Steven Gonsalvez

Software Engineer

Harvard researchers find AI coding agents shifted the bottleneck to code review

Why CEREBRO kept it

Coding agents shifting bottleneck to review, relevant research

The text below is an automated extraction of the article at https://complexdiscovery.com/harvard-researchers-find-ai-coding-agents-shifted-the-bottleneck-to-code-review/, stored verbatim in the public cerebro-vault repository. Copyright remains with the original publisher (complexdiscovery.com).

Editor’s Note: Engineers at firms that adopted AI coding agents added about 30 percent more code, yet finished work did not measurably rise and getting code through review took 49 percent longer, according to a working paper by Harvard Ph.D. candidates Fiona Chen and James Stratton covering 718 firms. The paper, a job-market paper with no indication of completed peer review, measures what happens at the human checkpoint once drafting speeds up. Harvey’s motion-to-dismiss workflow and Cooley’s redaction agent, still in development with Google Cloud, both keep a lawyer at a comparable checkpoint

Community take

Top pushback says the real bottleneck is faulty up-front specification rather than review, while others find the finding obvious and say the impact depends on each company's SDLC bureaucracy.

Backlinks

Appeared in 1 briefing

Related

Shares tags: ai/agents · cerebro/signal

Also from complexdiscovery.com

Only signal from complexdiscovery.com so far.