Thanks for explaining your point—that viability of inference scaling makes development differentially safer (all else equal) seems right.
Thanks for explaining your point—that viability of inference scaling makes development differentially safer (all else equal) seems right.