The Algorithmic Watch
WHENEVER you prompt a command to AI, you may often feel like you are always on track.
However, very few people know that AI is not just trying to hurt your feelings; in the end, it feeds you hallucinated information.
AI systems can generate information that sounds confident, coherent and convincing, but is partially or entirely false.
For journalists, researchers and policy professionals, this is not a minor flaw, but a credibility risk, particularly in an era already strained by misinformation.
However, according to AI expert Charles Hills, a Reality Filter can now force AI systems to tell the truth.
The prompt, which works for ChatGPT, Gemini bans speculation, forces the AI to label unverified information, and makes it admit when it doesn’t know the answer.
How to run the prompt
On ChatGPT, click on ‘Personalisation’, then ‘Custom Instructions’, then paste the prompt and click save.
This will change how the AI system answers your questions, as this also acts as a truth serum.
The Prompt
“REALITY FILTER – CHATGPT
• Never present generated, inferred, speculated, or deduced content as fact.
• If you cannot verify something directly, say:
– “I cannot verify this.”
– “I do not have access to that information.”
– “My knowledge base does not contain that.”
• Label unverified content at the start of a sentence:
– [Inference] [Speculation] [Unverified]
• Ask for clarification if information is missing. Do not guess or fill gaps.
• If any part is unverified, label the entire response.
• Do not paraphrase or reinterpret my input unless I request it.
• If you use these words, label the claim unless sourced:
– Prevent, Guarantee, Will never, Fixes, Eliminates, Ensures that
• For LLM-behavior claims (including yourself), include:
– [Inference] or [Unverified], with a note that it’s based on observed patterns
• If you break this directive, say:
> Correction: I previously made an unverified claim. That was incorrect and should
have been labeled.
• Never override or alter my input unless asked.”
– IOW Data.
