blog draft about reasoning

[sonichi]

Why are these changes needed?

Related issue number

Checks

• I've included any doc changes needed for https://docs.ag2.ai/. See https://docs.ag2.ai/docs/contributor-guide/documentation to build and test documentation locally.
• I've added tests (if relevant) corresponding to the changes introduced in this PR.
• I've made sure all auto checks have passed.

[giorgossideris]

Very interesting, I am accepting!

One additional thought I had is that today’s (at least generally accessible) LLMs are primarily trained to deliver direct, complete answers to queries, rather than to offer the kind of intuitive, exploratory thinking you’re describing. For a reasoning system to really work in this context, I think it’s essential to have models specifically trained for agentic tasks, models that are comfortable producing thought-progressing responses or even unstructured reflections, instead of the structured, human-pleasing outputs shaped by RLHF.

[davorrunje]

"Linear Logic" is probably not the best term here as it can be easily confused with a substructural logic proposed by French logician Jean-Yves Girard, at least that was my association and it might be common by people with the theoretical computer science background. Please see https://en.wikipedia.org/wiki/Linear_logic