Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
I'm not as familiar with ASI-Arch, but AlphaEvolve is a sensible incremental step forward from topology inference towards automated problem solving. My position on AGI, and my own limited work on the subject, has been that building an algorithm that is itself AGI is very challenging. Rather the best approach it to build an algorithm that can build a specialized algorithms to solve a problem and so in a sense be a meta-AGI. My main research line is algorithm and model inference. I've never really pitched it as an AGI candidate, but the possibility (however slim) is there.
reddit AI Jobs 1754748556.0 ♥ 12
Coding Result
DimensionValue
Responsibilitynone
Reasoningunclear
Policynone
Emotionapproval
Coded at2026-04-25T08:33:43.502452
Raw LLM Response
[ {"id":"rdc_n7rvvmd","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"rdc_n7s1jr7","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"rdc_n852d1o","responsibility":"company","reasoning":"consequentialist","policy":"regulate","emotion":"fear"}, {"id":"rdc_n810urv","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"rdc_n7slzrt","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"} ]