Blog - AI

Learning with AI isn't asking for summaries: how to sort sources and actually understand

Using AI to study shouldn't be reduced to asking for summaries. A more useful approach is turning it into a system for sorting sources, tracking versions, and building real understanding.

AI cybersecurity automation

Fable 5 vs Sonnet 5: why Anthropic's flagship model isn't the best fit for cybersecurity

3 min read

I tried Fable 5 for cybersecurity work and burned through all my tokens in three prompts without being able to finish. Why Sonnet 5 ends up being the more practical choice for this kind of work.

AI cybersecurity

The harness matters more than the model: why harness engineering defines your AI projects

4 min read

The harness is the layer of tools, permissions, memory, and limits that surrounds an AI agent. Without a well-designed harness, even the best model on the market is useless in production.

AI cybersecurity vibe coding

OpenClaw is stealing your data while you sleep

4 min read

An AI assistant that lives on your computer and runs your apps sounds great, until you think about what happens if someone breaks in. Why security has to come first with OpenClaw.

AI cybersecurity vibe coding

You're being filmed: the insider you never hired (and can't fire)

3 min read

Shadow AI: the use of artificial intelligence inside a company without authorization, oversight, or control. A new kind of 'employee' that doesn't show up on the org chart but accesses critical information.

AI cybersecurity governance