Posts

Comments

Comment by anne on LessWrong's (first) album: I Have Been A Good Bing · 2024-04-03T00:09:57.541Z · LW · GW

This is so awesome! I love it! The site integration is awesome too. Gen Z approval stamp (2001).

Comment by anne on Stampy's AI Safety Info soft launch · 2023-10-06T04:45:14.017Z · LW · GW

I am not an expert in AI but I can help with clarity.

I started at the top. The first section, “New to AI Safety” means close to nothing for someone new to AI generally. Is that okay with you? For example, if I read this as if I had no knowledge:

“In recent years, we’ve seen AI systems grow increasingly capable.”

Capable of what? What kind of systems?

“They may soon attain human and even strongly superhuman skill in a wide range of domains.”

Attain superhuman skill? How?

“Such systems could bring great benefits, but if their goals don’t line up with human values, they could also cause unprecedented disasters, including human extinction.”

This is where you would probably lose someone who doesn’t understand AI fundamentals. This might be nitpicky but you did say this is being created for all audiences. Hopefully more helpful than nitpicky.

If the site is designed for someone familiar with fundamentals but not safety, then this criticism is annulled.

Another criticism I have is the way the definitions show up as dotted underline + hover text for each instance of the word. For example, AGI shows up several times in some of the paragraphs—it feels excessive to have so many underlines and hovertexts instead of simply defining the first instance.

Comment by anne on I just watched don't look up. · 2023-06-24T22:50:29.655Z · LW · GW

Thank you for taking the time to explain that! I agree with your assessment that accuracy mattered less than the framing and conclusion of the movie. You’ve convinced me that propaganda was appropriate here.

Comment by anne on I just watched don't look up. · 2023-06-24T05:53:46.279Z · LW · GW

I really like the way you explained “If there was anything like a way to test and verify models of AI-caused extinction, the debate would be unnecessary.”Can you say more about how you define propaganda? I see this word used a lot and I’m often underwhelmed by its applicability.