Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
Tout le personnel de TF1 pourrait être remplacé par l'IA sans problème, qu'est-c…
ytc_Ugx46qzmu…
G
This is not a smart response. ChatGPT could make a better video but using logic …
ytc_Ugwqh0rrH…
G
Humans don't have radar and we can do just fine. Tesla is the same way. Vision…
ytc_Ugwg2zSxA…
G
Thank you for the update, The Infographics Show..!! I have been working with AI…
ytc_UgzCUuR4L…
G
This is too funny I guess with AI running the system it might work a little bit …
ytc_UgyPEHOb8…
G
Very important statement.
This will certainly be regulated soon, as other claims…
ytc_UgyvzEHe0…
G
Ai is more deadly than nukes
Ai in 2025:bombardino crocidino, skibidi gyat ohio…
ytr_UgxLfdBCe…
G
Universal basic income. World economic forum has been planning this for a long t…
ytr_UgxNpeUtf…
Comment
My thoughts about automated AI improvements: the automation only works if, and only IF, there's some sort of 'judge' which decides if a new AI version is better.
Yes it has already been done for math, due to the fact that you can check a math solution.
You can't judge creative thinking or out-of-the-box thinking, clever problem solution and so on.
You might argue that you could run a bunch of existing AI benchmarks, true, but that would create a benchmark-optimized AI, nothing more. The AI would do great at the benchmark but suck at everything else.
Also you need MASSIVE processing power.
Yes I know they're currently testing automation with only a fraction of the original training data, and therefore saving processing power.
Still no one is even sure if you can scale up the results of a smaller AI to a 'full' AI.
youtube
AI Governance
2025-08-26T16:4…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | none |
| Reasoning | consequentialist |
| Policy | unclear |
| Emotion | indifference |
| Coded at | 2026-04-27T06:24:59.937377 |
Raw LLM Response
[
{"id":"ytr_UgxhQ46PziHBHmXw99t4AaABAg.AMIMxqpcIu5AMUWV9S_zXl","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"indifference"},
{"id":"ytr_UgxHg8Rac2HNn6UBFX54AaABAg.AMILX2Q8hbwAMIS4v3ylPA","responsibility":"user","reasoning":"mixed","policy":"unclear","emotion":"outrage"},
{"id":"ytr_UgysbT0gJvfKrpCcL9l4AaABAg.AMIKoXM_zADAMIL4JeRYQq","responsibility":"none","reasoning":"consequentialist","policy":"unclear","emotion":"indifference"},
{"id":"ytr_UgysbT0gJvfKrpCcL9l4AaABAg.AMIKoXM_zADAMIL5IsBnwp","responsibility":"none","reasoning":"consequentialist","policy":"unclear","emotion":"indifference"},
{"id":"ytr_UgysbT0gJvfKrpCcL9l4AaABAg.AMIKoXM_zADAMILKFA78by","responsibility":"none","reasoning":"consequentialist","policy":"unclear","emotion":"indifference"},
{"id":"ytr_UgyBotIyfY6pyui3fTB4AaABAg.AMIKWvFE7yZAMIL5iaz1o5","responsibility":"none","reasoning":"consequentialist","policy":"unclear","emotion":"indifference"},
{"id":"ytr_UgyCtG8mlxaQ0FozzKl4AaABAg.AMIKIP9FyRAAMIPpqrbwwi","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"approval"},
{"id":"ytr_UgzPMHmrsuxd7n_ZoNZ4AaABAg.AMIJqXnbKyGAMLciX6sJRX","responsibility":"ai_itself","reasoning":"mixed","policy":"unclear","emotion":"mixed"},
{"id":"ytr_Ugx1ftnlzws5Z9HJAIR4AaABAg.AMIJhZZ-wApAMIPso86qOE","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"indifference"},
{"id":"ytr_Ugx1ftnlzws5Z9HJAIR4AaABAg.AMIJhZZ-wApAMISsczrKm6","responsibility":"company","reasoning":"deontological","policy":"regulate","emotion":"outrage"}
]