to state transition. Simulating time and in particular simulating relativistic space-time can be important for enhancing our understanding of modern physics. In the present paper we introduce the notion of the observer that is part of the simulated physical system. For this kind of observers we show that time and relativistic space-time (in the sense of special relativity) can emerge if the computation rules used to compute the state to state transition obey certain conditions. Thus, by taking into account the point of view of such observers, we can simulate the emergence of time, including relativistic space-time.
|
| |
|
|
| |
|
“Simulation: Methodology and A Need for Standards”
Vincent Wiegel, Delft University of Technology
Abstract: Computers simulations are more and more frequently used to validate theoretical models. The computer simulations are mostly idiosyncratic. There are two important, methodological and ontological issues that need to be addressed in order to progress further beyond the ‘private’ computer programs. The ontological issue has to do with the translation from theory to computer language. At the smallest level of abstraction in a theory the entities are still much more complex than the data structures and commands and functions in a computer language. This requires the reconstruction of those entities. This reconstruction is by no means self-evident. As long as there is not a standard ontology for basic entities and concepts such as agent, action, etc. it will remain unclear what exactly is simulated. This issue, in turn, has methodological implications. Computer simulations are hard to compare and validate. This is a symptom of a more general methodological weakness in most computer simulations: a lack of a coherent, standardized protocol for simulations. In this paper a argument is made for the development of a shared standards for simulations. These standards should include the publication of the data structures that underlie theory, model and implementation. LinkedData and DbPedia could provide a good framework for the implementation of these standards. This definition should be a part of a protocol for simulations that addresses requirements of Transparency, Repeatability, Verifiability.
|
| |
|
|
| |
4:30p – 5:00p |
Closing Announcements from the IACAP Executive Director (Erwin Steinberg Auditorium)
Tony Beavers, University of Evansville |
| |
|
|
|