Three prevalent themes in the discussion
-
DSPy’s relevance amid newer LM‑tool approaches – Some argue that DSPy is outdated now that models can reliably use tool calls and agents.
“We moved from structured decoding to tool calls and agents a long time ago. It was more useful when models were too weak to hold syntax reliably.” – visarga
-
Current limitations of language models – Contributors note that while models excel at syntactic tasks, they still struggle with attention to detail, creativity, planning, and decision‑making.
“Myself, a PhD in math, have less attention to detail than models. Now their main weakness is creativity, planning and decision making.” – vatsachak
-
Continued value of DSPy techniques – Defenders stress that DSPy’s methods (feedback, iterative improvement, deeper theory) remain useful despite claims that models alone are sufficient.
“Techniques in DSPy are still extremely valuable. Your comment is akin to saying, we don't need feedback because open loop control is so good now. It is until it isn't. I don't need a deep understanding of the underlying theory because my gut reaction is always right, until it isn't.” – genxy