John Horton
johnjhorton.bsky.social
John Horton
@johnjhorton.bsky.social
Economist researching online marketplaces. Mind and Hand. 🎲🦜- appreciator
March 10, 2025 at 10:42 PM
gave Claude Code (wow!) some work to do writing documentation. It fixed up a file "survey.py" then got interrupted and lost the context. When it went back, it started again with "survey.py" and said...
March 10, 2025 at 6:06 PM
Models you can run on @expectedparrot.bsky.social (w/ just one key); we also hit each one daily to make sure it's not deprecated/down and create this report: www.expectedparrot.com/getting-star...
March 10, 2025 at 5:00 PM
Two contrasting agent perspectives on DOGE:
March 5, 2025 at 12:58 PM
Creating two personas
March 5, 2025 at 12:58 PM
Extracting succinct descriptions of the policy
March 5, 2025 at 12:58 PM
Economic content of the speech last night (w/ GPT-4o & @expectedparrot.bsky.social ) 1/ NB:
www.expectedparrot.com/content/john...
March 5, 2025 at 12:58 PM
Note that I can parameterize the valuations the players have with a 'scenario' & that I can launch a bunch of different valuation scenarios easily:
March 4, 2025 at 3:17 PM
Using skip-logic & piping to implement bilateral bargaining in github.com/expectedparr... Notebook here: www.expectedparrot.com/content/john...
March 4, 2025 at 3:14 PM
EDSL uses these jinja2 templates + variables to figure out the right order for questions to be run in - it makes the API calls as parallel as possible
February 26, 2025 at 12:39 AM
Then, we have the survey. It has a bunch of named questions corresponding to the parts of the original prompt. They are all free text. Note the 'piping' of answers to earlier question to later questions:
February 26, 2025 at 12:39 AM
First, we go get the Senators from Claude - by using a QuestionList, we know we'll get back a list
February 26, 2025 at 12:39 AM
This is very cool & the point is push the model to its limits, but specifying this kind of complex flow is quite easy with github.com/expectedparr... - short thread 1/n
February 26, 2025 at 12:39 AM
What seems to be going on here is that (a) more people in student sample played exact EV in mirror than in lottery & (b) distributions are *both* roughly symmetric around EV. You can see why w/ means it looks PT-ish (& also at other quantiles). FWIW, I think everyone should CTFO & let Oprea respond
February 9, 2025 at 4:35 PM
I did the student sample plot w/ other quantiles & the mirrors look more like the lotteries - only the median is flat.
Seems like waiting to hear from Oprea would be prudent
February 8, 2025 at 3:11 AM
So it can live on here
November 24, 2024 at 2:45 AM
November 21, 2024 at 2:13 AM