It’s only recently that I’m starting to observe a totally unbounded sense of productivity at work, all attributed to AI acceleration in field of software engineering. Release of Claude Code marking it’s inception.
In the past seven days I have committed over 14,000 lines of code, in a programming language that I had only heard of before. Of course, line of code is not a very good signal of quality, but we have quality gates in place like test coverage, PR reviews, etcetera.
What really blocks this unbounded productivity is how can I get out of the way of AI agent and let it do the best work possible.