AI · Article
AWS Tests AI Agent Teams to Build a Cloud Emulator Without Human Coding
에너지경제신문 ·

At a glance
Amazon Web Services (AWS) let teams of AI agents write, test, and fix the code for a cloud emulator, a tool that lets people try AWS services on a personal computer. The AI agents wrote all of the code, and no human wrote or reviewed it. The project began as a weekend experiment by an AWS senior engineer.
Key points
- AWS used thousands of AI agents organized into three teams: one wrote tests, one ran them to find bugs, and one fixed the bugs.
- Fixed code was sent to other AI agents for checks on accuracy and security, and failed code was sent back for more work.
- The engineer first tested the AI coding tool Kiro by having it build part of the Amazon SQS messaging service, which took about 15 minutes.
- The first results were not ready for customers, and AWS said it still needs extra technical controls and engineering work before release.
Why it matters
It shows AI may take over much of software development work from people. It also shows that such code still needs more work before customers can use it.
Original source
“사람 대신 AI가 한 팀으로”…아마존의 별난 주말실험
에너지경제신문 · Nov 14, 06:00 KST
Read on 에너지경제신문 ↗AI summarized the original article. Check the original for accuracy. Copyright belongs to 에너지경제신문.
Other coverage
- AWS What's New · AWS Emulator is now available for building and testing AWS applications locally
- GeekNews · AWS가 마침내 자체 AWS 에뮬레이터를 만들었다
- 디지털투데이 · AWS, 클라우드 아닌 로컬서 코드 검증하는 오픈소스 ‘AWS 에뮬레이터' 공개
- Martin Cid Magazine · AWS Emulator, 아마존 200개 서비스 오프라인 재현… 코드는 AI 에이전트가 다 썼다
- AI 매터스 · AWS, 로컬 시험 도구 ‘AWS 에뮬레이터’ 공개…AI 에이전트가 개발
- 솔루션뉴스 · AWS, 노트북에서 클라우드 시험하는 '에뮬레이터' 공개…추가 요금 없다
Amazon Web Services · Ranking and more news
More from this field
- AWS Adds Shared Cache Skill to Cut Self-Hosted LLM Cost and Latency
AWS wrote about a shared cache that lets self-hosted AI models reuse saved work across servers, aiming to cut cost and wait time. The post shows how a coding agent can help set it up. Amazon says it can cut inference cost by up to 40% and first-token delay by up to 80%.
Amazon AWS Machine Learning blog · Nov 21, 09:14 KST - Can AI Really Find Black Friday TV Deals? A BBC Test Says Not Yet
A BBC team tested four popular AI chatbots to see if they could find a good Black Friday deal on a 55-inch TV. The results were mixed, with many links broken or out of date.
BBC News · Nov 21, 09:06 KST - Anthropic Mathematician Says Claude Helped Crack the Riemann Hypothesis
A video about a claim that an Anthropic mathematician, Levent Alpöge, used Claude to find a proof of the Riemann Hypothesis in about six weeks. The creator skips the heavy math and looks at what this could mean for AI and its future. The proof is still being checked.
TheAIGRID · Nov 21, 08:45 KST - Google Brings Duplex Calling to Gemini, Letting AI Phone Businesses for You
Google has moved its Duplex phone-calling technology into the Gemini app, so Gemini can now call businesses for you. The system has changed a lot since 2018, shifting from narrow custom models to a single stack built on Gemini.
Google Research blog · Nov 21, 08:30 KST