Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
You (and many skeptics) are missing two things. First, the rate of improvement: it's not just about what tools can do today, it's about extrapolating to where they will be in less than five years. [This graph](https://x.com/GarrisonLovely/status/1902450811835482253) might be a good intuition primer. Second, it seems you aren't even using current tools. Compare claude 3.7 (early 2025) to gpt-4o (early 2024) on the same tasks.
reddit AI Jobs 1743027482.0 ♥ 1
Coding Result
DimensionValue
Responsibilitynone
Reasoningunclear
Policynone
Emotionapproval
Coded at2026-04-25T08:33:43.502452
Raw LLM Response
[ {"id":"rdc_mjv8s61","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"rdc_mjw22wt","responsibility":"user","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"rdc_mjwgzbi","responsibility":"company","reasoning":"consequentialist","policy":"unclear","emotion":"fear"}, {"id":"rdc_mjwlckv","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"resignation"}, {"id":"rdc_mjww0wm","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"} ]