NASA’s new dark energy space telescope can also detect killer asteroids
technologyreview.com·2h ago
Software Engineering
The Pragmatic Engineer·Jun 23·5 min read
Slow down to speed up: so much has changed in 6 months’ time
TL;DR
Tech companies have faced significant challenges, such as Meta's major AI failure allowing unauthorized account changes. The keynote at Craft Conference emphasized the importance of slowing down to improve AI interactions and decision-making.
✦ Why It Matters
Engineers should prioritize thoughtful decision-making in AI development to avoid costly errors and enhance system reliability.
Key Takeaways
- Assess the balance between AI development and core product stability.
- Implement stricter code review processes for AI-generated changes.
- Monitor team productivity metrics to identify potential issues.
- Encourage open discussions about AI's impact on engineering practices.
How It Works
AI tools like Opus 4.5 and GPT-5.4 have become more capable, allowing developers to generate and integrate code at unprecedented rates. This has led to a significant increase in pull requests and code production, but also a decline in the thoroughness of code reviews, as teams struggle to keep up with the volume of changes.