[muglug]

For select tasks the latest LLMs can speed things up by an order of magnitude.

Best example I’ve found: translating code from one language to another where there’s a large corpus of existing acceptance tests.

[casey2]

Again, no silver bullet. You will have to know what tasks it's capable of and how to elicit that solution. The bottleneck was never code the bottleneck still is solving the right problem in the right way.