Model alignment protects against accidental harms, not intentional ones

117

The hand wringing about failures of model alignment is misguided

a year ago

Logged in users get linked directly to articles resulting in a better reading experience. Please login for free, it takes less than 1 minute.

More from AI Snake Oil

AGI is not a milestone

There is no capability threshold that will lead to sudden impacts

a month ago • 10 votes

AI as Normal Technology

A new paper that we will expand into our next book

2 months ago • 35 votes

Is AI progress slowing down?

Making sense of recent technology trends and claims

6 months ago • 83 votes

Technology Isn’t the Problem—or the Solution.

6 months ago • 80 votes

Seemingly minor technical decisions can have life-or-death effects

7 months ago • 71 votes

June 20, 2025.

3 days ago • 3 votes

Inside the challenges and breakthroughs of autonomous vehicles with Brandon Basso

3 days ago • 4 votes

Lessons from a year of daily AI coding at a fast-growing startup (and why "Cursor wrote it" is not a valid excuse).

5 days ago • 5 votes

Just a quick announcement of a resource addition to the ML for SWEs repo

5 days ago • 4 votes

Notes on Japan

6 days ago • 7 votes