technologyaiengineering
Notes on Building with LLMs: From Prototype to Production
Guardrails, evals, observability, and product loops that actually stick.
Read more →
Michael Pocress
I'm an AI Researcher, based in New York.
I'm passionate about developing responsible and interpretable AI systems and products to drive sustainable progress.
Featured Work
Here's some of my recent work—projects I've built, writing about AI, data, and how technology shapes the world, and resources you may find helpful.
Projects
Modeling Global GDP Determinants Using Health and Economic Indicators
Analyzes data from 179 countries (2000–2015) to identify which health and economic indicators are most closely associated with GDP per capita, using log-transformed regression models and mixed-effects analysis.
Data Science / ResearchRPythonStatistical ModelingRegression AnalysisData Analysis
Completed
Multi-Agent LLM Opinion Dynamics Simulator
A simulation system where multiple LLM-based agents with different political leanings and personalities discuss topics and update their opinions over time, revealing limitations in how current LLMs behave as social agents.
AI + ResearchPythonLLMsMulti-Agent SystemsSimulationNLP
Completed
AI & the Labor Market Dashboard / Job Automation Risk Map
Uses labor statistics and task-level data to score occupations by AI exposure. Visualizes risk levels, job changes, and emerging AI-driven opportunities.
AI + Careers / Labor MarketNext.jsTypeScriptPythonD3MapboxBLS APIONETEmbeddings
Writing
aitechnologypromptingllm
Are You Promoting the Language Model Correctly?
…Or Is Your Prompt Holding You Back?
We spend a lot of time talking about how to 'use' language models—but an equally important question often slips through the cracks: Are we promoting the language model correctly?
Read more →
roboticsautomationubieconomicstechnologypolicy
Robotics, Falling Costs, and Why UBI May Become Increasingly Relevant
Exploring how falling robotics costs and automation may make Universal Basic Income increasingly relevant as we enter an era where machines can perform tasks more cheaply and reliably than humans.
Read more →
Resources
EducationAI SafetyAlignmentCurriculumFellowship
AI Safety and Alignment Curriculum
An 8-week structured curriculum introducing fellows and students to AI Safety and Alignment.