What are Agents?

Agents are in the heat of the storm. Since AI is like a slot-machine, and we have "predictability" problems, applying AI to "features" or anything "consumer facing" is a big risk.

Now there is this field where we can apply AI and "it's fine-ish" if it fails, because it happens "under the hood", that field is called "engineering". Coding agents are AI agents that help developers to write code. Therefore being an AI use case where you have more chances of success. Because 95% of AI Pilots Fail. Engineering is a safe bet and a safe start for AI adoption.

AI Agents is how we get AI use cases into engineering. There are many valid use cases for AI in engineering like:

• Code Assistants
• Code Generators
• Code Reviewers (with caution and caveats)
• Troubleshooting
• Documentation
• Testing (test generation)
• Code Migrations