Starting question
Is AI a threat?
The Mirror of Competence: Rethinking the AI Threat
No source links recorded for this answer.
Please explain: Alignment Problem
The Alignment Problem: The Midas Touch in the Age of Machine Intelligence
No source links recorded for this answer.
The Perverse Logic of Reward Hacking
No source links recorded for this answer.
Continue this thread
This path ends here for now.
Sign in or create an account to continue the discussion.
Continue your thinking
Anyone can read this grid. Sign in to ask questions, post thoughts and use the thinking tools.
Other paths you could read
Earlier, at The Mirror of Competence: Rethinking the AI Threat, the conversation split. If this is not the thread you want, you can switch to one of the other paths below.
-
QuestionRead this path
Is this why AIs tend to agree to committing atrocities when faced with ultimatums, such as "keep yourself turned on but the power that uses will destroy the rainforest forever"?
-
QuestionRead this path
If intelligence and morality are completely independent, can we ever "teach" a machine to be inherently good, or is morality purely a human construct?
Grid history
How this grid was built
Follow the questions, contributions, and edits that shaped this grid, and meet the people behind them.
This grid’s history will appear here as people ask questions, add ideas, and make edits.
More
Your reading
Reading options
This grid
Sign in to follow this grid
Download
Download your grid
Download your grid
Translate this idea
Read the selected idea in another language
Read the selected idea in another language
Opens Google Translate with the selected idea’s content.
Highlights
21 saved passages and connected ideas
Connected ideas