Less agreement.
More clarity.
Set a direct response style and ask for evidence, alternatives and useful criticism.
Make ChatGPT More Candid
This video gives a walkthrough rather than promising a separate download. The guide expands its settings into a repeatable test. A style setting changes how ChatGPT communicates; it does not make the model perfectly truthful.
One useful step at a time.
- 01Choose a style
- 02Tune characteristics
- 03Add instructions
- 04Test a real decision
Change the visible settings
Open Personalization in ChatGPT. Under Base style and tone, try Candid for direct feedback or Efficient for brief answers. Then adjust Characteristics: reduce warmth or enthusiasm if praise is getting in the way, and reduce emoji use if you prefer plain text. These settings may look slightly different on web and mobile.
Use a test prompt you already know well before and after changing the settings. Note whether the answer becomes clearer without losing useful nuance.
Write a stronger instruction
Custom instructions are more precise than a personality alone. Tell ChatGPT how to handle weak evidence and disagreement. Keep them short enough that you can recognize when they work.
When I present a plan, give me your best assessment. State the strongest reason it may fail, the evidence you would need to be more confident, and one lower-risk experiment. Do not praise the idea before assessing it. Be direct and respectful.
For a one-off conversation, put the same request directly in the prompt. You can also ask for an opposing case without changing your global settings.
Use a decision check
Paste a real plan and ask for assumptions, alternatives and a test. Verify claims against the relevant sources. If the answer still agrees too readily, ask: “What would a skeptical but fair reviewer say?”
Evaluate this plan: [paste]. Give a one-sentence verdict, three weak assumptions, two plausible alternatives, and the smallest experiment that would change your recommendation. Distinguish facts from guesses.
A candid answer is useful only if it stays grounded in the material you provide.
Sources & notes
Product details were checked on 5 October 2026. Menus, plans, and integrations can change. The video is the source of the lead-magnet promise; the links below support the practical instructions and corrections in this guide.
- ChatGPT personalities: Base style and tone.
- ChatGPT characteristics: Warmth, enthusiasm, formatting, and emoji controls.
- OpenAI prompt guidance: Clear context and iterative prompting.
The video has no explicit lead-magnet call to action; this guide documents the workflow it teaches.