OREL Medium: Trajectories in Cognitive Science Session @ CogSci 2021

 

This content was originally posted to the Orthogonal Research and Education Lab blog on July 30.

Congratulations to Jesse Parent, Avery Lim, Bradly Alicea, and Anusha Sharma for heading up the “Trajectories in Cognitive Science” discussion group, held during the CogSci 2021  conference. The event happened in six parts (which we will recap) and is now on YouTube. We are also archiving the slides and reference list on the Open Science Framework (in progress).

Part I: Frontier Map and Cognitive Futures. Presenter: Jesse Parent.

The first part of the session involved an overview of Frontier Maps and the role of maps and visualizations in understanding how ideas form fields of study. Frontier Maps also enable casual learners to get an intellectual grasp on a certain area of study by learning its history and current trends.

Part II: Adjacent Futures. Presenter: Bradly Alicea.

The second talk was given by Bradly Alicea, and involved introducing the idea of Adjacent Futures, which is based on the notion of the adjacent possible. Our focus was on both the possibilities that define scientific discovery and interdisciplinary exploration and the factors (sometimes quite practical) that block combinatorial discovery that often define the boundaries of a given field.

Part III: The Place of Development in the History of CogSci. Presenters: Jesse Parent and Anusha Sharma.

This session (brought to us by Jesse Parent and Anusha Sharma) covered the rich history of developmental approaches in Cognitive Science, and how development has served as an alternative to the concept of “static adult minds”. The presentation was an in-depth presentation of the review article “The Place of Development in the History of Psychology and Cognitive Science” by Gabriella Airenti (Frontiers in Psychology, 10, 895). In the session, we explored the foundational concepts of Piaget as well as longstanding debates such as nature vs. nurture and representationalism vs. brain function.

Part IV: Neurodiversity and Cognitive Science. Presenter: Jesse Parent.

The fourth presentation was also by Jesse Parent, and covered the role of Neurodiversity in Cognitive Science. This was a short review of a book called “Neurodiversity Studies: a new critical paradigm”. Neurodiversity covers a number of alternative frameworks for understanding human cognition, including but not limited to queer, feminist, and critical race perspectives. Such perspectives contribute not only to the diversity of views in the field, but also lead to novel intellectual trajectories.

Part V: Trajectories of Interest in Developmental Psychobiology. Presenter: Avery Lim.

Avery Lim presented on a number of possible trajectories for developmental psychobiology considered broadly. This possibility space (discussed in Part II) includes building off of the study of phenomenology such as developmental critical periods or computational models of psychophysiology. A number of open questions were also posed that motivated our discussions in Part VI.

Part VI: Open Discussion!

If you are interested in discussing these topics further, you might be interested in joining the Orthogonal Lab Slack or the Computational Critical Periods Discord. You can also catch Saturday Morning NeuroSim weekly on YouTube, or get in contact to get on our mailing list and join in person.

Dispatches from the Emergent Interaction Workshop

 This content has been cross-posted at the Orthogonal Lab Medium.

Last month, the Orthogonal Lab was represented at the Emergent Interaction Workshop (part of SIGCHI 2021). We contributed a paper titled “Allostasis Machines: a model for understanding internal states and technological environments”, with a companion presentation now on YouTube. Thanks go to Bradly Alicea, Daniela Cialfi, Anson Lim, and Jesse Parent for their contribution. We are planning an expanded version of this work with Rishabh Chakraborty that will demonstrate Allostasis Machines as a Reinforcement Learning implementation.

The subtitle of this workshop was “Complexity, Dynamics, and Enaction in HCI”. Therefore, the focus was on advancing measurement and theory, in addition to better characterize complexity in the field of Human-Computer Interaction. The four-hour long session was summarized in our weekly meeting on May 22. I have also provided supplemental readings in four workshop-related categories at the bottom of this post.

Overview of the Emergent Interaction Miro board.

There were six other papers made available before the session. Two of the most interesting to the Orthogonal Lab group are “Fields of Affordances and Human Computer Interaction” by Jelle Bruineberg and “Simulating Social Acceptability With Agent-based Modeling” by Alarith Uhde and Marc Hassenzahl.

The Emergent Interaction utilized Zoom, Slack, and a Miro board to enable discussion during the session. Check out the overview paper titled “Emergent Interaction: Complexity, Dynamics, and Enaction in HCI” for more information.

Testing, 1, 2, Emergent T-shirt….

There were a number of interesting and innovative topics discussed in the workshop. Dynamical approaches came up several times, along with topics such as multifractality, attractor analysis, and co-evolutionary experimental design. For more information regarding the first two topics, check out Alan Dix’s blog on Making Sense of Quantitative Data, and Dan Bennett’s preprint “Multifractal Mice: Measuring Task Engagement and Readiness-to-hand via Hand Movement”.

Tom Froese presented on his Enactive Artificial Intelligence and HCI work. His Google Scholar profile features some really interesting work that cuts across the worlds of Artificial Life, Cybernetics, and Cognitive Science, but his workshop topic was how modern Machine Learning approaches are insufficiently embodied. I have posted references to two of his key works (workshop-wise) in the Further Readings section of this post.

Later, Parisa Eslambolchilar presented on first-order closed-loop feedback taking the form of sensor-based human interaction loops. She reviewed some of the things she developed in her Doctoral dissertation, then lead us into her more recent work. Learn more by reading “A Model-Based Approach to Analysis and Calibration of Sensor-based Human Interaction Loops”.

Then, Vassilis Kostakos discussed his work on modeling interactions between technology users (or users and interfaces) as a complex system. He utilized the “lynx-hare” predator-prey analogy, inspired by Lotka-Volterra co-evolutionary dynamics. Read more in this paper published last year in Human-Computer Interaction: “Modeling interaction as a complex system”.

Emergent Interaction is now on Twitter! Give them a follow to join the discussion.

Further Reading: Measurement techniques.

Rebout, N., Lone, J-C., De Marco, A., Cozzolino, R., Lemasson, A., and Thierry, B. (2021). Measuring complexity in organisms and organizations. Royal Society Open Science, 8, 200895.

Zhou, Q., Chua, C-C., Knibbe, J., Goncalves, J., and Velloso, E. (2021). Dance and Choreography in HCI: A Two-Decade Retrospective. Proceedings of CHI, 262, 1–14. Video

Further Reading: Enactive Approaches to Artificial Systems.

Froese, T. and Ziemke, T. (2009). Enactive artificial intelligence: Investigating the systemic organization of life and mind. Artificial Intelligence, 173, 466–500.

Froese, T., McGann, M., Bigge, W., Spiers, A., and Seth, A.K. (2012). The Enactive Torch: A New Tool for the Science of Perception. IEEE Transactions on Haptics, 5(4), 365–375.

Further Reading: Agent-based Modeling approaches.

Bonabeau, E. (2002). Agent-based modeling: Methods and techniques for
simulating human systems. PNAS, 99(3), 7280–7287.

Grimm, V., Revilla, E., Berger, U., Jeltsch, F., Mooij, W.M., Railsback, S.F., Thulke, H-H., Weiner, J., Wiegand, T., and DeAngelis, D.L. (2005).
Pattern-Oriented Modeling of Agent-Based Complex Systems: Lessons from Ecology. Science, 310, 987.

Further Reading: Criticalities and Characterizing Systems.

Dotov, D.G., Nie, L., and Chemero, A. (2010). A Demonstration of the Transition from Ready-to-Hand to Unready-to-Hand. PLoS One, 5(3), e9433.

Kelso, J.A.S. (2021). Unifying Large-and Small-Scale Theories of Coordination. Entropy, 23(5), 537.

DevoLearn (Open-source) Maintenance and Evangelism

 2021 has been a busy year for the DevoLearn initiative. Not only has Mayukh Deb been busy maintaining and generating new versions of the DevoLearn pre-trained model, but I (Bradly Alicea) has been engaging in technology evangelism to advance awareness and involvement in the initiative.  The DevoLearn pre-trained model software (for C. elegans embryogenesis) is now at version 0.3.0, and has garnered 12 contributors making 165 commits (largely since January 2021). Our involvement in Google Summer of Code has bolstered many of these contributions. While our popularity is currently limited, we are trying to spread the word.

To that end, we have presented two versions of a promotional talk on using DevoLearn for facilitating Computational Developmental Biology research and education. The first presentation (DevoLearn: Engaging learners with Computational Developmental Biology) is a flash talk given to the OSF Education Un-conference on Open Scholarship Practices in Education Research, held in February. The second presentation was a longer (15-minute) presentation to the INCF Assembly (DevoLearn: a platform for open Developmental Data Science, Machine Learning, and Education), held in April.

As for developing the broader platform, Ujjwal Singh and I will be working this Summer to develop algorithms for colony morphogenesis and behavior in the Diatom genus Bacillaria. This will be added to the platform in a manner similar to the DevoLearn pre-trained model. In addition to the software development activities, Mayukh, myself, and Krishna Katyal have been the main contributors to the DevoWorm Onboarding Guide. Looking forward to an exciting future!

Update, 8/19:

All of this work has paid off! From the #devolearn Slack channel (OpenWorm Slack team).

The World of Physical Intelligence

As part of the Embodied Intelligence workshop this past week, I saw a presentation by Metin Sitti of the Max Planck Institute for Intelligent Systems on the emerging paradigm of Physical Intelligence. What is Physical Intelligence? It is the intelligent behavior exhibited by motion and takes into account a morphology (embodiment) and action at multiple scales of spatial organization. The talk was wonderful and walked through a number of empirical studies involving both living and non-living systems. I will leave it up to the reader to appreciate all of the points raised in this talk.

Physical Intelligence as an emerging idea.

While Physical Intelligence is not a common paradigm, the general idea is actually not new. Other people have proposed very similar frameworks over the past 15 years or so, including yours truly. Physical Intelligence involves some form of motor or movement behavior, which is generated by an embodied agent, that in turn interacts with the physical environment that can be defined by features such as inertial and gravitational forces, surface textures, and even light energy. 

Physical stimuli for soft robots or other autonomous agents from Figure 1 in Shen et.al, Journal of Materials Chemistry B, 8, 8972-8991 (2020).

At its most human-centric, physical intelligence can be just another word for embodied intelligence, which is where the body shapes and determines what is experienced by the nervous system. In cases where the behaving agent has no nervous system, the body geometry shapes the agent's behavioral output. This is true both in cases of adaptive (intentional) behavior and reactive behavior. In some cases, physical intelligence is identical to Neuromechanics. In other cases, it resembles a range of fields, from Biophysics to Embodied Robotics. The important contribution of the Physical Intelligence paradigm is the principles that guide this diverse topical terrain. Both my talk and Metin's talk provide some of these potential principles, and Scott Grafton's book provides a few more.

One human-centric interpretation of Physical Intelligence (by Scott Grafton).

Metin brings up the example of the Strandbeest, which is a kinetic sculpture with no nervous system or centralized control. This is an example of a purely reactive system that also generates seemingly intelligent behavior. Closer to the human experience is the Passive Dynamic Walker, which produces human-like bipedalism without a central nervous system. This is the essence of the physical: a particular physical configuration can exhibit reactive behavior independently of a central controller.

While serving as part of the peripheral nervous system, and in fact being controlled by a central nervous system, muscles can also play a key role in physical intelligence. While muscle cells can be spontaneously active without being inputs by motor neurons, there is a elaborate coordination between the central nervous system and muscular control. Muscular control can produce both very fast and very slow adaptive movements. In addition, the overall shape of a body in relation to its muscles can constrain the behavior of the agent in question.

The main takeaway is that the brain, body, and environment all work interdependently to shape the behavior that emerges from this complex system. Even in cases where there is no brain (or neural network), the interactions between body and the environment are enough to generate reactive behaviors that appear to be intelligent. This is true for both individual morphologies and the collective behavior of many agents (e.g. swarm intelligence). In fact, the brain can be supplanted by control mechanisms that regulate the conformity and response to physical forces in the environment. Future work should focus on the differences between "neural" and "physical" behavior, as well as the necessity and sufficiency of each component in the triad.

A diagrammatic example of this relationship (with a brain in the feedback loop) from Chiel and Beer, Trends in Neuroscience, 20(12), P553-P557.

In conclusion, DARPA has also engaged in the idea of physical intelligence, and there was a proposer's conference in 2009. This version of Physical Intelligence has a strong cybernetics flavor, particularly in incorporating the EGRT (Every Good Regulator Theorem) into the mix.

Cybernetics of the firm (deemphasizing the role of individual morphology). COURTESY: New World Encyclopedia.

The Way of The Polymath

This content is cross-posted to the Orthogonal Research and Education Lab Medium.

What constitutes a polymath, and why are they so rare? Another way to ask this question is why are there so few foxes relative to hedgehogs? The occasional hyper-specialist would have you believe the term "polymath" is an epithet. However, there are a number of skills that the polymath possesses that translate into an advantage for advancing both theory and fundamental knowledge. Aside from the mastery of multiple intellectual areas, the most important of these is the ability to synthesize information from a number of sources. The advent of digital scholarship may enable this ability in the foxes among us [1].

One depiction (late 19th, early 20th century) of a polymath.

Back in 2015, Nature Careers released a list of recommendations to combat the hyper-specialist tendencies of PhD programs [2], but many of these are simply window-dressing. One view is that improving the state of interdisciplinary thinking is to improve the infrastructure for collaboration and disseminating big ideas. However, a more fundamental (and harder-to- implement) change that can be made is to reconfigure the epistemic landscape of science [3]. One aspect of this indeed involves the training of scientific generalists, but generalist training does not equate polymathism.


The other factor involves the potential zero-sum nature of generalized knowledge [4]. There is a constant tradeoff between deep expertise in one area versus more shallow expertise in a number of areas simultaneously. Society tends to reward deep expertise, and synthesis is rather expensive knowledge-wise. In any case, there is a game-theoretic interpretation of this scenario, but that is a topic for another post.  

Here is a semi-annotated reading list on singular (but multidisciplinary) academic activity:

Hossenfelder, S.   The loneliness of my notepad. Backreaction blog, July 8 (2015).


Issacson, W.   Myth of the Lone Genius. Aspen Journal of Ideas, July 24 (2015). 

These articles critically examine the myth of the lone genius. The first article points out that tools enabling collaboration (e.g. internet, large-scale consortia) are finally starting to bear fruit. Proportion of single-author papers has gone down over last 20-30 years, but that does not mean lone efforts are in absolute decline. In fact, "isolation" is a myth, given the social networks and information-sharing culture in academia.


Bateman, T.S. and Hess, A.M.   Different personal propensities among scientists relate to deeper vs. broader knowledge contributions. PNAS, 112(12), 3653-3658. (2015).


Kirkegaard, E.   Personality correlates of breadth vs. depth of research scholarship. Project Polymath blog, March 6 (2015).

* relates style of scientific investigation to type of contributions (specialized studies vs. broad interdisciplinary synthesis) made by scientists. Survey methodology does not assume that contributions can be both deep and broad, despite setting this up as a dichotomy. Is "deeper" vs. "broader" a major dichotomy in scientific exploration.

* suggests that the difference between scientific generalists and specialists is epistemic, not economic as traditionally assumed (e.g. a certain strategy is more or less risky).

* views differences in types of scientists (e.g. polymathy) as a matter of personality, not epistemic bias.


Palla, G., Tibely, G., Mones, E., Pollner, P., and Vicsek, T.   Hierarchical networks of scientific journals. arXiv, 1506.05661 (2015).

Presents a hierarchical network analysis of scientific journals and their relevance to measuring influence and the diffusion of ideas in specific scientific fields.


Muldoon, R. and Weisberg, M.  Robustness and idealization in models of cognitive labor. Synthese, 183, 161-174 (2011).

* introduce us to an economic optimization model called the marginal contribution/reward (MCR).

* motivation of individuals or groups of scientists is accomplished either through self-interest or epistemic norms.

* MCR assumes that cognitive labor can be optimally distributed across collaborations to solve hard problems. The problem is stated as a constrained maximization of "success" and "return". Model does not provide good approximations of success, return, or epistemic norms, not does it distinguish amongst different scientific skill-sets (generalists vs. specialists vs. hyper-specialists).


Sarma, G.P.   Should we train scientific generalists? arXiv, 1410.4422 (2014).

* how to introduce students to a vocabulary of multiple disciplines, and how this would encourage research breadth.

alexarje   Disciplinarities: intra, cross, multi, inter, trans. Alexander Refsum Jensenius blog, March 12. (2012).

NOTES:
[1] Alicea, B.   "Academic Connectivity and the Future of Scientific Ideas". Synthetic Daisies blog, September 9 (2011).

[2] Nurse, P.   To build a scientist. Nature, 523, 371-373 (2015).

[3] Weisberg, M. and Muldoon, R.   Epistemic Landscapes and the Division of Cognitive Labor. Philosophy of Science, 76(2), 225-252 (2009).

[4] Downey, G.   Interdisciplinarity, sub-disciplinarity, and inter-topicality. Uncovering Information Labor blog, March 31 (2006).