Blake Elias

LinkedIn | CV | Email | Twitter | Google Scholar

About

I am a researcher at the New England Complex Systems Institute, working under Yaneer Bar-Yam. My current work examines the game-theoretic and collective decision-making aspects of pandemic response.

My interests include machine learning, game theory and neural-symbolic reasoning. I am interested in applying AI for social good, via tools that amplify human collective intelligence and group decision-making, and via the design of better economic policy and social systems.

Previously, I was an AI Resident at Microsoft Research working on neural-symbolic computation and human-AI collaboration. Before this, I completed my SB and MEng degrees at MIT in Electrical Engineering and Computer Science, where I worked in the Media Lab and Synthetic Biology Center.

Recent Writing

Multi-Agent Cooperation via Intrinsic MotivationMarch 2, 2021

Intrinsic Reward, Biological Utility, and Saving the PlanetMarch 1, 2021

Open-Endedness, Multi-Agent Learning and Existential RiskOctober 10, 2020

Should the World Stay Home for a Month?March 11, 2020

 

See more...

Publications

Pandemic Response as Reinforcement Learning
Blake Elias, Alexander Siegenfeld, Yaneer Bar-Yam.
NeurIPS 2020 Workshop on Machine Learning for Economic Policy.
paper | code
Human-Machine Collaboration for Fast Land Cover Mapping
Caleb Robinson, Anthony Ortiz, Kolya Malkin, Blake Elias, Andi Peng, Dan Morris, Bistra Dilkina, Nebojsa Jojic.
Thirty-Fourth AAAI Conference on Artificial Intelligence, 2020. Oral.
pdf | arXiv | code | video
High Throughput Pin-Tool Based Automated DNA Assembly
Master of Engineering Thesis, Electrical Engineering and Computer Science, MIT, 2018.
pdf | MIT DSpace

Covid

Pandemic Response as Reinforcement Learning
Blake Elias, Alexander Siegenfeld, Yaneer Bar-Yam.
NeurIPS 2020 Workshop on Machine Learning for Economic Policy.
paper | code
Is Locking Down Worth It?
Blake Elias, Alexander Siegenfeld, Yaneer Bar-Yam.
Pandemic Lockdown Decision-Making
Bruno Felix, Chris Moreno-Stokoe, Sascha Frankel, Lennart Quispel, Ana Martinovici, Blake Elias, Yaneer Bar-Yam
Respiratory Health for Better COVID-19 Outcomes
Blake Elias, Chen Shen, Yaneer Bar-Yam. March 16, 2020.
New England Complex Systems Institute.
pdf | web
Should the World Stay Home for a Month?
March 11, 2020.
pdf
Could Air Filtration Reduce COVID-19 Severity and Spread?
Blake Elias, Yaneer Bar-Yam. March 9, 2020.
New England Complex Systems Institute.
pdf | web

Coding Projects

Neural-Symbolic Program Synthesis
poster
Toski (Scheme Debugger)
Kenneth Friedman, Blake Elias, Jared Pochtar
code | report | blog
Image credit: wired.com - 'Want to Remember Everything You'll Ever Learn? Surrender to This Algorithm' -- https://www.wired.com/2008/04/ff-wozniak/ Homer: Spaced Repetition Learning
code | report
Zero Robotics competition (team founder)
Team site | code (2010) | code (2011) | video (simulation) | video (ISS run)
FIRST Robotics Competition
code (2010-2011) | code (2012) | video
3D Rendering from Scratch in C
code | gallery

Writing

Intrinsic Reward, Biological Utility, and Saving the Planet
The Case for Open-Ended AI
Open-Endedness, Multi-Agent Learning and Existential Risk
Image credit: α-Rank: Multi-Agent Evaluation by Evolution -- Shayegan Omidshafiei, Christos Papadimitriou, Georgios Piliouras, Karl Tuyls, Mark Rowland, Jean-Baptiste Lespiau, Wojciech M. Czarnecki, Marc Lanctot, Julien Perolat, Remi Munos -- https://arxiv.org/pdf/1903.01373.pdfGradient Descent and Chain-Linked Systems
Image credit: [1] Vincent Dumoulin, Francesco Visin - A guide to convolution arithmetic for deep learning - https://arxiv.org/abs/1603.07285Softmax-Weighted Average Pooling
Source: https://arxiv.org/abs/1808.00508Machine Learning for Program Synthesis
On Hallucinatory Intelligence
My Path to AI Residency (Microsoft Research)

What I'm Reading / Watching

Problems of Knowledge and Freedom

Design as Participation

Behavior Without Utility