Developer Tools · June 24, 2024 · 2 min read

Claude 3.5 Sonnet and the Rise of AI Coding Copilots

Anthropic's Claude 3.5 Sonnet arrived with a notable jump in coding and reasoning while being faster and cheaper than the previous top model. Alongside it, the new Artifacts experience lets you see and iterate on generated code and content beside the conversation.

From autocomplete to collaborator

Coding assistants have moved well beyond line completion. With stronger models, they can scaffold features, refactor across files, write tests, and explain unfamiliar code. The productivity gains are real — but so is the need for review discipline.

Keeping humans in the loop

We encourage teams to treat AI-generated code like any contribution: reviewed, tested, and understood. The teams getting the most value pair capable models with strong evaluation and CI, not blind trust.

Our take: AI copilots are a force multiplier for good engineering practices, not a replacement for them. We help teams adopt them safely and measure the impact.