AI Coding
AI that writes and reviews software.
AI coding covers AI systems that write, review, and modify software — coding agents, code generation, and the changing developer workflow, from pull-request review to legacy modernization.
Writing a plausible code snippet and resolving an issue in an existing repository are different tasks. SWE-bench gives a model a codebase and a real GitHub issue, then asks it to edit the repository. The original paper describes problems that require coordinated changes across functions, classes and files (Jimenez et al., SWE-bench). For your own coding workflow, specify both the intended behavior and the checks that should demonstrate it before accepting generated changes.
For delegated development, compare software factories, the surrounding agent harness, and cost per PR. Follow the workflow, runtime, and measurement questions before treating a higher output count as success.
Laurie Voss in episode 74 argues that deciding what to build becomes central as agents write more code. Greg Foster in episode 40 discusses the development stack around generated code, including review. Who reviews AI-generated code? follows responsibility through that workflow, and can AI modernize legacy code? covers changes to existing applications. The AI engineering hub addresses building the AI product itself.
18 episodes
- How Cisco Secures Personal AI Assistants for 90,000 Employees | DJ Sampath
- For AI Agent Fleets, the Model Is Rarely the Bottleneck | Theory Ventures' Andy Triedman
- What Breaks When Agents Become the Customer | Svix CEO Tom Hacohen
- Your Best AI Engineer Might Have the Worst Metrics | Sonar CTO Andrea Malagodi
- Slack Wants to Be the Context Harness for Code | CPO Jaime DeLanghe
- Most of the Web Will Never Get APIs for AI Agents | Dhruv Batra, Yutori
- The AI Framework Era Is Over: Why Context Is the Moat | Jerry Liu of LlamaIndex
- Every AI Agent Has an Evaluation Gap | Alex Ratner, Snorkel AI
- 250,000 Lines of Code/Week: Inside an AMD VP's Agent-First Workflow | Anush Elangovan
- Gemini 3 & Robot Dogs: Inside Google DeepMind's AI Experiments | Paige Bailey
- After Code Gen: What Graphite Is Building for the Post-AI Dev Stack | Greg Foster
- Vercel's Playbook for AI Agents: From Vibe Check to Production | Malte Ubl
- From Demo to Defensibility: How to Build an AI Business that Lasts | Aurimas Griciūnas
- How AI Velocity is Rewriting the Rules for Engineering Leaders | ChatPRD's Claire Vo
- Building an AI-Native Startup | GrowthX's Marcel Santilli
- Mastering Multi-Agent Systems | MongoDB’s Mikiko Chandrasekhar
- First Code, Then AGI: Software’s Event Horizon with Poolside Founders Jason Warner & Eiso Kant
- Using AI to Modernize Your Legacy Applications | MongoDB’s Rachelle Palmer
Explainers on this topic
- How AI Coding Agents Work
- How Reinforcement Learning Improves AI Coding Models
- Can an AI Agent Run Your Workday?
- The AI Hiring Doom Loop
- Code Review Is the New Bottleneck
- Using AI to Modernize Legacy Code
- How Much Autonomy Should an AI Agent Have?
Terms on this topic
- Agent Instruction File
- Agent Skills
- Agentic Engineering
- AI Code Generation
- AI Code Review
- AI Copilot
- AI Slop
- Code Completion
- Coding Agent
- Pass at k
- Product engineer
- Software Factory
- Spec-Driven Development
- SWE-bench
- Technical Debt
- Unit Test
- Vibe Coding
Essays on this topic
- My Robot Army: How I review AI code
- Software Engineering is Dead. Long Live Product Engineering
- The Future of AI Development ft. Gemini, Robotics & Space
Guests on this topic
DJ SampathAndy TriedmanTom HacohenAndrea MalagodiJaime DeLangheDhruv BatraJerry LiuAlex RatnerAnush ElangovanPaige BaileyGreg FosterMalte UblAurimas GriciūnasClaire VoMarcel SantilliMikiko ChandrasekharJason WarnerEiso KantRachelle Palmer