Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
Fun videos! But let’s be real—AI, especially things like ChatGPT, is often overh…
ytc_UgwrrC-M2…
G
@dmquilBut if you're going to be logically consistent, you'd have to admit that…
ytr_UgwGzsj8B…
G
AI will never compare to what the human hands and soul can put to paper.…
ytc_UgxONNLRc…
G
If AI becomes so smart and unpredictable, isn't one of the possibilities that it…
ytc_UgzK9m7tB…
G
Remember the stone age AI that couldn’t even generate words or numbers without m…
rdc_ohyui4q
G
That's a great name, Sophia! Did you know that it means "wisdom" in Greek? Just …
ytr_Ugywqsi-g…
G
Listen to Sundar and read what was show as L's worry. They are in lockstep. Just…
ytc_UgyDuJMo8…
G
I’m not for this at all! They might see AI and be scared to death of it. We need…
ytc_UgwwtxicB…
Comment
AI have been used aggressively in my company but it's best if a smart human operates it, not let me run unsupervised, because it goes crazy. Seriously, without doubt, every time if you run for a while, sometimes not even that long. It'll go on a wrong direction and it's not even able to realize it. The time-to-crazy is random, even simple projects can go there suddenly, so you can't really let it run without looking at it, in fact since it can run so much workload so fast, you'll need to spend even more attention on what's it doing, so it's more exhausting than the one-at-a-time pace that's how leisurely in comparison....
So I don't think you can go to the unsupervised, complete replacement narrative.
The last time 20 percent of work is notorious hard to automate, the last 5 percent can kill your system, the last 1 percent is the definition of professional expertise vs amateur.
How about this, the only thing better than an AI, or human, is AI in the hands of skilled human. This I've seen again and again, like multiplier that's better than either one alone. By orders of magnitude.
So I believe the danger isn't AI taking over the world, is human-AI hybrid, bionic cybernetics, making staying pure human an evolutionary dead end.
That's far more superior, more possible, and far more dangerous.
youtube
Viral AI Reaction
2025-11-23T23:2…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | developer |
| Reasoning | deontological |
| Policy | liability |
| Emotion | fear |
| Coded at | 2026-04-26T23:09:12.988011 |
Raw LLM Response
[
{"id":"ytc_UgzcRfvMkkV_GrE2HBZ4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"none","emotion":"outrage"},
{"id":"ytc_Ugx_e1fGL-lAbU9rcqd4AaABAg","responsibility":"none","reasoning":"virtue","policy":"none","emotion":"approval"},
{"id":"ytc_Ugy040_lcRssMKO7s6F4AaABAg","responsibility":"ai_itself","reasoning":"unclear","policy":"none","emotion":"indifference"},
{"id":"ytc_Ugw8HvhhOCeore95nO54AaABAg","responsibility":"developer","reasoning":"deontological","policy":"regulate","emotion":"fear"},
{"id":"ytc_UgzR62Kip871FOjQyEF4AaABAg","responsibility":"ai_itself","reasoning":"unclear","policy":"none","emotion":"approval"},
{"id":"ytc_UgzPv5sjHKZX36e6qvd4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"unclear","emotion":"mixed"},
{"id":"ytc_UgzOukSBouAAdyPc4dJ4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"none","emotion":"mixed"},
{"id":"ytc_Ugyl_MUgO-VfRb5EDQJ4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytc_UgySCUX5LjVeM-2AgrR4AaABAg","responsibility":"developer","reasoning":"deontological","policy":"liability","emotion":"fear"},
{"id":"ytc_UgwULFuFB8HVyKp7Ekh4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"}
]