Tool Use Is the New Reasoning: Notes from fine-tuned distillation

For most of the last year, the conversation around long-context workflows has been louder than the evidence. That is starting to change.

Eval harnesses, once an afterthought, are becoming the most important piece of code in many AI projects. Duolingo's team treats theirs the way an SRE team treats a runbook.

Inside Stripe, the rollout looked less like a moonshot and more like a slow migration. A pilot, a champion, a quiet expansion, a budget line.

Eval harnesses, once an afterthought, are becoming the most important piece of code in many AI projects. Atlassian's team treats theirs the way an SRE team treats a runbook.

Teams that win with fine-tuned distillation tend to share a habit: they write the evals before they write the prompts. Everything else follows from that.

The skeptical read is that we are watching a feature, not a platform. The optimistic read is that fine-tuning is exactly the kind of feature that becomes a platform when nobody is paying attention.

What xAI actually shipped with Llama 4 is less a single capability and more a cluster of small, compounding improvements — the kind that only show up when you put a real workflow on top.

What Alibaba Qwen actually shipped with Command R+ 2 is less a single capability and more a cluster of small, compounding improvements — the kind that only show up when you put a real workflow on top.

None of this guarantees a clean story. Anthropic could ship a model next month that rearranges the assumptions in this piece. But the direction of travel, for now, is clear enough to plan around.