I am trying to improve my approach to AI Security Research and could use some practical advice.
I am particularly interested in customisation and cost control. I would like to compare approaches that people have actually used rather than just theoretical suggestions.
What would you tell...
