Raw LLM — Corpus Dashboard

Look up by comment ID

Random samples — click to inspect

G My theory on the 9.9 < 9.11 situation is the training data for an LLM is larg… rdc_mzyafjd G I cant believe we got to a point to saying "Grap a pen and learn" Like it was n… ytc_UgynJm-Mo… G I just think paying for a picture/art is stupid lol an industry that should not … ytc_Ugz9tffd9… G I miss my job as candle maker and horse doo bag manufacturer for horses that pul… ytc_UgzatM4-G… G Here's a question How did they get a building permit to build on your land It's … ytc_UgyMiuWkY… G They act as if feeding something to an AI and telling it to "improve" it is some… ytr_UgwtXUL1u… G AT 1:50 They will increase their already incredible wealth and power exponential… ytc_UgzmIXOF7… G Much of what he says is nonsense. Man simply won’t allow replacing mankind and a… ytc_Ugy2hLkqi…

Comment

@MM-24 I assume you're referring to engineers who are already skilled and experienced, because if they're not, I think we can agree they'll just generate bad code faster with AI, which will make reviewing code take much longer. So assuming they're experienced engineers, I would say from my personal experience, AI only speeds up the manual act of writing code that you already knew you were going to write. You still have to understand the codebase, know the right questions to ask, test your code, debug when it fails, refactor, and plan for future features. In fact, testing alone - like opening up the website, clicking a button, seeing that an event was triggered and was sent to AWS, checking the event store and event data, checking that something was updated in the database, etc - often takes longer than writing the code, especially for complex functionality, and AI can’t help with that, its all ad-hoc and not worth automating because its always changing with business requirements. ts not about replacing coding with reviewing code - its all these other bits no one talks about that make up way more of the engineering process. Trivialising software engineering as "reviewing code" makes sense in tiny side projects or to executives chasing AI hype. That might work in venture capital pitches, but not in real software companies that deliver value.

youtube 2025-03-13T02:3… ♥ 2

Coding Result

Dimension	Value
Responsibility	developer
Reasoning	mixed
Policy	none
Emotion	indifference
Coded at	2026-04-27T06:24:59.937377

Raw LLM Response

[
{"id":"ytr_UgzOVCb5QS591BfcpAd4AaABAg.AF_6_pqsH01AF_CgSXLKLi","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"resignation"},
{"id":"ytr_UgzwMyBhSBcVFwM8uMx4AaABAg.AF_65e0v-CtAFaN8yvJ23t","responsibility":"developer","reasoning":"mixed","policy":"none","emotion":"indifference"},
{"id":"ytr_UgzwMyBhSBcVFwM8uMx4AaABAg.AF_65e0v-CtAFaZLpua44z","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"},
{"id":"ytr_UgzwMyBhSBcVFwM8uMx4AaABAg.AF_65e0v-CtAFajGuZ_qzO","responsibility":"developer","reasoning":"mixed","policy":"none","emotion":"resignation"},
{"id":"ytr_UgxW7CMBtAqvEVQ2jUp4AaABAg.AF_5sDkocobAF_FOt3OrLC","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"mixed"},
{"id":"ytr_UgzCY_-43QpZJ9G6_y94AaABAg.AW0FMDRokCdAW2CURFY0PU","responsibility":"company","reasoning":"deontological","policy":"regulate","emotion":"outrage"},
{"id":"ytr_UgzaNscvSIXaic0adRh4AaABAg.APfCQ2kxOwcAPgVGyjHSgJ","responsibility":"ai_itself","reasoning":"virtue","policy":"none","emotion":"outrage"},
{"id":"ytr_UgyzUfyjRYoi4SiwNzp4AaABAg.APal8jQTiAwAPbB_2q01Ti","responsibility":"user","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytr_UgyZGTogXc6ZhulHbal4AaABAg.AP_h6wtOvTeAP_htHzi4Q7","responsibility":"company","reasoning":"consequentialist","policy":"liability","emotion":"outrage"},
{"id":"ytr_UgyWD4OtZhODqDcp2254AaABAg.AP_9OsNq9fdAPaBh4ml_E3","responsibility":"ai_itself","reasoning":"deontological","policy":"none","emotion":"outrage"}
]

Raw LLM Responses