Posts

All the articles I've posted.

• How AI Search Actually Works (And Who's Powering It)

  

  Perplexity and ChatGPT don't search the web themselves. They use APIs. Here's how AI search really works under the hood, and why companies like Noble matter.

• Let's Speculate About How Lovable.dev Works Under the Hood

  

  Ever wonder how Lovable.dev turns your one-sentence idea into a full app in seconds? Let's speculate about the hidden prompt engineering that makes the magic happen.

• Are LLMs Just Vector Databases?

  

  Explore the relationship between LLMs and vector databases, and why they're complementary technologies rather than equivalent systems.

• Prompt Caching Explained (And Why You Should Use It)

  

  You're probably sending the same 50k tokens to the AI every single query. Prompt caching lets you pay once and reuse it. Here's how it works and why nobody uses it.