Google's analysis of 15 million anonymized interactions with its Gemini App, AI Mode, and API finds that AI usage in the workplace remains shallow, with even adopting occupations touching a median of …
LeadDev's AI Impact Report 2026 found that 57% of respondents say tokenmaxxing is ineffective at gauging real value, and only 19% rate it as effective, as Meta, Amazon, Microsoft, and Uber have walked…
Engineering managers now need hands-on building experience to lead effectively, and the ability to spot AI-generated slop has become the critical review skill, according to LeadDev. The old speed-vs-q…
James Lowman, engineering team lead at trade infrastructure company Starboard, now asks candidates about the last three Claude Code skills they wrote, making access to leading agentic coding tools a h…
AI-powered development teams are shipping more code but delivering less business value, according to a LeadDev analysis warning that metrics like tokens, lines of code, and pull requests measure activ…
Vendict, a third-party risk management startup, rebuilt its software development lifecycle around the thesis that the SDLC itself is the context for AI agents, achieving reliable agent participation i…
Engineering managers are returning to hands-on coding, with the share of their time spent on technical work jumping from 20% in 2025 to 35% in 2026, according to LeadDev's Engineering Leadership Repor…
A new LeadDev survey finds that 81% of engineering teams give AI-coding agents write access to code repositories, while only 7% keep them read-only, and nearly a quarter reuse human credentials for ag…
A LeadDev.com article advises software engineers to delegate to AI only tasks that are verifiable, low risk, boring, and easy to define as done, warning that AI-generated work remains the engineer's r…
A controlled experiment by researchers testing Claude Opus 4.7 and Codex GPT-5.5 on 116 Python tasks found that pairing a weaker reviewer with a stronger writer can degrade performance: Claude's 91.4%…
McKinsey's survey found that 88% of organizations now use AI in at least one business function, but only 6% qualify as high performers and 50% have experienced negative consequences, highlighting a co…
AI adoption among engineering teams has surged 65%, but median pull request throughput has risen only 8%, according to data cited by LeadDev.com. The article argues that coding accounts for roughly 16…
Software engineering knowledge decay is structural, with fast cycles and churn stripping context even from teams that try to document, according to a LeadDev analysis. Most code work is upkeep—only 15…
A new analysis of AI-coding agents finds they are isolating individual developers rather than aiding teamwork, with 79% of agentic pull requests reviewed and modified by the same person without a seco…
Engineering manager roles have shifted from pure coordination to a 'player-coach' model, with 37% of leaders now deeply hands-on, according to a LeadDev article. AI tools enable managers to quickly re…
Senior engineers in 2026 now focus on 'system orchestration'—architecting and directing AI agents—rather than just coding, as AI-coding agents compress timelines from weeks to days, according to LeadD…
Synthetic LLM inference benchmarks misrepresent production performance because they use fixed prompt lengths, steady request rates, and single-model hardware, while real traffic is bursty and variable…
AI has broken traditional proxies for measuring engineering productivity, not productivity itself, according to Sherwin Wu, who leads engineering for OpenAI's API platform. At OpenAI, engineers using …
Influential leadership — the ability to move people and work forward without formal authority — is emerging as the key skill for engineering leaders as hierarchies flatten and AI enables more autonomo…
Engineering teams must stop measuring productivity by code output and start measuring validation throughput, argues a LeadDev.com analysis, because AI coding assistants produce far more code without s…