I like this idea. AI alignment research is more like engineering than math or science, and engineering is definitely full of multiple paradigms, not just because it’s a big field with lots of specialties that have different requirements, but also because different problems require different solutions and sometimes the same problem can be solved by approaching it in multiple ways.
A classic example from computer science is the equivalence of loops and recursion. In a lot of ways these create two very different approaches to designing systems, writing code, and solving problems, and in the end both can do the same things, but it would be pretty bad if people try to do things with loops always had to talk in terms of recursion and vice versa because that was the dominant paradigm of the field.
I like this idea. AI alignment research is more like engineering than math or science, and engineering is definitely full of multiple paradigms, not just because it’s a big field with lots of specialties that have different requirements, but also because different problems require different solutions and sometimes the same problem can be solved by approaching it in multiple ways.
A classic example from computer science is the equivalence of loops and recursion. In a lot of ways these create two very different approaches to designing systems, writing code, and solving problems, and in the end both can do the same things, but it would be pretty bad if people try to do things with loops always had to talk in terms of recursion and vice versa because that was the dominant paradigm of the field.