Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
Mad scientist pro A.I. nutter dresses up like a peace loving Hippy to hood wink …
ytc_Ugw_hzyBH…
G
Then people working at call centers will just keep their jobs. Of course the AI …
ytr_UgwyK5L4T…
G
I already insist that AI not referee to "itself" in subjective terms before answ…
ytc_UgwcvFBIu…
G
Q: "Is the glass half-full or half-empty?"
A: "The glass is twice as big as it n…
ytc_UgyhZiLbg…
G
Several years ago I read an article saying the thing that would defeat driverles…
ytc_UgxwKAj7A…
G
It is not AI. Greed. I work for Brown and Brown and they are doing layoffs qui…
ytc_UgzARj9lT…
G
I definitely think toasters would be the once that need that level of AI first:
…
ytc_Ugh76kssl…
G
We understand how seeing such advanced AI technology in action can evoke strong …
ytr_UgxxDWF-k…
Comment
Big fan of Yuval, but there is some misrepresentation in the way he speaks about how AI would lie from watching humans lie. This is not how AI training/learning works... at all.
If an LLM gets trained with curated data whereby outputs are honest to their inputs the AI will in no way learn to lie by itself. Not specially if a large enough set of cases are represented in the dataset. And specially not if the inputs in the data contemplate the case of observing users/humans lie, and then maintaining a honest attitude. If RL is done strictly in the same way, this, again, will have no chance of happening.
Indeed he has proven himself not to be an expert in AI, as anyone who works with DL will instantly rebate that claim.
This would only be possible in an active learning AI, and all labs have been very reluctant of implementing active learning, specially due to lack of performance. Frozen state AI is the norm for now.
youtube
Viral AI Reaction
2025-06-21T17:3…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | developer |
| Reasoning | deontological |
| Policy | industry_self |
| Emotion | indifference |
| Coded at | 2026-04-27T06:26:44.938723 |
Raw LLM Response
[
{"id":"ytc_UgwMjsmGswYL6LI_mEx4AaABAg","responsibility":"ai_itself","reasoning":"deontological","policy":"none","emotion":"indifference"},
{"id":"ytc_UgyhgJpQWxSHpTALsBV4AaABAg","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"approval"},
{"id":"ytc_UgzbiFb7giNduegeV3J4AaABAg","responsibility":"distributed","reasoning":"mixed","policy":"unclear","emotion":"fear"},
{"id":"ytc_UgwWH9S7Oa7JMdgznoB4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"regulate","emotion":"outrage"},
{"id":"ytc_UgwitQSOYF8xWUbW0HV4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"ban","emotion":"mixed"},
{"id":"ytc_Ugw3KFmBdAWf_cTlNTt4AaABAg","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"fear"},
{"id":"ytc_Ugx_J7i_Vh5miuP0b7J4AaABAg","responsibility":"developer","reasoning":"deontological","policy":"industry_self","emotion":"indifference"},
{"id":"ytc_Ugyo35uvgF9hF8dNSCJ4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"regulate","emotion":"approval"},
{"id":"ytc_UgzuBcMqn7Z9JEY_z2J4AaABAg","responsibility":"government","reasoning":"mixed","policy":"regulate","emotion":"fear"},
{"id":"ytc_UgwTuEXQjnMWWt2k-XZ4AaABAg","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"outrage"}
]