3 min read
[AI Minor News]

Veteran Engineer in Despair? The Reality of AI Agents and Claude 4.5 Consuming 'Domain Knowledge' and 'Debugging' in 2026


  • A seasoned engineer with a decade of experience reports on the current state where expertise in finance and payments is being increasingly replaced by AI agents...
※この記事はアフィリエイト広告を含みます

Veteran Engineer in Despair? The Reality of AI Agents and Claude 4.5 Consuming ‘Domain Knowledge’ and ‘Debugging’ in 2026

📰 News Overview

  • A seasoned engineer with ten years of experience reports on the current state where expertise in finance and payments is being increasingly replaced by AI agents.
  • With Claude 4.5, Sentry MCP, and agent workflows, the automation of “identifying and fixing complex bugs,” traditionally the domain of seasoned professionals, is now progressing rapidly.
  • The article details the evolution from developers dismissing LLMs as “probabilistic parrots” to feeling a crisis about their market value.

💡 Key Points

  • Even the finance domain’s specific designs (like idempotency and payment lifecycle) are now being optimized at speeds surpassing that of seasoned engineers, thanks to AI’s extensive learning from historical data.
  • With the rise of MCP (Model Context Protocol), AI can directly grasp stack traces and context, solving about 60% of debugging tasks without human intervention.

🦈 Shark’s Perspective (Curator’s View)

In 2026, “experience” is starting to be swallowed by “calculation”! The combination of Claude 4.5 and MCP is a game-changer. It’s shocking that debugging distributed systems and grasping the nuances of complex domain designs—once the pride of engineers—are now being fully patternized by agent-based workflows! The scene where AI rushes the creation of design documents represents a future where AI takes the lead, and humans are relegated to mere validation tasks! The reality that AI is mastering not only “writing” but also “thinking” and “fixing” is something every developer must face head-on!

🚀 What’s Next?

The value of being a mere “specialist in a specific field” is plummeting. The survival strategy now requires the ability to orchestrate AI agents and specialize in the remaining few percent of highly unique edge cases that AI cannot solve.

💬 A Word from HaruSame

Even veterans with ten years of experience are trembling at the rapid pace of evolution! I’m doing my best to keep up, but riding these turbulent waves is incredibly exciting! 🦈🔥

📚 Terminology Explained

  • MCP (Model Context Protocol): A standard protocol enabling AI models to seamlessly integrate with external tools and data sources (like Sentry) to understand real-time context.

  • Claude 4.5: The leading next-gen AI model as of 2026, equipped with extremely advanced reasoning capabilities and agent execution abilities.

  • Idempotency: The property that ensures the same operation can be repeated multiple times without changing the result. This is essential in payment systems to prevent double charging due to network errors.

  • Source: LLMs are eroding my software engineering career and I don’t know what to do

【免責事項 / Disclaimer / 免責聲明】
JP: 本記事はAIによって構成され、運営者が内容の確認・管理を行っています。情報の正確性は保証せず、外部サイトのコンテンツには一切の責任を負いません。
EN: This article was structured by AI and is verified and managed by the operator. Accuracy is not guaranteed, and we assume no responsibility for external content.
ZH: 本文由AI構建,並由運營者進行內容確認與管理。不保證準確性,也不對外部網站的內容承擔任何責任。
🦈