Power of Information Fellowships
The purpose of the Templeton Independent Research Fellowship in the Power of Information was to facilitate further research pertaining to the concept of information by exceptional early career researchers who are likely to establish themselves as prominent researchers in their field in the future. The Fellowship allowed the most talented postdoctoral fellows to pursue independent research projects that contributed to the overall goal of the Initiative. Fellowship holders had complete independence over the direction and methods of their research, and thereby got one step closer to becoming established independent researchers who were competitive for subsequent grants and fellowships, or other employment.
Power of Information Fellows
Larissa Albantakis (2016)
Larissa Albantakis is Assistant Scientist in the Department of Psychiatry, and researcher in the Wisconsin Institute for Sleep and Consciousness, at the University of Wisconsin Madison. She obtained her Diploma in physics from Ludwig-Maximilians University in Munich in 2007, and her PhD in Computational Neuroscience from Universitat Pompeu Fabra in Barcelona in 2011.
Larissa’s Fellowship project is entitled Informational Autonomy and Causal Responsibility. This is how she describes it:
“Both common sense and the law treat people as free and responsible agents. Yet, any deliberate action, such as raising one’s hand, is necessarily preceded by a chain of neural events, predicting the action even before the subject becomes aware of having initiated it. As many philosophers have argued, the fact that physical events are either determined by previous events, or emerge from (quantum) randomness, seem to preclude us from being autonomous and causally-responsible agents. My project aims to demonstrate that these seemingly unavoidable conclusions rest on an inadequate, reductionist notion of causation that lacks an informational basis and fails to distinguish “what happened” from “what caused what”. I will develop a theoretical framework to quantify causation at multiple spatiotemporal scales, in order to identify the actual, maximally-informative cause of an event. Combined with a formal account of autonomous systems as maxima of intrinsic, integrated information across space and time, one can determine when an autonomous agent, rather than external events, or the agent’s underlying micro-structure, deserves responsibility for an action and how much. Deliberate, spontaneous, forced, and reflex actions can thus be distinguished based on the origin (intrinsic vs. extrinsic) and specificity (information) of their actual cause. I will demonstrate this using simulated agents (“animats”) with small “brains” having varying autonomy from the environment, for which integrated information, actual causes, and external influences at different spatiotemporal levels can be assessed exhaustively. Adaptive animats will moreover be used to identify environmental conditions that are particularly suited to evolve freely-acting agents.”
Chiara Marletto (2016)
Chiara Marletto is a postdoctoral fellow in the Department of Materials at the University of Oxford. After dabbling in Italian literature, she obtained her Laurea in Physical Engineering from Politecnico di Torino in 2008, and her Laurea Magistrale from the same institution and in the same subject area in 2010. She was awarded her DPhil in quantum computing from the University of Oxford in 2013.
Chiara’s Fellowship project is entitled The Physics of Counterfactuals – Applications of the Constructor Theory of Information to Thermodynamics, the Physics of Time and Quantum Information. In this project, Chiara says, she will further develop Constructor Theory by applying a “constructor-information-theoretic approach to fundamental problems in three new areas: (1) thermodynamics, thereby providing an exact information-theoretic understanding of single-particle heat engines, thermal equilibrium and thermodynamic irreversibility; (2) the physics of time, specifically developing an exact, information-theoretic operational approach; and (3) quantum information, focusing particularly on expressing quantum-information properties, (e.g., entanglement and coherence) within CT’s local, non-probabilistic framework.” Chiara will formulate new information-theoretic laws about ‘counterfactual entities’ in those areas, especially those—such as heat—that have long been considered emergent and not fundamental. This will demonstrate the power of information in fundamental physics, while also establishing CT both as a new field of physics, and—beyond physics alone–as a profoundly positive framework for science, one in which everything that is not physically impossible can be achieved.
Lee Rozema (2016)
Lee Rozema is a postdoctoral fellow in the Quantum Information Science and Quantum Computation Group at the University of Vienna. Lee was awarded the Bachelor of Science degree in physics and computer science from Brock University in 2008, and then completed a PhD in physics at the University of Toronto in 2014.
Lee’s Fellowship project is entitled Experimental Investigation of Indefinite Causal Orders in Quantum Mechanics. He describes the project as follows:
“A causal order determines the order in which events occur, and how these different events can affect each other. Recently, a description of quantum mechanics making no explicit assumption about the order of events was developed. In this formalism, events are treated in terms of the ability to exchange information between each other.
Applying this new formalism to quantum mechanics has shown that processes without a predefined causal order can exist. Information can nonetheless be transferred through these processes. Extraordinarily, these processes can accomplish tasks which are not possible when the causal order is definite. The goal of my project is to investigate these processes in more detail, finding practical ways to build them in a laboratory, efficient ways to characterize them, and new applications for them.
This project will begin with a novel construction of a process—the quantum SWITCH—which lacks a definite causal order. In this process, quantum information will be encoded in a photonic system travelling in a network of optical fibres. The quantum SWITCH will be built using fibre-optic technologies, as this will result in a flexible architecture. I will use this fibre-based quantum device as a test bed to demonstrate that information can be processed and transmitted more efficiently when there is not a well-defined causal order. I will develop theoretical methods to characterize such processes, and will test these methods using the quantum SWITCH. In addition, I will search for other processes lacking a definite causal order.”
Natalia Ares (2017)
Natalia Ares is a postdoctoral researcher in the Department of Materials at the University of Oxford. She was awarded a Masters degree (equivalent) in Physics by the Universidad de Buenos Aires in 2009, followed by a PhD in physics from the Université de Grenoble in 2013. Her recent research has been supported by a Marie Sklodowska-Curie Fellowship.
During her Fellowship Natalia will work on a project entitled Quantum Information Thermodynamics in Nanoscale Devices. She describes the project as follows:
“Maxwell’s demon, born in 1867, still thrives in modern physics. This intelligent being had information about the velocities and positions of the particles in a gas, and could therefore transfer the fast, hot particles from a cold reservoir to a hot one, in apparent violation of the second law of thermodynamics.
In the classical regime, the connection between information and thermodynamics has been studied extensively, but very little is known of how it is revised in the quantum world. The time is right now to dive in this unexplored field; I will draw on the fast-paced development of nanodevices operating at cryogenic temperatures to build an experimental platform to explore how thermodynamics and information relate in a quantum mechanical picture. For example, in classical mechanics and thermodynamics, work is determined by the force needed to move a particle along a trajectory. For quantum systems the situation is much more subtle, since unique trajectories do not exist. Quantum thermodynamics has thus remained somewhat elusive with many peculiar features and open questions. These questions range from definitions of work, for which there is no consensus in the community, as well as unknowns in the thermalisation of quantum systems (approached by quantum information theory), to the efficiency and power of quantum engines.
My project aims at pushing the limits of conventional thermodynamics to include quantum effects. I will develop hybrid nanodevices that combine electronic and mechanical degrees of freedom to give access to thermodynamic experiments at the nanoscale.”
Celia Herrera-Rincon (2017)
Celia Herrera-Rincon is a postdoctoral fellow in developmental and regenerative biology at Tufts University. She was educated at the Complutense University of Madrid, obtaining her Bachelor of Science degree in biology, Master of Science in neurobiology, and PhD in neuroscience in 2008, 2010, and 2014 respectively.
For her Fellowship Celia will work on a project entitled From Microbes to Minds: Using a Neural-Bacteria Interface to Discover a Universal Code for Information Processing Across Scales of Biological Organization. In this project, Celia says, “We will probe the universality of information processing, using prokaryote biofilms and vertebrate brains, the two ‘ends’ of the biological spectrum. Current studies on the brain-gut-microbiota axis have revealed several molecular mechanisms by which the components can communicate, but these studies have largely neglected the informational aspects of this communication. My hypotheses are that (1) there is a universal ancestral communication code, which would allow effective information transfer across biological kingdoms, and that (2) a computational analysis of cross-talk between bacteria and neural cells will reveal unique aspects of how Information-processing underlies emergence of complex systems.
To test these hypotheses, I will design and construct the first integrated electrical-optical neural-bacteria interface, a versatile, multi-site stimulation and recording platform specifically suited to extract information in real time across biological entities. The experiments will validate the health of the two biological components and will verify that we can read (monitor information transfer) and write (provide stimuli) information into the system. I will then characterize the communication channel and generate a quantitative model of cross-level communication by computationally analyzing communication between the two biological components in real time using information metrics. Subsequent experiments will manipulate and re-write active communication between the neural and bacterial components to test and refine our specific models, and integrate the existing data into a wholistic picture of Information dynamics. This platform will enable our laboratory and many others to probe fundamental questions of Information embodiment and action.”
Arnaud Pocheville (2017)
Arnaud Pocheville is a research fellow in the Department of Philosophy and the Charles Perkins Centre at the University of Sydney. He received his university training in France, having obtained a Bachelor of Life Sciences degree at the University of Paris 11 in 2002, a Masters degree in ecology, biodiversity, and evolution at the University of Paris 6 and Ecole Normale Supérieure in 2004, a Masters degree in logic, philosophy, history, and sociology of sciences from University of Paris 7 in 2006, and finally a PhD in life sciences from Ecole Normale Supérieure in 2010.
Arnaud’s Fellowship project is called Biological Information and Biological Function. He describes the project as follows:
“Biologists routinely speak of living systems as if they were information-processing systems and many think this is what distinguishes life from non-life. I and my collaborators previously developed quantitative measures of biological information that move this idea from the realm of metaphor to the realm of scientific model building.
Information has two very different aspects. Formal measures of information concern how objects covary. Here, information is causal. The second aspect of information concerns how objects can have function, meaning, and intentionality. This aspect is semantic. My previous work with collaborators focused on causal information. In this new project I will explore how semantic information emerges from causal information.
This will first require showing how living systems process causal information, and then exploring how this processed causal information can lead to biological function and purpose. Starting with cell biology, I will simulate and measure the transfer and processing of causal information between molecules during cellular functioning. I will evaluate whether more complex, informational objects are more likely to be functional, using data on the catalytic properties of RNA sequences.
At the other end of the biological hierarchy, I will look at how causal information can lead to representation and intentionality in signalling networks. I will use agent-based models developed in collaboration with behavioural ecologists to simulate the evolution of these networks.
The overall aim is to show that biological information per se can bear a biological function. Information gives rise to purpose.”
Fabio Anza (2018)
Fabio Anza is a postdoctoral fellow in the Complexity Science Center in the Department of Physics at the University of California at Davis. He holds a Bachelor’s degree in physics from the University of Palermo, a Master of Science degree in theoretical physics from the University of Pisa, and a D.Phil. in quantum information from the University of Oxford.
Fabio’s Fellowship project is entitled Information Transport in Quantum Materials: Harnessing Dynamics to Master Physical Complexity. He describes the project as follows:
“Transport phenomena are both ubiquitous and heterogeneous. In this project, I will tackle the issue of understanding their phenomenology and providing a unified theory to model their behaviour. This is a key step for the purpose of mastering the dynamical behaviour of complex systems. The ultimate goal of this research is to build a unified theory of the transport phenomena occurring on different physical substrates.
My approach would build on the unifying power of information theory, making a concrete distinction between “carriers of information” (e.g. charge, mass or spin) and the underlying “transport of information” mediated by the carriers. Indeed, so far, the different transport theories (charge, mass, spin and the like) rely heavily on the properties of the carriers. The result of such medium-dependent approach is a set of separate theories with common elements but no unifying structure. Such state of things is far from satisfying and masks many commonalities. I will exploit the universal character of Information Theory to reveal the carrier-independent mechanisms which underly transport. To do so, the tools from Information Theory will be augmented with ideas and results from computational mechanics: a theory that successfully describes how systems process information. While originally developed in the classical domain, its recent success in addressing the structural complexity of quantum states, together with the rise of Quantum Information, makes these theories particularly suitable to address the transport of information in natural systems. Success will establish the power of Information theory to tackle the dynamics of complex physical systems.”
Alec Boyd (2018)
Alec Boyd is a postdoctoral fellow in the Complexity Institute at Nanyang Technological University. He holds a B.A. in physics from Pomona College, and a Ph.D. in physics from the University of California, Davis. Alec’s Fellowship project is entitled Thermodynamics of Computational Structure in Complex Information Processing. He describes it as follows:
“Rolf Landauer set firm lower bounds on the work required to erase a bit, known as Landauer’s Principle. Building on this work, we found new mechanisms of energy dissipation beyond erasure. Complex computations are usually composed of simpler modular components, and this modularity surprisingly leads to excess work costs beyond Landauer’s principle. We called this modularity dissipation. We then proved that thermodynamically efficient modular information processors must match the complexity of their environment -- the principle of requisite complexity. Thus, we saw that, in addition to Landauer’s relationship between information and energy, complexity itself has real, well-defined energetic consequences.
I am collaborating with Professor Mile Gu, an expert in complexity, information theory, and computation at the Complexity Institute in Nanyang Technological University. We are using our modularity dissipation results to understand how to design energetically efficient complex computations. First, to understand fundamental inefficiencies in conventional computing, we are examining digital logic, which is composed of networks of modular elements. Second, we will go beyond conventional computers, extending our previous work on modular thermodynamic computers known as transducers, which can perform much more general computations, as thermal randomness is an essential part of their functionality.
This project will illuminate thermodynamic structural principles of complex information processing. Beyond basic research insights into the foundational physical relationships between information, energy, and complexity, we hope this work will also help us engineer better conventional computers at the nanoscale and better understand nature's nano-computing designs, as in living systems.”
Patrice Camati (2018)
Patrice Camati is a postdoctoral fellow at Institut Néel. He holds a Bachelor’s degree in physics from Universidade Federal do Rio Grande do Sul (UFRGS), a Master’s degree in physics from Instituto de Física Teórica - Univeridade Estadual Paulista (IFT-UNESP), and a Ph.D. in physics from Universidade Federal do ABC (UFABC), all in Brazil.
Patrice’s Fellowship project is entitled Information in Nonequilibrium Quantum Thermodynamics. He describes it as follows:
“The first and second laws of thermodynamics are two pillars of modern physics. The first law is usually stated as the conservation of energy, whereas the second law is stated in a more mathematical way, as the impossibility to decrease the entropy of a system. These laws are fundamental to study all macroscopic phenomena, from phase transitions to efficiency of heat engines.
In the microscopic scale, our world is composed of minute particles swirling around. In this regime the well-known laws of thermodynamics are slightly changed to comply with the behavior of the microworld. In the second half of the 19th century, Maxwell realized that if one had information and control over the particle dynamics, then the second law would be violated. Nowadays, we know that the second law is not violated, but this extracted information used to control the system has to be accounted together with the entropy itself.
In this research project, I will investigate the multifaceted role information plays in the nonequilibrium thermodynamics of quantum systems. This theory of quantum thermodynamics of information will not only study the relation of information to energy and entropy, but entwine information to nonintuitive quantum phenomena, such as interference, entanglement, and the strong influence of the external observer. This theory will elucidate how information is connected to fundamental physics as well as provide the mathematical tools to address the thermodynamics of the second-generation quantum technologies, such as a quantum computer and the quantum internet.”
Alex Ng (2018)
Alex Ng is a postdoctoral fellow at the California Institute of Technology (Caltech). He holds a Bachelor’s degree in statistics and biochemistry from the University of British Columbia, and a Ph.D. in systems biology from Harvard University.
Alex’s Fellowship project is entitled A Synthetic Biology Approach to Multicellular Patterning. He describes it as follows:
“Multicellular development is perhaps the most astonishing process in biology, as it involves the transformation of a relatively simple individual cell into a complex, highly organized structure of trillions of cells in diverse states. To achieve this, cells must continually produce and respond to a constantly changing stream of information in their local environments. Understanding this information processing in natural embryos is difficult because of the cross-talk between concurrent processes, pleiotropic effects of genetic perturbations, and the difficulty of controlling circuit architectures.
Ideally, one would like to be able to design, construct, and analyze a broad range of genetic circuits operating within and between individual cells, and observe what kinds of multicellular patterns each one can produce, and what tradeoffs exist between different designs. Previously, this “synthetic biology” approach to development would have been infeasible. Recently, however, advances in genome engineering, reconstitution of morphogen gradients, and quantitative live cell analysis make it possible.
Here, I propose to develop such a synthetic biology approach to multicellular patterning and use it to engineer cells that are able to self-generate specific spatial patterns de novo through engineered morphogenetic information processing systems. I will design genetic circuits that implement three classic multicellular patterning behaviors, which represent crucial elementary steps that can be combined in the longer term to enable the engineering of more complex spatial structures, resembling digit patterning. This approach should reveal fundamental new information processing and design principles for multicellular pattern formation equally relevant to natural development and tissue engineering.”