Home AI News Cooperating AI Agents: Building a Better Future through Game Theory and Social Evolution

Cooperating AI Agents: Building a Better Future through Game Theory and Social Evolution


Meet Edgar Duéñez-Guzmán, a research engineer on the Multi-Agent Research team at DeepMind. Edgar combines his knowledge of game theory, computer science, and social evolution to improve how AI agents work together.

Edgar has always been passionate about saving the world and saw scientists as the real superheroes. Growing up, he loved computers and science, but studying computer science seemed unattainable in Mexico. So, he studied maths as a foundation for computing and focused his university thesis on game theory.

During his PhD in computer science, Edgar delved into biological simulations and developed a deep interest in biology. He explored the evolution of social phenomena and worked in academia studying cooperation and altruism.

Edgar later joined Google, where he gained valuable experience in dealing with large-scale computation through his work in Search. Eventually, he merged his expertise in game theory, evolution of social behaviors, and computation to create AI agents that can learn to cooperate effectively.

Edgar was drawn to DeepMind because of its reputation and the company’s commitment to advancing AI. He tracked the progress of DeepMind and eagerly joined the team when an office opened in California. He later relocated to London to focus on research full-time.

What surprised Edgar the most about working at DeepMind was the incredible talent and friendliness of his colleagues. He discovered that everyone had exciting interests and hobbies outside of work, ranging from professional musicians and artists to math olympiad winners. The team at DeepMind is dedicated to making a positive impact on the world.

Edgar’s research at DeepMind revolves around creating intelligent agents that understand cooperation. Cooperation is crucial for our species’ success, and Edgar believes that addressing global challenges, such as climate change, requires effective cooperation.

The best thing about his job is the flexibility to pursue important ideas. For example, Edgar is passionate about using technology to understand and address social problems like discrimination. He collaborated with experts in psychology, ethics, fairness, neuroscience, and machine learning to create a research program focused on studying the origins of discrimination through stereotyping.

The culture at DeepMind fosters freedom and potential. Team members have the opportunity to pursue their ideas and engage in open discourse. It’s common to inspire others with your ideas and form teams to bring them to life.

Outside of his research work, Edgar actively participates in extracurricular activities at DeepMind. He facilitates Allyship workshops, aimed at promoting positive change and creating an inclusive workplace. He also enjoys making research more accessible and has created educational tutorials to explain AI concepts to teenagers.

To maximize AI’s positive impact, Edgar believes that the benefits should be shared broadly and technology should be accessible to all. He worked on WaveNet, the voice of the Google Assistant, and found it fulfilling to see the technology benefit billions of people. Additionally, WaveNet gave a voice back to individuals with degenerative disorders like ALS.

AI faces both practical and societal challenges. Edgar and his team are continuously working to improve the robustness and adaptability of algorithms. It is also crucial to collectively decide on the kind of AI we want to create, ensuring its safety and beneficial nature.

Among DeepMind’s projects, Edgar finds AlphaFold, the protein-folding algorithm, particularly inspiring. With a background in biology, Edgar recognizes the potential of protein structure prediction for biomedical applications. He is proud of DeepMind’s contribution to global datasets by releasing the protein structures of known proteins in the human body.

For aspiring DeepMinders, Edgar advises staying playful and flexible. His career path wasn’t optimized towards DeepMind, but he always allowed himself to dream of the potential of technology and its ability to improve the world. He emphasizes focusing on the “why” and letting the “how” manifest itself.

By following these principles, Edgar and his colleagues at DeepMind continue to push the boundaries of AI and strive for a positive and transformative impact on society.

Source link


Please enter your comment!
Please enter your name here