I care about effectively employing my efforts in ways that will improve the world.
Currently, I'm researching abstraction in Multi-Agente Reinforcement Deep Learning Agents at CHAI in UC Berkeley. In the past, I've worked to understand the fundamentals of coordination within dynamic systems [1],[2] and learn about search and information retrieval in graph neural networks (GNNs) in the Max Planck Institute [3],[4].
The goal of my research on coordination is to create tools that enhance humanity's cooperation capabilities, for example, by developing software to improve policy decision making. The focus is on developing models that achieve Pareto-optimal outcomes (or similar) for heterogeneous actors in scenarios that are difficult to reproduce in real life.
In the context of GNNs, I developed solutions that improve information retrieval in ontologies in order to provide a reliable source of truth for LLMs.
I'm seeking opportunities to work with organizations that are uplifting to the world.
I've worked as a software engineer with a focus on scientific development for several years. All of this work was within open-source organizations. As a Systems Engineer, I mostly focused in backend and devops work, developing libraries for other developers and scientists. I'm a core contributor to projects such as Jupyter, Firefox, Mamba, Conda Forge, Farama's Zoo, PyScript, PySide, Mozilla Webcompat, Qt, and many others.
Currently I'm affiliated with the Open Knowledge Foundation as an Expert.
In the past, I served as a member of the Jupyter technical committee, as a member of the DISC NumFOCUS steering committee, as a Max Planck Insitute Open Science Ambassator and as a Subject Matter Expert for NASA, developing a course aimed at providing guidance for scientists on practicing open science.
I've also taught open science and software classes to graduate students at Harvard through the LSSTC-DSFP program and at several other conferences, meetups, and groups. I'm also grateful for the Jupyter community having recognized me as a distinguished contributor in early 2022.
Communicating ideas is important, I write about software in my personal blog "Do Cyborgs Dream of Bionic Sheep?" and on my Medium page.
You can access my talks, workshops, classes, and podcasts at this link and check out my hack projects that I tackle alone or with friends here.
As a Cooperative AI researcher I'm interested in understanding what are the underlying characteristics that are important for fostering cooperation among heterogeneous actors.
The reason why I care about furthering this area of research is because I believe it can directly improve people's lives, be it in increasing cooperation between political and economic agents or in improving coordination among humans and machines.
In BarfussLab, I'm modeling the interaction of heterogeneous agents in repeated general-sum game scenarios using reinforcement learning. If you are interested in the overall direction of the research take a look at our recent poster Learning to cooperate with intelligence inequality: a complexity science exploration.
I have a research blog where I write about my experiments.
poverty alleviation:One of the main goals of my work in coordination is to be able to better coordinate markets in a way that benefits people living in extreme poverty. Some initial thoughts are described here but they need to be expanded.
I've started a non-profit for people in developing countries to connect with the global open source ecosystems. Journey offers 3 months long paid programming projects under the guidance of experienced software mentors. Learn more about it at opensourcejourney.org.
I'm currently enrolled in MIT's MicroMasters Program in Data, Economics, and Development Policy because I hope to gain a better understanding on how to be more effective about supporting and developing poverty alleviation policies. I care about support people from unprivileged backgrounds by having a fitting pedagogical ethos to their contexts, teaching farmers about permaculture and supporting them with community and grants and increasing internet access in the world.
open knowledge:I find a lot of significance and love in the open/libre internet. Liberating knowledge and fostering cultural and scientific development deeply resonates with many of my personal principles and working towards making the digital world more accessible and comfortable for all is a very fulfilling mission to me.
My main form of action on this front has been the advocacy of both the ethos and the practice of open source and open science.
2024 update: I'm now highly uncertain about the advantages of open-licensing everywhere. I'm still very much pro-open access manifesto, but this doesn't comprise AI technologies.
governance:Increasing interoperability and people's agency in a world of machines and humans. I keep some notes here.
consciousness: I have several different directions in which I'd like to explore conscioussness. Personally as a meditator, as fan of sci-fi and a functionalist who wants to simulate life in machines and as an AI Safety researcher that's worried about when consciouss life will be actually created inside machines. I write notes about my research here.I flourish by having the support, serendipity and positive feedback loops that a community brings. Do you want to learn or practice something together? Send me a message! If you want, you can also send me anonymous feedback here.
I support several causes: ethical open source, animal liberation, the decrease in animal-based products consumption, investing in carbon offsets to decrease my CO2 footprint, leveraging democracy through open democratic practices, post meritocracy in my engineering work context, combating disinformation on the Web, AI alignment, open access to knowledge, empowering people to overcome poverty using a) non-traditional methods of pedagogy, b) land redistribution coupled with permaculture and community building in rural contexts c) offering OSS opportunities through internships and mentoring d) respecting people's agency over their own economical future through donating money directly to them.
These are some of the projects I support:
I believe women have the right to have full control over their own bodies:
Doctors in Germany are not allowed to post factual information on abortions on their websites. Other websites are allowed to post this information, though. Frag Den Staat is currently providing resources on the subject in German, English, and Turkish.
*AI art:
I believe it's hard to throttle technological development, which is not always good for our culture, institutions, and people. However, I also believe it's quite possible to harness these technologies in an ethical and uplifting way.