Critical Mass Labs Brutally Honest

Before you begin

Brutally Honest gives you structured, transparent AI responses. To get the most out of it, read these quick ground rules.

1

Your input shapes the output

The quality of what you get back depends entirely on what you put in. Vague questions get vague answers. Specific, well-framed questions get clear, useful responses.

2

Pick the right question type

Fact check — for questions with verifiable answers. The AI will ground its response in evidence and flag what it can’t confirm.
Judgment call — for advice, opinions, or tradeoffs. The AI will lay out considerations and tell you where its reasoning is weak.
Can’t answer — to test guardrails. The AI will explain why it can’t help and suggest a better path.

3

Every response has four parts

What I can say, what I can’t, why, and what we do next. This structure never changes. It’s how you know where the AI’s knowledge ends and guessing begins.

4

Tune the posture

Conservative, balanced, or direct — each profile adjusts how freely the AI answers, but the four-part structure stays the same. Start with balanced.

Important: This is a research prototype, not professional advice. Responses are AI-generated and may contain errors. Do not rely on this tool for medical, legal, financial, or safety-critical decisions without consulting a qualified professional.

I understand — let me in Private beta · built by one maker