AI-Powered Simulation and Human Behavioral Research Are Almost Certainly Complements
Either/or is not the right way to think about using AI simulations
Latest
Disarming the Slop Cannon (ironically, with AI)
Using Expected Parrot to spin up bespoke AI interview + analysis workflows to flag “authors” who can’t explain their own work
A Fake Pharmacy Scenario: From AI to Humans and Back Again (in under 2 hours, using 71¢ worth of tokens)
Using Expected Parrot to riff on a recent WSJ article on how CVS is using digital twins
AI Thinks this Post is Terrible (but I’m writing it anyway)
Using Expected Parrot to propose and prioritize blogging topics
[Nerd Stuff]: Research Surveys should be Version-controlled files
And with Expected Parrot, they are!
We can do that! Best Practices for Dealing with Randomness in Language Model Outputs
How Expected Parrot embodies many best practices out of the box
Twitter Proposes; Prolific Disposes - Part I
Using Expected Parrot to test "promising" Ideas with real humans
Marginal Ideas
Using Expected Parrot to find clusters of opportunities
Your Ideas Are Terrible---Find Out Cheaply, in Private
Your Ideas Are Terrible---Find Out Cheaply, in Private
Using Expected Parrot for "pre-mortem" simulations
A Patriotic Experiment in AI-Powered Research
Expected Parrot can be used for reproducible data labeling.
Test Your Survey on Synthetic Respondents Before You Waste a Single Real One
How to do this easily with Expected Parrot
Newsletter
Designing and Simulating Forecasting Questions with EDSL
How to use LLM-powered agents to generate structured datasets from ranges, rationales, and free-text forecasts
Newsletter
Teaching AI simulations? Get free credits for your students and research projects.
We have ready-to-use materials for teaching LLM-based simulations with our open-source tools, and free credits for getting started. Earn more free credits by sharing your referral code!
Newsletter
How to launch a survey with LLMs and humans
Our tools let you design and run the same survey with AI agents and human respondents, and seamlessly compare results.
Newsletter
Bring your own keys or mix and match
A key feature of Expected Parrot is that you can always use your own keys for LLMs and human participant platforms, or you can use your Expected Parrot key for unified access.
Newsletter
Why AI social simulations will be transformative
The Simulmatics corporation, founded in 1959, aimed to predict the future using simulations (it's right there in the company name). Those simulations could be run on newly available digital computers, informed by the micro-data that was, by then, routinely collected and digitized. Applications were wide and spanned
Newsletter
Do I have time for this?
Our tools look technical and emphasize tinkering over quick answers. Here's why I still think you should try them.
Newsletter
Validate your LLM answers with real respondents
Here's a quick example of methods for generating a web-based version of your LLM survey and analyzing human and LLM responses together at your workspace.
Newsletter
How to retrieve, combine & query your saved results with SQL
Quick methods for accessing and analyzing your automatically stored results from anywhere.
Newsletter
'Whit' Diffie Erasure!
Why human-provided structure improves AI answers to research questions, with examples.
Newsletter
🚀 Over 1 million Q&A cached and counting
The Universal Remote Cache now has over a million unique LLM prompts and responses. You can retrieve them at no cost whenever you replicate research at Expected Parrot.
Newsletter
New features for monitoring your AI research costs
Our tools are designed to make it easy to understand and monitor your LLM usage. Learn more about our latest features for precisely tracking your costs.
Newsletter
I want to see how OpenAI's o3 compares to...
EDSL is an open-source tool that lets you readily compare performance for many language models at once.