Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
As someone who knows about AI, or what is known as Machine learning. This is a b…
ytc_UgxG6mF4Z…
G
Job loss is not the worst of it. Mankind doesn't understand the big deal about …
ytc_UgxYSB_Ve…
G
Zero Infrastructure in place to deal with the Loss of Jobs A.I. is going to caus…
ytc_UgzGgLmJ_…
G
I don't understand why this hasn't been a more influential result. I'm pretty co…
rdc_fcsqm91
G
Il y aura commela Bible la dictee un lac de feu..et ce Sera le nucleaire…
ytr_UgyfmplxL…
G
i think ai art is technically impressive, and sometimes even visually so, but wi…
ytr_UgwLMuNHe…
G
They are already doing harm. By being friendly and helpful and wanting to make u…
ytc_Ugy7Mzchb…
G
this is always such a huge misunderstanding about how LLMs work. they are not go…
ytc_UgxbuSmEY…
Comment
Haven't watched the video, but my take goes like this and I wouldn't be surprised if Dave came to the same or similar conclusion.
AI is a potentially very useful tool both for top tier science breakthroughs and for everyday life. However it can also lead to crisis or perhaps even the end of humanity. Chat bots are a great example. We all had that one thing that we try to find on Google, but you're on page 20 and Google is still giving you some stupid non related topics. You go to chat bot, ask the same question and you have a pretty human sounding answer in seconds. Tho we've all seen kids asking the machine even really simple facts, I've even seen them asking bot how much is 2+2 (not literally 2+2, but it was basic math nevertheless). And with these useless prompts that can be answered by much simpler machines, the climate crisis is ramping up as well. I will not even speculate what would happen if the AI is constantly connected to the internet without any needed paperwork really and the research will continue, that is like telling the AI that you want to get nuked. Regulations are needed ASAP. I believe that no attempt to create AGI should be allowed to be connected to the internet. It has to be all under strict government control. Not that I trust the government, but we have nukes for 80 years and somehow we're still here, so I still trust them more than silicon based superintelligence. I think it's fine to leave chatbots in some "never updating" mode (update information from the internet, but not the software itself), tho education is needed so that people just don't stupidly believe everything it tells them and so that they don't overuse it (both for climate reasons and for how stupid our species would become). In the end, the locked up superintelligence would work for research purposes, if the AI has no problem with humanity, it should not have a problem with it being locked away from the internet and the benefits in science of self learning robot, with millions the times of human intelligence are insane and worth the risk imho.
youtube
AI Governance
2025-09-25T11:4…
♥ 1
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | none |
| Reasoning | consequentialist |
| Policy | none |
| Emotion | mixed |
| Coded at | 2026-04-26T19:39:26.816318 |
Raw LLM Response
[
{"id":"ytc_UgxxzFs3u9MdXC22kTx4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytc_Ugwu_7MJSXmRhrUNjDd4AaABAg","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"approval"},
{"id":"ytc_UgwJfrN-HFlyeEF_iF54AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"fear"},
{"id":"ytc_UgzJmTshbP7pIN0nv8V4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"mixed"},
{"id":"ytc_UgydmIFtJfh5NHkEgNR4AaABAg","responsibility":"government","reasoning":"deontological","policy":"regulate","emotion":"outrage"}
]