The archive.

On 15 September, a start-up released an AI model that can't write a single sentence. Five weeks later, it has USD 870 million in fresh funding, an OpenAI copy and more than 30 rivals - so we put it to the big durian test.

Most GenAI pilots fail before they matter. Beat the odds: pick the right tasks to automate, demand the right framework and platform, and ship apps that get better with every conversation.

LLMs don't think; they predict the next token. A walkthrough of the engineering tricks — from prompting and RAG to tool use and Tree of Thoughts — that force a simple autocomplete to behave intelligently.

Let's say your "super-intelligent" agentic chatbot — the one with access to sensitive customer data — is hijacked. You've effectively welcomed a genius-level saboteur behind your own defense lines.