{"slug": "amazon-tests-a-safer-path-to-self-improving-ai", "title": "Amazon tests a safer path to self-improving AI", "summary": "Amazon unveiled Kiro-trace, an internal tracing system that improves Kiro, Amazon's AI coding agent, by mining its own runs for mistakes and testing proposed fixes before a human reviewer approves them. In its first week of operation, Kiro-trace opened eight changes to the agent and seven shipped, including one that raised the agent's ability to perform long tasks from 82% to 92%; the system originated from a single incorrect tool description that produced more than 61,000 errors. Shweta Garg, applied science lead for Kiro at AWS, said the proposing and judging roles are kept apart: \"The part that proposes a change never decides whether it worked. That decision runs in deterministic code the model cannot reach.", "body_md": "One mistake may have gotten Amazon researchers a step closer to self-improving AI.\n\nOn Tuesday, Amazon unveiled Kiro-trace, an internal tracing system that improves Kiro, Amazon's AI coding agent, from its own runs. Put simply, this system sifts through Kiro's previous work to find mistakes or wasted effort and proposes changes, then tests whether those changes actually improved the coding agent.\n\nThe idea originated from a mistake in Kiro's file-read and search tool, in which one incorrect description resulted in more than 61,000 errors. Though these errors were inconsequential, as they didn't cause anything to break down, because they didn't cause any larger problems, they likely would have gone undetected. Kiro-trace, however, spotted the error and removed it.\n\nThe researchers claim that Kiro needed to solve four problems simultaneously: Find problems that are worth fixing; create a solution that won't cause cascading errors in other places; make the solution general rather than \"tuned to a benchmark;\" and finally prove that the change actually worked.\n\nAdditionally, Amazon's researchers found a solution to a common problem in self-improving AI: Separating the system that's proposing the change from the system that's judging its efficacy.\n\n- The tracing system will find problems in Kiro's runs, propose changes one at a time, and only keep changes when a separate \"deterministic check\" proves it's a good change. After that, the change is sent to a human reviewer who approves it.\n- Most Kiro-trace change candidates don't actually make it to a human reviewer before the deterministic test rejects it, the research notes. \"The ones that do are the changes that can meaningfully improve the developer experience,\" the company said.\n- And it's already starting to show results: In the first week of operation, Kiro-trace opened eight changes to the agent, seven of which shipped. One of those changes improved the agent's ability to perform long tasks from 82% to 92%.\n\n\"When the same model proposes a change and scores it, it can optimize for its own score system, instead of the real goal, and the system drifts without anyone seeing it,\" Shweta Garg, applied science lead for Kiro at AWS, told The Deep View. \"Kiro-trace is built around keeping those two jobs apart. The part that proposes a change never decides whether it worked. That decision runs in deterministic code the model cannot reach.\"\n\nThe result for the Kiro user is an agent that simply works better, the researchers said. For instance, users can expect fewer wasted calls, more careful task delegation, and better tracking of long context conversations.\n\n## Our Deeper *View*\n\nThe conversation around recursive self-improvement is often paired with a growing fear of a loss of control, begging the question: If AI is capable of endlessly improving itself, will it eventually grow so advanced that humans cannot understand it? But Amazon's research may suggest a way to develop self-improving AI while mitigating that fear. By keeping humans in the loop and separating the checks and balances from a system that may be incentivized to improve in nonoptimal ways, we are automatically putting a harness on the system before we arrive at a point where it's too late to control it.", "url": "https://wpnews.pro/news/amazon-tests-a-safer-path-to-self-improving-ai", "canonical_source": "https://www.thedeepview.com/articles/amazon-tests-a-safer-path-to-self-improving-ai", "published_at": "2026-10-20 13:00:00+00:00", "updated_at": "2026-10-20 13:16:43.801316+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-agents", "ai-tools", "ai-safety", "developer-tools"], "entities": ["Amazon", "Kiro", "Kiro-trace", "AWS", "Shweta Garg"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/amazon-tests-a-safer-path-to-self-improving-ai", "markdown": "https://wpnews.pro/news/amazon-tests-a-safer-path-to-self-improving-ai.md", "text": "https://wpnews.pro/news/amazon-tests-a-safer-path-to-self-improving-ai.txt", "jsonld": "https://wpnews.pro/news/amazon-tests-a-safer-path-to-self-improving-ai.jsonld"}}