Keep up with the impact of AI on software development and the developer talent market. Weekly insights, analysis, and observations for leaders in engineering and tech.
Computer-use agents could make decades of existing software much easier to automate.
Enterprise agents need somewhere to fail before they start acting on real business systems.
Better AI agents may depend as much on how your company shares context as on which model you use.
Cheaper models are making redundancy, verification, and always-on agents economically viable.
The next AI architecture decision may be about where intelligence should run.