Deep Culture - Living with Difference in the Age of Deep Learning (Deep Culture)

Deep learning technologies have taken the worlds of commercial and academic AI by storm. ChatGPT has generated a lot of excitement about our new relations with AI but has also revealed much public anxiety around deep learning. There is too little understanding of the fundamental shift in cultural relations deep learning has brought about. Blanke’s project coins the term ‘deep culture’ to describe the global transformations that deep learning has wrought on culture and how culture is in turn key to deep learning. The project proposes reshaping our relationships with it to address the complexities of cultures and values of difference. 

Validating Biodegradation Rates and Reactions Applying Novel Technologies and Systems Ecology Approaches (ViBRANT-SEA) 

Prof. Linda Amaral-Zettler, UvA professor by special appointment of Marine Microbiology, will head a project based at the Royal Netherlands Institute for Sea Research (NIOZ) in collaboration with the UvA. The project is entitled Validating Biodegradation Rates and Reactions Applying Novel Technologies and Systems Ecology Approaches (ViBRANT-SEA) and will look at the effects of plastic marine debris on the world’s oceans.

Precision Rayleigh jet sprays

Daniel Bonn, professor of complex fluids at the Institute of Physics, has received an ERC Advanced Research Grant of 2.5 million euros. Bonn will use the grant to make the droplet size of sprays much more controllable. This is relevant for a wide range of applications, both in agriculture and medicine.

ADDI: Advancing Digital Democratic Innovation

Computer scientist Ulle Endriss is awarded a ERC Synergy Grant and receives €2.4 for research into using technology to better engage citizens in democratic decision-making. The research team will design technological applications and procedures for citizen participation and assessing the impact of digital platforms. The team is also working to understand the structure of preferences formed by citizens when they participate in digitally supported forms of civic engagement.

Animating Metamaterials using Non-Reciprocity (ANIMETA)

The European Research Council has awarded a three million euro Consolidator grant for the project 'Animating Metamaterials using Non-Reciprocity' (ANIMETA) of UvA physicist Corentin Coulais. With the project, Coulais aims to take inspiration from nature and design synthetic animate materials in the lab.

Using short radio flashes to probe the remnants of neutron star mergers (QuickBlitz)

Rowlinson’s project aims to find an answer to the fundamental question of what happens after neutron star mergers by using LOFAR, the largest low frequency radio telescope in the world. LOFAR is based in the Netherlands and operated by ASTRON (the Netherlands Institute for Radio Astronomy). QuickBlitz will also build a new instrument on LOFAR that will be able to search the whole visible radio sky above the Netherlands for the tell-tale radio flashes.

Towards globally accessible language technology and its alignment to cultural contexts (CulturAL)

Taking a step towards a more inclusive and equitable language technology, Shutova’s project will develop a novel methodology for cross-lingual transfer of LLMs to a wide-range of (low-resource, understudied) languages and dialects, and their alignment to diverse cultural contexts. Her project will, therefore, advance multilingual natural language processing technology, extending its reach to populations currently underserved by it and making it safe for them to use. 

Entanglement Theory: a Quantum Odyssey – From the generalised quantum stein’s lemma to quantum gravity (ETQO)

In quantum computing, the notion of entanglement – the typical quantum phenomenon where disconnected systems can share information that goes beyond traditional 'bits' – plays a crucial role. Quantum entanglement fuels the most important quantum information processing protocols and is central to modern physics. However, the most profound questions concerning its operational meaning remain unanswered. A single key statement connects all of these questions: the so-called Generalised Quantum Stein’s Lemma (GQSL). However, previously Lami has uncovered a fatal flaw in the original proof of the GQSL, and this is now regarded as one of the most important open problems in entanglement theory. His project provides a road map to address all of the above questions in a unified way through the GQSL. 

SpatioTemporal Reconstruction of Interacting People for pErceiving Systems (STRIPES)

How can we teach computers to understand people and their actions, so that they can help people perform tasks? To develop smart systems that can do this, computers must be able to comprehend and process different interactions between people and objects. In his research project Tzionas will reconstruct 3D (or 4D, with time as an extra dimension) representations of what is shown in 2D images by using new techniques. This will be an important step towards the development of new intelligent systems that can better support people in their tasks, such as assistance robots and virtual 3D assistants.

Topology in Mathematics and Physics

Topology is the study of shapes, focusing on persistent features unchanged by deformations. Cheng’s project is inspired by topology’s fundamental role in mathematics and physics. Cheng will adopt an innovative approach facilitated by rapid advances in theories and computing. Employing artificial intelligence (AI), topological invariants will be decoded, revealing hidden correlations. AI-driven computational tools for systems will be created with non-trivial topology, set to impact fields from mathematics to fundamental physics and engineering.

Neural ideograms: shaping AI with geometry-grounded learning

This project aims to develop artificial intelligence (AI) systems that learn and utilize 'neural ideograms'—geometric symbols for abstract concepts. These symbols will function similarly to how pictograms and ideograms simplify and convey complex information, thus enabling AI to interpret and organize complicated data through geometric constructs efficiently. Recognizing that most data is rooted in our physical world—and inherently grounded in geometry and physics—it's clear that neural ideograms must preserve this grounding to reality in order to be meaningful. 

Living levers: unveiling the adaptive and active elasto-viscoplasticity of living systems

The materials from which life is built are typically soft, adaptable and active. In order to study them, researchers need to model materials with similar properties. Based on new insights into a complex material class called elasto-viscoplastics, which includes toothpaste, Jalaal's project develops a new class of active elasto-viscoplastics. These materials could mimic natural life processes like cell migration and embryo development.

Small fibres, big legacies?

Our clothes release very small plastic fibres (microfibres) during washing, which break down extremely slowly in the environment. We do not know what happens if they continue to accumulate over a very long time in our rivers. In this project, researchers will use a unique dataset on microfibre release from textiles collected by citizen scientists, to predict current and future emissions of microfibres into Dutch rivers. Combining models and experiments, they will investigate how many microfibres are needed to disrupt natural sediment processes. Interactive maps will visualise microfibre accumulation in the Netherlands under different scenarios.

Improving social media with Artificial Intelligence

Social media have become invaluable spaces where we share ideas and debate important societal issues. But our current social media platforms make such productive conversations difficult by amplifying conflict and driving polarization. How can we redesign social media to help foster more productive political conversations? With the recent emergence of AI chatbots like ChatGPT, we finally have a way of answering this crucial question. This project creates synthetic social media platforms with thousands of interacting chatbots, allowing researchers to experimentally test how different platforms shape political discourse. This will help us design social media with more positive societal outcomes.

PLASTIC-JUNC: Predicting nanoPLASTIC risks to animal JUNCtional protein integrity

Nanoplastics are specks of plastic a million times smaller than a millimeter. Recently, researchers found nanoplastics in our food, organs and even in the placenta. Nanoplastic accumulation can lead to organ failure and reduced coordination and memory in different animals. Scientists do not yet understand how nanoplastics trespass the protective cell layer around our organs, the endothelium. Pérez will use computer simulations to discover how nanoplastics weaken key adhesive, i.e., junctional, proteins in the endothelium. Using artificial intelligence Pérez can obtain insights about these processes and identify the most damaging nanoplastics; providing valuable information for biomedicine and public health.

Stereoselective Synthesis of Amines via Photoexcitation of Enzyme-Substrate Complexes
Amines are crucial building blocks in the pharmaceutical and agrochemical industries but controlling their geometry during synthesis, especially when multiple stereocenters must be formed simultaneously, is intricate and resource-demanding, thereby affecting practicality and economic feasibility. Tseliou will use light to unlock new reaction pathways within the active sites of enzymes, enabling the one-step synthesis of amine compounds featuring multiple stereocenters by using a single catalyst. Unlike established photoenzymatic processes, the new artificial photoenzyme will operate without the need for any cofactor, thus opening new avenues for reaction invention.

A diagrammatic toolbox for quantum circuit simulation

Van de Wetering will use a language of diagrams, the ZX-calculus, to unify different approaches to simulating quantum computations, which will allow for larger simulations. He will then use the connection of quantum simulation to condensed-matter physics and logical-formula solving to find areas where these improved techniques will lead to better algorithms. In this way we bring these fields closer together.

Neuromodulatory weighing of prior beliefs in perception and decision-making

Predictions inform almost all our decisions. For example, when choosing the bike over the bus, based on the predicted duration and weather conditions. But what happens when unexpected changes in the world render our predictions invalid? For example, nearby roadworks result in a longer-than-predicted trip duration. This research project uses a new and multidisciplinary approach to uncover the brain circuits that mediate the flexible updating of prior beliefs. This will lead to fundamental insights about consciousness and decision-making as well as applications in the realm of medical psychology.

Aiming for the massive stars: understanding the progenitors of black holes and supernovae in the Gaia, LIGO, and LSST era

Through their strong winds and explosions, massive stars enrich the Universe with the building blocks of life, including the carbon in our cells and the oxygen we breathe. However, not every massive star explodes; some form black holes. Moreover, we now know that most massive stars live with close companion stars that can change their fate and could lead to universal signals. With the recent discovery of gravitational waves, stars with invisible companions, and automated observations of exploding stars, astrophysicists can finally test the existence of these signals and better understand the final fate of massive stars.

Disorder put to good use

Quantum effects are fascinating, but can only be observed at extremely low temperatures. One approach to push the quantum world to higher temperatures is to use a newly discovered class of crystalline materials: magnetic topological insulators (MTIs). In Anna Isaeva's project, the search is on for new MTIs with exactly the right properties to be used for next-generation spintronics applications, energy-saving and quantum technologies. Where traditionally disorder is the enemy of crystalline materials, the researchers will exploit intermixing disorder, swapping magnetically active atoms around in the crystalline lattice to create robust designer magnets that accomodate quantum effects.

What is the origin of the Universe?

In an attempt to address this question, the research project that Daan Meerburg and Christoph Weniger propose will use clever and very advanced computer algorithms. By simulating and analysing radiowaves from the early universe, the researchers will explore whether these algorithms can distinguish a signal from the early universe from the noise and provide clues about its origin. Results from the work will be of direct importance to existing and upcoming radio experiments with direct involvement of the Dutch astronomy community.

Trichome truckers: how self-defence metabolites are transported and stored in tomato gland cells

Plants produce a large variety of specialised metabolites involved in stress response and interaction with the environment. Wild ancestors use specialised metabolites to protect themselves against pest and pathogens. In tomato self-defence metabolites are produced and stored in special glands, glandular trichomes. It is not yet known how they are transported from the cells that make them into the storage cavity for accumulation until needed. Bleeker and her team will elucidate this defence-metabolite transport. Understanding how wild tomato ancestors do this, might enable breeders to rearm cultivars with their own defence metabolites reducing their reliance on chemicals.

Reputation as a new route to cooperation in multi-agent reinforcement learning.

Artificial agents are likely to face the same dilemmas of cooperation that humans evolved to solve. To design efficient systems where collective gains are maximised, artificial agents must learn to forego their self-interest and spend effort to help others. How to design adaptive agents that autonomously learn to cooperate? In this project we study how to use reputation systems to sustain cooperation among agents adapting through trial-and-error (that is, through reinforcement learning). We will explore how reputations should be assigned to sustain cooperation in the long-run, and how to design reputation systems in increasingly complex environments.

The mystery of the missing mixing angle

Quarks are elementary particles that combine to form atomic nuclei. Quarks come in six flavors and it is possible for one quark flavor to mix into another, causing nuclear radioactive decays. Our current best model of how nature works at the smallest scales predicts how this mixing should work, but these predictions disagree with the most precise experiments. In Jordy de Vries' proposal, new calculational methods are developed to resolve the quark mixing mystery and to find out who or what is the culprit behind it.

Compost effects: balancing Carbon, Microbes and Pesticides in African soils

The increased usage of compost, following the shift towards more sustainable soil management in agriculture, shows promising benefits for increasing yields. However, the effects of its application are not fully understood. It may introduce pollutants and alter existing soil dynamics. In this study, focusing on Kenya’s rose cultivation, we will research the long-term effects of compost addition on carbon stabilization, pesticide dynamics, and bacterial communities. By combining (ongoing) field- and lab experiments we want to provide insights for sustainable composting management, aiming to improve global practices.

To realize cancer treatments with reduced side effects, I will study a new synthetic methodology based on shuttle catalysis to locally activate drugs. This methodology allows the conversion of inactive drugs into active anticancer drugs. The local performance of this conversion in cancer cells will help to prevent side effects.