Updated
Updated · InfoWorld · Jul 15
AI Agents Target 60%-70% of Developer Work as Coding Bottleneck Shifts to Production
Updated
Updated · InfoWorld · Jul 15

AI Agents Target 60%-70% of Developer Work as Coding Bottleneck Shifts to Production

3 articles · Updated · InfoWorld · Jul 15

Summary

  • AI agents are emerging as the next tool for production troubleshooting, as agentic coding makes code generation fast enough that maintenance and support now dominate developers’ workload.
  • 60% to 70% of a developer’s job could be the next target for automation, with post-coding work such as reading logs, reproducing failures and tracing complex settings becoming the new choke point.
  • Production incidents are a strong fit for agents because they can process thousands of log entries in seconds or minutes and keep far more context than human engineers.
  • Resolve AI CEO Spiros Xanthos said purpose-built AI can take on more of the cognitive load of production work, freeing developers to spend more time improving systems.
  • The shift suggests software teams will increasingly apply AI beyond writing code, focusing on the harder, less bounded problem of diagnosing and fixing live-system failures.

Insights

Is 'AI brain fry' the new developer burnout as humans oversee untrustworthy, AI-generated code?
AI writes 4x more code, so why do 90% of CEOs report zero productivity gains from it?
When 98% of AI agents can access private data and the web, is a catastrophic security failure inevitable?