About me

Hi! I’m Ben, a fourth-year PhD candidate in Computer Science at Brown University advised by George Konidaris. My research focuses on the language grounding problem, and I draw inspiration from linguistics, cognitive science, philosophy of mind/language, and semiotics.

I’m honored to be a recipient of the NSF Graduate Research Fellowship and the Brown University Morgan Edwards Fellowship. In 2023, I had the privilege of serving as the lead organizer for the Brown Robotics Talk Series. I was also a founding member of the Brown AI Safety Team.

Research Philosophy

Human perception and action systems induce a rich and highly structured decision process — natural language serves as a medium to express information about this decision process. My earlier works involve highly engineered solutions to the language grounding problem using a formal language for describing Markov Decision Processes as a semantic representation, my newer work leverages multi-agent RL to generate synthetic languages about decision processes.

I maintain a (not always up to date) record of my academic readings, with papers cataloged here and books listed here.

News

• October, 2025: Talks at MIT in the CoCoSci and CPL labs. “A Path to Language Understanding: Grounding Language to Markov Decision Processes”.
• Jul 31st, 2025: I gave my first oral presentation in the Computational Modeling 1 Talks session at CogSci 2025! Here’s a quick summary of the paper on substack.
• Jun 4th, 2025: My first position piece titled “AGI Is Not Multimodal” was published at The Gradient. Here’s a short tweet thread about it, and a Hacker News thread that popped up. It was also written about in TLDR AI. I wrote a short substack post describing how it connects to my broader research arc.
• Apr 24th, 2025: My recent CogSci paper was covered by Perplexity and mentioned in WIRED, after I tweeted about it:

  Why did only humans invent graphical systems like writing? 🧠✍️
  
  In our new paper at @cogsci_soc, we explore how agents learn to communicate using a model of pictographic signification similar to human proto-writing. 🧵👇 pic.twitter.com/3NtveUk4hu

  — Benjamin Spiegel (@superspeeg) April 22, 2025

• Apr 4th, 2025: My CogSci 2025 submission titled Visual Theory of Mind Enables the Invention of Proto-Writing was accepted for oral presentation.

Personal Interests

I enjoy reading, watching movies, playing ultimate frisbee, playing piano, listening to music, and solving Rubik’s Cubes (my personal best for a 3-by-3 is 9.58s, and my best Ao5 was ~12s). I organize and facilitate a bi-weekly Salon-style discussion group for grad students at Brown. I also practice mindfulness meditation, enjoy the Making Sense podcast (though do not agree with Sam on all issues), and was briefly a frequenter of LessWrong.