{
  "id": 14411263,
  "title": "Your Best AI Asset Has Been Working Here for Years | Opinion",
  "url": "https://urgent.news/2026/10/14/your-best-ai-asset-has-been-working-here-for-years-opinion",
  "topic": "ai",
  "section": "AI",
  "published": "2026-10-14T09:00:00.000Z",
  "source": {
    "name": "Newsweek",
    "slug": "newsweek",
    "url": "https://www.newsweek.com/your-best-ai-asset-has-been-working-here-for-years-opinion-12560490"
  },
  "original_language": "en",
  "account": "The story highlights the challenges of implementing AI in real-world operations. A chief executive officer shared an example of an AI dispatch system predicting a storm's impact on a utility's operations. The AI suggested stationing crews at a specific depot to be ready for the storm. However, the CEO overrode this recommendation and sent crews to another location, which led to faster restoration of power. This scenario illustrates the main issue with AI: without a deep understanding of how things work in the real world, even the best AI model can be ineffective. Many AI pilots and proofs of concept fail to transition to production due to this lack of practical knowledge. A global study found that the percentage of companies abandoning most AI initiatives before reaching production increased from 17% to 42% year over year, with approximately 46% of projects scrapped between the proof of concept and full adoption. Despite advancements in AI technology, businesses struggle to translate promising demos into fully functional solutions. AI systems need more than just data; they require an understanding of the business context, business semantics, and human expertise. Experienced employees can help define what the AI system should know, do, and stop doing. This partnership between AI and human experts enables faster resolutions, better decisions, and reduced errors. A U.S. specialty insurer collaborated with Wipro Intelligence to build an AI system for underwriters processing complex commercial-risk submissions. While the AI could extract relevant data and identify anomalies, it was the underwriters' expertise that distinguished a harmless inconsistency from a material exposure. The AI system was successful not by becoming an underwriter itself but by working alongside human underwriters, handling tasks faster while still relying on human judgment when necessary. For CEOs, this story emphasizes that AI should not be treated as a standalone technology project, but rather as a business redesign effort requiring top-level ownership. To successfully implement AI, organizations should start with workflows where a better decision or faster resolution holds clear value, involve the people who know that work best in the technology development process, establish a shared definition of success, and tailor the AI models to the specific tasks at hand based on operational expertise. The key to making AI work lies in recognizing the immense value that resides in the heads of employees who truly understand how things get done.",
  "summary": "Without a deep, nuanced understanding of how things work in the real world, even the best AI model will be functionally hamstrung.",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}