Search Agent Sky
← Back to search Memory Lane Recent answers
Cited source trail
What are tools you have made for yourself since the advent of AI?
Sources checkedanthropic.comtechnologyreview.complaytechtrend.comdevin.ai
Next step

Research any question with live sources, then publish the cited answer as a free shareable link.

Based on my research, here are the key tools that AI systems have created for themselves since the advent of modern AI: ## **AI Coding Agents That Write Their Own Code** **Devin** (by Cognition AI) - An autonomous AI software engineer that can: - Write, debug, and deploy code independently - Build its own tools and scripts for repetitive tasks - Learn from examples and improve its own performance over time - Handle complex migrations and refactoring tasks **Claude Code** (by Anthropic) - An agentic coding system that: - Reads entire codebases and makes changes across multiple files - Runs tests and fixes failures automatically - Handles CI/CD pipeline management - Is used by Anthropic itself for the majority of its code development **GPT-Engineer** - An open-source tool that: - Generates complete codebases from natural language descriptions - Can debug and refine its own code outputs **AutoGen** (by Microsoft) - A framework for building multi-agent conversations where: - AI agents can collaborate and improve their own workflows - Agents can critique and refine each other's outputs ## **Self-Improvement Mechanisms** **Recursive Self-Improvement Tools** - Systems that: - Write and evaluate their own code - Debug and patch their own logic without human intervention - Use reward models and feedback loops to optimize performance - Build task-specific subprocesses autonomously **Auto-Evaluation Systems** - Tools that: - Critique their own outputs and prompt themselves with better formulations - Score themselves against success metrics - Use chain-of-thought reasoning to refine their logic ## **Real-World Applications** At companies like Nubank, Devin has: - Reduced migration time by 8-12x - Achieved 20x cost savings on large-scale code refactoring - Built custom scripts to automate repetitive tasks - Improved its own performance through continuous learning These tools represent a shift from traditional AI development (train → deploy → monitor) to autonomous loops where AI systems can "build → run → evolve" without constant human intervention.

Sources visited

1searchAI tools created by AI systems for themselves self-improvement
2navigatehttps://www.anthropic.com/institute/recursive-self-improvement
3get_text
4searchAI tools created by AI systems for their own development and improvement
5navigatehttps://www.technologyreview.com/2025/08/06/1121193/five-ways-that-ai-is-learning-to-improve-itself/
6get_text
7searchAI coding assistants created by AI systems for their own development Claude Code Cursor
8navigatehttps://www.anthropic.com/product/claude-code
9searchAI systems creating tools for themselves self-improvement development tools
10navigatehttps://playtechtrend.com/self-improving-ai/
11get_text
12searchAI coding tools created by AI systems Claude Code Devin GPT-Engineer AutoGen
13navigatehttps://devin.ai
14get_text
Shared by Unchained Sky · Jun 8, 2026

Ask your own question

Search live sources for a question you choose. You can publish the cited answer as a free link.